diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14526 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2069,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00048344210780759005,
+      "grad_norm": 2.4757904153326913,
+      "learning_rate": 0.0,
+      "loss": 0.9924,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009668842156151801,
+      "grad_norm": 2.3793618820940923,
+      "learning_rate": 9.615384615384617e-08,
+      "loss": 0.9738,
+      "step": 2
+    },
+    {
+      "epoch": 0.0014503263234227702,
+      "grad_norm": 2.375691201697703,
+      "learning_rate": 1.9230769230769234e-07,
+      "loss": 0.9588,
+      "step": 3
+    },
+    {
+      "epoch": 0.0019337684312303602,
+      "grad_norm": 2.3403619553808497,
+      "learning_rate": 2.884615384615385e-07,
+      "loss": 0.9862,
+      "step": 4
+    },
+    {
+      "epoch": 0.00241721053903795,
+      "grad_norm": 2.3613475552419394,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 0.9758,
+      "step": 5
+    },
+    {
+      "epoch": 0.0029006526468455403,
+      "grad_norm": 2.374422358129782,
+      "learning_rate": 4.807692307692308e-07,
+      "loss": 0.9716,
+      "step": 6
+    },
+    {
+      "epoch": 0.00338409475465313,
+      "grad_norm": 2.478706471115894,
+      "learning_rate": 5.76923076923077e-07,
+      "loss": 0.976,
+      "step": 7
+    },
+    {
+      "epoch": 0.0038675368624607204,
+      "grad_norm": 2.3811968693026198,
+      "learning_rate": 6.730769230769231e-07,
+      "loss": 0.9873,
+      "step": 8
+    },
+    {
+      "epoch": 0.00435097897026831,
+      "grad_norm": 2.2147341913024956,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 0.9286,
+      "step": 9
+    },
+    {
+      "epoch": 0.0048344210780759,
+      "grad_norm": 2.335255162349414,
+      "learning_rate": 8.653846153846154e-07,
+      "loss": 0.9845,
+      "step": 10
+    },
+    {
+      "epoch": 0.005317863185883491,
+      "grad_norm": 2.218894644037587,
+      "learning_rate": 9.615384615384617e-07,
+      "loss": 0.95,
+      "step": 11
+    },
+    {
+      "epoch": 0.005801305293691081,
+      "grad_norm": 2.2403773948516226,
+      "learning_rate": 1.0576923076923078e-06,
+      "loss": 0.9715,
+      "step": 12
+    },
+    {
+      "epoch": 0.0062847474014986705,
+      "grad_norm": 2.164785866254398,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 0.9422,
+      "step": 13
+    },
+    {
+      "epoch": 0.00676818950930626,
+      "grad_norm": 2.2075110374685947,
+      "learning_rate": 1.25e-06,
+      "loss": 0.965,
+      "step": 14
+    },
+    {
+      "epoch": 0.007251631617113851,
+      "grad_norm": 1.9308798071113116,
+      "learning_rate": 1.3461538461538462e-06,
+      "loss": 0.9372,
+      "step": 15
+    },
+    {
+      "epoch": 0.007735073724921441,
+      "grad_norm": 1.8705357350667309,
+      "learning_rate": 1.4423076923076922e-06,
+      "loss": 0.9443,
+      "step": 16
+    },
+    {
+      "epoch": 0.00821851583272903,
+      "grad_norm": 1.775691766227149,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 0.9362,
+      "step": 17
+    },
+    {
+      "epoch": 0.00870195794053662,
+      "grad_norm": 1.7290053738093054,
+      "learning_rate": 1.6346153846153848e-06,
+      "loss": 0.9298,
+      "step": 18
+    },
+    {
+      "epoch": 0.00918540004834421,
+      "grad_norm": 1.6541389612298973,
+      "learning_rate": 1.7307692307692308e-06,
+      "loss": 0.9336,
+      "step": 19
+    },
+    {
+      "epoch": 0.0096688421561518,
+      "grad_norm": 1.2338607620225968,
+      "learning_rate": 1.826923076923077e-06,
+      "loss": 0.9055,
+      "step": 20
+    },
+    {
+      "epoch": 0.01015228426395939,
+      "grad_norm": 1.1808086522456918,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 0.8962,
+      "step": 21
+    },
+    {
+      "epoch": 0.010635726371766982,
+      "grad_norm": 1.090531117286559,
+      "learning_rate": 2.0192307692307692e-06,
+      "loss": 0.8702,
+      "step": 22
+    },
+    {
+      "epoch": 0.011119168479574571,
+      "grad_norm": 1.095517820053717,
+      "learning_rate": 2.1153846153846155e-06,
+      "loss": 0.8816,
+      "step": 23
+    },
+    {
+      "epoch": 0.011602610587382161,
+      "grad_norm": 1.0208393908518454,
+      "learning_rate": 2.211538461538462e-06,
+      "loss": 0.8699,
+      "step": 24
+    },
+    {
+      "epoch": 0.012086052695189751,
+      "grad_norm": 1.004109121666044,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 0.8669,
+      "step": 25
+    },
+    {
+      "epoch": 0.012569494802997341,
+      "grad_norm": 0.98169412760157,
+      "learning_rate": 2.403846153846154e-06,
+      "loss": 0.8371,
+      "step": 26
+    },
+    {
+      "epoch": 0.01305293691080493,
+      "grad_norm": 0.9209444270757048,
+      "learning_rate": 2.5e-06,
+      "loss": 0.8388,
+      "step": 27
+    },
+    {
+      "epoch": 0.01353637901861252,
+      "grad_norm": 0.8619822284316448,
+      "learning_rate": 2.5961538461538465e-06,
+      "loss": 0.8041,
+      "step": 28
+    },
+    {
+      "epoch": 0.01401982112642011,
+      "grad_norm": 0.9241232197315488,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 0.8091,
+      "step": 29
+    },
+    {
+      "epoch": 0.014503263234227702,
+      "grad_norm": 0.917429451582305,
+      "learning_rate": 2.7884615384615386e-06,
+      "loss": 0.7749,
+      "step": 30
+    },
+    {
+      "epoch": 0.014986705342035292,
+      "grad_norm": 0.9043786370452085,
+      "learning_rate": 2.8846153846153845e-06,
+      "loss": 0.8144,
+      "step": 31
+    },
+    {
+      "epoch": 0.015470147449842882,
+      "grad_norm": 0.820031258272968,
+      "learning_rate": 2.980769230769231e-06,
+      "loss": 0.7931,
+      "step": 32
+    },
+    {
+      "epoch": 0.01595358955765047,
+      "grad_norm": 0.7726701862119408,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.7903,
+      "step": 33
+    },
+    {
+      "epoch": 0.01643703166545806,
+      "grad_norm": 0.6871638945331215,
+      "learning_rate": 3.1730769230769233e-06,
+      "loss": 0.7256,
+      "step": 34
+    },
+    {
+      "epoch": 0.01692047377326565,
+      "grad_norm": 0.7302580243312591,
+      "learning_rate": 3.2692307692307696e-06,
+      "loss": 0.7956,
+      "step": 35
+    },
+    {
+      "epoch": 0.01740391588107324,
+      "grad_norm": 0.6737498928543134,
+      "learning_rate": 3.365384615384616e-06,
+      "loss": 0.7478,
+      "step": 36
+    },
+    {
+      "epoch": 0.01788735798888083,
+      "grad_norm": 0.6914440787905148,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 0.7621,
+      "step": 37
+    },
+    {
+      "epoch": 0.01837080009668842,
+      "grad_norm": 0.6869329802424697,
+      "learning_rate": 3.557692307692308e-06,
+      "loss": 0.7706,
+      "step": 38
+    },
+    {
+      "epoch": 0.01885424220449601,
+      "grad_norm": 0.6549740713679569,
+      "learning_rate": 3.653846153846154e-06,
+      "loss": 0.7237,
+      "step": 39
+    },
+    {
+      "epoch": 0.0193376843123036,
+      "grad_norm": 0.6922145753217636,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.7537,
+      "step": 40
+    },
+    {
+      "epoch": 0.01982112642011119,
+      "grad_norm": 0.665714225887781,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.7656,
+      "step": 41
+    },
+    {
+      "epoch": 0.02030456852791878,
+      "grad_norm": 0.6390200477155564,
+      "learning_rate": 3.942307692307692e-06,
+      "loss": 0.7558,
+      "step": 42
+    },
+    {
+      "epoch": 0.020788010635726373,
+      "grad_norm": 0.6558268717213803,
+      "learning_rate": 4.0384615384615385e-06,
+      "loss": 0.7408,
+      "step": 43
+    },
+    {
+      "epoch": 0.021271452743533963,
+      "grad_norm": 0.6191043915893901,
+      "learning_rate": 4.134615384615385e-06,
+      "loss": 0.7482,
+      "step": 44
+    },
+    {
+      "epoch": 0.021754894851341553,
+      "grad_norm": 0.6193196066166552,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 0.7424,
+      "step": 45
+    },
+    {
+      "epoch": 0.022238336959149143,
+      "grad_norm": 0.6141046550876093,
+      "learning_rate": 4.326923076923077e-06,
+      "loss": 0.7372,
+      "step": 46
+    },
+    {
+      "epoch": 0.022721779066956733,
+      "grad_norm": 0.6265315399192994,
+      "learning_rate": 4.423076923076924e-06,
+      "loss": 0.7362,
+      "step": 47
+    },
+    {
+      "epoch": 0.023205221174764323,
+      "grad_norm": 0.6704953048927751,
+      "learning_rate": 4.51923076923077e-06,
+      "loss": 0.7326,
+      "step": 48
+    },
+    {
+      "epoch": 0.023688663282571912,
+      "grad_norm": 0.6544522629648533,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 0.7275,
+      "step": 49
+    },
+    {
+      "epoch": 0.024172105390379502,
+      "grad_norm": 0.6277879949612973,
+      "learning_rate": 4.711538461538462e-06,
+      "loss": 0.7311,
+      "step": 50
+    },
+    {
+      "epoch": 0.024655547498187092,
+      "grad_norm": 0.5924725867824154,
+      "learning_rate": 4.807692307692308e-06,
+      "loss": 0.7261,
+      "step": 51
+    },
+    {
+      "epoch": 0.025138989605994682,
+      "grad_norm": 0.591545350722231,
+      "learning_rate": 4.903846153846154e-06,
+      "loss": 0.7092,
+      "step": 52
+    },
+    {
+      "epoch": 0.025622431713802272,
+      "grad_norm": 0.5698079528908845,
+      "learning_rate": 5e-06,
+      "loss": 0.7093,
+      "step": 53
+    },
+    {
+      "epoch": 0.02610587382160986,
+      "grad_norm": 0.6162650749995418,
+      "learning_rate": 5.096153846153846e-06,
+      "loss": 0.687,
+      "step": 54
+    },
+    {
+      "epoch": 0.02658931592941745,
+      "grad_norm": 0.5577930499697958,
+      "learning_rate": 5.192307692307693e-06,
+      "loss": 0.7143,
+      "step": 55
+    },
+    {
+      "epoch": 0.02707275803722504,
+      "grad_norm": 0.5640209708881836,
+      "learning_rate": 5.288461538461539e-06,
+      "loss": 0.7059,
+      "step": 56
+    },
+    {
+      "epoch": 0.02755620014503263,
+      "grad_norm": 0.5430126891143467,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 0.704,
+      "step": 57
+    },
+    {
+      "epoch": 0.02803964225284022,
+      "grad_norm": 0.6009517822786309,
+      "learning_rate": 5.480769230769232e-06,
+      "loss": 0.7169,
+      "step": 58
+    },
+    {
+      "epoch": 0.02852308436064781,
+      "grad_norm": 0.6167733540891279,
+      "learning_rate": 5.576923076923077e-06,
+      "loss": 0.7065,
+      "step": 59
+    },
+    {
+      "epoch": 0.029006526468455404,
+      "grad_norm": 0.5731458421774205,
+      "learning_rate": 5.6730769230769235e-06,
+      "loss": 0.6373,
+      "step": 60
+    },
+    {
+      "epoch": 0.029489968576262994,
+      "grad_norm": 0.5489403473955915,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 0.7018,
+      "step": 61
+    },
+    {
+      "epoch": 0.029973410684070584,
+      "grad_norm": 0.5325599545502842,
+      "learning_rate": 5.865384615384616e-06,
+      "loss": 0.6959,
+      "step": 62
+    },
+    {
+      "epoch": 0.030456852791878174,
+      "grad_norm": 0.5478537783639954,
+      "learning_rate": 5.961538461538462e-06,
+      "loss": 0.6896,
+      "step": 63
+    },
+    {
+      "epoch": 0.030940294899685764,
+      "grad_norm": 0.5363515063211778,
+      "learning_rate": 6.057692307692308e-06,
+      "loss": 0.7014,
+      "step": 64
+    },
+    {
+      "epoch": 0.03142373700749335,
+      "grad_norm": 0.5641946867306303,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.6903,
+      "step": 65
+    },
+    {
+      "epoch": 0.03190717911530094,
+      "grad_norm": 0.5481835775113026,
+      "learning_rate": 6.25e-06,
+      "loss": 0.6893,
+      "step": 66
+    },
+    {
+      "epoch": 0.03239062122310853,
+      "grad_norm": 0.5143476489389097,
+      "learning_rate": 6.3461538461538466e-06,
+      "loss": 0.6946,
+      "step": 67
+    },
+    {
+      "epoch": 0.03287406333091612,
+      "grad_norm": 0.590656354467126,
+      "learning_rate": 6.442307692307693e-06,
+      "loss": 0.6788,
+      "step": 68
+    },
+    {
+      "epoch": 0.03335750543872371,
+      "grad_norm": 0.5203563663337313,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 0.6847,
+      "step": 69
+    },
+    {
+      "epoch": 0.0338409475465313,
+      "grad_norm": 0.5084500426939229,
+      "learning_rate": 6.6346153846153846e-06,
+      "loss": 0.7086,
+      "step": 70
+    },
+    {
+      "epoch": 0.03432438965433889,
+      "grad_norm": 0.5032784264719405,
+      "learning_rate": 6.730769230769232e-06,
+      "loss": 0.6724,
+      "step": 71
+    },
+    {
+      "epoch": 0.03480783176214648,
+      "grad_norm": 0.5205048813341548,
+      "learning_rate": 6.826923076923078e-06,
+      "loss": 0.6592,
+      "step": 72
+    },
+    {
+      "epoch": 0.03529127386995407,
+      "grad_norm": 0.5066251849073853,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 0.6674,
+      "step": 73
+    },
+    {
+      "epoch": 0.03577471597776166,
+      "grad_norm": 0.5305760257061701,
+      "learning_rate": 7.01923076923077e-06,
+      "loss": 0.6665,
+      "step": 74
+    },
+    {
+      "epoch": 0.03625815808556925,
+      "grad_norm": 0.5583636863825877,
+      "learning_rate": 7.115384615384616e-06,
+      "loss": 0.6685,
+      "step": 75
+    },
+    {
+      "epoch": 0.03674160019337684,
+      "grad_norm": 0.6055857508188283,
+      "learning_rate": 7.211538461538462e-06,
+      "loss": 0.6826,
+      "step": 76
+    },
+    {
+      "epoch": 0.03722504230118443,
+      "grad_norm": 0.5576393446552599,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 0.6811,
+      "step": 77
+    },
+    {
+      "epoch": 0.03770848440899202,
+      "grad_norm": 0.5550469150359895,
+      "learning_rate": 7.403846153846155e-06,
+      "loss": 0.6752,
+      "step": 78
+    },
+    {
+      "epoch": 0.03819192651679961,
+      "grad_norm": 0.5095358853416947,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.657,
+      "step": 79
+    },
+    {
+      "epoch": 0.0386753686246072,
+      "grad_norm": 0.519449515803278,
+      "learning_rate": 7.5961538461538465e-06,
+      "loss": 0.6326,
+      "step": 80
+    },
+    {
+      "epoch": 0.03915881073241479,
+      "grad_norm": 0.5360371671954463,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.6577,
+      "step": 81
+    },
+    {
+      "epoch": 0.03964225284022238,
+      "grad_norm": 0.5220981103197152,
+      "learning_rate": 7.78846153846154e-06,
+      "loss": 0.6803,
+      "step": 82
+    },
+    {
+      "epoch": 0.04012569494802997,
+      "grad_norm": 0.5357442529489778,
+      "learning_rate": 7.884615384615384e-06,
+      "loss": 0.662,
+      "step": 83
+    },
+    {
+      "epoch": 0.04060913705583756,
+      "grad_norm": 0.5922567510802571,
+      "learning_rate": 7.980769230769232e-06,
+      "loss": 0.6784,
+      "step": 84
+    },
+    {
+      "epoch": 0.04109257916364515,
+      "grad_norm": 0.5471555288958341,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 0.663,
+      "step": 85
+    },
+    {
+      "epoch": 0.04157602127145275,
+      "grad_norm": 0.5234614832210157,
+      "learning_rate": 8.173076923076923e-06,
+      "loss": 0.6633,
+      "step": 86
+    },
+    {
+      "epoch": 0.04205946337926034,
+      "grad_norm": 0.5014680527453607,
+      "learning_rate": 8.26923076923077e-06,
+      "loss": 0.6345,
+      "step": 87
+    },
+    {
+      "epoch": 0.04254290548706793,
+      "grad_norm": 0.5541925854592269,
+      "learning_rate": 8.365384615384616e-06,
+      "loss": 0.661,
+      "step": 88
+    },
+    {
+      "epoch": 0.043026347594875516,
+      "grad_norm": 0.5556737957241218,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 0.6476,
+      "step": 89
+    },
+    {
+      "epoch": 0.043509789702683106,
+      "grad_norm": 0.5358812925229628,
+      "learning_rate": 8.557692307692308e-06,
+      "loss": 0.6667,
+      "step": 90
+    },
+    {
+      "epoch": 0.043993231810490696,
+      "grad_norm": 0.5285944473021625,
+      "learning_rate": 8.653846153846155e-06,
+      "loss": 0.6558,
+      "step": 91
+    },
+    {
+      "epoch": 0.044476673918298286,
+      "grad_norm": 0.6130129115794695,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.6662,
+      "step": 92
+    },
+    {
+      "epoch": 0.044960116026105876,
+      "grad_norm": 0.6086871477606206,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.6768,
+      "step": 93
+    },
+    {
+      "epoch": 0.045443558133913466,
+      "grad_norm": 0.534737794998822,
+      "learning_rate": 8.942307692307693e-06,
+      "loss": 0.633,
+      "step": 94
+    },
+    {
+      "epoch": 0.045927000241721055,
+      "grad_norm": 0.5048674854153722,
+      "learning_rate": 9.03846153846154e-06,
+      "loss": 0.6075,
+      "step": 95
+    },
+    {
+      "epoch": 0.046410442349528645,
+      "grad_norm": 0.5516912026027078,
+      "learning_rate": 9.134615384615384e-06,
+      "loss": 0.623,
+      "step": 96
+    },
+    {
+      "epoch": 0.046893884457336235,
+      "grad_norm": 0.6083291149980872,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.6556,
+      "step": 97
+    },
+    {
+      "epoch": 0.047377326565143825,
+      "grad_norm": 0.5460750932826393,
+      "learning_rate": 9.326923076923079e-06,
+      "loss": 0.6524,
+      "step": 98
+    },
+    {
+      "epoch": 0.047860768672951415,
+      "grad_norm": 0.5459534721301705,
+      "learning_rate": 9.423076923076923e-06,
+      "loss": 0.6449,
+      "step": 99
+    },
+    {
+      "epoch": 0.048344210780759005,
+      "grad_norm": 0.5622412415254093,
+      "learning_rate": 9.51923076923077e-06,
+      "loss": 0.6517,
+      "step": 100
+    },
+    {
+      "epoch": 0.048827652888566594,
+      "grad_norm": 0.6148179967646931,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 0.636,
+      "step": 101
+    },
+    {
+      "epoch": 0.049311094996374184,
+      "grad_norm": 0.5377477077942675,
+      "learning_rate": 9.711538461538462e-06,
+      "loss": 0.6569,
+      "step": 102
+    },
+    {
+      "epoch": 0.049794537104181774,
+      "grad_norm": 0.535881794576154,
+      "learning_rate": 9.807692307692308e-06,
+      "loss": 0.6515,
+      "step": 103
+    },
+    {
+      "epoch": 0.050277979211989364,
+      "grad_norm": 0.5554528998874018,
+      "learning_rate": 9.903846153846155e-06,
+      "loss": 0.6471,
+      "step": 104
+    },
+    {
+      "epoch": 0.050761421319796954,
+      "grad_norm": 0.5472055318440415,
+      "learning_rate": 1e-05,
+      "loss": 0.6212,
+      "step": 105
+    },
+    {
+      "epoch": 0.051244863427604544,
+      "grad_norm": 0.5562423079812571,
+      "learning_rate": 9.99999360979851e-06,
+      "loss": 0.6483,
+      "step": 106
+    },
+    {
+      "epoch": 0.051728305535412133,
+      "grad_norm": 0.6203972023036308,
+      "learning_rate": 9.999974439210376e-06,
+      "loss": 0.6474,
+      "step": 107
+    },
+    {
+      "epoch": 0.05221174764321972,
+      "grad_norm": 0.5879110259866966,
+      "learning_rate": 9.999942488284598e-06,
+      "loss": 0.6506,
+      "step": 108
+    },
+    {
+      "epoch": 0.05269518975102731,
+      "grad_norm": 0.5415023727684817,
+      "learning_rate": 9.999897757102843e-06,
+      "loss": 0.641,
+      "step": 109
+    },
+    {
+      "epoch": 0.0531786318588349,
+      "grad_norm": 0.5661080832571289,
+      "learning_rate": 9.99984024577945e-06,
+      "loss": 0.6561,
+      "step": 110
+    },
+    {
+      "epoch": 0.05366207396664249,
+      "grad_norm": 0.6384080684659277,
+      "learning_rate": 9.999769954461425e-06,
+      "loss": 0.6181,
+      "step": 111
+    },
+    {
+      "epoch": 0.05414551607445008,
+      "grad_norm": 0.5321086465207798,
+      "learning_rate": 9.999686883328433e-06,
+      "loss": 0.6269,
+      "step": 112
+    },
+    {
+      "epoch": 0.05462895818225767,
+      "grad_norm": 0.5658443861351884,
+      "learning_rate": 9.999591032592813e-06,
+      "loss": 0.6317,
+      "step": 113
+    },
+    {
+      "epoch": 0.05511240029006526,
+      "grad_norm": 0.5759014415066968,
+      "learning_rate": 9.999482402499569e-06,
+      "loss": 0.6468,
+      "step": 114
+    },
+    {
+      "epoch": 0.05559584239787285,
+      "grad_norm": 0.5480587710988183,
+      "learning_rate": 9.999360993326366e-06,
+      "loss": 0.6359,
+      "step": 115
+    },
+    {
+      "epoch": 0.05607928450568044,
+      "grad_norm": 0.6380718424826206,
+      "learning_rate": 9.999226805383534e-06,
+      "loss": 0.6349,
+      "step": 116
+    },
+    {
+      "epoch": 0.05656272661348803,
+      "grad_norm": 0.5246861209498886,
+      "learning_rate": 9.999079839014074e-06,
+      "loss": 0.6399,
+      "step": 117
+    },
+    {
+      "epoch": 0.05704616872129562,
+      "grad_norm": 0.515686613549135,
+      "learning_rate": 9.998920094593637e-06,
+      "loss": 0.5984,
+      "step": 118
+    },
+    {
+      "epoch": 0.05752961082910321,
+      "grad_norm": 0.5607127828178857,
+      "learning_rate": 9.998747572530548e-06,
+      "loss": 0.6398,
+      "step": 119
+    },
+    {
+      "epoch": 0.05801305293691081,
+      "grad_norm": 0.5459763409466101,
+      "learning_rate": 9.998562273265786e-06,
+      "loss": 0.626,
+      "step": 120
+    },
+    {
+      "epoch": 0.0584964950447184,
+      "grad_norm": 0.5525418327052581,
+      "learning_rate": 9.998364197272988e-06,
+      "loss": 0.6537,
+      "step": 121
+    },
+    {
+      "epoch": 0.05897993715252599,
+      "grad_norm": 6.74083445541264,
+      "learning_rate": 9.998153345058454e-06,
+      "loss": 0.9475,
+      "step": 122
+    },
+    {
+      "epoch": 0.05946337926033358,
+      "grad_norm": 0.6141628359508349,
+      "learning_rate": 9.997929717161142e-06,
+      "loss": 0.6473,
+      "step": 123
+    },
+    {
+      "epoch": 0.05994682136814117,
+      "grad_norm": 0.6652958169663876,
+      "learning_rate": 9.997693314152658e-06,
+      "loss": 0.6342,
+      "step": 124
+    },
+    {
+      "epoch": 0.06043026347594876,
+      "grad_norm": 0.606711498986106,
+      "learning_rate": 9.99744413663727e-06,
+      "loss": 0.623,
+      "step": 125
+    },
+    {
+      "epoch": 0.06091370558375635,
+      "grad_norm": 0.5175309840849823,
+      "learning_rate": 9.997182185251896e-06,
+      "loss": 0.6221,
+      "step": 126
+    },
+    {
+      "epoch": 0.06139714769156394,
+      "grad_norm": 0.5341566674562975,
+      "learning_rate": 9.996907460666104e-06,
+      "loss": 0.6357,
+      "step": 127
+    },
+    {
+      "epoch": 0.06188058979937153,
+      "grad_norm": 0.5922792510376619,
+      "learning_rate": 9.996619963582113e-06,
+      "loss": 0.6043,
+      "step": 128
+    },
+    {
+      "epoch": 0.06236403190717912,
+      "grad_norm": 0.5694036510960461,
+      "learning_rate": 9.996319694734787e-06,
+      "loss": 0.6311,
+      "step": 129
+    },
+    {
+      "epoch": 0.0628474740149867,
+      "grad_norm": 0.5540844850790518,
+      "learning_rate": 9.99600665489164e-06,
+      "loss": 0.6411,
+      "step": 130
+    },
+    {
+      "epoch": 0.06333091612279429,
+      "grad_norm": 0.5371960793753483,
+      "learning_rate": 9.995680844852824e-06,
+      "loss": 0.6403,
+      "step": 131
+    },
+    {
+      "epoch": 0.06381435823060189,
+      "grad_norm": 0.5225384791967033,
+      "learning_rate": 9.995342265451138e-06,
+      "loss": 0.6269,
+      "step": 132
+    },
+    {
+      "epoch": 0.06429780033840947,
+      "grad_norm": 0.6035451474536077,
+      "learning_rate": 9.994990917552017e-06,
+      "loss": 0.6321,
+      "step": 133
+    },
+    {
+      "epoch": 0.06478124244621707,
+      "grad_norm": 0.6507380493478006,
+      "learning_rate": 9.994626802053536e-06,
+      "loss": 0.6236,
+      "step": 134
+    },
+    {
+      "epoch": 0.06526468455402465,
+      "grad_norm": 0.5456651842881993,
+      "learning_rate": 9.994249919886402e-06,
+      "loss": 0.6258,
+      "step": 135
+    },
+    {
+      "epoch": 0.06574812666183225,
+      "grad_norm": 0.5172506944070536,
+      "learning_rate": 9.993860272013958e-06,
+      "loss": 0.6162,
+      "step": 136
+    },
+    {
+      "epoch": 0.06623156876963984,
+      "grad_norm": 0.6233262394445207,
+      "learning_rate": 9.993457859432172e-06,
+      "loss": 0.6261,
+      "step": 137
+    },
+    {
+      "epoch": 0.06671501087744743,
+      "grad_norm": 0.6073445562745826,
+      "learning_rate": 9.993042683169647e-06,
+      "loss": 0.6371,
+      "step": 138
+    },
+    {
+      "epoch": 0.06719845298525502,
+      "grad_norm": 0.5857241687958673,
+      "learning_rate": 9.992614744287605e-06,
+      "loss": 0.6275,
+      "step": 139
+    },
+    {
+      "epoch": 0.0676818950930626,
+      "grad_norm": 0.5304150460003405,
+      "learning_rate": 9.992174043879893e-06,
+      "loss": 0.6175,
+      "step": 140
+    },
+    {
+      "epoch": 0.0681653372008702,
+      "grad_norm": 0.5933722892089892,
+      "learning_rate": 9.991720583072975e-06,
+      "loss": 0.6255,
+      "step": 141
+    },
+    {
+      "epoch": 0.06864877930867778,
+      "grad_norm": 0.561723953482763,
+      "learning_rate": 9.991254363025935e-06,
+      "loss": 0.6257,
+      "step": 142
+    },
+    {
+      "epoch": 0.06913222141648538,
+      "grad_norm": 0.532228224452236,
+      "learning_rate": 9.99077538493047e-06,
+      "loss": 0.6301,
+      "step": 143
+    },
+    {
+      "epoch": 0.06961566352429296,
+      "grad_norm": 0.541783938730816,
+      "learning_rate": 9.990283650010883e-06,
+      "loss": 0.619,
+      "step": 144
+    },
+    {
+      "epoch": 0.07009910563210056,
+      "grad_norm": 0.5606995950440783,
+      "learning_rate": 9.989779159524091e-06,
+      "loss": 0.5818,
+      "step": 145
+    },
+    {
+      "epoch": 0.07058254773990814,
+      "grad_norm": 0.5286741282148979,
+      "learning_rate": 9.989261914759612e-06,
+      "loss": 0.6105,
+      "step": 146
+    },
+    {
+      "epoch": 0.07106598984771574,
+      "grad_norm": 0.525375741245272,
+      "learning_rate": 9.988731917039564e-06,
+      "loss": 0.6154,
+      "step": 147
+    },
+    {
+      "epoch": 0.07154943195552332,
+      "grad_norm": 0.5132546936158348,
+      "learning_rate": 9.988189167718665e-06,
+      "loss": 0.5533,
+      "step": 148
+    },
+    {
+      "epoch": 0.07203287406333092,
+      "grad_norm": 0.5330232205089095,
+      "learning_rate": 9.987633668184227e-06,
+      "loss": 0.6281,
+      "step": 149
+    },
+    {
+      "epoch": 0.0725163161711385,
+      "grad_norm": 0.5459730729112252,
+      "learning_rate": 9.98706541985615e-06,
+      "loss": 0.5836,
+      "step": 150
+    },
+    {
+      "epoch": 0.0729997582789461,
+      "grad_norm": 0.5818263727750432,
+      "learning_rate": 9.986484424186922e-06,
+      "loss": 0.6246,
+      "step": 151
+    },
+    {
+      "epoch": 0.07348320038675368,
+      "grad_norm": 0.5754133435232375,
+      "learning_rate": 9.985890682661616e-06,
+      "loss": 0.6038,
+      "step": 152
+    },
+    {
+      "epoch": 0.07396664249456128,
+      "grad_norm": 0.5528911744587542,
+      "learning_rate": 9.985284196797884e-06,
+      "loss": 0.6246,
+      "step": 153
+    },
+    {
+      "epoch": 0.07445008460236886,
+      "grad_norm": 0.5484687585797547,
+      "learning_rate": 9.984664968145953e-06,
+      "loss": 0.6318,
+      "step": 154
+    },
+    {
+      "epoch": 0.07493352671017646,
+      "grad_norm": 0.5351986552762329,
+      "learning_rate": 9.984032998288617e-06,
+      "loss": 0.6184,
+      "step": 155
+    },
+    {
+      "epoch": 0.07541696881798404,
+      "grad_norm": 0.519416066205614,
+      "learning_rate": 9.983388288841246e-06,
+      "loss": 0.6185,
+      "step": 156
+    },
+    {
+      "epoch": 0.07590041092579164,
+      "grad_norm": 0.5470449402548487,
+      "learning_rate": 9.982730841451768e-06,
+      "loss": 0.625,
+      "step": 157
+    },
+    {
+      "epoch": 0.07638385303359922,
+      "grad_norm": 0.5887016805140373,
+      "learning_rate": 9.982060657800672e-06,
+      "loss": 0.6183,
+      "step": 158
+    },
+    {
+      "epoch": 0.07686729514140682,
+      "grad_norm": 0.5522566946881194,
+      "learning_rate": 9.981377739601002e-06,
+      "loss": 0.6137,
+      "step": 159
+    },
+    {
+      "epoch": 0.0773507372492144,
+      "grad_norm": 0.5411997809451911,
+      "learning_rate": 9.980682088598349e-06,
+      "loss": 0.6229,
+      "step": 160
+    },
+    {
+      "epoch": 0.077834179357022,
+      "grad_norm": 0.52840707851752,
+      "learning_rate": 9.979973706570856e-06,
+      "loss": 0.614,
+      "step": 161
+    },
+    {
+      "epoch": 0.07831762146482958,
+      "grad_norm": 0.6047062373713257,
+      "learning_rate": 9.979252595329204e-06,
+      "loss": 0.6222,
+      "step": 162
+    },
+    {
+      "epoch": 0.07880106357263718,
+      "grad_norm": 0.5420471794760692,
+      "learning_rate": 9.978518756716611e-06,
+      "loss": 0.5856,
+      "step": 163
+    },
+    {
+      "epoch": 0.07928450568044476,
+      "grad_norm": 0.5857386315586672,
+      "learning_rate": 9.977772192608827e-06,
+      "loss": 0.6291,
+      "step": 164
+    },
+    {
+      "epoch": 0.07976794778825236,
+      "grad_norm": 0.5691356356316107,
+      "learning_rate": 9.977012904914133e-06,
+      "loss": 0.6149,
+      "step": 165
+    },
+    {
+      "epoch": 0.08025138989605994,
+      "grad_norm": 0.5823273363045892,
+      "learning_rate": 9.976240895573326e-06,
+      "loss": 0.6147,
+      "step": 166
+    },
+    {
+      "epoch": 0.08073483200386754,
+      "grad_norm": 0.538212010864403,
+      "learning_rate": 9.975456166559725e-06,
+      "loss": 0.6002,
+      "step": 167
+    },
+    {
+      "epoch": 0.08121827411167512,
+      "grad_norm": 0.601371610274862,
+      "learning_rate": 9.974658719879163e-06,
+      "loss": 0.606,
+      "step": 168
+    },
+    {
+      "epoch": 0.08170171621948272,
+      "grad_norm": 0.588104162701253,
+      "learning_rate": 9.973848557569974e-06,
+      "loss": 0.6226,
+      "step": 169
+    },
+    {
+      "epoch": 0.0821851583272903,
+      "grad_norm": 0.5316828963553285,
+      "learning_rate": 9.973025681703e-06,
+      "loss": 0.6144,
+      "step": 170
+    },
+    {
+      "epoch": 0.0826686004350979,
+      "grad_norm": 0.5405916050680715,
+      "learning_rate": 9.972190094381578e-06,
+      "loss": 0.6148,
+      "step": 171
+    },
+    {
+      "epoch": 0.0831520425429055,
+      "grad_norm": 0.5102891757426009,
+      "learning_rate": 9.971341797741538e-06,
+      "loss": 0.616,
+      "step": 172
+    },
+    {
+      "epoch": 0.08363548465071308,
+      "grad_norm": 0.5551757535954606,
+      "learning_rate": 9.970480793951194e-06,
+      "loss": 0.6196,
+      "step": 173
+    },
+    {
+      "epoch": 0.08411892675852067,
+      "grad_norm": 0.5349760515746151,
+      "learning_rate": 9.96960708521134e-06,
+      "loss": 0.5902,
+      "step": 174
+    },
+    {
+      "epoch": 0.08460236886632826,
+      "grad_norm": 0.5713299053870873,
+      "learning_rate": 9.968720673755246e-06,
+      "loss": 0.6039,
+      "step": 175
+    },
+    {
+      "epoch": 0.08508581097413585,
+      "grad_norm": 0.5886201187493544,
+      "learning_rate": 9.96782156184865e-06,
+      "loss": 0.6128,
+      "step": 176
+    },
+    {
+      "epoch": 0.08556925308194344,
+      "grad_norm": 0.52487297166769,
+      "learning_rate": 9.966909751789758e-06,
+      "loss": 0.6201,
+      "step": 177
+    },
+    {
+      "epoch": 0.08605269518975103,
+      "grad_norm": 0.47488673856360863,
+      "learning_rate": 9.965985245909226e-06,
+      "loss": 0.581,
+      "step": 178
+    },
+    {
+      "epoch": 0.08653613729755862,
+      "grad_norm": 0.5390345004627665,
+      "learning_rate": 9.96504804657017e-06,
+      "loss": 0.5748,
+      "step": 179
+    },
+    {
+      "epoch": 0.08701957940536621,
+      "grad_norm": 0.5030595297893009,
+      "learning_rate": 9.964098156168143e-06,
+      "loss": 0.6025,
+      "step": 180
+    },
+    {
+      "epoch": 0.0875030215131738,
+      "grad_norm": 0.5468598312459072,
+      "learning_rate": 9.963135577131144e-06,
+      "loss": 0.6086,
+      "step": 181
+    },
+    {
+      "epoch": 0.08798646362098139,
+      "grad_norm": 0.48113219800404783,
+      "learning_rate": 9.962160311919601e-06,
+      "loss": 0.5759,
+      "step": 182
+    },
+    {
+      "epoch": 0.08846990572878898,
+      "grad_norm": 0.5498772940672643,
+      "learning_rate": 9.96117236302637e-06,
+      "loss": 0.6009,
+      "step": 183
+    },
+    {
+      "epoch": 0.08895334783659657,
+      "grad_norm": 0.572150853367621,
+      "learning_rate": 9.960171732976731e-06,
+      "loss": 0.5891,
+      "step": 184
+    },
+    {
+      "epoch": 0.08943678994440415,
+      "grad_norm": 0.5440182913032069,
+      "learning_rate": 9.959158424328373e-06,
+      "loss": 0.6126,
+      "step": 185
+    },
+    {
+      "epoch": 0.08992023205221175,
+      "grad_norm": 0.5124606491120447,
+      "learning_rate": 9.958132439671392e-06,
+      "loss": 0.6113,
+      "step": 186
+    },
+    {
+      "epoch": 0.09040367416001933,
+      "grad_norm": 0.5122426086233111,
+      "learning_rate": 9.957093781628294e-06,
+      "loss": 0.5585,
+      "step": 187
+    },
+    {
+      "epoch": 0.09088711626782693,
+      "grad_norm": 0.5466339032920954,
+      "learning_rate": 9.956042452853967e-06,
+      "loss": 0.5829,
+      "step": 188
+    },
+    {
+      "epoch": 0.09137055837563451,
+      "grad_norm": 0.5319185267267565,
+      "learning_rate": 9.954978456035695e-06,
+      "loss": 0.6014,
+      "step": 189
+    },
+    {
+      "epoch": 0.09185400048344211,
+      "grad_norm": 0.5439360347029544,
+      "learning_rate": 9.953901793893137e-06,
+      "loss": 0.6135,
+      "step": 190
+    },
+    {
+      "epoch": 0.0923374425912497,
+      "grad_norm": 0.5572467498872743,
+      "learning_rate": 9.95281246917833e-06,
+      "loss": 0.6126,
+      "step": 191
+    },
+    {
+      "epoch": 0.09282088469905729,
+      "grad_norm": 0.5541110285684123,
+      "learning_rate": 9.951710484675677e-06,
+      "loss": 0.6077,
+      "step": 192
+    },
+    {
+      "epoch": 0.09330432680686487,
+      "grad_norm": 0.4850481807152515,
+      "learning_rate": 9.950595843201936e-06,
+      "loss": 0.6052,
+      "step": 193
+    },
+    {
+      "epoch": 0.09378776891467247,
+      "grad_norm": 0.4982494369774088,
+      "learning_rate": 9.949468547606222e-06,
+      "loss": 0.608,
+      "step": 194
+    },
+    {
+      "epoch": 0.09427121102248005,
+      "grad_norm": 0.5222210926075901,
+      "learning_rate": 9.948328600769996e-06,
+      "loss": 0.5725,
+      "step": 195
+    },
+    {
+      "epoch": 0.09475465313028765,
+      "grad_norm": 0.5156665548407187,
+      "learning_rate": 9.94717600560705e-06,
+      "loss": 0.5981,
+      "step": 196
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 0.4789398218595176,
+      "learning_rate": 9.946010765063512e-06,
+      "loss": 0.6163,
+      "step": 197
+    },
+    {
+      "epoch": 0.09572153734590283,
+      "grad_norm": 0.5066106303118647,
+      "learning_rate": 9.94483288211783e-06,
+      "loss": 0.6049,
+      "step": 198
+    },
+    {
+      "epoch": 0.09620497945371041,
+      "grad_norm": 0.519086410125638,
+      "learning_rate": 9.943642359780767e-06,
+      "loss": 0.6034,
+      "step": 199
+    },
+    {
+      "epoch": 0.09668842156151801,
+      "grad_norm": 0.5726309849663989,
+      "learning_rate": 9.942439201095398e-06,
+      "loss": 0.5977,
+      "step": 200
+    },
+    {
+      "epoch": 0.09717186366932559,
+      "grad_norm": 0.5149014744932526,
+      "learning_rate": 9.941223409137088e-06,
+      "loss": 0.6147,
+      "step": 201
+    },
+    {
+      "epoch": 0.09765530577713319,
+      "grad_norm": 0.5009166664227639,
+      "learning_rate": 9.939994987013505e-06,
+      "loss": 0.595,
+      "step": 202
+    },
+    {
+      "epoch": 0.09813874788494077,
+      "grad_norm": 0.5677576117209191,
+      "learning_rate": 9.93875393786459e-06,
+      "loss": 0.5825,
+      "step": 203
+    },
+    {
+      "epoch": 0.09862218999274837,
+      "grad_norm": 0.5705628060741978,
+      "learning_rate": 9.937500264862567e-06,
+      "loss": 0.6106,
+      "step": 204
+    },
+    {
+      "epoch": 0.09910563210055595,
+      "grad_norm": 0.5166084751955315,
+      "learning_rate": 9.936233971211926e-06,
+      "loss": 0.5724,
+      "step": 205
+    },
+    {
+      "epoch": 0.09958907420836355,
+      "grad_norm": 0.4998369485071646,
+      "learning_rate": 9.934955060149413e-06,
+      "loss": 0.5702,
+      "step": 206
+    },
+    {
+      "epoch": 0.10007251631617115,
+      "grad_norm": 0.4773861112208611,
+      "learning_rate": 9.933663534944029e-06,
+      "loss": 0.5976,
+      "step": 207
+    },
+    {
+      "epoch": 0.10055595842397873,
+      "grad_norm": 0.5142399648385931,
+      "learning_rate": 9.932359398897018e-06,
+      "loss": 0.5662,
+      "step": 208
+    },
+    {
+      "epoch": 0.10103940053178632,
+      "grad_norm": 0.5152331134346968,
+      "learning_rate": 9.931042655341856e-06,
+      "loss": 0.5987,
+      "step": 209
+    },
+    {
+      "epoch": 0.10152284263959391,
+      "grad_norm": 0.5697107336495173,
+      "learning_rate": 9.929713307644245e-06,
+      "loss": 0.5956,
+      "step": 210
+    },
+    {
+      "epoch": 0.1020062847474015,
+      "grad_norm": 0.5437386464507225,
+      "learning_rate": 9.928371359202103e-06,
+      "loss": 0.6023,
+      "step": 211
+    },
+    {
+      "epoch": 0.10248972685520909,
+      "grad_norm": 0.48930400532530816,
+      "learning_rate": 9.927016813445562e-06,
+      "loss": 0.5941,
+      "step": 212
+    },
+    {
+      "epoch": 0.10297316896301668,
+      "grad_norm": 0.5691895002113943,
+      "learning_rate": 9.925649673836949e-06,
+      "loss": 0.5977,
+      "step": 213
+    },
+    {
+      "epoch": 0.10345661107082427,
+      "grad_norm": 0.518358121778254,
+      "learning_rate": 9.924269943870781e-06,
+      "loss": 0.599,
+      "step": 214
+    },
+    {
+      "epoch": 0.10394005317863186,
+      "grad_norm": 0.5179203447080591,
+      "learning_rate": 9.922877627073763e-06,
+      "loss": 0.565,
+      "step": 215
+    },
+    {
+      "epoch": 0.10442349528643945,
+      "grad_norm": 0.5348210146349037,
+      "learning_rate": 9.921472727004765e-06,
+      "loss": 0.6038,
+      "step": 216
+    },
+    {
+      "epoch": 0.10490693739424704,
+      "grad_norm": 0.5011388091471438,
+      "learning_rate": 9.920055247254827e-06,
+      "loss": 0.5951,
+      "step": 217
+    },
+    {
+      "epoch": 0.10539037950205463,
+      "grad_norm": 0.5706178448892886,
+      "learning_rate": 9.91862519144714e-06,
+      "loss": 0.604,
+      "step": 218
+    },
+    {
+      "epoch": 0.10587382160986222,
+      "grad_norm": 0.5667257328777994,
+      "learning_rate": 9.917182563237045e-06,
+      "loss": 0.6006,
+      "step": 219
+    },
+    {
+      "epoch": 0.1063572637176698,
+      "grad_norm": 0.5402529870671051,
+      "learning_rate": 9.915727366312012e-06,
+      "loss": 0.591,
+      "step": 220
+    },
+    {
+      "epoch": 0.1068407058254774,
+      "grad_norm": 0.5123066262170495,
+      "learning_rate": 9.914259604391642e-06,
+      "loss": 0.5818,
+      "step": 221
+    },
+    {
+      "epoch": 0.10732414793328499,
+      "grad_norm": 0.5104812232878251,
+      "learning_rate": 9.912779281227656e-06,
+      "loss": 0.5991,
+      "step": 222
+    },
+    {
+      "epoch": 0.10780759004109258,
+      "grad_norm": 0.5073553912253322,
+      "learning_rate": 9.911286400603878e-06,
+      "loss": 0.5783,
+      "step": 223
+    },
+    {
+      "epoch": 0.10829103214890017,
+      "grad_norm": 0.5516103650201469,
+      "learning_rate": 9.90978096633623e-06,
+      "loss": 0.6007,
+      "step": 224
+    },
+    {
+      "epoch": 0.10877447425670776,
+      "grad_norm": 0.5241670992889956,
+      "learning_rate": 9.908262982272724e-06,
+      "loss": 0.5865,
+      "step": 225
+    },
+    {
+      "epoch": 0.10925791636451535,
+      "grad_norm": 0.4894067875331202,
+      "learning_rate": 9.906732452293448e-06,
+      "loss": 0.5635,
+      "step": 226
+    },
+    {
+      "epoch": 0.10974135847232294,
+      "grad_norm": 0.5079732216995924,
+      "learning_rate": 9.905189380310564e-06,
+      "loss": 0.5982,
+      "step": 227
+    },
+    {
+      "epoch": 0.11022480058013052,
+      "grad_norm": 0.47288266380376864,
+      "learning_rate": 9.903633770268286e-06,
+      "loss": 0.5734,
+      "step": 228
+    },
+    {
+      "epoch": 0.11070824268793812,
+      "grad_norm": 0.5195973051222883,
+      "learning_rate": 9.902065626142876e-06,
+      "loss": 0.6021,
+      "step": 229
+    },
+    {
+      "epoch": 0.1111916847957457,
+      "grad_norm": 0.5159734590151601,
+      "learning_rate": 9.900484951942642e-06,
+      "loss": 0.5847,
+      "step": 230
+    },
+    {
+      "epoch": 0.1116751269035533,
+      "grad_norm": 0.5157347113387764,
+      "learning_rate": 9.89889175170791e-06,
+      "loss": 0.5946,
+      "step": 231
+    },
+    {
+      "epoch": 0.11215856901136088,
+      "grad_norm": 0.49409523170190334,
+      "learning_rate": 9.89728602951103e-06,
+      "loss": 0.5941,
+      "step": 232
+    },
+    {
+      "epoch": 0.11264201111916848,
+      "grad_norm": 0.5699641967141135,
+      "learning_rate": 9.89566778945636e-06,
+      "loss": 0.5965,
+      "step": 233
+    },
+    {
+      "epoch": 0.11312545322697606,
+      "grad_norm": 0.5565932357020583,
+      "learning_rate": 9.894037035680246e-06,
+      "loss": 0.6076,
+      "step": 234
+    },
+    {
+      "epoch": 0.11360889533478366,
+      "grad_norm": 0.4762368359891958,
+      "learning_rate": 9.892393772351033e-06,
+      "loss": 0.5749,
+      "step": 235
+    },
+    {
+      "epoch": 0.11409233744259124,
+      "grad_norm": 0.5226269336653058,
+      "learning_rate": 9.890738003669029e-06,
+      "loss": 0.5882,
+      "step": 236
+    },
+    {
+      "epoch": 0.11457577955039884,
+      "grad_norm": 0.5893232226185929,
+      "learning_rate": 9.889069733866515e-06,
+      "loss": 0.5978,
+      "step": 237
+    },
+    {
+      "epoch": 0.11505922165820642,
+      "grad_norm": 0.5556325697280562,
+      "learning_rate": 9.887388967207722e-06,
+      "loss": 0.6,
+      "step": 238
+    },
+    {
+      "epoch": 0.11554266376601402,
+      "grad_norm": 0.48160661753964396,
+      "learning_rate": 9.885695707988825e-06,
+      "loss": 0.5977,
+      "step": 239
+    },
+    {
+      "epoch": 0.11602610587382162,
+      "grad_norm": 0.5122405505133801,
+      "learning_rate": 9.883989960537934e-06,
+      "loss": 0.6044,
+      "step": 240
+    },
+    {
+      "epoch": 0.1165095479816292,
+      "grad_norm": 0.5812889541684825,
+      "learning_rate": 9.882271729215071e-06,
+      "loss": 0.5849,
+      "step": 241
+    },
+    {
+      "epoch": 0.1169929900894368,
+      "grad_norm": 0.4906401332764143,
+      "learning_rate": 9.880541018412179e-06,
+      "loss": 0.5986,
+      "step": 242
+    },
+    {
+      "epoch": 0.11747643219724438,
+      "grad_norm": 0.48951055967126716,
+      "learning_rate": 9.878797832553093e-06,
+      "loss": 0.5646,
+      "step": 243
+    },
+    {
+      "epoch": 0.11795987430505198,
+      "grad_norm": 0.4836474446158179,
+      "learning_rate": 9.877042176093537e-06,
+      "loss": 0.5998,
+      "step": 244
+    },
+    {
+      "epoch": 0.11844331641285956,
+      "grad_norm": 0.4962973453940785,
+      "learning_rate": 9.875274053521107e-06,
+      "loss": 0.5846,
+      "step": 245
+    },
+    {
+      "epoch": 0.11892675852066716,
+      "grad_norm": 0.45261755838242107,
+      "learning_rate": 9.873493469355271e-06,
+      "loss": 0.5912,
+      "step": 246
+    },
+    {
+      "epoch": 0.11941020062847474,
+      "grad_norm": 0.49934013758424506,
+      "learning_rate": 9.871700428147342e-06,
+      "loss": 0.5836,
+      "step": 247
+    },
+    {
+      "epoch": 0.11989364273628234,
+      "grad_norm": 0.48318779237357384,
+      "learning_rate": 9.86989493448048e-06,
+      "loss": 0.5898,
+      "step": 248
+    },
+    {
+      "epoch": 0.12037708484408992,
+      "grad_norm": 0.4877998807669757,
+      "learning_rate": 9.868076992969672e-06,
+      "loss": 0.5933,
+      "step": 249
+    },
+    {
+      "epoch": 0.12086052695189752,
+      "grad_norm": 0.4697579805390032,
+      "learning_rate": 9.866246608261725e-06,
+      "loss": 0.5855,
+      "step": 250
+    },
+    {
+      "epoch": 0.1213439690597051,
+      "grad_norm": 0.512552573820198,
+      "learning_rate": 9.864403785035246e-06,
+      "loss": 0.5989,
+      "step": 251
+    },
+    {
+      "epoch": 0.1218274111675127,
+      "grad_norm": 0.4865753568683563,
+      "learning_rate": 9.862548528000644e-06,
+      "loss": 0.5722,
+      "step": 252
+    },
+    {
+      "epoch": 0.12231085327532028,
+      "grad_norm": 0.5276925045930954,
+      "learning_rate": 9.860680841900101e-06,
+      "loss": 0.5879,
+      "step": 253
+    },
+    {
+      "epoch": 0.12279429538312787,
+      "grad_norm": 0.5213718677505005,
+      "learning_rate": 9.858800731507575e-06,
+      "loss": 0.5999,
+      "step": 254
+    },
+    {
+      "epoch": 0.12327773749093546,
+      "grad_norm": 0.5180845494091726,
+      "learning_rate": 9.85690820162878e-06,
+      "loss": 0.586,
+      "step": 255
+    },
+    {
+      "epoch": 0.12376117959874305,
+      "grad_norm": 0.5698025401421347,
+      "learning_rate": 9.855003257101177e-06,
+      "loss": 0.6011,
+      "step": 256
+    },
+    {
+      "epoch": 0.12424462170655064,
+      "grad_norm": 0.562343589994959,
+      "learning_rate": 9.853085902793952e-06,
+      "loss": 0.5894,
+      "step": 257
+    },
+    {
+      "epoch": 0.12472806381435823,
+      "grad_norm": 0.5160827286882833,
+      "learning_rate": 9.851156143608025e-06,
+      "loss": 0.5897,
+      "step": 258
+    },
+    {
+      "epoch": 0.12521150592216582,
+      "grad_norm": 0.5407107287832078,
+      "learning_rate": 9.84921398447601e-06,
+      "loss": 0.59,
+      "step": 259
+    },
+    {
+      "epoch": 0.1256949480299734,
+      "grad_norm": 0.4828245059112851,
+      "learning_rate": 9.847259430362222e-06,
+      "loss": 0.5642,
+      "step": 260
+    },
+    {
+      "epoch": 0.126178390137781,
+      "grad_norm": 0.5766667340207283,
+      "learning_rate": 9.845292486262664e-06,
+      "loss": 0.6016,
+      "step": 261
+    },
+    {
+      "epoch": 0.12666183224558858,
+      "grad_norm": 0.5818866932241936,
+      "learning_rate": 9.843313157204999e-06,
+      "loss": 0.5807,
+      "step": 262
+    },
+    {
+      "epoch": 0.12714527435339618,
+      "grad_norm": 0.5140923007570054,
+      "learning_rate": 9.841321448248552e-06,
+      "loss": 0.5858,
+      "step": 263
+    },
+    {
+      "epoch": 0.12762871646120377,
+      "grad_norm": 0.513399510660716,
+      "learning_rate": 9.839317364484295e-06,
+      "loss": 0.5847,
+      "step": 264
+    },
+    {
+      "epoch": 0.12811215856901137,
+      "grad_norm": 0.5227642580781724,
+      "learning_rate": 9.837300911034824e-06,
+      "loss": 0.5888,
+      "step": 265
+    },
+    {
+      "epoch": 0.12859560067681894,
+      "grad_norm": 0.5579358896097371,
+      "learning_rate": 9.83527209305436e-06,
+      "loss": 0.5928,
+      "step": 266
+    },
+    {
+      "epoch": 0.12907904278462654,
+      "grad_norm": 0.5145348442577231,
+      "learning_rate": 9.83323091572872e-06,
+      "loss": 0.5872,
+      "step": 267
+    },
+    {
+      "epoch": 0.12956248489243413,
+      "grad_norm": 0.5112821410236051,
+      "learning_rate": 9.831177384275323e-06,
+      "loss": 0.5805,
+      "step": 268
+    },
+    {
+      "epoch": 0.13004592700024173,
+      "grad_norm": 0.5497912960403669,
+      "learning_rate": 9.829111503943159e-06,
+      "loss": 0.5837,
+      "step": 269
+    },
+    {
+      "epoch": 0.1305293691080493,
+      "grad_norm": 0.5226743950335115,
+      "learning_rate": 9.827033280012783e-06,
+      "loss": 0.5539,
+      "step": 270
+    },
+    {
+      "epoch": 0.1310128112158569,
+      "grad_norm": 0.5713921241049837,
+      "learning_rate": 9.824942717796304e-06,
+      "loss": 0.5881,
+      "step": 271
+    },
+    {
+      "epoch": 0.1314962533236645,
+      "grad_norm": 0.5241764388189555,
+      "learning_rate": 9.822839822637369e-06,
+      "loss": 0.6032,
+      "step": 272
+    },
+    {
+      "epoch": 0.1319796954314721,
+      "grad_norm": 0.5162440352522167,
+      "learning_rate": 9.820724599911147e-06,
+      "loss": 0.5842,
+      "step": 273
+    },
+    {
+      "epoch": 0.13246313753927969,
+      "grad_norm": 0.5431692492650363,
+      "learning_rate": 9.818597055024315e-06,
+      "loss": 0.585,
+      "step": 274
+    },
+    {
+      "epoch": 0.13294657964708725,
+      "grad_norm": 0.5124783198553914,
+      "learning_rate": 9.816457193415055e-06,
+      "loss": 0.5779,
+      "step": 275
+    },
+    {
+      "epoch": 0.13343002175489485,
+      "grad_norm": 0.5257695390265421,
+      "learning_rate": 9.81430502055302e-06,
+      "loss": 0.5798,
+      "step": 276
+    },
+    {
+      "epoch": 0.13391346386270245,
+      "grad_norm": 0.49781008962990064,
+      "learning_rate": 9.812140541939338e-06,
+      "loss": 0.5836,
+      "step": 277
+    },
+    {
+      "epoch": 0.13439690597051004,
+      "grad_norm": 0.5327804269781539,
+      "learning_rate": 9.809963763106593e-06,
+      "loss": 0.5733,
+      "step": 278
+    },
+    {
+      "epoch": 0.1348803480783176,
+      "grad_norm": 0.5167258655366103,
+      "learning_rate": 9.807774689618806e-06,
+      "loss": 0.58,
+      "step": 279
+    },
+    {
+      "epoch": 0.1353637901861252,
+      "grad_norm": 0.5321175943512093,
+      "learning_rate": 9.805573327071428e-06,
+      "loss": 0.5911,
+      "step": 280
+    },
+    {
+      "epoch": 0.1358472322939328,
+      "grad_norm": 0.49961117510050285,
+      "learning_rate": 9.803359681091313e-06,
+      "loss": 0.5737,
+      "step": 281
+    },
+    {
+      "epoch": 0.1363306744017404,
+      "grad_norm": 0.5314962622355859,
+      "learning_rate": 9.801133757336726e-06,
+      "loss": 0.593,
+      "step": 282
+    },
+    {
+      "epoch": 0.13681411650954797,
+      "grad_norm": 0.48173417582091976,
+      "learning_rate": 9.798895561497299e-06,
+      "loss": 0.5818,
+      "step": 283
+    },
+    {
+      "epoch": 0.13729755861735557,
+      "grad_norm": 0.5127693228983886,
+      "learning_rate": 9.796645099294049e-06,
+      "loss": 0.6024,
+      "step": 284
+    },
+    {
+      "epoch": 0.13778100072516317,
+      "grad_norm": 0.5128313174228813,
+      "learning_rate": 9.794382376479334e-06,
+      "loss": 0.5837,
+      "step": 285
+    },
+    {
+      "epoch": 0.13826444283297076,
+      "grad_norm": 0.502862882638082,
+      "learning_rate": 9.792107398836859e-06,
+      "loss": 0.5781,
+      "step": 286
+    },
+    {
+      "epoch": 0.13874788494077833,
+      "grad_norm": 0.5169656633134686,
+      "learning_rate": 9.789820172181648e-06,
+      "loss": 0.5821,
+      "step": 287
+    },
+    {
+      "epoch": 0.13923132704858593,
+      "grad_norm": 6.7246508188992,
+      "learning_rate": 9.787520702360035e-06,
+      "loss": 1.0972,
+      "step": 288
+    },
+    {
+      "epoch": 0.13971476915639353,
+      "grad_norm": 0.6005251051430991,
+      "learning_rate": 9.785208995249655e-06,
+      "loss": 0.5803,
+      "step": 289
+    },
+    {
+      "epoch": 0.14019821126420112,
+      "grad_norm": 0.5531574758650235,
+      "learning_rate": 9.782885056759413e-06,
+      "loss": 0.563,
+      "step": 290
+    },
+    {
+      "epoch": 0.1406816533720087,
+      "grad_norm": 0.5273779406180227,
+      "learning_rate": 9.780548892829486e-06,
+      "loss": 0.5872,
+      "step": 291
+    },
+    {
+      "epoch": 0.1411650954798163,
+      "grad_norm": 0.5063770192301159,
+      "learning_rate": 9.778200509431297e-06,
+      "loss": 0.5782,
+      "step": 292
+    },
+    {
+      "epoch": 0.14164853758762389,
+      "grad_norm": 0.5401099132225082,
+      "learning_rate": 9.775839912567502e-06,
+      "loss": 0.5804,
+      "step": 293
+    },
+    {
+      "epoch": 0.14213197969543148,
+      "grad_norm": 0.607784811294971,
+      "learning_rate": 9.773467108271978e-06,
+      "loss": 0.5831,
+      "step": 294
+    },
+    {
+      "epoch": 0.14261542180323905,
+      "grad_norm": 0.5051370116219928,
+      "learning_rate": 9.771082102609803e-06,
+      "loss": 0.5597,
+      "step": 295
+    },
+    {
+      "epoch": 0.14309886391104665,
+      "grad_norm": 0.5723810352863865,
+      "learning_rate": 9.768684901677245e-06,
+      "loss": 0.5779,
+      "step": 296
+    },
+    {
+      "epoch": 0.14358230601885424,
+      "grad_norm": 0.529491415132923,
+      "learning_rate": 9.766275511601742e-06,
+      "loss": 0.5849,
+      "step": 297
+    },
+    {
+      "epoch": 0.14406574812666184,
+      "grad_norm": 0.6275998382003428,
+      "learning_rate": 9.763853938541887e-06,
+      "loss": 0.5915,
+      "step": 298
+    },
+    {
+      "epoch": 0.1445491902344694,
+      "grad_norm": 0.5906428033404255,
+      "learning_rate": 9.76142018868742e-06,
+      "loss": 0.5816,
+      "step": 299
+    },
+    {
+      "epoch": 0.145032632342277,
+      "grad_norm": 0.597638837356143,
+      "learning_rate": 9.7589742682592e-06,
+      "loss": 0.5578,
+      "step": 300
+    },
+    {
+      "epoch": 0.1455160744500846,
+      "grad_norm": 0.5365546900890564,
+      "learning_rate": 9.756516183509198e-06,
+      "loss": 0.5833,
+      "step": 301
+    },
+    {
+      "epoch": 0.1459995165578922,
+      "grad_norm": 0.554155920273677,
+      "learning_rate": 9.754045940720471e-06,
+      "loss": 0.581,
+      "step": 302
+    },
+    {
+      "epoch": 0.14648295866569977,
+      "grad_norm": 0.5290449152773149,
+      "learning_rate": 9.751563546207167e-06,
+      "loss": 0.5879,
+      "step": 303
+    },
+    {
+      "epoch": 0.14696640077350737,
+      "grad_norm": 0.5303051981230842,
+      "learning_rate": 9.749069006314481e-06,
+      "loss": 0.557,
+      "step": 304
+    },
+    {
+      "epoch": 0.14744984288131496,
+      "grad_norm": 0.4750712434505446,
+      "learning_rate": 9.74656232741866e-06,
+      "loss": 0.5236,
+      "step": 305
+    },
+    {
+      "epoch": 0.14793328498912256,
+      "grad_norm": 0.515780571537496,
+      "learning_rate": 9.744043515926975e-06,
+      "loss": 0.5827,
+      "step": 306
+    },
+    {
+      "epoch": 0.14841672709693013,
+      "grad_norm": 0.5886066507830542,
+      "learning_rate": 9.741512578277715e-06,
+      "loss": 0.5741,
+      "step": 307
+    },
+    {
+      "epoch": 0.14890016920473773,
+      "grad_norm": 0.5712616310834069,
+      "learning_rate": 9.738969520940158e-06,
+      "loss": 0.587,
+      "step": 308
+    },
+    {
+      "epoch": 0.14938361131254532,
+      "grad_norm": 0.5883909446108012,
+      "learning_rate": 9.736414350414564e-06,
+      "loss": 0.5836,
+      "step": 309
+    },
+    {
+      "epoch": 0.14986705342035292,
+      "grad_norm": 0.49300111186175044,
+      "learning_rate": 9.733847073232156e-06,
+      "loss": 0.583,
+      "step": 310
+    },
+    {
+      "epoch": 0.15035049552816052,
+      "grad_norm": 0.47057695692490953,
+      "learning_rate": 9.7312676959551e-06,
+      "loss": 0.5433,
+      "step": 311
+    },
+    {
+      "epoch": 0.15083393763596809,
+      "grad_norm": 0.5647156070035382,
+      "learning_rate": 9.72867622517649e-06,
+      "loss": 0.5859,
+      "step": 312
+    },
+    {
+      "epoch": 0.15131737974377568,
+      "grad_norm": 0.5698749374107666,
+      "learning_rate": 9.726072667520338e-06,
+      "loss": 0.5759,
+      "step": 313
+    },
+    {
+      "epoch": 0.15180082185158328,
+      "grad_norm": 0.4935935341959304,
+      "learning_rate": 9.723457029641547e-06,
+      "loss": 0.5883,
+      "step": 314
+    },
+    {
+      "epoch": 0.15228426395939088,
+      "grad_norm": 0.5040267732247843,
+      "learning_rate": 9.720829318225897e-06,
+      "loss": 0.5723,
+      "step": 315
+    },
+    {
+      "epoch": 0.15276770606719844,
+      "grad_norm": 0.5390674583456238,
+      "learning_rate": 9.718189539990029e-06,
+      "loss": 0.5748,
+      "step": 316
+    },
+    {
+      "epoch": 0.15325114817500604,
+      "grad_norm": 0.5449958057788811,
+      "learning_rate": 9.715537701681431e-06,
+      "loss": 0.5831,
+      "step": 317
+    },
+    {
+      "epoch": 0.15373459028281364,
+      "grad_norm": 0.48895966772949706,
+      "learning_rate": 9.712873810078415e-06,
+      "loss": 0.5505,
+      "step": 318
+    },
+    {
+      "epoch": 0.15421803239062123,
+      "grad_norm": 0.5694877152526486,
+      "learning_rate": 9.710197871990101e-06,
+      "loss": 0.5789,
+      "step": 319
+    },
+    {
+      "epoch": 0.1547014744984288,
+      "grad_norm": 0.5390854150150773,
+      "learning_rate": 9.707509894256406e-06,
+      "loss": 0.5699,
+      "step": 320
+    },
+    {
+      "epoch": 0.1551849166062364,
+      "grad_norm": 0.5339825765060972,
+      "learning_rate": 9.704809883748012e-06,
+      "loss": 0.5841,
+      "step": 321
+    },
+    {
+      "epoch": 0.155668358714044,
+      "grad_norm": 0.5691147363910026,
+      "learning_rate": 9.70209784736637e-06,
+      "loss": 0.5791,
+      "step": 322
+    },
+    {
+      "epoch": 0.1561518008218516,
+      "grad_norm": 0.5098897525025804,
+      "learning_rate": 9.699373792043658e-06,
+      "loss": 0.5789,
+      "step": 323
+    },
+    {
+      "epoch": 0.15663524292965916,
+      "grad_norm": 0.5233093422091403,
+      "learning_rate": 9.696637724742785e-06,
+      "loss": 0.5791,
+      "step": 324
+    },
+    {
+      "epoch": 0.15711868503746676,
+      "grad_norm": 0.4951608627676522,
+      "learning_rate": 9.693889652457359e-06,
+      "loss": 0.5664,
+      "step": 325
+    },
+    {
+      "epoch": 0.15760212714527436,
+      "grad_norm": 0.5085606430384619,
+      "learning_rate": 9.691129582211671e-06,
+      "loss": 0.5777,
+      "step": 326
+    },
+    {
+      "epoch": 0.15808556925308195,
+      "grad_norm": 0.5137102450781047,
+      "learning_rate": 9.688357521060685e-06,
+      "loss": 0.5843,
+      "step": 327
+    },
+    {
+      "epoch": 0.15856901136088952,
+      "grad_norm": 0.4769071854330559,
+      "learning_rate": 9.685573476090015e-06,
+      "loss": 0.578,
+      "step": 328
+    },
+    {
+      "epoch": 0.15905245346869712,
+      "grad_norm": 0.542975418114207,
+      "learning_rate": 9.6827774544159e-06,
+      "loss": 0.5859,
+      "step": 329
+    },
+    {
+      "epoch": 0.15953589557650472,
+      "grad_norm": 0.4926718305346952,
+      "learning_rate": 9.6799694631852e-06,
+      "loss": 0.5871,
+      "step": 330
+    },
+    {
+      "epoch": 0.1600193376843123,
+      "grad_norm": 0.5010989320404932,
+      "learning_rate": 9.677149509575365e-06,
+      "loss": 0.5841,
+      "step": 331
+    },
+    {
+      "epoch": 0.16050277979211988,
+      "grad_norm": 0.5446382005351177,
+      "learning_rate": 9.674317600794426e-06,
+      "loss": 0.5762,
+      "step": 332
+    },
+    {
+      "epoch": 0.16098622189992748,
+      "grad_norm": 0.5406240370145704,
+      "learning_rate": 9.67147374408097e-06,
+      "loss": 0.5685,
+      "step": 333
+    },
+    {
+      "epoch": 0.16146966400773508,
+      "grad_norm": 0.5171074604025283,
+      "learning_rate": 9.66861794670412e-06,
+      "loss": 0.5856,
+      "step": 334
+    },
+    {
+      "epoch": 0.16195310611554267,
+      "grad_norm": 0.5545080974369176,
+      "learning_rate": 9.665750215963528e-06,
+      "loss": 0.5789,
+      "step": 335
+    },
+    {
+      "epoch": 0.16243654822335024,
+      "grad_norm": 0.49939805294647144,
+      "learning_rate": 9.662870559189344e-06,
+      "loss": 0.5702,
+      "step": 336
+    },
+    {
+      "epoch": 0.16291999033115784,
+      "grad_norm": 0.49295646596373777,
+      "learning_rate": 9.6599789837422e-06,
+      "loss": 0.5742,
+      "step": 337
+    },
+    {
+      "epoch": 0.16340343243896543,
+      "grad_norm": 0.5522231456414357,
+      "learning_rate": 9.657075497013202e-06,
+      "loss": 0.5752,
+      "step": 338
+    },
+    {
+      "epoch": 0.16388687454677303,
+      "grad_norm": 0.5606395929711875,
+      "learning_rate": 9.654160106423891e-06,
+      "loss": 0.5854,
+      "step": 339
+    },
+    {
+      "epoch": 0.1643703166545806,
+      "grad_norm": 0.5086990809592122,
+      "learning_rate": 9.651232819426242e-06,
+      "loss": 0.5764,
+      "step": 340
+    },
+    {
+      "epoch": 0.1648537587623882,
+      "grad_norm": 0.4984930367771814,
+      "learning_rate": 9.648293643502636e-06,
+      "loss": 0.5619,
+      "step": 341
+    },
+    {
+      "epoch": 0.1653372008701958,
+      "grad_norm": 0.5217470426797576,
+      "learning_rate": 9.645342586165845e-06,
+      "loss": 0.5833,
+      "step": 342
+    },
+    {
+      "epoch": 0.1658206429780034,
+      "grad_norm": 0.546389261380125,
+      "learning_rate": 9.642379654959006e-06,
+      "loss": 0.5381,
+      "step": 343
+    },
+    {
+      "epoch": 0.166304085085811,
+      "grad_norm": 0.5439151860872452,
+      "learning_rate": 9.639404857455614e-06,
+      "loss": 0.5674,
+      "step": 344
+    },
+    {
+      "epoch": 0.16678752719361856,
+      "grad_norm": 0.5469688158149608,
+      "learning_rate": 9.63641820125949e-06,
+      "loss": 0.5705,
+      "step": 345
+    },
+    {
+      "epoch": 0.16727096930142615,
+      "grad_norm": 0.4994352161741759,
+      "learning_rate": 9.633419694004767e-06,
+      "loss": 0.555,
+      "step": 346
+    },
+    {
+      "epoch": 0.16775441140923375,
+      "grad_norm": 0.5270157823994652,
+      "learning_rate": 9.63040934335587e-06,
+      "loss": 0.5741,
+      "step": 347
+    },
+    {
+      "epoch": 0.16823785351704135,
+      "grad_norm": 0.5302701119307424,
+      "learning_rate": 9.627387157007502e-06,
+      "loss": 0.5775,
+      "step": 348
+    },
+    {
+      "epoch": 0.16872129562484892,
+      "grad_norm": 0.5005904286760833,
+      "learning_rate": 9.624353142684611e-06,
+      "loss": 0.5724,
+      "step": 349
+    },
+    {
+      "epoch": 0.1692047377326565,
+      "grad_norm": 0.5035595085634601,
+      "learning_rate": 9.621307308142385e-06,
+      "loss": 0.5794,
+      "step": 350
+    },
+    {
+      "epoch": 0.1696881798404641,
+      "grad_norm": 0.521381746170865,
+      "learning_rate": 9.618249661166218e-06,
+      "loss": 0.5764,
+      "step": 351
+    },
+    {
+      "epoch": 0.1701716219482717,
+      "grad_norm": 0.48214165657815927,
+      "learning_rate": 9.615180209571709e-06,
+      "loss": 0.5804,
+      "step": 352
+    },
+    {
+      "epoch": 0.17065506405607928,
+      "grad_norm": 0.47552991671065514,
+      "learning_rate": 9.612098961204617e-06,
+      "loss": 0.5581,
+      "step": 353
+    },
+    {
+      "epoch": 0.17113850616388687,
+      "grad_norm": 0.46097880469562935,
+      "learning_rate": 9.609005923940865e-06,
+      "loss": 0.5618,
+      "step": 354
+    },
+    {
+      "epoch": 0.17162194827169447,
+      "grad_norm": 0.5629931104502605,
+      "learning_rate": 9.605901105686503e-06,
+      "loss": 0.5694,
+      "step": 355
+    },
+    {
+      "epoch": 0.17210539037950207,
+      "grad_norm": 0.5179757776717347,
+      "learning_rate": 9.602784514377701e-06,
+      "loss": 0.5897,
+      "step": 356
+    },
+    {
+      "epoch": 0.17258883248730963,
+      "grad_norm": 0.5355839686571028,
+      "learning_rate": 9.599656157980715e-06,
+      "loss": 0.5724,
+      "step": 357
+    },
+    {
+      "epoch": 0.17307227459511723,
+      "grad_norm": 0.6350286695754506,
+      "learning_rate": 9.596516044491873e-06,
+      "loss": 0.577,
+      "step": 358
+    },
+    {
+      "epoch": 0.17355571670292483,
+      "grad_norm": 0.5295601313068036,
+      "learning_rate": 9.593364181937563e-06,
+      "loss": 0.5834,
+      "step": 359
+    },
+    {
+      "epoch": 0.17403915881073242,
+      "grad_norm": 0.5016272467409,
+      "learning_rate": 9.590200578374198e-06,
+      "loss": 0.5848,
+      "step": 360
+    },
+    {
+      "epoch": 0.17452260091854,
+      "grad_norm": 0.4734403734457174,
+      "learning_rate": 9.587025241888202e-06,
+      "loss": 0.5629,
+      "step": 361
+    },
+    {
+      "epoch": 0.1750060430263476,
+      "grad_norm": 0.5345541955737336,
+      "learning_rate": 9.583838180595993e-06,
+      "loss": 0.5619,
+      "step": 362
+    },
+    {
+      "epoch": 0.1754894851341552,
+      "grad_norm": 0.5159159294276754,
+      "learning_rate": 9.580639402643957e-06,
+      "loss": 0.5788,
+      "step": 363
+    },
+    {
+      "epoch": 0.17597292724196278,
+      "grad_norm": 0.5475730953848408,
+      "learning_rate": 9.577428916208426e-06,
+      "loss": 0.5758,
+      "step": 364
+    },
+    {
+      "epoch": 0.17645636934977035,
+      "grad_norm": 0.5065491502971655,
+      "learning_rate": 9.574206729495662e-06,
+      "loss": 0.5739,
+      "step": 365
+    },
+    {
+      "epoch": 0.17693981145757795,
+      "grad_norm": 0.5385122338140608,
+      "learning_rate": 9.570972850741839e-06,
+      "loss": 0.5646,
+      "step": 366
+    },
+    {
+      "epoch": 0.17742325356538555,
+      "grad_norm": 0.5282114345918013,
+      "learning_rate": 9.567727288213005e-06,
+      "loss": 0.5809,
+      "step": 367
+    },
+    {
+      "epoch": 0.17790669567319314,
+      "grad_norm": 0.5183724179001736,
+      "learning_rate": 9.564470050205084e-06,
+      "loss": 0.5745,
+      "step": 368
+    },
+    {
+      "epoch": 0.1783901377810007,
+      "grad_norm": 0.501228022506401,
+      "learning_rate": 9.561201145043835e-06,
+      "loss": 0.5759,
+      "step": 369
+    },
+    {
+      "epoch": 0.1788735798888083,
+      "grad_norm": 0.5161478035704796,
+      "learning_rate": 9.557920581084848e-06,
+      "loss": 0.5716,
+      "step": 370
+    },
+    {
+      "epoch": 0.1793570219966159,
+      "grad_norm": 0.5508440640900468,
+      "learning_rate": 9.554628366713506e-06,
+      "loss": 0.5681,
+      "step": 371
+    },
+    {
+      "epoch": 0.1798404641044235,
+      "grad_norm": 0.4958022642187558,
+      "learning_rate": 9.551324510344972e-06,
+      "loss": 0.5674,
+      "step": 372
+    },
+    {
+      "epoch": 0.18032390621223107,
+      "grad_norm": 0.5211800045547449,
+      "learning_rate": 9.548009020424172e-06,
+      "loss": 0.5759,
+      "step": 373
+    },
+    {
+      "epoch": 0.18080734832003867,
+      "grad_norm": 0.5234346072417955,
+      "learning_rate": 9.544681905425767e-06,
+      "loss": 0.5761,
+      "step": 374
+    },
+    {
+      "epoch": 0.18129079042784627,
+      "grad_norm": 0.5277623761050696,
+      "learning_rate": 9.541343173854128e-06,
+      "loss": 0.5846,
+      "step": 375
+    },
+    {
+      "epoch": 0.18177423253565386,
+      "grad_norm": 0.5159488960453931,
+      "learning_rate": 9.537992834243323e-06,
+      "loss": 0.5655,
+      "step": 376
+    },
+    {
+      "epoch": 0.18225767464346146,
+      "grad_norm": 0.5036893425002033,
+      "learning_rate": 9.53463089515709e-06,
+      "loss": 0.578,
+      "step": 377
+    },
+    {
+      "epoch": 0.18274111675126903,
+      "grad_norm": 0.5500694186101432,
+      "learning_rate": 9.531257365188818e-06,
+      "loss": 0.5683,
+      "step": 378
+    },
+    {
+      "epoch": 0.18322455885907662,
+      "grad_norm": 0.4446123327167339,
+      "learning_rate": 9.527872252961518e-06,
+      "loss": 0.5112,
+      "step": 379
+    },
+    {
+      "epoch": 0.18370800096688422,
+      "grad_norm": 0.49646226307611685,
+      "learning_rate": 9.524475567127813e-06,
+      "loss": 0.5799,
+      "step": 380
+    },
+    {
+      "epoch": 0.18419144307469182,
+      "grad_norm": 0.5455620647014985,
+      "learning_rate": 9.521067316369903e-06,
+      "loss": 0.5601,
+      "step": 381
+    },
+    {
+      "epoch": 0.1846748851824994,
+      "grad_norm": 0.5073331374598753,
+      "learning_rate": 9.517647509399555e-06,
+      "loss": 0.5399,
+      "step": 382
+    },
+    {
+      "epoch": 0.18515832729030698,
+      "grad_norm": 0.5171824333562809,
+      "learning_rate": 9.514216154958067e-06,
+      "loss": 0.5754,
+      "step": 383
+    },
+    {
+      "epoch": 0.18564176939811458,
+      "grad_norm": 0.5085818096253197,
+      "learning_rate": 9.510773261816261e-06,
+      "loss": 0.5623,
+      "step": 384
+    },
+    {
+      "epoch": 0.18612521150592218,
+      "grad_norm": 0.50056273177622,
+      "learning_rate": 9.507318838774448e-06,
+      "loss": 0.5774,
+      "step": 385
+    },
+    {
+      "epoch": 0.18660865361372975,
+      "grad_norm": 0.5493241761943409,
+      "learning_rate": 9.50385289466241e-06,
+      "loss": 0.5698,
+      "step": 386
+    },
+    {
+      "epoch": 0.18709209572153734,
+      "grad_norm": 0.48083872272472233,
+      "learning_rate": 9.500375438339384e-06,
+      "loss": 0.5634,
+      "step": 387
+    },
+    {
+      "epoch": 0.18757553782934494,
+      "grad_norm": 0.48598643847981954,
+      "learning_rate": 9.496886478694025e-06,
+      "loss": 0.5642,
+      "step": 388
+    },
+    {
+      "epoch": 0.18805897993715254,
+      "grad_norm": 0.4945695421669264,
+      "learning_rate": 9.493386024644396e-06,
+      "loss": 0.5763,
+      "step": 389
+    },
+    {
+      "epoch": 0.1885424220449601,
+      "grad_norm": 0.4630609260733735,
+      "learning_rate": 9.48987408513794e-06,
+      "loss": 0.5667,
+      "step": 390
+    },
+    {
+      "epoch": 0.1890258641527677,
+      "grad_norm": 0.5178132025025237,
+      "learning_rate": 9.486350669151455e-06,
+      "loss": 0.5633,
+      "step": 391
+    },
+    {
+      "epoch": 0.1895093062605753,
+      "grad_norm": 0.4855261545618926,
+      "learning_rate": 9.482815785691082e-06,
+      "loss": 0.5705,
+      "step": 392
+    },
+    {
+      "epoch": 0.1899927483683829,
+      "grad_norm": 0.48580056178653924,
+      "learning_rate": 9.47926944379226e-06,
+      "loss": 0.5703,
+      "step": 393
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 0.5308237684959329,
+      "learning_rate": 9.475711652519732e-06,
+      "loss": 0.5583,
+      "step": 394
+    },
+    {
+      "epoch": 0.19095963258399806,
+      "grad_norm": 0.5127712618313278,
+      "learning_rate": 9.472142420967496e-06,
+      "loss": 0.5674,
+      "step": 395
+    },
+    {
+      "epoch": 0.19144307469180566,
+      "grad_norm": 0.4833488281294125,
+      "learning_rate": 9.468561758258795e-06,
+      "loss": 0.578,
+      "step": 396
+    },
+    {
+      "epoch": 0.19192651679961326,
+      "grad_norm": 0.5644191416840888,
+      "learning_rate": 9.464969673546092e-06,
+      "loss": 0.582,
+      "step": 397
+    },
+    {
+      "epoch": 0.19240995890742082,
+      "grad_norm": 0.5561428050479044,
+      "learning_rate": 9.461366176011047e-06,
+      "loss": 0.5762,
+      "step": 398
+    },
+    {
+      "epoch": 0.19289340101522842,
+      "grad_norm": 0.49800634280761286,
+      "learning_rate": 9.457751274864486e-06,
+      "loss": 0.5786,
+      "step": 399
+    },
+    {
+      "epoch": 0.19337684312303602,
+      "grad_norm": 0.464098426014889,
+      "learning_rate": 9.454124979346392e-06,
+      "loss": 0.531,
+      "step": 400
+    },
+    {
+      "epoch": 0.19386028523084362,
+      "grad_norm": 0.5317711530861378,
+      "learning_rate": 9.450487298725866e-06,
+      "loss": 0.5735,
+      "step": 401
+    },
+    {
+      "epoch": 0.19434372733865118,
+      "grad_norm": 0.5700860255634325,
+      "learning_rate": 9.446838242301113e-06,
+      "loss": 0.5736,
+      "step": 402
+    },
+    {
+      "epoch": 0.19482716944645878,
+      "grad_norm": 0.5415575586047788,
+      "learning_rate": 9.443177819399416e-06,
+      "loss": 0.5682,
+      "step": 403
+    },
+    {
+      "epoch": 0.19531061155426638,
+      "grad_norm": 0.45162964809703743,
+      "learning_rate": 9.439506039377111e-06,
+      "loss": 0.5457,
+      "step": 404
+    },
+    {
+      "epoch": 0.19579405366207397,
+      "grad_norm": 0.48073200361222107,
+      "learning_rate": 9.435822911619564e-06,
+      "loss": 0.5452,
+      "step": 405
+    },
+    {
+      "epoch": 0.19627749576988154,
+      "grad_norm": 0.5218011226870963,
+      "learning_rate": 9.432128445541147e-06,
+      "loss": 0.5569,
+      "step": 406
+    },
+    {
+      "epoch": 0.19676093787768914,
+      "grad_norm": 0.5241766492312198,
+      "learning_rate": 9.42842265058521e-06,
+      "loss": 0.5791,
+      "step": 407
+    },
+    {
+      "epoch": 0.19724437998549674,
+      "grad_norm": 0.4747479232641684,
+      "learning_rate": 9.424705536224065e-06,
+      "loss": 0.572,
+      "step": 408
+    },
+    {
+      "epoch": 0.19772782209330433,
+      "grad_norm": 0.4892195750767198,
+      "learning_rate": 9.420977111958957e-06,
+      "loss": 0.577,
+      "step": 409
+    },
+    {
+      "epoch": 0.1982112642011119,
+      "grad_norm": 0.49625147154018395,
+      "learning_rate": 9.41723738732004e-06,
+      "loss": 0.5673,
+      "step": 410
+    },
+    {
+      "epoch": 0.1986947063089195,
+      "grad_norm": 0.553969116933997,
+      "learning_rate": 9.41348637186635e-06,
+      "loss": 0.5805,
+      "step": 411
+    },
+    {
+      "epoch": 0.1991781484167271,
+      "grad_norm": 0.5271833056864474,
+      "learning_rate": 9.409724075185782e-06,
+      "loss": 0.5811,
+      "step": 412
+    },
+    {
+      "epoch": 0.1996615905245347,
+      "grad_norm": 0.541152410560869,
+      "learning_rate": 9.405950506895074e-06,
+      "loss": 0.5539,
+      "step": 413
+    },
+    {
+      "epoch": 0.2001450326323423,
+      "grad_norm": 0.4827367980584999,
+      "learning_rate": 9.40216567663977e-06,
+      "loss": 0.5754,
+      "step": 414
+    },
+    {
+      "epoch": 0.20062847474014986,
+      "grad_norm": 0.49177545628835745,
+      "learning_rate": 9.398369594094198e-06,
+      "loss": 0.508,
+      "step": 415
+    },
+    {
+      "epoch": 0.20111191684795746,
+      "grad_norm": 0.50467312755319,
+      "learning_rate": 9.394562268961454e-06,
+      "loss": 0.5681,
+      "step": 416
+    },
+    {
+      "epoch": 0.20159535895576505,
+      "grad_norm": 0.4916777572033636,
+      "learning_rate": 9.390743710973366e-06,
+      "loss": 0.575,
+      "step": 417
+    },
+    {
+      "epoch": 0.20207880106357265,
+      "grad_norm": 0.5183550927798377,
+      "learning_rate": 9.386913929890478e-06,
+      "loss": 0.57,
+      "step": 418
+    },
+    {
+      "epoch": 0.20256224317138022,
+      "grad_norm": 0.47362092706218123,
+      "learning_rate": 9.383072935502018e-06,
+      "loss": 0.5644,
+      "step": 419
+    },
+    {
+      "epoch": 0.20304568527918782,
+      "grad_norm": 0.49530019201729136,
+      "learning_rate": 9.379220737625877e-06,
+      "loss": 0.564,
+      "step": 420
+    },
+    {
+      "epoch": 0.2035291273869954,
+      "grad_norm": 0.5108751966700111,
+      "learning_rate": 9.375357346108583e-06,
+      "loss": 0.5602,
+      "step": 421
+    },
+    {
+      "epoch": 0.204012569494803,
+      "grad_norm": 0.5230318233484302,
+      "learning_rate": 9.371482770825277e-06,
+      "loss": 0.5695,
+      "step": 422
+    },
+    {
+      "epoch": 0.20449601160261058,
+      "grad_norm": 0.4802393361720882,
+      "learning_rate": 9.367597021679686e-06,
+      "loss": 0.5661,
+      "step": 423
+    },
+    {
+      "epoch": 0.20497945371041817,
+      "grad_norm": 0.541773712373739,
+      "learning_rate": 9.363700108604096e-06,
+      "loss": 0.5582,
+      "step": 424
+    },
+    {
+      "epoch": 0.20546289581822577,
+      "grad_norm": 0.5110126727655455,
+      "learning_rate": 9.359792041559334e-06,
+      "loss": 0.5645,
+      "step": 425
+    },
+    {
+      "epoch": 0.20594633792603337,
+      "grad_norm": 0.5486480496411716,
+      "learning_rate": 9.35587283053473e-06,
+      "loss": 0.5677,
+      "step": 426
+    },
+    {
+      "epoch": 0.20642978003384094,
+      "grad_norm": 0.5379779057549923,
+      "learning_rate": 9.351942485548109e-06,
+      "loss": 0.5435,
+      "step": 427
+    },
+    {
+      "epoch": 0.20691322214164853,
+      "grad_norm": 0.5341397558862222,
+      "learning_rate": 9.348001016645744e-06,
+      "loss": 0.5599,
+      "step": 428
+    },
+    {
+      "epoch": 0.20739666424945613,
+      "grad_norm": 0.44238086682442823,
+      "learning_rate": 9.344048433902351e-06,
+      "loss": 0.541,
+      "step": 429
+    },
+    {
+      "epoch": 0.20788010635726373,
+      "grad_norm": 0.5213851954927032,
+      "learning_rate": 9.340084747421048e-06,
+      "loss": 0.5366,
+      "step": 430
+    },
+    {
+      "epoch": 0.2083635484650713,
+      "grad_norm": 0.5349032988779688,
+      "learning_rate": 9.336109967333337e-06,
+      "loss": 0.5571,
+      "step": 431
+    },
+    {
+      "epoch": 0.2088469905728789,
+      "grad_norm": 0.4554230771685569,
+      "learning_rate": 9.332124103799075e-06,
+      "loss": 0.5516,
+      "step": 432
+    },
+    {
+      "epoch": 0.2093304326806865,
+      "grad_norm": 0.5021585721937876,
+      "learning_rate": 9.328127167006457e-06,
+      "loss": 0.5679,
+      "step": 433
+    },
+    {
+      "epoch": 0.2098138747884941,
+      "grad_norm": 0.5025134126056662,
+      "learning_rate": 9.324119167171967e-06,
+      "loss": 0.5659,
+      "step": 434
+    },
+    {
+      "epoch": 0.21029731689630166,
+      "grad_norm": 0.48977518403096176,
+      "learning_rate": 9.320100114540382e-06,
+      "loss": 0.5753,
+      "step": 435
+    },
+    {
+      "epoch": 0.21078075900410925,
+      "grad_norm": 0.4789181842167065,
+      "learning_rate": 9.316070019384722e-06,
+      "loss": 0.558,
+      "step": 436
+    },
+    {
+      "epoch": 0.21126420111191685,
+      "grad_norm": 0.48417362744631853,
+      "learning_rate": 9.312028892006233e-06,
+      "loss": 0.5637,
+      "step": 437
+    },
+    {
+      "epoch": 0.21174764321972445,
+      "grad_norm": 0.5040441298097904,
+      "learning_rate": 9.307976742734366e-06,
+      "loss": 0.5603,
+      "step": 438
+    },
+    {
+      "epoch": 0.21223108532753202,
+      "grad_norm": 0.5003182083782678,
+      "learning_rate": 9.30391358192674e-06,
+      "loss": 0.5583,
+      "step": 439
+    },
+    {
+      "epoch": 0.2127145274353396,
+      "grad_norm": 0.5188458903874932,
+      "learning_rate": 9.299839419969119e-06,
+      "loss": 0.5614,
+      "step": 440
+    },
+    {
+      "epoch": 0.2131979695431472,
+      "grad_norm": 0.4990120996823676,
+      "learning_rate": 9.295754267275393e-06,
+      "loss": 0.5732,
+      "step": 441
+    },
+    {
+      "epoch": 0.2136814116509548,
+      "grad_norm": 0.43407580533296863,
+      "learning_rate": 9.291658134287537e-06,
+      "loss": 0.5451,
+      "step": 442
+    },
+    {
+      "epoch": 0.21416485375876237,
+      "grad_norm": 0.487299832131986,
+      "learning_rate": 9.287551031475604e-06,
+      "loss": 0.5486,
+      "step": 443
+    },
+    {
+      "epoch": 0.21464829586656997,
+      "grad_norm": 0.4748601209022523,
+      "learning_rate": 9.283432969337672e-06,
+      "loss": 0.5568,
+      "step": 444
+    },
+    {
+      "epoch": 0.21513173797437757,
+      "grad_norm": 0.5116954397180901,
+      "learning_rate": 9.279303958399846e-06,
+      "loss": 0.5561,
+      "step": 445
+    },
+    {
+      "epoch": 0.21561518008218516,
+      "grad_norm": 0.5103832796562369,
+      "learning_rate": 9.275164009216205e-06,
+      "loss": 0.5653,
+      "step": 446
+    },
+    {
+      "epoch": 0.21609862218999276,
+      "grad_norm": 0.453674255766726,
+      "learning_rate": 9.271013132368799e-06,
+      "loss": 0.5359,
+      "step": 447
+    },
+    {
+      "epoch": 0.21658206429780033,
+      "grad_norm": 0.4865827031825044,
+      "learning_rate": 9.266851338467598e-06,
+      "loss": 0.5627,
+      "step": 448
+    },
+    {
+      "epoch": 0.21706550640560793,
+      "grad_norm": 0.4474998958247519,
+      "learning_rate": 9.262678638150486e-06,
+      "loss": 0.5372,
+      "step": 449
+    },
+    {
+      "epoch": 0.21754894851341552,
+      "grad_norm": 0.5312817145455567,
+      "learning_rate": 9.258495042083222e-06,
+      "loss": 0.583,
+      "step": 450
+    },
+    {
+      "epoch": 0.21803239062122312,
+      "grad_norm": 0.5326646088756841,
+      "learning_rate": 9.254300560959413e-06,
+      "loss": 0.5641,
+      "step": 451
+    },
+    {
+      "epoch": 0.2185158327290307,
+      "grad_norm": 0.47741110714076435,
+      "learning_rate": 9.25009520550049e-06,
+      "loss": 0.5692,
+      "step": 452
+    },
+    {
+      "epoch": 0.2189992748368383,
+      "grad_norm": 0.4992778758439529,
+      "learning_rate": 9.245878986455684e-06,
+      "loss": 0.5732,
+      "step": 453
+    },
+    {
+      "epoch": 0.21948271694464588,
+      "grad_norm": 0.5067531688765293,
+      "learning_rate": 9.241651914601986e-06,
+      "loss": 0.5684,
+      "step": 454
+    },
+    {
+      "epoch": 0.21996615905245348,
+      "grad_norm": 0.5259329600281596,
+      "learning_rate": 9.237414000744134e-06,
+      "loss": 0.5728,
+      "step": 455
+    },
+    {
+      "epoch": 0.22044960116026105,
+      "grad_norm": 0.4912112930780334,
+      "learning_rate": 9.23316525571458e-06,
+      "loss": 0.5543,
+      "step": 456
+    },
+    {
+      "epoch": 0.22093304326806865,
+      "grad_norm": 0.4325116439857764,
+      "learning_rate": 9.228905690373456e-06,
+      "loss": 0.5109,
+      "step": 457
+    },
+    {
+      "epoch": 0.22141648537587624,
+      "grad_norm": 0.5251969417490432,
+      "learning_rate": 9.224635315608554e-06,
+      "loss": 0.5613,
+      "step": 458
+    },
+    {
+      "epoch": 0.22189992748368384,
+      "grad_norm": 0.5371164613513753,
+      "learning_rate": 9.2203541423353e-06,
+      "loss": 0.5758,
+      "step": 459
+    },
+    {
+      "epoch": 0.2223833695914914,
+      "grad_norm": 0.49879877094748626,
+      "learning_rate": 9.216062181496712e-06,
+      "loss": 0.5656,
+      "step": 460
+    },
+    {
+      "epoch": 0.222866811699299,
+      "grad_norm": 0.4666739038962981,
+      "learning_rate": 9.211759444063392e-06,
+      "loss": 0.5643,
+      "step": 461
+    },
+    {
+      "epoch": 0.2233502538071066,
+      "grad_norm": 0.5019702713381807,
+      "learning_rate": 9.207445941033483e-06,
+      "loss": 0.5645,
+      "step": 462
+    },
+    {
+      "epoch": 0.2238336959149142,
+      "grad_norm": 0.560484985437826,
+      "learning_rate": 9.203121683432646e-06,
+      "loss": 0.5622,
+      "step": 463
+    },
+    {
+      "epoch": 0.22431713802272177,
+      "grad_norm": 0.501701537299382,
+      "learning_rate": 9.19878668231403e-06,
+      "loss": 0.5686,
+      "step": 464
+    },
+    {
+      "epoch": 0.22480058013052936,
+      "grad_norm": 0.48640275847390047,
+      "learning_rate": 9.19444094875825e-06,
+      "loss": 0.5617,
+      "step": 465
+    },
+    {
+      "epoch": 0.22528402223833696,
+      "grad_norm": 0.5066662929437282,
+      "learning_rate": 9.190084493873353e-06,
+      "loss": 0.5733,
+      "step": 466
+    },
+    {
+      "epoch": 0.22576746434614456,
+      "grad_norm": 0.5297511031777309,
+      "learning_rate": 9.185717328794784e-06,
+      "loss": 0.5632,
+      "step": 467
+    },
+    {
+      "epoch": 0.22625090645395213,
+      "grad_norm": 0.5778692323663056,
+      "learning_rate": 9.18133946468537e-06,
+      "loss": 0.5684,
+      "step": 468
+    },
+    {
+      "epoch": 0.22673434856175972,
+      "grad_norm": 0.5148715492097395,
+      "learning_rate": 9.176950912735287e-06,
+      "loss": 0.5559,
+      "step": 469
+    },
+    {
+      "epoch": 0.22721779066956732,
+      "grad_norm": 0.5157447753884506,
+      "learning_rate": 9.172551684162025e-06,
+      "loss": 0.5731,
+      "step": 470
+    },
+    {
+      "epoch": 0.22770123277737492,
+      "grad_norm": 0.4783319000473412,
+      "learning_rate": 9.16814179021037e-06,
+      "loss": 0.5671,
+      "step": 471
+    },
+    {
+      "epoch": 0.2281846748851825,
+      "grad_norm": 0.5017422895280137,
+      "learning_rate": 9.163721242152362e-06,
+      "loss": 0.5661,
+      "step": 472
+    },
+    {
+      "epoch": 0.22866811699299008,
+      "grad_norm": 0.49272124386072536,
+      "learning_rate": 9.159290051287282e-06,
+      "loss": 0.5627,
+      "step": 473
+    },
+    {
+      "epoch": 0.22915155910079768,
+      "grad_norm": 0.47471736533769476,
+      "learning_rate": 9.154848228941607e-06,
+      "loss": 0.5615,
+      "step": 474
+    },
+    {
+      "epoch": 0.22963500120860528,
+      "grad_norm": 0.5071884927272643,
+      "learning_rate": 9.150395786468998e-06,
+      "loss": 0.5645,
+      "step": 475
+    },
+    {
+      "epoch": 0.23011844331641285,
+      "grad_norm": 0.48690399925776484,
+      "learning_rate": 9.14593273525025e-06,
+      "loss": 0.5647,
+      "step": 476
+    },
+    {
+      "epoch": 0.23060188542422044,
+      "grad_norm": 0.5041235784595942,
+      "learning_rate": 9.14145908669329e-06,
+      "loss": 0.5729,
+      "step": 477
+    },
+    {
+      "epoch": 0.23108532753202804,
+      "grad_norm": 0.5265161224054821,
+      "learning_rate": 9.136974852233118e-06,
+      "loss": 0.5587,
+      "step": 478
+    },
+    {
+      "epoch": 0.23156876963983564,
+      "grad_norm": 0.4778337324840926,
+      "learning_rate": 9.132480043331801e-06,
+      "loss": 0.5646,
+      "step": 479
+    },
+    {
+      "epoch": 0.23205221174764323,
+      "grad_norm": 0.5036800160533508,
+      "learning_rate": 9.127974671478432e-06,
+      "loss": 0.5655,
+      "step": 480
+    },
+    {
+      "epoch": 0.2325356538554508,
+      "grad_norm": 0.4915164507750186,
+      "learning_rate": 9.123458748189105e-06,
+      "loss": 0.5608,
+      "step": 481
+    },
+    {
+      "epoch": 0.2330190959632584,
+      "grad_norm": 0.4447947403953834,
+      "learning_rate": 9.118932285006886e-06,
+      "loss": 0.5254,
+      "step": 482
+    },
+    {
+      "epoch": 0.233502538071066,
+      "grad_norm": 0.4936810479165672,
+      "learning_rate": 9.114395293501775e-06,
+      "loss": 0.5751,
+      "step": 483
+    },
+    {
+      "epoch": 0.2339859801788736,
+      "grad_norm": 0.4933009245810686,
+      "learning_rate": 9.10984778527069e-06,
+      "loss": 0.5603,
+      "step": 484
+    },
+    {
+      "epoch": 0.23446942228668116,
+      "grad_norm": 0.4720549987110232,
+      "learning_rate": 9.10528977193743e-06,
+      "loss": 0.5703,
+      "step": 485
+    },
+    {
+      "epoch": 0.23495286439448876,
+      "grad_norm": 0.5362136689894559,
+      "learning_rate": 9.100721265152644e-06,
+      "loss": 0.5635,
+      "step": 486
+    },
+    {
+      "epoch": 0.23543630650229636,
+      "grad_norm": 0.47602005538977166,
+      "learning_rate": 9.096142276593802e-06,
+      "loss": 0.5721,
+      "step": 487
+    },
+    {
+      "epoch": 0.23591974861010395,
+      "grad_norm": 0.48887012727323886,
+      "learning_rate": 9.09155281796517e-06,
+      "loss": 0.5502,
+      "step": 488
+    },
+    {
+      "epoch": 0.23640319071791152,
+      "grad_norm": 0.5468866437635687,
+      "learning_rate": 9.086952900997774e-06,
+      "loss": 0.5628,
+      "step": 489
+    },
+    {
+      "epoch": 0.23688663282571912,
+      "grad_norm": 0.468285091758703,
+      "learning_rate": 9.082342537449369e-06,
+      "loss": 0.5649,
+      "step": 490
+    },
+    {
+      "epoch": 0.23737007493352671,
+      "grad_norm": 0.49449575173177474,
+      "learning_rate": 9.07772173910442e-06,
+      "loss": 0.5363,
+      "step": 491
+    },
+    {
+      "epoch": 0.2378535170413343,
+      "grad_norm": 0.5665277859908898,
+      "learning_rate": 9.073090517774057e-06,
+      "loss": 0.5679,
+      "step": 492
+    },
+    {
+      "epoch": 0.23833695914914188,
+      "grad_norm": 0.559218042712036,
+      "learning_rate": 9.068448885296057e-06,
+      "loss": 0.5598,
+      "step": 493
+    },
+    {
+      "epoch": 0.23882040125694948,
+      "grad_norm": 0.5572180299965971,
+      "learning_rate": 9.063796853534808e-06,
+      "loss": 0.5606,
+      "step": 494
+    },
+    {
+      "epoch": 0.23930384336475707,
+      "grad_norm": 0.4852501650353095,
+      "learning_rate": 9.059134434381274e-06,
+      "loss": 0.5614,
+      "step": 495
+    },
+    {
+      "epoch": 0.23978728547256467,
+      "grad_norm": 0.5235782249928449,
+      "learning_rate": 9.054461639752976e-06,
+      "loss": 0.5637,
+      "step": 496
+    },
+    {
+      "epoch": 0.24027072758037224,
+      "grad_norm": 0.5028533022976227,
+      "learning_rate": 9.049778481593954e-06,
+      "loss": 0.5718,
+      "step": 497
+    },
+    {
+      "epoch": 0.24075416968817984,
+      "grad_norm": 0.508045864936268,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.5651,
+      "step": 498
+    },
+    {
+      "epoch": 0.24123761179598743,
+      "grad_norm": 0.5575870011120908,
+      "learning_rate": 9.040381122592317e-06,
+      "loss": 0.565,
+      "step": 499
+    },
+    {
+      "epoch": 0.24172105390379503,
+      "grad_norm": 0.5201685839473924,
+      "learning_rate": 9.035666945770107e-06,
+      "loss": 0.5593,
+      "step": 500
+    },
+    {
+      "epoch": 0.2422044960116026,
+      "grad_norm": 0.48179233555943923,
+      "learning_rate": 9.030942453457928e-06,
+      "loss": 0.5199,
+      "step": 501
+    },
+    {
+      "epoch": 0.2426879381194102,
+      "grad_norm": 0.4867208952029737,
+      "learning_rate": 9.02620765773196e-06,
+      "loss": 0.5548,
+      "step": 502
+    },
+    {
+      "epoch": 0.2431713802272178,
+      "grad_norm": 0.5240394440690106,
+      "learning_rate": 9.02146257069472e-06,
+      "loss": 0.5611,
+      "step": 503
+    },
+    {
+      "epoch": 0.2436548223350254,
+      "grad_norm": 0.48307750050965703,
+      "learning_rate": 9.01670720447504e-06,
+      "loss": 0.5577,
+      "step": 504
+    },
+    {
+      "epoch": 0.24413826444283296,
+      "grad_norm": 0.5034030614527921,
+      "learning_rate": 9.011941571228015e-06,
+      "loss": 0.5608,
+      "step": 505
+    },
+    {
+      "epoch": 0.24462170655064056,
+      "grad_norm": 0.46379490536223517,
+      "learning_rate": 9.007165683134986e-06,
+      "loss": 0.5315,
+      "step": 506
+    },
+    {
+      "epoch": 0.24510514865844815,
+      "grad_norm": 0.5103811282689319,
+      "learning_rate": 9.00237955240351e-06,
+      "loss": 0.5613,
+      "step": 507
+    },
+    {
+      "epoch": 0.24558859076625575,
+      "grad_norm": 0.47564392120255755,
+      "learning_rate": 8.997583191267326e-06,
+      "loss": 0.5764,
+      "step": 508
+    },
+    {
+      "epoch": 0.24607203287406332,
+      "grad_norm": 0.4811799201923712,
+      "learning_rate": 8.992776611986313e-06,
+      "loss": 0.5704,
+      "step": 509
+    },
+    {
+      "epoch": 0.24655547498187091,
+      "grad_norm": 0.4799439081762819,
+      "learning_rate": 8.987959826846479e-06,
+      "loss": 0.5573,
+      "step": 510
+    },
+    {
+      "epoch": 0.2470389170896785,
+      "grad_norm": 0.5219349618857427,
+      "learning_rate": 8.983132848159916e-06,
+      "loss": 0.5583,
+      "step": 511
+    },
+    {
+      "epoch": 0.2475223591974861,
+      "grad_norm": 0.5010818591918965,
+      "learning_rate": 8.978295688264768e-06,
+      "loss": 0.5699,
+      "step": 512
+    },
+    {
+      "epoch": 0.2480058013052937,
+      "grad_norm": 0.5282819201955711,
+      "learning_rate": 8.973448359525207e-06,
+      "loss": 0.5641,
+      "step": 513
+    },
+    {
+      "epoch": 0.24848924341310127,
+      "grad_norm": 0.5025819972323563,
+      "learning_rate": 8.968590874331395e-06,
+      "loss": 0.5649,
+      "step": 514
+    },
+    {
+      "epoch": 0.24897268552090887,
+      "grad_norm": 0.4880024154213522,
+      "learning_rate": 8.963723245099456e-06,
+      "loss": 0.5533,
+      "step": 515
+    },
+    {
+      "epoch": 0.24945612762871647,
+      "grad_norm": 0.4844265343558768,
+      "learning_rate": 8.958845484271443e-06,
+      "loss": 0.5571,
+      "step": 516
+    },
+    {
+      "epoch": 0.24993956973652406,
+      "grad_norm": 0.4918270286134992,
+      "learning_rate": 8.953957604315306e-06,
+      "loss": 0.5612,
+      "step": 517
+    },
+    {
+      "epoch": 0.25042301184433163,
+      "grad_norm": 0.4532098318099568,
+      "learning_rate": 8.949059617724859e-06,
+      "loss": 0.5532,
+      "step": 518
+    },
+    {
+      "epoch": 0.25090645395213923,
+      "grad_norm": 0.4784777680132966,
+      "learning_rate": 8.944151537019752e-06,
+      "loss": 0.5314,
+      "step": 519
+    },
+    {
+      "epoch": 0.2513898960599468,
+      "grad_norm": 0.49834032614411844,
+      "learning_rate": 8.939233374745432e-06,
+      "loss": 0.561,
+      "step": 520
+    },
+    {
+      "epoch": 0.2518733381677544,
+      "grad_norm": 0.43922831313439964,
+      "learning_rate": 8.934305143473123e-06,
+      "loss": 0.5229,
+      "step": 521
+    },
+    {
+      "epoch": 0.252356780275562,
+      "grad_norm": 0.4774051999235377,
+      "learning_rate": 8.929366855799777e-06,
+      "loss": 0.5584,
+      "step": 522
+    },
+    {
+      "epoch": 0.2528402223833696,
+      "grad_norm": 0.4860585540987837,
+      "learning_rate": 8.924418524348058e-06,
+      "loss": 0.5722,
+      "step": 523
+    },
+    {
+      "epoch": 0.25332366449117716,
+      "grad_norm": 0.476115105724116,
+      "learning_rate": 8.919460161766299e-06,
+      "loss": 0.5527,
+      "step": 524
+    },
+    {
+      "epoch": 0.25380710659898476,
+      "grad_norm": 0.49670836036646415,
+      "learning_rate": 8.914491780728471e-06,
+      "loss": 0.565,
+      "step": 525
+    },
+    {
+      "epoch": 0.25429054870679235,
+      "grad_norm": 0.49705890206049747,
+      "learning_rate": 8.909513393934162e-06,
+      "loss": 0.5562,
+      "step": 526
+    },
+    {
+      "epoch": 0.25477399081459995,
+      "grad_norm": 0.5118474736649574,
+      "learning_rate": 8.904525014108529e-06,
+      "loss": 0.5536,
+      "step": 527
+    },
+    {
+      "epoch": 0.25525743292240755,
+      "grad_norm": 0.5301718242423505,
+      "learning_rate": 8.899526654002268e-06,
+      "loss": 0.5612,
+      "step": 528
+    },
+    {
+      "epoch": 0.25574087503021514,
+      "grad_norm": 0.4796891269551852,
+      "learning_rate": 8.894518326391595e-06,
+      "loss": 0.5578,
+      "step": 529
+    },
+    {
+      "epoch": 0.25622431713802274,
+      "grad_norm": 0.4825310469483714,
+      "learning_rate": 8.889500044078199e-06,
+      "loss": 0.5554,
+      "step": 530
+    },
+    {
+      "epoch": 0.25670775924583034,
+      "grad_norm": 0.474416307358851,
+      "learning_rate": 8.88447181988921e-06,
+      "loss": 0.5466,
+      "step": 531
+    },
+    {
+      "epoch": 0.2571912013536379,
+      "grad_norm": 0.4754427571901456,
+      "learning_rate": 8.87943366667718e-06,
+      "loss": 0.5232,
+      "step": 532
+    },
+    {
+      "epoch": 0.2576746434614455,
+      "grad_norm": 0.5283380707149146,
+      "learning_rate": 8.87438559732003e-06,
+      "loss": 0.5575,
+      "step": 533
+    },
+    {
+      "epoch": 0.25815808556925307,
+      "grad_norm": 0.49022668890084664,
+      "learning_rate": 8.869327624721033e-06,
+      "loss": 0.5584,
+      "step": 534
+    },
+    {
+      "epoch": 0.25864152767706067,
+      "grad_norm": 0.45329648879294543,
+      "learning_rate": 8.864259761808778e-06,
+      "loss": 0.5557,
+      "step": 535
+    },
+    {
+      "epoch": 0.25912496978486826,
+      "grad_norm": 0.5150923796193744,
+      "learning_rate": 8.859182021537126e-06,
+      "loss": 0.5672,
+      "step": 536
+    },
+    {
+      "epoch": 0.25960841189267586,
+      "grad_norm": 0.49475203737919254,
+      "learning_rate": 8.854094416885192e-06,
+      "loss": 0.5513,
+      "step": 537
+    },
+    {
+      "epoch": 0.26009185400048346,
+      "grad_norm": 0.48640723658571816,
+      "learning_rate": 8.848996960857308e-06,
+      "loss": 0.5542,
+      "step": 538
+    },
+    {
+      "epoch": 0.26057529610829105,
+      "grad_norm": 0.5011403090647114,
+      "learning_rate": 8.843889666482977e-06,
+      "loss": 0.5503,
+      "step": 539
+    },
+    {
+      "epoch": 0.2610587382160986,
+      "grad_norm": 0.45868293065964316,
+      "learning_rate": 8.838772546816857e-06,
+      "loss": 0.5245,
+      "step": 540
+    },
+    {
+      "epoch": 0.2615421803239062,
+      "grad_norm": 0.5028670832415251,
+      "learning_rate": 8.833645614938716e-06,
+      "loss": 0.563,
+      "step": 541
+    },
+    {
+      "epoch": 0.2620256224317138,
+      "grad_norm": 0.4652958998559184,
+      "learning_rate": 8.82850888395341e-06,
+      "loss": 0.5214,
+      "step": 542
+    },
+    {
+      "epoch": 0.2625090645395214,
+      "grad_norm": 0.48814680090193757,
+      "learning_rate": 8.823362366990833e-06,
+      "loss": 0.5539,
+      "step": 543
+    },
+    {
+      "epoch": 0.262992506647329,
+      "grad_norm": 0.4686742850265713,
+      "learning_rate": 8.818206077205899e-06,
+      "loss": 0.5432,
+      "step": 544
+    },
+    {
+      "epoch": 0.2634759487551366,
+      "grad_norm": 0.44177435010013455,
+      "learning_rate": 8.8130400277785e-06,
+      "loss": 0.5432,
+      "step": 545
+    },
+    {
+      "epoch": 0.2639593908629442,
+      "grad_norm": 0.48745119596264225,
+      "learning_rate": 8.807864231913475e-06,
+      "loss": 0.5609,
+      "step": 546
+    },
+    {
+      "epoch": 0.2644428329707518,
+      "grad_norm": 0.5387031701921053,
+      "learning_rate": 8.802678702840575e-06,
+      "loss": 0.5608,
+      "step": 547
+    },
+    {
+      "epoch": 0.26492627507855937,
+      "grad_norm": 0.47706550642594997,
+      "learning_rate": 8.79748345381443e-06,
+      "loss": 0.5487,
+      "step": 548
+    },
+    {
+      "epoch": 0.2654097171863669,
+      "grad_norm": 0.4694250929319588,
+      "learning_rate": 8.792278498114517e-06,
+      "loss": 0.549,
+      "step": 549
+    },
+    {
+      "epoch": 0.2658931592941745,
+      "grad_norm": 0.4937111232536657,
+      "learning_rate": 8.78706384904512e-06,
+      "loss": 0.5564,
+      "step": 550
+    },
+    {
+      "epoch": 0.2663766014019821,
+      "grad_norm": 0.468357648344204,
+      "learning_rate": 8.7818395199353e-06,
+      "loss": 0.5546,
+      "step": 551
+    },
+    {
+      "epoch": 0.2668600435097897,
+      "grad_norm": 0.4713572915315673,
+      "learning_rate": 8.77660552413887e-06,
+      "loss": 0.5512,
+      "step": 552
+    },
+    {
+      "epoch": 0.2673434856175973,
+      "grad_norm": 0.4740909938486332,
+      "learning_rate": 8.77136187503434e-06,
+      "loss": 0.5631,
+      "step": 553
+    },
+    {
+      "epoch": 0.2678269277254049,
+      "grad_norm": 0.4472174307551216,
+      "learning_rate": 8.766108586024904e-06,
+      "loss": 0.5222,
+      "step": 554
+    },
+    {
+      "epoch": 0.2683103698332125,
+      "grad_norm": 0.4548110236983466,
+      "learning_rate": 8.760845670538387e-06,
+      "loss": 0.5485,
+      "step": 555
+    },
+    {
+      "epoch": 0.2687938119410201,
+      "grad_norm": 0.5173119662805489,
+      "learning_rate": 8.755573142027228e-06,
+      "loss": 0.5624,
+      "step": 556
+    },
+    {
+      "epoch": 0.26927725404882763,
+      "grad_norm": 0.4812632123799694,
+      "learning_rate": 8.750291013968432e-06,
+      "loss": 0.5562,
+      "step": 557
+    },
+    {
+      "epoch": 0.2697606961566352,
+      "grad_norm": 0.472663174890125,
+      "learning_rate": 8.744999299863549e-06,
+      "loss": 0.5669,
+      "step": 558
+    },
+    {
+      "epoch": 0.2702441382644428,
+      "grad_norm": 0.463122081686998,
+      "learning_rate": 8.739698013238625e-06,
+      "loss": 0.557,
+      "step": 559
+    },
+    {
+      "epoch": 0.2707275803722504,
+      "grad_norm": 0.5188284707009508,
+      "learning_rate": 8.734387167644171e-06,
+      "loss": 0.5202,
+      "step": 560
+    },
+    {
+      "epoch": 0.271211022480058,
+      "grad_norm": 0.49659334079030504,
+      "learning_rate": 8.729066776655144e-06,
+      "loss": 0.5605,
+      "step": 561
+    },
+    {
+      "epoch": 0.2716944645878656,
+      "grad_norm": 0.4433765304016552,
+      "learning_rate": 8.723736853870888e-06,
+      "loss": 0.5193,
+      "step": 562
+    },
+    {
+      "epoch": 0.2721779066956732,
+      "grad_norm": 0.46285084832800716,
+      "learning_rate": 8.718397412915114e-06,
+      "loss": 0.5583,
+      "step": 563
+    },
+    {
+      "epoch": 0.2726613488034808,
+      "grad_norm": 0.4406166472711255,
+      "learning_rate": 8.713048467435865e-06,
+      "loss": 0.5365,
+      "step": 564
+    },
+    {
+      "epoch": 0.27314479091128835,
+      "grad_norm": 0.49591339367367465,
+      "learning_rate": 8.707690031105478e-06,
+      "loss": 0.5638,
+      "step": 565
+    },
+    {
+      "epoch": 0.27362823301909595,
+      "grad_norm": 0.47717175741546425,
+      "learning_rate": 8.702322117620547e-06,
+      "loss": 0.5375,
+      "step": 566
+    },
+    {
+      "epoch": 0.27411167512690354,
+      "grad_norm": 0.49399204569955096,
+      "learning_rate": 8.696944740701891e-06,
+      "loss": 0.5502,
+      "step": 567
+    },
+    {
+      "epoch": 0.27459511723471114,
+      "grad_norm": 0.462084403002843,
+      "learning_rate": 8.69155791409452e-06,
+      "loss": 0.549,
+      "step": 568
+    },
+    {
+      "epoch": 0.27507855934251874,
+      "grad_norm": 0.4733870628371529,
+      "learning_rate": 8.686161651567596e-06,
+      "loss": 0.5479,
+      "step": 569
+    },
+    {
+      "epoch": 0.27556200145032633,
+      "grad_norm": 0.4586305030542931,
+      "learning_rate": 8.6807559669144e-06,
+      "loss": 0.517,
+      "step": 570
+    },
+    {
+      "epoch": 0.27604544355813393,
+      "grad_norm": 0.4786603573138254,
+      "learning_rate": 8.6753408739523e-06,
+      "loss": 0.5449,
+      "step": 571
+    },
+    {
+      "epoch": 0.2765288856659415,
+      "grad_norm": 0.4923356872720239,
+      "learning_rate": 8.669916386522708e-06,
+      "loss": 0.5516,
+      "step": 572
+    },
+    {
+      "epoch": 0.27701232777374907,
+      "grad_norm": 0.47497918747290174,
+      "learning_rate": 8.664482518491053e-06,
+      "loss": 0.5527,
+      "step": 573
+    },
+    {
+      "epoch": 0.27749576988155666,
+      "grad_norm": 0.5463551243922615,
+      "learning_rate": 8.659039283746738e-06,
+      "loss": 0.5528,
+      "step": 574
+    },
+    {
+      "epoch": 0.27797921198936426,
+      "grad_norm": 0.5125817786426824,
+      "learning_rate": 8.653586696203111e-06,
+      "loss": 0.5428,
+      "step": 575
+    },
+    {
+      "epoch": 0.27846265409717186,
+      "grad_norm": 0.46930359618316736,
+      "learning_rate": 8.648124769797424e-06,
+      "loss": 0.5566,
+      "step": 576
+    },
+    {
+      "epoch": 0.27894609620497945,
+      "grad_norm": 0.48871061545968875,
+      "learning_rate": 8.6426535184908e-06,
+      "loss": 0.5517,
+      "step": 577
+    },
+    {
+      "epoch": 0.27942953831278705,
+      "grad_norm": 0.49517469382405177,
+      "learning_rate": 8.637172956268203e-06,
+      "loss": 0.5537,
+      "step": 578
+    },
+    {
+      "epoch": 0.27991298042059465,
+      "grad_norm": 0.46885534036424203,
+      "learning_rate": 8.631683097138386e-06,
+      "loss": 0.5455,
+      "step": 579
+    },
+    {
+      "epoch": 0.28039642252840224,
+      "grad_norm": 0.4404595141316285,
+      "learning_rate": 8.626183955133876e-06,
+      "loss": 0.5216,
+      "step": 580
+    },
+    {
+      "epoch": 0.2808798646362098,
+      "grad_norm": 0.45805738086780906,
+      "learning_rate": 8.620675544310921e-06,
+      "loss": 0.5483,
+      "step": 581
+    },
+    {
+      "epoch": 0.2813633067440174,
+      "grad_norm": 0.43283074014496,
+      "learning_rate": 8.615157878749462e-06,
+      "loss": 0.546,
+      "step": 582
+    },
+    {
+      "epoch": 0.281846748851825,
+      "grad_norm": 0.45714306807295113,
+      "learning_rate": 8.609630972553098e-06,
+      "loss": 0.5521,
+      "step": 583
+    },
+    {
+      "epoch": 0.2823301909596326,
+      "grad_norm": 0.47311376331001226,
+      "learning_rate": 8.604094839849047e-06,
+      "loss": 0.5586,
+      "step": 584
+    },
+    {
+      "epoch": 0.2828136330674402,
+      "grad_norm": 0.4424955765808361,
+      "learning_rate": 8.598549494788111e-06,
+      "loss": 0.5384,
+      "step": 585
+    },
+    {
+      "epoch": 0.28329707517524777,
+      "grad_norm": 0.467505480407099,
+      "learning_rate": 8.592994951544637e-06,
+      "loss": 0.5368,
+      "step": 586
+    },
+    {
+      "epoch": 0.28378051728305537,
+      "grad_norm": 0.48553503600686004,
+      "learning_rate": 8.587431224316488e-06,
+      "loss": 0.5475,
+      "step": 587
+    },
+    {
+      "epoch": 0.28426395939086296,
+      "grad_norm": 0.46529868946828945,
+      "learning_rate": 8.581858327324996e-06,
+      "loss": 0.5212,
+      "step": 588
+    },
+    {
+      "epoch": 0.28474740149867056,
+      "grad_norm": 0.4898248932325677,
+      "learning_rate": 8.576276274814936e-06,
+      "loss": 0.553,
+      "step": 589
+    },
+    {
+      "epoch": 0.2852308436064781,
+      "grad_norm": 0.4455201034159363,
+      "learning_rate": 8.570685081054487e-06,
+      "loss": 0.5216,
+      "step": 590
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.46516198660507346,
+      "learning_rate": 8.565084760335188e-06,
+      "loss": 0.5505,
+      "step": 591
+    },
+    {
+      "epoch": 0.2861977278220933,
+      "grad_norm": 0.5006509760317717,
+      "learning_rate": 8.559475326971907e-06,
+      "loss": 0.551,
+      "step": 592
+    },
+    {
+      "epoch": 0.2866811699299009,
+      "grad_norm": 0.4994980078510237,
+      "learning_rate": 8.553856795302815e-06,
+      "loss": 0.5421,
+      "step": 593
+    },
+    {
+      "epoch": 0.2871646120377085,
+      "grad_norm": 0.5029150812228765,
+      "learning_rate": 8.548229179689325e-06,
+      "loss": 0.5519,
+      "step": 594
+    },
+    {
+      "epoch": 0.2876480541455161,
+      "grad_norm": 0.5073270485472724,
+      "learning_rate": 8.54259249451608e-06,
+      "loss": 0.5537,
+      "step": 595
+    },
+    {
+      "epoch": 0.2881314962533237,
+      "grad_norm": 0.45709172284548705,
+      "learning_rate": 8.536946754190903e-06,
+      "loss": 0.564,
+      "step": 596
+    },
+    {
+      "epoch": 0.2886149383611313,
+      "grad_norm": 0.47174309410425874,
+      "learning_rate": 8.531291973144755e-06,
+      "loss": 0.5452,
+      "step": 597
+    },
+    {
+      "epoch": 0.2890983804689388,
+      "grad_norm": 0.46639713589843634,
+      "learning_rate": 8.52562816583172e-06,
+      "loss": 0.5509,
+      "step": 598
+    },
+    {
+      "epoch": 0.2895818225767464,
+      "grad_norm": 0.4508036851803557,
+      "learning_rate": 8.519955346728939e-06,
+      "loss": 0.5428,
+      "step": 599
+    },
+    {
+      "epoch": 0.290065264684554,
+      "grad_norm": 0.44468353218524803,
+      "learning_rate": 8.5142735303366e-06,
+      "loss": 0.5205,
+      "step": 600
+    },
+    {
+      "epoch": 0.2905487067923616,
+      "grad_norm": 0.4850164433619974,
+      "learning_rate": 8.50858273117788e-06,
+      "loss": 0.5476,
+      "step": 601
+    },
+    {
+      "epoch": 0.2910321489001692,
+      "grad_norm": 0.49303870805000655,
+      "learning_rate": 8.502882963798923e-06,
+      "loss": 0.545,
+      "step": 602
+    },
+    {
+      "epoch": 0.2915155910079768,
+      "grad_norm": 0.47572858582093197,
+      "learning_rate": 8.497174242768792e-06,
+      "loss": 0.5515,
+      "step": 603
+    },
+    {
+      "epoch": 0.2919990331157844,
+      "grad_norm": 0.5284607359345597,
+      "learning_rate": 8.49145658267944e-06,
+      "loss": 0.5453,
+      "step": 604
+    },
+    {
+      "epoch": 0.292482475223592,
+      "grad_norm": 0.47829654266425203,
+      "learning_rate": 8.485729998145665e-06,
+      "loss": 0.5452,
+      "step": 605
+    },
+    {
+      "epoch": 0.29296591733139954,
+      "grad_norm": 0.4503645291799449,
+      "learning_rate": 8.479994503805079e-06,
+      "loss": 0.5536,
+      "step": 606
+    },
+    {
+      "epoch": 0.29344935943920714,
+      "grad_norm": 0.4693738299713831,
+      "learning_rate": 8.474250114318066e-06,
+      "loss": 0.5216,
+      "step": 607
+    },
+    {
+      "epoch": 0.29393280154701473,
+      "grad_norm": 0.4988674830387375,
+      "learning_rate": 8.468496844367752e-06,
+      "loss": 0.5582,
+      "step": 608
+    },
+    {
+      "epoch": 0.29441624365482233,
+      "grad_norm": 0.47627140431869974,
+      "learning_rate": 8.462734708659959e-06,
+      "loss": 0.5511,
+      "step": 609
+    },
+    {
+      "epoch": 0.2948996857626299,
+      "grad_norm": 0.43233992742433075,
+      "learning_rate": 8.456963721923166e-06,
+      "loss": 0.5279,
+      "step": 610
+    },
+    {
+      "epoch": 0.2953831278704375,
+      "grad_norm": 0.4908070311501362,
+      "learning_rate": 8.451183898908484e-06,
+      "loss": 0.5546,
+      "step": 611
+    },
+    {
+      "epoch": 0.2958665699782451,
+      "grad_norm": 0.4519643584485447,
+      "learning_rate": 8.445395254389605e-06,
+      "loss": 0.5221,
+      "step": 612
+    },
+    {
+      "epoch": 0.2963500120860527,
+      "grad_norm": 0.48396713453490725,
+      "learning_rate": 8.439597803162773e-06,
+      "loss": 0.5489,
+      "step": 613
+    },
+    {
+      "epoch": 0.29683345419386026,
+      "grad_norm": 0.4611763742603572,
+      "learning_rate": 8.433791560046737e-06,
+      "loss": 0.5457,
+      "step": 614
+    },
+    {
+      "epoch": 0.29731689630166785,
+      "grad_norm": 0.472544396347692,
+      "learning_rate": 8.427976539882725e-06,
+      "loss": 0.5553,
+      "step": 615
+    },
+    {
+      "epoch": 0.29780033840947545,
+      "grad_norm": 0.5058827141310254,
+      "learning_rate": 8.422152757534395e-06,
+      "loss": 0.5435,
+      "step": 616
+    },
+    {
+      "epoch": 0.29828378051728305,
+      "grad_norm": 0.4766589825937423,
+      "learning_rate": 8.416320227887805e-06,
+      "loss": 0.5526,
+      "step": 617
+    },
+    {
+      "epoch": 0.29876722262509064,
+      "grad_norm": 0.47223702801719897,
+      "learning_rate": 8.410478965851371e-06,
+      "loss": 0.5542,
+      "step": 618
+    },
+    {
+      "epoch": 0.29925066473289824,
+      "grad_norm": 0.4819039683875086,
+      "learning_rate": 8.404628986355832e-06,
+      "loss": 0.5546,
+      "step": 619
+    },
+    {
+      "epoch": 0.29973410684070584,
+      "grad_norm": 0.49462386708237827,
+      "learning_rate": 8.398770304354203e-06,
+      "loss": 0.5566,
+      "step": 620
+    },
+    {
+      "epoch": 0.30021754894851344,
+      "grad_norm": 0.4977376021667819,
+      "learning_rate": 8.39290293482175e-06,
+      "loss": 0.5508,
+      "step": 621
+    },
+    {
+      "epoch": 0.30070099105632103,
+      "grad_norm": 0.506465713525892,
+      "learning_rate": 8.387026892755942e-06,
+      "loss": 0.5568,
+      "step": 622
+    },
+    {
+      "epoch": 0.3011844331641286,
+      "grad_norm": 0.4712688161265267,
+      "learning_rate": 8.381142193176414e-06,
+      "loss": 0.5489,
+      "step": 623
+    },
+    {
+      "epoch": 0.30166787527193617,
+      "grad_norm": 0.4615215470431895,
+      "learning_rate": 8.375248851124937e-06,
+      "loss": 0.5554,
+      "step": 624
+    },
+    {
+      "epoch": 0.30215131737974377,
+      "grad_norm": 0.5110895222198079,
+      "learning_rate": 8.369346881665364e-06,
+      "loss": 0.5466,
+      "step": 625
+    },
+    {
+      "epoch": 0.30263475948755136,
+      "grad_norm": 0.47157470051165545,
+      "learning_rate": 8.363436299883604e-06,
+      "loss": 0.5644,
+      "step": 626
+    },
+    {
+      "epoch": 0.30311820159535896,
+      "grad_norm": 0.4789841431133952,
+      "learning_rate": 8.357517120887586e-06,
+      "loss": 0.5493,
+      "step": 627
+    },
+    {
+      "epoch": 0.30360164370316656,
+      "grad_norm": 0.47629409809645545,
+      "learning_rate": 8.351589359807204e-06,
+      "loss": 0.5523,
+      "step": 628
+    },
+    {
+      "epoch": 0.30408508581097415,
+      "grad_norm": 0.4618925314784255,
+      "learning_rate": 8.345653031794292e-06,
+      "loss": 0.5348,
+      "step": 629
+    },
+    {
+      "epoch": 0.30456852791878175,
+      "grad_norm": 0.500646417496574,
+      "learning_rate": 8.339708152022586e-06,
+      "loss": 0.554,
+      "step": 630
+    },
+    {
+      "epoch": 0.3050519700265893,
+      "grad_norm": 0.4553680483630167,
+      "learning_rate": 8.333754735687677e-06,
+      "loss": 0.5489,
+      "step": 631
+    },
+    {
+      "epoch": 0.3055354121343969,
+      "grad_norm": 0.4591000336987377,
+      "learning_rate": 8.327792798006977e-06,
+      "loss": 0.5508,
+      "step": 632
+    },
+    {
+      "epoch": 0.3060188542422045,
+      "grad_norm": 0.48696012413599493,
+      "learning_rate": 8.321822354219677e-06,
+      "loss": 0.5505,
+      "step": 633
+    },
+    {
+      "epoch": 0.3065022963500121,
+      "grad_norm": 0.4952413093498077,
+      "learning_rate": 8.315843419586717e-06,
+      "loss": 0.5574,
+      "step": 634
+    },
+    {
+      "epoch": 0.3069857384578197,
+      "grad_norm": 0.4716786308005616,
+      "learning_rate": 8.309856009390732e-06,
+      "loss": 0.5281,
+      "step": 635
+    },
+    {
+      "epoch": 0.3074691805656273,
+      "grad_norm": 0.5207738583309734,
+      "learning_rate": 8.303860138936027e-06,
+      "loss": 0.5607,
+      "step": 636
+    },
+    {
+      "epoch": 0.3079526226734349,
+      "grad_norm": 0.5226978234399785,
+      "learning_rate": 8.297855823548528e-06,
+      "loss": 0.5565,
+      "step": 637
+    },
+    {
+      "epoch": 0.30843606478124247,
+      "grad_norm": 0.49251100209183046,
+      "learning_rate": 8.291843078575752e-06,
+      "loss": 0.5485,
+      "step": 638
+    },
+    {
+      "epoch": 0.30891950688905,
+      "grad_norm": 0.4769824051475033,
+      "learning_rate": 8.285821919386758e-06,
+      "loss": 0.5456,
+      "step": 639
+    },
+    {
+      "epoch": 0.3094029489968576,
+      "grad_norm": 0.503019530780954,
+      "learning_rate": 8.279792361372114e-06,
+      "loss": 0.5602,
+      "step": 640
+    },
+    {
+      "epoch": 0.3098863911046652,
+      "grad_norm": 0.48405162661408385,
+      "learning_rate": 8.273754419943856e-06,
+      "loss": 0.5536,
+      "step": 641
+    },
+    {
+      "epoch": 0.3103698332124728,
+      "grad_norm": 0.4657304337869963,
+      "learning_rate": 8.267708110535449e-06,
+      "loss": 0.5477,
+      "step": 642
+    },
+    {
+      "epoch": 0.3108532753202804,
+      "grad_norm": 0.5106373655355231,
+      "learning_rate": 8.26165344860175e-06,
+      "loss": 0.571,
+      "step": 643
+    },
+    {
+      "epoch": 0.311336717428088,
+      "grad_norm": 0.4854760780132044,
+      "learning_rate": 8.255590449618958e-06,
+      "loss": 0.546,
+      "step": 644
+    },
+    {
+      "epoch": 0.3118201595358956,
+      "grad_norm": 0.4817908473273075,
+      "learning_rate": 8.24951912908459e-06,
+      "loss": 0.5446,
+      "step": 645
+    },
+    {
+      "epoch": 0.3123036016437032,
+      "grad_norm": 0.4792564537130554,
+      "learning_rate": 8.243439502517432e-06,
+      "loss": 0.5352,
+      "step": 646
+    },
+    {
+      "epoch": 0.31278704375151073,
+      "grad_norm": 0.5188490831185355,
+      "learning_rate": 8.237351585457499e-06,
+      "loss": 0.5298,
+      "step": 647
+    },
+    {
+      "epoch": 0.3132704858593183,
+      "grad_norm": 0.5232755336111542,
+      "learning_rate": 8.231255393465993e-06,
+      "loss": 0.5387,
+      "step": 648
+    },
+    {
+      "epoch": 0.3137539279671259,
+      "grad_norm": 0.48933101067554713,
+      "learning_rate": 8.225150942125278e-06,
+      "loss": 0.5156,
+      "step": 649
+    },
+    {
+      "epoch": 0.3142373700749335,
+      "grad_norm": 0.47579138598403903,
+      "learning_rate": 8.21903824703882e-06,
+      "loss": 0.552,
+      "step": 650
+    },
+    {
+      "epoch": 0.3147208121827411,
+      "grad_norm": 0.47742859766681844,
+      "learning_rate": 8.21291732383116e-06,
+      "loss": 0.5498,
+      "step": 651
+    },
+    {
+      "epoch": 0.3152042542905487,
+      "grad_norm": 0.5282098462854927,
+      "learning_rate": 8.206788188147874e-06,
+      "loss": 0.5327,
+      "step": 652
+    },
+    {
+      "epoch": 0.3156876963983563,
+      "grad_norm": 0.4655983753785802,
+      "learning_rate": 8.200650855655525e-06,
+      "loss": 0.5523,
+      "step": 653
+    },
+    {
+      "epoch": 0.3161711385061639,
+      "grad_norm": 0.46598993965098007,
+      "learning_rate": 8.19450534204163e-06,
+      "loss": 0.5428,
+      "step": 654
+    },
+    {
+      "epoch": 0.3166545806139715,
+      "grad_norm": 0.48320040727215685,
+      "learning_rate": 8.188351663014615e-06,
+      "loss": 0.5511,
+      "step": 655
+    },
+    {
+      "epoch": 0.31713802272177904,
+      "grad_norm": 0.4851268795547935,
+      "learning_rate": 8.182189834303783e-06,
+      "loss": 0.5515,
+      "step": 656
+    },
+    {
+      "epoch": 0.31762146482958664,
+      "grad_norm": 0.4829311813743368,
+      "learning_rate": 8.176019871659263e-06,
+      "loss": 0.5425,
+      "step": 657
+    },
+    {
+      "epoch": 0.31810490693739424,
+      "grad_norm": 0.4268110510337058,
+      "learning_rate": 8.169841790851976e-06,
+      "loss": 0.5192,
+      "step": 658
+    },
+    {
+      "epoch": 0.31858834904520184,
+      "grad_norm": 0.46970357309915234,
+      "learning_rate": 8.163655607673594e-06,
+      "loss": 0.5516,
+      "step": 659
+    },
+    {
+      "epoch": 0.31907179115300943,
+      "grad_norm": 0.4688205789040297,
+      "learning_rate": 8.157461337936506e-06,
+      "loss": 0.5398,
+      "step": 660
+    },
+    {
+      "epoch": 0.31955523326081703,
+      "grad_norm": 0.49966496418563966,
+      "learning_rate": 8.151258997473757e-06,
+      "loss": 0.5501,
+      "step": 661
+    },
+    {
+      "epoch": 0.3200386753686246,
+      "grad_norm": 0.4715831395525512,
+      "learning_rate": 8.145048602139031e-06,
+      "loss": 0.5473,
+      "step": 662
+    },
+    {
+      "epoch": 0.3205221174764322,
+      "grad_norm": 0.46025918766438206,
+      "learning_rate": 8.138830167806601e-06,
+      "loss": 0.5481,
+      "step": 663
+    },
+    {
+      "epoch": 0.32100555958423976,
+      "grad_norm": 0.49250849769551697,
+      "learning_rate": 8.132603710371287e-06,
+      "loss": 0.5563,
+      "step": 664
+    },
+    {
+      "epoch": 0.32148900169204736,
+      "grad_norm": 0.46277397720994495,
+      "learning_rate": 8.126369245748413e-06,
+      "loss": 0.5418,
+      "step": 665
+    },
+    {
+      "epoch": 0.32197244379985496,
+      "grad_norm": 0.44842320811529324,
+      "learning_rate": 8.120126789873775e-06,
+      "loss": 0.549,
+      "step": 666
+    },
+    {
+      "epoch": 0.32245588590766255,
+      "grad_norm": 0.4487718178782243,
+      "learning_rate": 8.113876358703593e-06,
+      "loss": 0.5515,
+      "step": 667
+    },
+    {
+      "epoch": 0.32293932801547015,
+      "grad_norm": 0.49737040438900676,
+      "learning_rate": 8.10761796821447e-06,
+      "loss": 0.5529,
+      "step": 668
+    },
+    {
+      "epoch": 0.32342277012327775,
+      "grad_norm": 0.5088088437400782,
+      "learning_rate": 8.10135163440336e-06,
+      "loss": 0.5507,
+      "step": 669
+    },
+    {
+      "epoch": 0.32390621223108534,
+      "grad_norm": 0.5221100660415426,
+      "learning_rate": 8.095077373287517e-06,
+      "loss": 0.5363,
+      "step": 670
+    },
+    {
+      "epoch": 0.32438965433889294,
+      "grad_norm": 0.5098038198929602,
+      "learning_rate": 8.088795200904457e-06,
+      "loss": 0.5443,
+      "step": 671
+    },
+    {
+      "epoch": 0.3248730964467005,
+      "grad_norm": 0.5299548080054053,
+      "learning_rate": 8.08250513331192e-06,
+      "loss": 0.5547,
+      "step": 672
+    },
+    {
+      "epoch": 0.3253565385545081,
+      "grad_norm": 0.47991648628747413,
+      "learning_rate": 8.076207186587826e-06,
+      "loss": 0.552,
+      "step": 673
+    },
+    {
+      "epoch": 0.3258399806623157,
+      "grad_norm": 0.4928995313967277,
+      "learning_rate": 8.069901376830232e-06,
+      "loss": 0.5449,
+      "step": 674
+    },
+    {
+      "epoch": 0.3263234227701233,
+      "grad_norm": 0.526245201002504,
+      "learning_rate": 8.063587720157298e-06,
+      "loss": 0.5544,
+      "step": 675
+    },
+    {
+      "epoch": 0.32680686487793087,
+      "grad_norm": 0.5169185895561939,
+      "learning_rate": 8.057266232707239e-06,
+      "loss": 0.5388,
+      "step": 676
+    },
+    {
+      "epoch": 0.32729030698573847,
+      "grad_norm": 0.45862190884382065,
+      "learning_rate": 8.050936930638285e-06,
+      "loss": 0.5523,
+      "step": 677
+    },
+    {
+      "epoch": 0.32777374909354606,
+      "grad_norm": 0.4791194354627634,
+      "learning_rate": 8.044599830128643e-06,
+      "loss": 0.5498,
+      "step": 678
+    },
+    {
+      "epoch": 0.32825719120135366,
+      "grad_norm": 0.5040011739287719,
+      "learning_rate": 8.038254947376454e-06,
+      "loss": 0.5378,
+      "step": 679
+    },
+    {
+      "epoch": 0.3287406333091612,
+      "grad_norm": 0.42346684737245893,
+      "learning_rate": 8.03190229859975e-06,
+      "loss": 0.5541,
+      "step": 680
+    },
+    {
+      "epoch": 0.3292240754169688,
+      "grad_norm": 0.48225697444636256,
+      "learning_rate": 8.02554190003641e-06,
+      "loss": 0.5505,
+      "step": 681
+    },
+    {
+      "epoch": 0.3297075175247764,
+      "grad_norm": 0.480200233217211,
+      "learning_rate": 8.019173767944128e-06,
+      "loss": 0.5563,
+      "step": 682
+    },
+    {
+      "epoch": 0.330190959632584,
+      "grad_norm": 0.4406037883552503,
+      "learning_rate": 8.012797918600363e-06,
+      "loss": 0.5241,
+      "step": 683
+    },
+    {
+      "epoch": 0.3306744017403916,
+      "grad_norm": 0.4838913486529156,
+      "learning_rate": 8.006414368302297e-06,
+      "loss": 0.5251,
+      "step": 684
+    },
+    {
+      "epoch": 0.3311578438481992,
+      "grad_norm": 0.45454190895682295,
+      "learning_rate": 8.000023133366804e-06,
+      "loss": 0.5449,
+      "step": 685
+    },
+    {
+      "epoch": 0.3316412859560068,
+      "grad_norm": 0.49869890620532237,
+      "learning_rate": 7.99362423013039e-06,
+      "loss": 0.5401,
+      "step": 686
+    },
+    {
+      "epoch": 0.3321247280638144,
+      "grad_norm": 0.4727231220514769,
+      "learning_rate": 7.98721767494917e-06,
+      "loss": 0.5381,
+      "step": 687
+    },
+    {
+      "epoch": 0.332608170171622,
+      "grad_norm": 0.46944667758244535,
+      "learning_rate": 7.980803484198817e-06,
+      "loss": 0.5542,
+      "step": 688
+    },
+    {
+      "epoch": 0.3330916122794295,
+      "grad_norm": 0.4643616722232514,
+      "learning_rate": 7.974381674274517e-06,
+      "loss": 0.5394,
+      "step": 689
+    },
+    {
+      "epoch": 0.3335750543872371,
+      "grad_norm": 0.4529493856728362,
+      "learning_rate": 7.967952261590936e-06,
+      "loss": 0.5478,
+      "step": 690
+    },
+    {
+      "epoch": 0.3340584964950447,
+      "grad_norm": 0.4497900124215144,
+      "learning_rate": 7.961515262582168e-06,
+      "loss": 0.5387,
+      "step": 691
+    },
+    {
+      "epoch": 0.3345419386028523,
+      "grad_norm": 0.4613195703294155,
+      "learning_rate": 7.955070693701704e-06,
+      "loss": 0.5488,
+      "step": 692
+    },
+    {
+      "epoch": 0.3350253807106599,
+      "grad_norm": 0.45208853687907335,
+      "learning_rate": 7.94861857142238e-06,
+      "loss": 0.5161,
+      "step": 693
+    },
+    {
+      "epoch": 0.3355088228184675,
+      "grad_norm": 0.45338462953665065,
+      "learning_rate": 7.942158912236339e-06,
+      "loss": 0.5504,
+      "step": 694
+    },
+    {
+      "epoch": 0.3359922649262751,
+      "grad_norm": 0.45784135957705213,
+      "learning_rate": 7.935691732654995e-06,
+      "loss": 0.5525,
+      "step": 695
+    },
+    {
+      "epoch": 0.3364757070340827,
+      "grad_norm": 0.4745455134248678,
+      "learning_rate": 7.929217049208977e-06,
+      "loss": 0.5549,
+      "step": 696
+    },
+    {
+      "epoch": 0.33695914914189024,
+      "grad_norm": 0.46788843343497605,
+      "learning_rate": 7.922734878448099e-06,
+      "loss": 0.5543,
+      "step": 697
+    },
+    {
+      "epoch": 0.33744259124969783,
+      "grad_norm": 0.4894111106267614,
+      "learning_rate": 7.916245236941311e-06,
+      "loss": 0.5456,
+      "step": 698
+    },
+    {
+      "epoch": 0.33792603335750543,
+      "grad_norm": 0.4818527781927651,
+      "learning_rate": 7.90974814127666e-06,
+      "loss": 0.5436,
+      "step": 699
+    },
+    {
+      "epoch": 0.338409475465313,
+      "grad_norm": 0.48230512049955104,
+      "learning_rate": 7.903243608061246e-06,
+      "loss": 0.5569,
+      "step": 700
+    },
+    {
+      "epoch": 0.3388929175731206,
+      "grad_norm": 0.4651013778967097,
+      "learning_rate": 7.89673165392118e-06,
+      "loss": 0.5497,
+      "step": 701
+    },
+    {
+      "epoch": 0.3393763596809282,
+      "grad_norm": 0.5263037891579944,
+      "learning_rate": 7.890212295501542e-06,
+      "loss": 0.5489,
+      "step": 702
+    },
+    {
+      "epoch": 0.3398598017887358,
+      "grad_norm": 0.47525750483933155,
+      "learning_rate": 7.883685549466337e-06,
+      "loss": 0.5438,
+      "step": 703
+    },
+    {
+      "epoch": 0.3403432438965434,
+      "grad_norm": 0.48435256135519467,
+      "learning_rate": 7.877151432498456e-06,
+      "loss": 0.5506,
+      "step": 704
+    },
+    {
+      "epoch": 0.34082668600435095,
+      "grad_norm": 0.49040296450298604,
+      "learning_rate": 7.870609961299627e-06,
+      "loss": 0.536,
+      "step": 705
+    },
+    {
+      "epoch": 0.34131012811215855,
+      "grad_norm": 0.4437135993163076,
+      "learning_rate": 7.864061152590376e-06,
+      "loss": 0.5539,
+      "step": 706
+    },
+    {
+      "epoch": 0.34179357021996615,
+      "grad_norm": 0.48585487486606105,
+      "learning_rate": 7.857505023109989e-06,
+      "loss": 0.5461,
+      "step": 707
+    },
+    {
+      "epoch": 0.34227701232777374,
+      "grad_norm": 0.48202275018795376,
+      "learning_rate": 7.850941589616458e-06,
+      "loss": 0.5371,
+      "step": 708
+    },
+    {
+      "epoch": 0.34276045443558134,
+      "grad_norm": 0.4716406712767161,
+      "learning_rate": 7.844370868886452e-06,
+      "loss": 0.5557,
+      "step": 709
+    },
+    {
+      "epoch": 0.34324389654338894,
+      "grad_norm": 0.49083867550017374,
+      "learning_rate": 7.83779287771526e-06,
+      "loss": 0.5459,
+      "step": 710
+    },
+    {
+      "epoch": 0.34372733865119653,
+      "grad_norm": 0.49576681886311147,
+      "learning_rate": 7.831207632916757e-06,
+      "loss": 0.5466,
+      "step": 711
+    },
+    {
+      "epoch": 0.34421078075900413,
+      "grad_norm": 0.4533195955521626,
+      "learning_rate": 7.824615151323363e-06,
+      "loss": 0.519,
+      "step": 712
+    },
+    {
+      "epoch": 0.3446942228668117,
+      "grad_norm": 0.4638295505748454,
+      "learning_rate": 7.818015449785987e-06,
+      "loss": 0.5485,
+      "step": 713
+    },
+    {
+      "epoch": 0.34517766497461927,
+      "grad_norm": 0.4802273717901249,
+      "learning_rate": 7.811408545174001e-06,
+      "loss": 0.5453,
+      "step": 714
+    },
+    {
+      "epoch": 0.34566110708242687,
+      "grad_norm": 0.4633815927205105,
+      "learning_rate": 7.804794454375189e-06,
+      "loss": 0.5504,
+      "step": 715
+    },
+    {
+      "epoch": 0.34614454919023446,
+      "grad_norm": 0.4455507327132057,
+      "learning_rate": 7.798173194295693e-06,
+      "loss": 0.5425,
+      "step": 716
+    },
+    {
+      "epoch": 0.34662799129804206,
+      "grad_norm": 0.4555461305882042,
+      "learning_rate": 7.791544781859993e-06,
+      "loss": 0.5402,
+      "step": 717
+    },
+    {
+      "epoch": 0.34711143340584966,
+      "grad_norm": 0.4459519768062681,
+      "learning_rate": 7.784909234010843e-06,
+      "loss": 0.5448,
+      "step": 718
+    },
+    {
+      "epoch": 0.34759487551365725,
+      "grad_norm": 0.43036102684437805,
+      "learning_rate": 7.778266567709239e-06,
+      "loss": 0.5532,
+      "step": 719
+    },
+    {
+      "epoch": 0.34807831762146485,
+      "grad_norm": 0.4640780423848208,
+      "learning_rate": 7.771616799934372e-06,
+      "loss": 0.5403,
+      "step": 720
+    },
+    {
+      "epoch": 0.34856175972927245,
+      "grad_norm": 0.4785048230046999,
+      "learning_rate": 7.764959947683581e-06,
+      "loss": 0.5484,
+      "step": 721
+    },
+    {
+      "epoch": 0.34904520183708,
+      "grad_norm": 0.5054523460781126,
+      "learning_rate": 7.758296027972324e-06,
+      "loss": 0.5367,
+      "step": 722
+    },
+    {
+      "epoch": 0.3495286439448876,
+      "grad_norm": 0.46292028395178175,
+      "learning_rate": 7.751625057834107e-06,
+      "loss": 0.5221,
+      "step": 723
+    },
+    {
+      "epoch": 0.3500120860526952,
+      "grad_norm": 0.5080355944834025,
+      "learning_rate": 7.744947054320475e-06,
+      "loss": 0.552,
+      "step": 724
+    },
+    {
+      "epoch": 0.3504955281605028,
+      "grad_norm": 0.4692605361826857,
+      "learning_rate": 7.73826203450094e-06,
+      "loss": 0.5516,
+      "step": 725
+    },
+    {
+      "epoch": 0.3509789702683104,
+      "grad_norm": 0.4452014990295793,
+      "learning_rate": 7.731570015462953e-06,
+      "loss": 0.5385,
+      "step": 726
+    },
+    {
+      "epoch": 0.35146241237611797,
+      "grad_norm": 0.4407238797603078,
+      "learning_rate": 7.724871014311853e-06,
+      "loss": 0.5512,
+      "step": 727
+    },
+    {
+      "epoch": 0.35194585448392557,
+      "grad_norm": 0.4578450461185458,
+      "learning_rate": 7.718165048170827e-06,
+      "loss": 0.5436,
+      "step": 728
+    },
+    {
+      "epoch": 0.35242929659173317,
+      "grad_norm": 0.47134089664050416,
+      "learning_rate": 7.711452134180865e-06,
+      "loss": 0.5439,
+      "step": 729
+    },
+    {
+      "epoch": 0.3529127386995407,
+      "grad_norm": 0.45807802586279717,
+      "learning_rate": 7.704732289500717e-06,
+      "loss": 0.535,
+      "step": 730
+    },
+    {
+      "epoch": 0.3533961808073483,
+      "grad_norm": 0.46923078006226726,
+      "learning_rate": 7.698005531306844e-06,
+      "loss": 0.5438,
+      "step": 731
+    },
+    {
+      "epoch": 0.3538796229151559,
+      "grad_norm": 0.47163216368627525,
+      "learning_rate": 7.691271876793387e-06,
+      "loss": 0.5412,
+      "step": 732
+    },
+    {
+      "epoch": 0.3543630650229635,
+      "grad_norm": 0.43982060964801745,
+      "learning_rate": 7.684531343172108e-06,
+      "loss": 0.5326,
+      "step": 733
+    },
+    {
+      "epoch": 0.3548465071307711,
+      "grad_norm": 0.49898422329655673,
+      "learning_rate": 7.677783947672352e-06,
+      "loss": 0.5352,
+      "step": 734
+    },
+    {
+      "epoch": 0.3553299492385787,
+      "grad_norm": 0.4426035951569431,
+      "learning_rate": 7.67102970754101e-06,
+      "loss": 0.5083,
+      "step": 735
+    },
+    {
+      "epoch": 0.3558133913463863,
+      "grad_norm": 0.473881609856312,
+      "learning_rate": 7.664268640042459e-06,
+      "loss": 0.5493,
+      "step": 736
+    },
+    {
+      "epoch": 0.3562968334541939,
+      "grad_norm": 0.48477702288906854,
+      "learning_rate": 7.657500762458536e-06,
+      "loss": 0.5415,
+      "step": 737
+    },
+    {
+      "epoch": 0.3567802755620014,
+      "grad_norm": 0.4673684560489235,
+      "learning_rate": 7.65072609208848e-06,
+      "loss": 0.5402,
+      "step": 738
+    },
+    {
+      "epoch": 0.357263717669809,
+      "grad_norm": 0.45922228645390506,
+      "learning_rate": 7.643944646248898e-06,
+      "loss": 0.5523,
+      "step": 739
+    },
+    {
+      "epoch": 0.3577471597776166,
+      "grad_norm": 0.5023203702238386,
+      "learning_rate": 7.637156442273705e-06,
+      "loss": 0.5472,
+      "step": 740
+    },
+    {
+      "epoch": 0.3582306018854242,
+      "grad_norm": 0.5158066743775931,
+      "learning_rate": 7.630361497514104e-06,
+      "loss": 0.5409,
+      "step": 741
+    },
+    {
+      "epoch": 0.3587140439932318,
+      "grad_norm": 0.4279389055361383,
+      "learning_rate": 7.6235598293385184e-06,
+      "loss": 0.548,
+      "step": 742
+    },
+    {
+      "epoch": 0.3591974861010394,
+      "grad_norm": 0.48124896416843527,
+      "learning_rate": 7.616751455132561e-06,
+      "loss": 0.5061,
+      "step": 743
+    },
+    {
+      "epoch": 0.359680928208847,
+      "grad_norm": 0.45130820986839676,
+      "learning_rate": 7.6099363922989845e-06,
+      "loss": 0.5408,
+      "step": 744
+    },
+    {
+      "epoch": 0.3601643703166546,
+      "grad_norm": 0.43968032097493187,
+      "learning_rate": 7.60311465825764e-06,
+      "loss": 0.5419,
+      "step": 745
+    },
+    {
+      "epoch": 0.36064781242446214,
+      "grad_norm": 0.4638052394642039,
+      "learning_rate": 7.596286270445429e-06,
+      "loss": 0.5474,
+      "step": 746
+    },
+    {
+      "epoch": 0.36113125453226974,
+      "grad_norm": 0.47215533812036253,
+      "learning_rate": 7.5894512463162595e-06,
+      "loss": 0.5481,
+      "step": 747
+    },
+    {
+      "epoch": 0.36161469664007734,
+      "grad_norm": 0.4910077404120728,
+      "learning_rate": 7.5826096033410056e-06,
+      "loss": 0.5483,
+      "step": 748
+    },
+    {
+      "epoch": 0.36209813874788493,
+      "grad_norm": 0.49435519282302404,
+      "learning_rate": 7.575761359007459e-06,
+      "loss": 0.5375,
+      "step": 749
+    },
+    {
+      "epoch": 0.36258158085569253,
+      "grad_norm": 0.4645080520487796,
+      "learning_rate": 7.568906530820281e-06,
+      "loss": 0.5406,
+      "step": 750
+    },
+    {
+      "epoch": 0.36306502296350013,
+      "grad_norm": 0.48056836362105476,
+      "learning_rate": 7.562045136300969e-06,
+      "loss": 0.547,
+      "step": 751
+    },
+    {
+      "epoch": 0.3635484650713077,
+      "grad_norm": 0.4648855493783244,
+      "learning_rate": 7.555177192987797e-06,
+      "loss": 0.5372,
+      "step": 752
+    },
+    {
+      "epoch": 0.3640319071791153,
+      "grad_norm": 0.43851001194612105,
+      "learning_rate": 7.5483027184357825e-06,
+      "loss": 0.5484,
+      "step": 753
+    },
+    {
+      "epoch": 0.3645153492869229,
+      "grad_norm": 0.4234049796935857,
+      "learning_rate": 7.541421730216638e-06,
+      "loss": 0.4914,
+      "step": 754
+    },
+    {
+      "epoch": 0.36499879139473046,
+      "grad_norm": 0.4886945785128111,
+      "learning_rate": 7.534534245918723e-06,
+      "loss": 0.5362,
+      "step": 755
+    },
+    {
+      "epoch": 0.36548223350253806,
+      "grad_norm": 0.47490196043064764,
+      "learning_rate": 7.527640283147003e-06,
+      "loss": 0.5387,
+      "step": 756
+    },
+    {
+      "epoch": 0.36596567561034565,
+      "grad_norm": 0.4422931000418374,
+      "learning_rate": 7.520739859523001e-06,
+      "loss": 0.5334,
+      "step": 757
+    },
+    {
+      "epoch": 0.36644911771815325,
+      "grad_norm": 0.45103949345201827,
+      "learning_rate": 7.513832992684758e-06,
+      "loss": 0.5423,
+      "step": 758
+    },
+    {
+      "epoch": 0.36693255982596085,
+      "grad_norm": 0.5023542886793314,
+      "learning_rate": 7.50691970028678e-06,
+      "loss": 0.5371,
+      "step": 759
+    },
+    {
+      "epoch": 0.36741600193376844,
+      "grad_norm": 0.48961990962706975,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.5602,
+      "step": 760
+    },
+    {
+      "epoch": 0.36789944404157604,
+      "grad_norm": 0.47244479352550756,
+      "learning_rate": 7.493073909511732e-06,
+      "loss": 0.5399,
+      "step": 761
+    },
+    {
+      "epoch": 0.36838288614938364,
+      "grad_norm": 0.47262373426445514,
+      "learning_rate": 7.486141446525619e-06,
+      "loss": 0.5465,
+      "step": 762
+    },
+    {
+      "epoch": 0.3688663282571912,
+      "grad_norm": 0.47837508137309714,
+      "learning_rate": 7.479202628761597e-06,
+      "loss": 0.5412,
+      "step": 763
+    },
+    {
+      "epoch": 0.3693497703649988,
+      "grad_norm": 0.47771254343171743,
+      "learning_rate": 7.472257473955841e-06,
+      "loss": 0.5429,
+      "step": 764
+    },
+    {
+      "epoch": 0.36983321247280637,
+      "grad_norm": 0.45892920422210776,
+      "learning_rate": 7.465305999860728e-06,
+      "loss": 0.5358,
+      "step": 765
+    },
+    {
+      "epoch": 0.37031665458061397,
+      "grad_norm": 0.4636269912834914,
+      "learning_rate": 7.4583482242447856e-06,
+      "loss": 0.528,
+      "step": 766
+    },
+    {
+      "epoch": 0.37080009668842157,
+      "grad_norm": 0.45196732778688614,
+      "learning_rate": 7.45138416489265e-06,
+      "loss": 0.5466,
+      "step": 767
+    },
+    {
+      "epoch": 0.37128353879622916,
+      "grad_norm": 0.44240214579051484,
+      "learning_rate": 7.444413839605017e-06,
+      "loss": 0.5315,
+      "step": 768
+    },
+    {
+      "epoch": 0.37176698090403676,
+      "grad_norm": 0.45295775865600874,
+      "learning_rate": 7.437437266198602e-06,
+      "loss": 0.5443,
+      "step": 769
+    },
+    {
+      "epoch": 0.37225042301184436,
+      "grad_norm": 0.44966146652009026,
+      "learning_rate": 7.430454462506085e-06,
+      "loss": 0.5417,
+      "step": 770
+    },
+    {
+      "epoch": 0.3727338651196519,
+      "grad_norm": 0.451144144721217,
+      "learning_rate": 7.423465446376079e-06,
+      "loss": 0.5389,
+      "step": 771
+    },
+    {
+      "epoch": 0.3732173072274595,
+      "grad_norm": 0.45620688667035586,
+      "learning_rate": 7.416470235673069e-06,
+      "loss": 0.538,
+      "step": 772
+    },
+    {
+      "epoch": 0.3737007493352671,
+      "grad_norm": 0.44604596469243557,
+      "learning_rate": 7.40946884827738e-06,
+      "loss": 0.5293,
+      "step": 773
+    },
+    {
+      "epoch": 0.3741841914430747,
+      "grad_norm": 0.4476638425696451,
+      "learning_rate": 7.402461302085121e-06,
+      "loss": 0.5402,
+      "step": 774
+    },
+    {
+      "epoch": 0.3746676335508823,
+      "grad_norm": 0.46401364076227924,
+      "learning_rate": 7.395447615008147e-06,
+      "loss": 0.5377,
+      "step": 775
+    },
+    {
+      "epoch": 0.3751510756586899,
+      "grad_norm": 0.46250351228418424,
+      "learning_rate": 7.388427804974003e-06,
+      "loss": 0.5455,
+      "step": 776
+    },
+    {
+      "epoch": 0.3756345177664975,
+      "grad_norm": 0.448294668881327,
+      "learning_rate": 7.381401889925894e-06,
+      "loss": 0.5311,
+      "step": 777
+    },
+    {
+      "epoch": 0.3761179598743051,
+      "grad_norm": 0.4490599419042619,
+      "learning_rate": 7.374369887822623e-06,
+      "loss": 0.5416,
+      "step": 778
+    },
+    {
+      "epoch": 0.3766014019821126,
+      "grad_norm": 0.44904272890455516,
+      "learning_rate": 7.367331816638554e-06,
+      "loss": 0.5464,
+      "step": 779
+    },
+    {
+      "epoch": 0.3770848440899202,
+      "grad_norm": 0.4731428544902919,
+      "learning_rate": 7.360287694363566e-06,
+      "loss": 0.5415,
+      "step": 780
+    },
+    {
+      "epoch": 0.3775682861977278,
+      "grad_norm": 0.4760085542862297,
+      "learning_rate": 7.353237539002999e-06,
+      "loss": 0.5388,
+      "step": 781
+    },
+    {
+      "epoch": 0.3780517283055354,
+      "grad_norm": 0.4817431722421546,
+      "learning_rate": 7.346181368577624e-06,
+      "loss": 0.5513,
+      "step": 782
+    },
+    {
+      "epoch": 0.378535170413343,
+      "grad_norm": 0.462287277146555,
+      "learning_rate": 7.3391192011235764e-06,
+      "loss": 0.5393,
+      "step": 783
+    },
+    {
+      "epoch": 0.3790186125211506,
+      "grad_norm": 0.44812435609118556,
+      "learning_rate": 7.3320510546923285e-06,
+      "loss": 0.5509,
+      "step": 784
+    },
+    {
+      "epoch": 0.3795020546289582,
+      "grad_norm": 0.48025260306275075,
+      "learning_rate": 7.324976947350631e-06,
+      "loss": 0.5387,
+      "step": 785
+    },
+    {
+      "epoch": 0.3799854967367658,
+      "grad_norm": 0.4639545538957294,
+      "learning_rate": 7.317896897180472e-06,
+      "loss": 0.5298,
+      "step": 786
+    },
+    {
+      "epoch": 0.3804689388445734,
+      "grad_norm": 0.46088677266135386,
+      "learning_rate": 7.31081092227903e-06,
+      "loss": 0.5371,
+      "step": 787
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 0.4446814202000039,
+      "learning_rate": 7.303719040758631e-06,
+      "loss": 0.5368,
+      "step": 788
+    },
+    {
+      "epoch": 0.38143582306018853,
+      "grad_norm": 0.4628164716114684,
+      "learning_rate": 7.296621270746691e-06,
+      "loss": 0.5439,
+      "step": 789
+    },
+    {
+      "epoch": 0.3819192651679961,
+      "grad_norm": 0.46916898249294825,
+      "learning_rate": 7.289517630385687e-06,
+      "loss": 0.5188,
+      "step": 790
+    },
+    {
+      "epoch": 0.3824027072758037,
+      "grad_norm": 0.44187034218765375,
+      "learning_rate": 7.282408137833093e-06,
+      "loss": 0.5404,
+      "step": 791
+    },
+    {
+      "epoch": 0.3828861493836113,
+      "grad_norm": 0.47877844839355055,
+      "learning_rate": 7.275292811261346e-06,
+      "loss": 0.5377,
+      "step": 792
+    },
+    {
+      "epoch": 0.3833695914914189,
+      "grad_norm": 0.46034254680546527,
+      "learning_rate": 7.268171668857794e-06,
+      "loss": 0.5489,
+      "step": 793
+    },
+    {
+      "epoch": 0.3838530335992265,
+      "grad_norm": 0.4557903207271804,
+      "learning_rate": 7.261044728824652e-06,
+      "loss": 0.5415,
+      "step": 794
+    },
+    {
+      "epoch": 0.3843364757070341,
+      "grad_norm": 0.477573699663043,
+      "learning_rate": 7.253912009378953e-06,
+      "loss": 0.5526,
+      "step": 795
+    },
+    {
+      "epoch": 0.38481991781484165,
+      "grad_norm": 0.5200587123977845,
+      "learning_rate": 7.246773528752501e-06,
+      "loss": 0.5452,
+      "step": 796
+    },
+    {
+      "epoch": 0.38530335992264925,
+      "grad_norm": 0.4504279257869106,
+      "learning_rate": 7.239629305191828e-06,
+      "loss": 0.528,
+      "step": 797
+    },
+    {
+      "epoch": 0.38578680203045684,
+      "grad_norm": 0.47332344951633437,
+      "learning_rate": 7.2324793569581474e-06,
+      "loss": 0.5413,
+      "step": 798
+    },
+    {
+      "epoch": 0.38627024413826444,
+      "grad_norm": 0.4658360771399747,
+      "learning_rate": 7.2253237023273e-06,
+      "loss": 0.5111,
+      "step": 799
+    },
+    {
+      "epoch": 0.38675368624607204,
+      "grad_norm": 0.4820288472202763,
+      "learning_rate": 7.21816235958972e-06,
+      "loss": 0.5472,
+      "step": 800
+    },
+    {
+      "epoch": 0.38723712835387963,
+      "grad_norm": 0.4713168204495061,
+      "learning_rate": 7.210995347050372e-06,
+      "loss": 0.5441,
+      "step": 801
+    },
+    {
+      "epoch": 0.38772057046168723,
+      "grad_norm": 0.45204543208032005,
+      "learning_rate": 7.203822683028721e-06,
+      "loss": 0.5403,
+      "step": 802
+    },
+    {
+      "epoch": 0.3882040125694948,
+      "grad_norm": 0.4577372851640406,
+      "learning_rate": 7.196644385858673e-06,
+      "loss": 0.5303,
+      "step": 803
+    },
+    {
+      "epoch": 0.38868745467730237,
+      "grad_norm": 0.4458565390228467,
+      "learning_rate": 7.189460473888535e-06,
+      "loss": 0.5453,
+      "step": 804
+    },
+    {
+      "epoch": 0.38917089678510997,
+      "grad_norm": 0.4735785239686837,
+      "learning_rate": 7.182270965480963e-06,
+      "loss": 0.5491,
+      "step": 805
+    },
+    {
+      "epoch": 0.38965433889291756,
+      "grad_norm": 0.44995709460533084,
+      "learning_rate": 7.17507587901292e-06,
+      "loss": 0.5328,
+      "step": 806
+    },
+    {
+      "epoch": 0.39013778100072516,
+      "grad_norm": 0.45714294309440695,
+      "learning_rate": 7.167875232875632e-06,
+      "loss": 0.5401,
+      "step": 807
+    },
+    {
+      "epoch": 0.39062122310853276,
+      "grad_norm": 0.45625350898261685,
+      "learning_rate": 7.160669045474524e-06,
+      "loss": 0.5198,
+      "step": 808
+    },
+    {
+      "epoch": 0.39110466521634035,
+      "grad_norm": 0.45109568818047574,
+      "learning_rate": 7.153457335229196e-06,
+      "loss": 0.5396,
+      "step": 809
+    },
+    {
+      "epoch": 0.39158810732414795,
+      "grad_norm": 0.4362283889114229,
+      "learning_rate": 7.146240120573358e-06,
+      "loss": 0.5421,
+      "step": 810
+    },
+    {
+      "epoch": 0.39207154943195555,
+      "grad_norm": 0.44454460051068084,
+      "learning_rate": 7.1390174199547945e-06,
+      "loss": 0.5495,
+      "step": 811
+    },
+    {
+      "epoch": 0.3925549915397631,
+      "grad_norm": 0.47028424443890243,
+      "learning_rate": 7.131789251835309e-06,
+      "loss": 0.5528,
+      "step": 812
+    },
+    {
+      "epoch": 0.3930384336475707,
+      "grad_norm": 0.44600311061331627,
+      "learning_rate": 7.124555634690684e-06,
+      "loss": 0.546,
+      "step": 813
+    },
+    {
+      "epoch": 0.3935218757553783,
+      "grad_norm": 0.42127806705345067,
+      "learning_rate": 7.117316587010625e-06,
+      "loss": 0.5164,
+      "step": 814
+    },
+    {
+      "epoch": 0.3940053178631859,
+      "grad_norm": 0.4919804442725305,
+      "learning_rate": 7.110072127298722e-06,
+      "loss": 0.5405,
+      "step": 815
+    },
+    {
+      "epoch": 0.3944887599709935,
+      "grad_norm": 0.433673881020081,
+      "learning_rate": 7.1028222740724e-06,
+      "loss": 0.5474,
+      "step": 816
+    },
+    {
+      "epoch": 0.39497220207880107,
+      "grad_norm": 0.44154650103792475,
+      "learning_rate": 7.095567045862867e-06,
+      "loss": 0.537,
+      "step": 817
+    },
+    {
+      "epoch": 0.39545564418660867,
+      "grad_norm": 0.4862239172154185,
+      "learning_rate": 7.0883064612150684e-06,
+      "loss": 0.5418,
+      "step": 818
+    },
+    {
+      "epoch": 0.39593908629441626,
+      "grad_norm": 0.44482243313717656,
+      "learning_rate": 7.081040538687649e-06,
+      "loss": 0.5421,
+      "step": 819
+    },
+    {
+      "epoch": 0.3964225284022238,
+      "grad_norm": 0.4674091065145529,
+      "learning_rate": 7.073769296852888e-06,
+      "loss": 0.5322,
+      "step": 820
+    },
+    {
+      "epoch": 0.3969059705100314,
+      "grad_norm": 0.42008927529357837,
+      "learning_rate": 7.066492754296668e-06,
+      "loss": 0.5021,
+      "step": 821
+    },
+    {
+      "epoch": 0.397389412617839,
+      "grad_norm": 0.4604995654095531,
+      "learning_rate": 7.059210929618416e-06,
+      "loss": 0.5449,
+      "step": 822
+    },
+    {
+      "epoch": 0.3978728547256466,
+      "grad_norm": 0.45094698267399413,
+      "learning_rate": 7.051923841431063e-06,
+      "loss": 0.5376,
+      "step": 823
+    },
+    {
+      "epoch": 0.3983562968334542,
+      "grad_norm": 0.46562198473083916,
+      "learning_rate": 7.044631508360996e-06,
+      "loss": 0.5449,
+      "step": 824
+    },
+    {
+      "epoch": 0.3988397389412618,
+      "grad_norm": 0.45370053407350325,
+      "learning_rate": 7.037333949048005e-06,
+      "loss": 0.5443,
+      "step": 825
+    },
+    {
+      "epoch": 0.3993231810490694,
+      "grad_norm": 0.45490375066885613,
+      "learning_rate": 7.03003118214524e-06,
+      "loss": 0.4994,
+      "step": 826
+    },
+    {
+      "epoch": 0.399806623156877,
+      "grad_norm": 0.4689025885486327,
+      "learning_rate": 7.022723226319159e-06,
+      "loss": 0.5249,
+      "step": 827
+    },
+    {
+      "epoch": 0.4002900652646846,
+      "grad_norm": 0.4587224558807211,
+      "learning_rate": 7.0154101002494914e-06,
+      "loss": 0.531,
+      "step": 828
+    },
+    {
+      "epoch": 0.4007735073724921,
+      "grad_norm": 0.43840918685603564,
+      "learning_rate": 7.008091822629172e-06,
+      "loss": 0.5331,
+      "step": 829
+    },
+    {
+      "epoch": 0.4012569494802997,
+      "grad_norm": 0.5046588696515812,
+      "learning_rate": 7.00076841216431e-06,
+      "loss": 0.5302,
+      "step": 830
+    },
+    {
+      "epoch": 0.4017403915881073,
+      "grad_norm": 0.485313377425035,
+      "learning_rate": 6.993439887574133e-06,
+      "loss": 0.5418,
+      "step": 831
+    },
+    {
+      "epoch": 0.4022238336959149,
+      "grad_norm": 0.47268076488419086,
+      "learning_rate": 6.986106267590942e-06,
+      "loss": 0.535,
+      "step": 832
+    },
+    {
+      "epoch": 0.4027072758037225,
+      "grad_norm": 0.4528644527398112,
+      "learning_rate": 6.978767570960057e-06,
+      "loss": 0.5362,
+      "step": 833
+    },
+    {
+      "epoch": 0.4031907179115301,
+      "grad_norm": 0.458241227473586,
+      "learning_rate": 6.971423816439782e-06,
+      "loss": 0.5346,
+      "step": 834
+    },
+    {
+      "epoch": 0.4036741600193377,
+      "grad_norm": 0.476721951726036,
+      "learning_rate": 6.964075022801341e-06,
+      "loss": 0.541,
+      "step": 835
+    },
+    {
+      "epoch": 0.4041576021271453,
+      "grad_norm": 0.4634216130492689,
+      "learning_rate": 6.956721208828847e-06,
+      "loss": 0.5441,
+      "step": 836
+    },
+    {
+      "epoch": 0.40464104423495284,
+      "grad_norm": 0.46202808494163927,
+      "learning_rate": 6.949362393319239e-06,
+      "loss": 0.5416,
+      "step": 837
+    },
+    {
+      "epoch": 0.40512448634276044,
+      "grad_norm": 0.44981029617918733,
+      "learning_rate": 6.941998595082243e-06,
+      "loss": 0.5438,
+      "step": 838
+    },
+    {
+      "epoch": 0.40560792845056803,
+      "grad_norm": 0.44835769630421096,
+      "learning_rate": 6.934629832940322e-06,
+      "loss": 0.5322,
+      "step": 839
+    },
+    {
+      "epoch": 0.40609137055837563,
+      "grad_norm": 0.4879526087044361,
+      "learning_rate": 6.927256125728624e-06,
+      "loss": 0.544,
+      "step": 840
+    },
+    {
+      "epoch": 0.4065748126661832,
+      "grad_norm": 0.47319742567660167,
+      "learning_rate": 6.91987749229494e-06,
+      "loss": 0.5401,
+      "step": 841
+    },
+    {
+      "epoch": 0.4070582547739908,
+      "grad_norm": 0.4492955794688954,
+      "learning_rate": 6.91249395149965e-06,
+      "loss": 0.5393,
+      "step": 842
+    },
+    {
+      "epoch": 0.4075416968817984,
+      "grad_norm": 0.4436463500280526,
+      "learning_rate": 6.905105522215684e-06,
+      "loss": 0.5384,
+      "step": 843
+    },
+    {
+      "epoch": 0.408025138989606,
+      "grad_norm": 0.43897712052796284,
+      "learning_rate": 6.897712223328457e-06,
+      "loss": 0.5297,
+      "step": 844
+    },
+    {
+      "epoch": 0.40850858109741356,
+      "grad_norm": 0.5141765277378779,
+      "learning_rate": 6.89031407373584e-06,
+      "loss": 0.5386,
+      "step": 845
+    },
+    {
+      "epoch": 0.40899202320522116,
+      "grad_norm": 0.48662407879743685,
+      "learning_rate": 6.8829110923481e-06,
+      "loss": 0.5429,
+      "step": 846
+    },
+    {
+      "epoch": 0.40947546531302875,
+      "grad_norm": 0.4671606204232888,
+      "learning_rate": 6.875503298087853e-06,
+      "loss": 0.5339,
+      "step": 847
+    },
+    {
+      "epoch": 0.40995890742083635,
+      "grad_norm": 0.47636170298906977,
+      "learning_rate": 6.868090709890016e-06,
+      "loss": 0.5392,
+      "step": 848
+    },
+    {
+      "epoch": 0.41044234952864395,
+      "grad_norm": 0.4249603079226182,
+      "learning_rate": 6.8606733467017675e-06,
+      "loss": 0.5046,
+      "step": 849
+    },
+    {
+      "epoch": 0.41092579163645154,
+      "grad_norm": 0.4669023850804772,
+      "learning_rate": 6.85325122748248e-06,
+      "loss": 0.5331,
+      "step": 850
+    },
+    {
+      "epoch": 0.41140923374425914,
+      "grad_norm": 0.43795275267260886,
+      "learning_rate": 6.845824371203691e-06,
+      "loss": 0.5409,
+      "step": 851
+    },
+    {
+      "epoch": 0.41189267585206674,
+      "grad_norm": 0.44363617551756607,
+      "learning_rate": 6.838392796849042e-06,
+      "loss": 0.5371,
+      "step": 852
+    },
+    {
+      "epoch": 0.4123761179598743,
+      "grad_norm": 0.4451816120752011,
+      "learning_rate": 6.830956523414239e-06,
+      "loss": 0.5304,
+      "step": 853
+    },
+    {
+      "epoch": 0.4128595600676819,
+      "grad_norm": 0.4684744852800341,
+      "learning_rate": 6.8235155699069944e-06,
+      "loss": 0.5316,
+      "step": 854
+    },
+    {
+      "epoch": 0.41334300217548947,
+      "grad_norm": 0.44092290998190986,
+      "learning_rate": 6.816069955346986e-06,
+      "loss": 0.5127,
+      "step": 855
+    },
+    {
+      "epoch": 0.41382644428329707,
+      "grad_norm": 0.49751628964469147,
+      "learning_rate": 6.808619698765804e-06,
+      "loss": 0.5459,
+      "step": 856
+    },
+    {
+      "epoch": 0.41430988639110466,
+      "grad_norm": 0.45953817226939175,
+      "learning_rate": 6.8011648192069045e-06,
+      "loss": 0.5316,
+      "step": 857
+    },
+    {
+      "epoch": 0.41479332849891226,
+      "grad_norm": 0.470694123027967,
+      "learning_rate": 6.7937053357255585e-06,
+      "loss": 0.5341,
+      "step": 858
+    },
+    {
+      "epoch": 0.41527677060671986,
+      "grad_norm": 0.482654546386677,
+      "learning_rate": 6.786241267388812e-06,
+      "loss": 0.5392,
+      "step": 859
+    },
+    {
+      "epoch": 0.41576021271452746,
+      "grad_norm": 0.49418422954918506,
+      "learning_rate": 6.778772633275421e-06,
+      "loss": 0.5259,
+      "step": 860
+    },
+    {
+      "epoch": 0.41624365482233505,
+      "grad_norm": 0.4508128718503279,
+      "learning_rate": 6.771299452475818e-06,
+      "loss": 0.5439,
+      "step": 861
+    },
+    {
+      "epoch": 0.4167270969301426,
+      "grad_norm": 0.483501145929134,
+      "learning_rate": 6.763821744092054e-06,
+      "loss": 0.521,
+      "step": 862
+    },
+    {
+      "epoch": 0.4172105390379502,
+      "grad_norm": 0.45484324344169746,
+      "learning_rate": 6.756339527237756e-06,
+      "loss": 0.5282,
+      "step": 863
+    },
+    {
+      "epoch": 0.4176939811457578,
+      "grad_norm": 0.47800071067084154,
+      "learning_rate": 6.748852821038075e-06,
+      "loss": 0.5362,
+      "step": 864
+    },
+    {
+      "epoch": 0.4181774232535654,
+      "grad_norm": 0.4561384481336352,
+      "learning_rate": 6.741361644629629e-06,
+      "loss": 0.5452,
+      "step": 865
+    },
+    {
+      "epoch": 0.418660865361373,
+      "grad_norm": 0.4974806458018085,
+      "learning_rate": 6.733866017160475e-06,
+      "loss": 0.5374,
+      "step": 866
+    },
+    {
+      "epoch": 0.4191443074691806,
+      "grad_norm": 0.4696802681175673,
+      "learning_rate": 6.7263659577900375e-06,
+      "loss": 0.5368,
+      "step": 867
+    },
+    {
+      "epoch": 0.4196277495769882,
+      "grad_norm": 0.46421987800289705,
+      "learning_rate": 6.718861485689077e-06,
+      "loss": 0.5361,
+      "step": 868
+    },
+    {
+      "epoch": 0.42011119168479577,
+      "grad_norm": 0.42817223127930704,
+      "learning_rate": 6.711352620039623e-06,
+      "loss": 0.5132,
+      "step": 869
+    },
+    {
+      "epoch": 0.4205946337926033,
+      "grad_norm": 0.4191221628297101,
+      "learning_rate": 6.703839380034945e-06,
+      "loss": 0.5282,
+      "step": 870
+    },
+    {
+      "epoch": 0.4210780759004109,
+      "grad_norm": 0.4151086659722137,
+      "learning_rate": 6.6963217848794895e-06,
+      "loss": 0.5046,
+      "step": 871
+    },
+    {
+      "epoch": 0.4215615180082185,
+      "grad_norm": 0.4638906293888461,
+      "learning_rate": 6.6887998537888354e-06,
+      "loss": 0.5365,
+      "step": 872
+    },
+    {
+      "epoch": 0.4220449601160261,
+      "grad_norm": 0.46061668652958593,
+      "learning_rate": 6.681273605989643e-06,
+      "loss": 0.5315,
+      "step": 873
+    },
+    {
+      "epoch": 0.4225284022238337,
+      "grad_norm": 0.5190170389663172,
+      "learning_rate": 6.673743060719613e-06,
+      "loss": 0.5328,
+      "step": 874
+    },
+    {
+      "epoch": 0.4230118443316413,
+      "grad_norm": 0.45460321890251315,
+      "learning_rate": 6.666208237227421e-06,
+      "loss": 0.5359,
+      "step": 875
+    },
+    {
+      "epoch": 0.4234952864394489,
+      "grad_norm": 0.4292973087733905,
+      "learning_rate": 6.6586691547726855e-06,
+      "loss": 0.5139,
+      "step": 876
+    },
+    {
+      "epoch": 0.4239787285472565,
+      "grad_norm": 0.47082103025419264,
+      "learning_rate": 6.651125832625908e-06,
+      "loss": 0.5455,
+      "step": 877
+    },
+    {
+      "epoch": 0.42446217065506403,
+      "grad_norm": 0.5113179606142492,
+      "learning_rate": 6.6435782900684284e-06,
+      "loss": 0.5528,
+      "step": 878
+    },
+    {
+      "epoch": 0.4249456127628716,
+      "grad_norm": 0.48001824284611705,
+      "learning_rate": 6.636026546392374e-06,
+      "loss": 0.5391,
+      "step": 879
+    },
+    {
+      "epoch": 0.4254290548706792,
+      "grad_norm": 0.4190847750909328,
+      "learning_rate": 6.628470620900611e-06,
+      "loss": 0.5309,
+      "step": 880
+    },
+    {
+      "epoch": 0.4259124969784868,
+      "grad_norm": 0.4429086362697282,
+      "learning_rate": 6.620910532906692e-06,
+      "loss": 0.5194,
+      "step": 881
+    },
+    {
+      "epoch": 0.4263959390862944,
+      "grad_norm": 0.4425962079317176,
+      "learning_rate": 6.613346301734813e-06,
+      "loss": 0.5132,
+      "step": 882
+    },
+    {
+      "epoch": 0.426879381194102,
+      "grad_norm": 0.4614396271141222,
+      "learning_rate": 6.605777946719757e-06,
+      "loss": 0.5358,
+      "step": 883
+    },
+    {
+      "epoch": 0.4273628233019096,
+      "grad_norm": 7.793396607210369,
+      "learning_rate": 6.59820548720685e-06,
+      "loss": 0.74,
+      "step": 884
+    },
+    {
+      "epoch": 0.4278462654097172,
+      "grad_norm": 0.47916638143499957,
+      "learning_rate": 6.590628942551909e-06,
+      "loss": 0.5401,
+      "step": 885
+    },
+    {
+      "epoch": 0.42832970751752475,
+      "grad_norm": 0.5010520173856451,
+      "learning_rate": 6.583048332121193e-06,
+      "loss": 0.5384,
+      "step": 886
+    },
+    {
+      "epoch": 0.42881314962533235,
+      "grad_norm": 0.43637840250511184,
+      "learning_rate": 6.5754636752913535e-06,
+      "loss": 0.5018,
+      "step": 887
+    },
+    {
+      "epoch": 0.42929659173313994,
+      "grad_norm": 0.4347273985958766,
+      "learning_rate": 6.567874991449383e-06,
+      "loss": 0.5303,
+      "step": 888
+    },
+    {
+      "epoch": 0.42978003384094754,
+      "grad_norm": 0.470696307591686,
+      "learning_rate": 6.560282299992571e-06,
+      "loss": 0.5454,
+      "step": 889
+    },
+    {
+      "epoch": 0.43026347594875514,
+      "grad_norm": 0.43949603254219816,
+      "learning_rate": 6.552685620328447e-06,
+      "loss": 0.5115,
+      "step": 890
+    },
+    {
+      "epoch": 0.43074691805656273,
+      "grad_norm": 0.5002179505262337,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.5386,
+      "step": 891
+    },
+    {
+      "epoch": 0.43123036016437033,
+      "grad_norm": 0.48898211514461637,
+      "learning_rate": 6.537480374059313e-06,
+      "loss": 0.5464,
+      "step": 892
+    },
+    {
+      "epoch": 0.4317138022721779,
+      "grad_norm": 0.47111181264654595,
+      "learning_rate": 6.529871846320138e-06,
+      "loss": 0.5225,
+      "step": 893
+    },
+    {
+      "epoch": 0.4321972443799855,
+      "grad_norm": 0.47111234243294525,
+      "learning_rate": 6.522259408105223e-06,
+      "loss": 0.5363,
+      "step": 894
+    },
+    {
+      "epoch": 0.43268068648779306,
+      "grad_norm": 0.4867092311804253,
+      "learning_rate": 6.514643078872571e-06,
+      "loss": 0.533,
+      "step": 895
+    },
+    {
+      "epoch": 0.43316412859560066,
+      "grad_norm": 0.4583905211154658,
+      "learning_rate": 6.507022878090137e-06,
+      "loss": 0.5428,
+      "step": 896
+    },
+    {
+      "epoch": 0.43364757070340826,
+      "grad_norm": 0.4950332559469317,
+      "learning_rate": 6.499398825235767e-06,
+      "loss": 0.5337,
+      "step": 897
+    },
+    {
+      "epoch": 0.43413101281121586,
+      "grad_norm": 0.46856791974797646,
+      "learning_rate": 6.491770939797152e-06,
+      "loss": 0.5323,
+      "step": 898
+    },
+    {
+      "epoch": 0.43461445491902345,
+      "grad_norm": 0.4646364739679311,
+      "learning_rate": 6.4841392412717864e-06,
+      "loss": 0.5407,
+      "step": 899
+    },
+    {
+      "epoch": 0.43509789702683105,
+      "grad_norm": 0.43558921933796657,
+      "learning_rate": 6.476503749166903e-06,
+      "loss": 0.5347,
+      "step": 900
+    },
+    {
+      "epoch": 0.43558133913463865,
+      "grad_norm": 0.4961569425990492,
+      "learning_rate": 6.4688644829994385e-06,
+      "loss": 0.5295,
+      "step": 901
+    },
+    {
+      "epoch": 0.43606478124244624,
+      "grad_norm": 0.4698493360586227,
+      "learning_rate": 6.4612214622959705e-06,
+      "loss": 0.5457,
+      "step": 902
+    },
+    {
+      "epoch": 0.4365482233502538,
+      "grad_norm": 0.435939008717009,
+      "learning_rate": 6.453574706592676e-06,
+      "loss": 0.521,
+      "step": 903
+    },
+    {
+      "epoch": 0.4370316654580614,
+      "grad_norm": 0.5347621942726917,
+      "learning_rate": 6.44592423543528e-06,
+      "loss": 0.5354,
+      "step": 904
+    },
+    {
+      "epoch": 0.437515107565869,
+      "grad_norm": 0.49607229965381217,
+      "learning_rate": 6.4382700683790025e-06,
+      "loss": 0.544,
+      "step": 905
+    },
+    {
+      "epoch": 0.4379985496736766,
+      "grad_norm": 0.4645864217211478,
+      "learning_rate": 6.4306122249885105e-06,
+      "loss": 0.5192,
+      "step": 906
+    },
+    {
+      "epoch": 0.43848199178148417,
+      "grad_norm": 0.46522668277788565,
+      "learning_rate": 6.422950724837872e-06,
+      "loss": 0.526,
+      "step": 907
+    },
+    {
+      "epoch": 0.43896543388929177,
+      "grad_norm": 0.46658657241667495,
+      "learning_rate": 6.415285587510495e-06,
+      "loss": 0.5088,
+      "step": 908
+    },
+    {
+      "epoch": 0.43944887599709936,
+      "grad_norm": 0.49352648435992785,
+      "learning_rate": 6.407616832599091e-06,
+      "loss": 0.5291,
+      "step": 909
+    },
+    {
+      "epoch": 0.43993231810490696,
+      "grad_norm": 0.5022475233173346,
+      "learning_rate": 6.399944479705615e-06,
+      "loss": 0.5349,
+      "step": 910
+    },
+    {
+      "epoch": 0.4404157602127145,
+      "grad_norm": 0.4645078975290342,
+      "learning_rate": 6.392268548441218e-06,
+      "loss": 0.5356,
+      "step": 911
+    },
+    {
+      "epoch": 0.4408992023205221,
+      "grad_norm": 0.4656283146754065,
+      "learning_rate": 6.384589058426201e-06,
+      "loss": 0.5297,
+      "step": 912
+    },
+    {
+      "epoch": 0.4413826444283297,
+      "grad_norm": 0.4778447743673107,
+      "learning_rate": 6.3769060292899585e-06,
+      "loss": 0.531,
+      "step": 913
+    },
+    {
+      "epoch": 0.4418660865361373,
+      "grad_norm": 0.5067362493290994,
+      "learning_rate": 6.3692194806709326e-06,
+      "loss": 0.5266,
+      "step": 914
+    },
+    {
+      "epoch": 0.4423495286439449,
+      "grad_norm": 0.4583050351500669,
+      "learning_rate": 6.36152943221656e-06,
+      "loss": 0.5068,
+      "step": 915
+    },
+    {
+      "epoch": 0.4428329707517525,
+      "grad_norm": 0.44674303828563183,
+      "learning_rate": 6.353835903583225e-06,
+      "loss": 0.5135,
+      "step": 916
+    },
+    {
+      "epoch": 0.4433164128595601,
+      "grad_norm": 0.4845057402223313,
+      "learning_rate": 6.346138914436207e-06,
+      "loss": 0.53,
+      "step": 917
+    },
+    {
+      "epoch": 0.4437998549673677,
+      "grad_norm": 0.46927160898270703,
+      "learning_rate": 6.338438484449632e-06,
+      "loss": 0.5282,
+      "step": 918
+    },
+    {
+      "epoch": 0.4442832970751752,
+      "grad_norm": 0.46844884470128584,
+      "learning_rate": 6.330734633306415e-06,
+      "loss": 0.5205,
+      "step": 919
+    },
+    {
+      "epoch": 0.4447667391829828,
+      "grad_norm": 0.46793830654029384,
+      "learning_rate": 6.3230273806982254e-06,
+      "loss": 0.5354,
+      "step": 920
+    },
+    {
+      "epoch": 0.4452501812907904,
+      "grad_norm": 0.46904655646766624,
+      "learning_rate": 6.31531674632542e-06,
+      "loss": 0.5403,
+      "step": 921
+    },
+    {
+      "epoch": 0.445733623398598,
+      "grad_norm": 0.4797829974325548,
+      "learning_rate": 6.307602749897001e-06,
+      "loss": 0.5285,
+      "step": 922
+    },
+    {
+      "epoch": 0.4462170655064056,
+      "grad_norm": 0.5126549879385887,
+      "learning_rate": 6.299885411130566e-06,
+      "loss": 0.5293,
+      "step": 923
+    },
+    {
+      "epoch": 0.4467005076142132,
+      "grad_norm": 0.47873452298166375,
+      "learning_rate": 6.292164749752256e-06,
+      "loss": 0.5358,
+      "step": 924
+    },
+    {
+      "epoch": 0.4471839497220208,
+      "grad_norm": 0.4697820734594197,
+      "learning_rate": 6.284440785496701e-06,
+      "loss": 0.5303,
+      "step": 925
+    },
+    {
+      "epoch": 0.4476673918298284,
+      "grad_norm": 0.48049048675766015,
+      "learning_rate": 6.27671353810698e-06,
+      "loss": 0.5384,
+      "step": 926
+    },
+    {
+      "epoch": 0.448150833937636,
+      "grad_norm": 0.44829305517954404,
+      "learning_rate": 6.268983027334557e-06,
+      "loss": 0.5349,
+      "step": 927
+    },
+    {
+      "epoch": 0.44863427604544354,
+      "grad_norm": 0.47046261605559686,
+      "learning_rate": 6.2612492729392396e-06,
+      "loss": 0.5445,
+      "step": 928
+    },
+    {
+      "epoch": 0.44911771815325113,
+      "grad_norm": 0.43943250727373434,
+      "learning_rate": 6.25351229468913e-06,
+      "loss": 0.5243,
+      "step": 929
+    },
+    {
+      "epoch": 0.44960116026105873,
+      "grad_norm": 0.457243800658822,
+      "learning_rate": 6.245772112360568e-06,
+      "loss": 0.5335,
+      "step": 930
+    },
+    {
+      "epoch": 0.4500846023688663,
+      "grad_norm": 0.4558342802943274,
+      "learning_rate": 6.2380287457380814e-06,
+      "loss": 0.5295,
+      "step": 931
+    },
+    {
+      "epoch": 0.4505680444766739,
+      "grad_norm": 0.46490304474994104,
+      "learning_rate": 6.230282214614342e-06,
+      "loss": 0.5277,
+      "step": 932
+    },
+    {
+      "epoch": 0.4510514865844815,
+      "grad_norm": 0.46808167357734604,
+      "learning_rate": 6.222532538790107e-06,
+      "loss": 0.5038,
+      "step": 933
+    },
+    {
+      "epoch": 0.4515349286922891,
+      "grad_norm": 0.44746189473984266,
+      "learning_rate": 6.214779738074169e-06,
+      "loss": 0.5482,
+      "step": 934
+    },
+    {
+      "epoch": 0.4520183708000967,
+      "grad_norm": 0.4749348946440431,
+      "learning_rate": 6.2070238322833165e-06,
+      "loss": 0.5408,
+      "step": 935
+    },
+    {
+      "epoch": 0.45250181290790426,
+      "grad_norm": 0.4105832023778957,
+      "learning_rate": 6.199264841242267e-06,
+      "loss": 0.507,
+      "step": 936
+    },
+    {
+      "epoch": 0.45298525501571185,
+      "grad_norm": 0.45569820188618954,
+      "learning_rate": 6.191502784783627e-06,
+      "loss": 0.5361,
+      "step": 937
+    },
+    {
+      "epoch": 0.45346869712351945,
+      "grad_norm": 0.4487391558411953,
+      "learning_rate": 6.183737682747839e-06,
+      "loss": 0.5404,
+      "step": 938
+    },
+    {
+      "epoch": 0.45395213923132705,
+      "grad_norm": 0.38508572249224443,
+      "learning_rate": 6.17596955498313e-06,
+      "loss": 0.475,
+      "step": 939
+    },
+    {
+      "epoch": 0.45443558133913464,
+      "grad_norm": 0.47287816273000344,
+      "learning_rate": 6.16819842134546e-06,
+      "loss": 0.5293,
+      "step": 940
+    },
+    {
+      "epoch": 0.45491902344694224,
+      "grad_norm": 0.4650031849319042,
+      "learning_rate": 6.160424301698472e-06,
+      "loss": 0.5315,
+      "step": 941
+    },
+    {
+      "epoch": 0.45540246555474984,
+      "grad_norm": 0.4613271364601288,
+      "learning_rate": 6.1526472159134454e-06,
+      "loss": 0.5398,
+      "step": 942
+    },
+    {
+      "epoch": 0.45588590766255743,
+      "grad_norm": 0.42626048007656836,
+      "learning_rate": 6.1448671838692365e-06,
+      "loss": 0.5246,
+      "step": 943
+    },
+    {
+      "epoch": 0.456369349770365,
+      "grad_norm": 0.45709658015590027,
+      "learning_rate": 6.1370842254522325e-06,
+      "loss": 0.5392,
+      "step": 944
+    },
+    {
+      "epoch": 0.45685279187817257,
+      "grad_norm": 0.4969316742066114,
+      "learning_rate": 6.129298360556304e-06,
+      "loss": 0.5216,
+      "step": 945
+    },
+    {
+      "epoch": 0.45733623398598017,
+      "grad_norm": 0.48874038272824616,
+      "learning_rate": 6.1215096090827485e-06,
+      "loss": 0.5341,
+      "step": 946
+    },
+    {
+      "epoch": 0.45781967609378776,
+      "grad_norm": 0.43416511670164404,
+      "learning_rate": 6.1137179909402445e-06,
+      "loss": 0.5357,
+      "step": 947
+    },
+    {
+      "epoch": 0.45830311820159536,
+      "grad_norm": 0.4398289245515299,
+      "learning_rate": 6.105923526044794e-06,
+      "loss": 0.5312,
+      "step": 948
+    },
+    {
+      "epoch": 0.45878656030940296,
+      "grad_norm": 0.506729531362168,
+      "learning_rate": 6.098126234319679e-06,
+      "loss": 0.5164,
+      "step": 949
+    },
+    {
+      "epoch": 0.45927000241721055,
+      "grad_norm": 0.49661020376021475,
+      "learning_rate": 6.0903261356954035e-06,
+      "loss": 0.5406,
+      "step": 950
+    },
+    {
+      "epoch": 0.45975344452501815,
+      "grad_norm": 0.49213602324131056,
+      "learning_rate": 6.08252325010965e-06,
+      "loss": 0.5369,
+      "step": 951
+    },
+    {
+      "epoch": 0.4602368866328257,
+      "grad_norm": 0.4522403269153531,
+      "learning_rate": 6.074717597507223e-06,
+      "loss": 0.5318,
+      "step": 952
+    },
+    {
+      "epoch": 0.4607203287406333,
+      "grad_norm": 0.46334490270574613,
+      "learning_rate": 6.066909197839996e-06,
+      "loss": 0.5053,
+      "step": 953
+    },
+    {
+      "epoch": 0.4612037708484409,
+      "grad_norm": 0.4555176211572043,
+      "learning_rate": 6.059098071066874e-06,
+      "loss": 0.5313,
+      "step": 954
+    },
+    {
+      "epoch": 0.4616872129562485,
+      "grad_norm": 0.4697386359927019,
+      "learning_rate": 6.051284237153723e-06,
+      "loss": 0.5304,
+      "step": 955
+    },
+    {
+      "epoch": 0.4621706550640561,
+      "grad_norm": 0.47606054578122203,
+      "learning_rate": 6.043467716073333e-06,
+      "loss": 0.5392,
+      "step": 956
+    },
+    {
+      "epoch": 0.4626540971718637,
+      "grad_norm": 0.4504863880612236,
+      "learning_rate": 6.035648527805359e-06,
+      "loss": 0.5333,
+      "step": 957
+    },
+    {
+      "epoch": 0.4631375392796713,
+      "grad_norm": 0.47075391826579777,
+      "learning_rate": 6.0278266923362805e-06,
+      "loss": 0.5331,
+      "step": 958
+    },
+    {
+      "epoch": 0.46362098138747887,
+      "grad_norm": 0.4403281566352786,
+      "learning_rate": 6.0200022296593375e-06,
+      "loss": 0.5432,
+      "step": 959
+    },
+    {
+      "epoch": 0.46410442349528647,
+      "grad_norm": 0.4572023392519702,
+      "learning_rate": 6.012175159774488e-06,
+      "loss": 0.5323,
+      "step": 960
+    },
+    {
+      "epoch": 0.464587865603094,
+      "grad_norm": 0.44773102967818523,
+      "learning_rate": 6.004345502688353e-06,
+      "loss": 0.5299,
+      "step": 961
+    },
+    {
+      "epoch": 0.4650713077109016,
+      "grad_norm": 0.4371461582728146,
+      "learning_rate": 5.996513278414166e-06,
+      "loss": 0.5385,
+      "step": 962
+    },
+    {
+      "epoch": 0.4655547498187092,
+      "grad_norm": 0.42182149099377253,
+      "learning_rate": 5.988678506971726e-06,
+      "loss": 0.5303,
+      "step": 963
+    },
+    {
+      "epoch": 0.4660381919265168,
+      "grad_norm": 0.4594942911793403,
+      "learning_rate": 5.980841208387338e-06,
+      "loss": 0.503,
+      "step": 964
+    },
+    {
+      "epoch": 0.4665216340343244,
+      "grad_norm": 0.4501324492723119,
+      "learning_rate": 5.973001402693769e-06,
+      "loss": 0.5253,
+      "step": 965
+    },
+    {
+      "epoch": 0.467005076142132,
+      "grad_norm": 0.4373460975849127,
+      "learning_rate": 5.965159109930196e-06,
+      "loss": 0.5386,
+      "step": 966
+    },
+    {
+      "epoch": 0.4674885182499396,
+      "grad_norm": 0.4418857621109752,
+      "learning_rate": 5.957314350142149e-06,
+      "loss": 0.529,
+      "step": 967
+    },
+    {
+      "epoch": 0.4679719603577472,
+      "grad_norm": 0.485651415733174,
+      "learning_rate": 5.94946714338147e-06,
+      "loss": 0.538,
+      "step": 968
+    },
+    {
+      "epoch": 0.4684554024655547,
+      "grad_norm": 0.44827050412474007,
+      "learning_rate": 5.941617509706247e-06,
+      "loss": 0.5333,
+      "step": 969
+    },
+    {
+      "epoch": 0.4689388445733623,
+      "grad_norm": 0.44674045160443804,
+      "learning_rate": 5.933765469180779e-06,
+      "loss": 0.5329,
+      "step": 970
+    },
+    {
+      "epoch": 0.4694222866811699,
+      "grad_norm": 0.47047904115025807,
+      "learning_rate": 5.925911041875514e-06,
+      "loss": 0.5304,
+      "step": 971
+    },
+    {
+      "epoch": 0.4699057287889775,
+      "grad_norm": 0.47618192657266745,
+      "learning_rate": 5.9180542478670025e-06,
+      "loss": 0.5339,
+      "step": 972
+    },
+    {
+      "epoch": 0.4703891708967851,
+      "grad_norm": 0.4372302817189476,
+      "learning_rate": 5.910195107237842e-06,
+      "loss": 0.5311,
+      "step": 973
+    },
+    {
+      "epoch": 0.4708726130045927,
+      "grad_norm": 0.4599705922186888,
+      "learning_rate": 5.902333640076627e-06,
+      "loss": 0.5294,
+      "step": 974
+    },
+    {
+      "epoch": 0.4713560551124003,
+      "grad_norm": 0.4668491954105507,
+      "learning_rate": 5.894469866477905e-06,
+      "loss": 0.5319,
+      "step": 975
+    },
+    {
+      "epoch": 0.4718394972202079,
+      "grad_norm": 0.4298713287991765,
+      "learning_rate": 5.886603806542114e-06,
+      "loss": 0.5308,
+      "step": 976
+    },
+    {
+      "epoch": 0.47232293932801545,
+      "grad_norm": 0.44120727064086296,
+      "learning_rate": 5.878735480375537e-06,
+      "loss": 0.5271,
+      "step": 977
+    },
+    {
+      "epoch": 0.47280638143582304,
+      "grad_norm": 0.45517085979612587,
+      "learning_rate": 5.87086490809025e-06,
+      "loss": 0.5363,
+      "step": 978
+    },
+    {
+      "epoch": 0.47328982354363064,
+      "grad_norm": 0.41718667843710106,
+      "learning_rate": 5.862992109804071e-06,
+      "loss": 0.5208,
+      "step": 979
+    },
+    {
+      "epoch": 0.47377326565143824,
+      "grad_norm": 0.4698208514148652,
+      "learning_rate": 5.855117105640503e-06,
+      "loss": 0.5045,
+      "step": 980
+    },
+    {
+      "epoch": 0.47425670775924583,
+      "grad_norm": 0.44330430106357593,
+      "learning_rate": 5.847239915728695e-06,
+      "loss": 0.5213,
+      "step": 981
+    },
+    {
+      "epoch": 0.47474014986705343,
+      "grad_norm": 0.46348151704515983,
+      "learning_rate": 5.839360560203379e-06,
+      "loss": 0.5416,
+      "step": 982
+    },
+    {
+      "epoch": 0.475223591974861,
+      "grad_norm": 0.4369134554030048,
+      "learning_rate": 5.831479059204822e-06,
+      "loss": 0.5397,
+      "step": 983
+    },
+    {
+      "epoch": 0.4757070340826686,
+      "grad_norm": 0.434332595795343,
+      "learning_rate": 5.823595432878775e-06,
+      "loss": 0.5253,
+      "step": 984
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 0.45066877191979876,
+      "learning_rate": 5.815709701376424e-06,
+      "loss": 0.501,
+      "step": 985
+    },
+    {
+      "epoch": 0.47667391829828376,
+      "grad_norm": 0.43680722385313575,
+      "learning_rate": 5.8078218848543326e-06,
+      "loss": 0.5425,
+      "step": 986
+    },
+    {
+      "epoch": 0.47715736040609136,
+      "grad_norm": 0.462367177083934,
+      "learning_rate": 5.799932003474398e-06,
+      "loss": 0.5293,
+      "step": 987
+    },
+    {
+      "epoch": 0.47764080251389895,
+      "grad_norm": 0.46046005650848004,
+      "learning_rate": 5.7920400774037884e-06,
+      "loss": 0.5273,
+      "step": 988
+    },
+    {
+      "epoch": 0.47812424462170655,
+      "grad_norm": 0.4325701614014005,
+      "learning_rate": 5.784146126814909e-06,
+      "loss": 0.523,
+      "step": 989
+    },
+    {
+      "epoch": 0.47860768672951415,
+      "grad_norm": 0.42688145845821907,
+      "learning_rate": 5.776250171885329e-06,
+      "loss": 0.5289,
+      "step": 990
+    },
+    {
+      "epoch": 0.47909112883732174,
+      "grad_norm": 0.4553909518433745,
+      "learning_rate": 5.768352232797748e-06,
+      "loss": 0.5354,
+      "step": 991
+    },
+    {
+      "epoch": 0.47957457094512934,
+      "grad_norm": 0.4612709888611227,
+      "learning_rate": 5.760452329739933e-06,
+      "loss": 0.5346,
+      "step": 992
+    },
+    {
+      "epoch": 0.48005801305293694,
+      "grad_norm": 0.4671180697912862,
+      "learning_rate": 5.752550482904674e-06,
+      "loss": 0.5381,
+      "step": 993
+    },
+    {
+      "epoch": 0.4805414551607445,
+      "grad_norm": 0.44306247126150655,
+      "learning_rate": 5.744646712489729e-06,
+      "loss": 0.5347,
+      "step": 994
+    },
+    {
+      "epoch": 0.4810248972685521,
+      "grad_norm": 0.47845590337101257,
+      "learning_rate": 5.736741038697771e-06,
+      "loss": 0.546,
+      "step": 995
+    },
+    {
+      "epoch": 0.4815083393763597,
+      "grad_norm": 0.42271596693767594,
+      "learning_rate": 5.728833481736339e-06,
+      "loss": 0.5189,
+      "step": 996
+    },
+    {
+      "epoch": 0.48199178148416727,
+      "grad_norm": 0.4538365308944498,
+      "learning_rate": 5.720924061817786e-06,
+      "loss": 0.5405,
+      "step": 997
+    },
+    {
+      "epoch": 0.48247522359197487,
+      "grad_norm": 0.4492861025701022,
+      "learning_rate": 5.71301279915923e-06,
+      "loss": 0.5317,
+      "step": 998
+    },
+    {
+      "epoch": 0.48295866569978246,
+      "grad_norm": 0.4745326397538611,
+      "learning_rate": 5.705099713982491e-06,
+      "loss": 0.532,
+      "step": 999
+    },
+    {
+      "epoch": 0.48344210780759006,
+      "grad_norm": 0.44214626511814337,
+      "learning_rate": 5.697184826514058e-06,
+      "loss": 0.5305,
+      "step": 1000
+    },
+    {
+      "epoch": 0.48392554991539766,
+      "grad_norm": 0.46953984488621225,
+      "learning_rate": 5.689268156985015e-06,
+      "loss": 0.5385,
+      "step": 1001
+    },
+    {
+      "epoch": 0.4844089920232052,
+      "grad_norm": 0.421787366978404,
+      "learning_rate": 5.6813497256310124e-06,
+      "loss": 0.5468,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4848924341310128,
+      "grad_norm": 0.44994209423666665,
+      "learning_rate": 5.673429552692196e-06,
+      "loss": 0.5259,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4853758762388204,
+      "grad_norm": 0.4777649344533823,
+      "learning_rate": 5.66550765841317e-06,
+      "loss": 0.5312,
+      "step": 1004
+    },
+    {
+      "epoch": 0.485859318346628,
+      "grad_norm": 0.4450135880363142,
+      "learning_rate": 5.6575840630429295e-06,
+      "loss": 0.5234,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4863427604544356,
+      "grad_norm": 0.48478431421225965,
+      "learning_rate": 5.649658786834825e-06,
+      "loss": 0.5337,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4868262025622432,
+      "grad_norm": 0.4600376201705413,
+      "learning_rate": 5.641731850046503e-06,
+      "loss": 0.5292,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4873096446700508,
+      "grad_norm": 0.46358685580492737,
+      "learning_rate": 5.633803272939851e-06,
+      "loss": 0.5033,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4877930867778584,
+      "grad_norm": 0.4424655348417247,
+      "learning_rate": 5.62587307578095e-06,
+      "loss": 0.5199,
+      "step": 1009
+    },
+    {
+      "epoch": 0.4882765288856659,
+      "grad_norm": 0.44958190643210255,
+      "learning_rate": 5.6179412788400255e-06,
+      "loss": 0.5285,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4887599709934735,
+      "grad_norm": 0.46154354908053036,
+      "learning_rate": 5.610007902391387e-06,
+      "loss": 0.5302,
+      "step": 1011
+    },
+    {
+      "epoch": 0.4892434131012811,
+      "grad_norm": 0.4787830389554928,
+      "learning_rate": 5.602072966713389e-06,
+      "loss": 0.5319,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4897268552090887,
+      "grad_norm": 0.46247126201641375,
+      "learning_rate": 5.594136492088363e-06,
+      "loss": 0.533,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4902102973168963,
+      "grad_norm": 0.4497969101747785,
+      "learning_rate": 5.586198498802577e-06,
+      "loss": 0.5207,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4906937394247039,
+      "grad_norm": 0.4711750674532031,
+      "learning_rate": 5.578259007146183e-06,
+      "loss": 0.5182,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4911771815325115,
+      "grad_norm": 0.44280948751548693,
+      "learning_rate": 5.570318037413162e-06,
+      "loss": 0.5335,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4916606236403191,
+      "grad_norm": 0.4140593593091692,
+      "learning_rate": 5.562375609901273e-06,
+      "loss": 0.498,
+      "step": 1017
+    },
+    {
+      "epoch": 0.49214406574812664,
+      "grad_norm": 0.5035482968527856,
+      "learning_rate": 5.5544317449119975e-06,
+      "loss": 0.535,
+      "step": 1018
+    },
+    {
+      "epoch": 0.49262750785593423,
+      "grad_norm": 0.470430802340302,
+      "learning_rate": 5.546486462750499e-06,
+      "loss": 0.529,
+      "step": 1019
+    },
+    {
+      "epoch": 0.49311094996374183,
+      "grad_norm": 0.4614608636774156,
+      "learning_rate": 5.538539783725556e-06,
+      "loss": 0.5415,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4935943920715494,
+      "grad_norm": 0.43439106514800707,
+      "learning_rate": 5.530591728149522e-06,
+      "loss": 0.5237,
+      "step": 1021
+    },
+    {
+      "epoch": 0.494077834179357,
+      "grad_norm": 0.4532581332185661,
+      "learning_rate": 5.522642316338268e-06,
+      "loss": 0.5275,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4945612762871646,
+      "grad_norm": 0.44400011265861933,
+      "learning_rate": 5.51469156861113e-06,
+      "loss": 0.5279,
+      "step": 1023
+    },
+    {
+      "epoch": 0.4950447183949722,
+      "grad_norm": 0.464713276215896,
+      "learning_rate": 5.50673950529086e-06,
+      "loss": 0.5261,
+      "step": 1024
+    },
+    {
+      "epoch": 0.4955281605027798,
+      "grad_norm": 0.46801515065857147,
+      "learning_rate": 5.498786146703575e-06,
+      "loss": 0.5392,
+      "step": 1025
+    },
+    {
+      "epoch": 0.4960116026105874,
+      "grad_norm": 0.43826397020984875,
+      "learning_rate": 5.490831513178698e-06,
+      "loss": 0.513,
+      "step": 1026
+    },
+    {
+      "epoch": 0.49649504471839495,
+      "grad_norm": 0.4545609138302334,
+      "learning_rate": 5.482875625048916e-06,
+      "loss": 0.5342,
+      "step": 1027
+    },
+    {
+      "epoch": 0.49697848682620255,
+      "grad_norm": 0.45528220157554766,
+      "learning_rate": 5.474918502650116e-06,
+      "loss": 0.5371,
+      "step": 1028
+    },
+    {
+      "epoch": 0.49746192893401014,
+      "grad_norm": 0.4918245291291218,
+      "learning_rate": 5.466960166321348e-06,
+      "loss": 0.5248,
+      "step": 1029
+    },
+    {
+      "epoch": 0.49794537104181774,
+      "grad_norm": 0.43603238272665606,
+      "learning_rate": 5.459000636404759e-06,
+      "loss": 0.5236,
+      "step": 1030
+    },
+    {
+      "epoch": 0.49842881314962534,
+      "grad_norm": 0.43802907063700564,
+      "learning_rate": 5.451039933245551e-06,
+      "loss": 0.5342,
+      "step": 1031
+    },
+    {
+      "epoch": 0.49891225525743294,
+      "grad_norm": 0.4660487433831574,
+      "learning_rate": 5.44307807719192e-06,
+      "loss": 0.5249,
+      "step": 1032
+    },
+    {
+      "epoch": 0.49939569736524053,
+      "grad_norm": 0.4349479566561577,
+      "learning_rate": 5.435115088595016e-06,
+      "loss": 0.4997,
+      "step": 1033
+    },
+    {
+      "epoch": 0.49987913947304813,
+      "grad_norm": 0.43592525172206065,
+      "learning_rate": 5.4271509878088755e-06,
+      "loss": 0.5263,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5003625815808557,
+      "grad_norm": 0.42380458268711035,
+      "learning_rate": 5.4191857951903825e-06,
+      "loss": 0.503,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5008460236886633,
+      "grad_norm": 0.47347721204466,
+      "learning_rate": 5.4112195310992144e-06,
+      "loss": 0.5228,
+      "step": 1036
+    },
+    {
+      "epoch": 0.5013294657964709,
+      "grad_norm": 0.4523581952116975,
+      "learning_rate": 5.403252215897781e-06,
+      "loss": 0.5295,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5018129079042785,
+      "grad_norm": 0.44718772307460525,
+      "learning_rate": 5.395283869951184e-06,
+      "loss": 0.5402,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5022963500120861,
+      "grad_norm": 0.42574725470561453,
+      "learning_rate": 5.387314513627156e-06,
+      "loss": 0.5228,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5027797921198937,
+      "grad_norm": 0.4816073830916942,
+      "learning_rate": 5.379344167296017e-06,
+      "loss": 0.5302,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5032632342277013,
+      "grad_norm": 0.454535172796951,
+      "learning_rate": 5.371372851330612e-06,
+      "loss": 0.5337,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5037466763355088,
+      "grad_norm": 0.44304402033518747,
+      "learning_rate": 5.3634005861062675e-06,
+      "loss": 0.5348,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5042301184433164,
+      "grad_norm": 0.4647077624164304,
+      "learning_rate": 5.355427392000736e-06,
+      "loss": 0.5367,
+      "step": 1043
+    },
+    {
+      "epoch": 0.504713560551124,
+      "grad_norm": 0.47204921033701974,
+      "learning_rate": 5.347453289394146e-06,
+      "loss": 0.5236,
+      "step": 1044
+    },
+    {
+      "epoch": 0.5051970026589316,
+      "grad_norm": 0.444269913904137,
+      "learning_rate": 5.339478298668943e-06,
+      "loss": 0.5374,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5056804447667392,
+      "grad_norm": 0.437749383717649,
+      "learning_rate": 5.331502440209849e-06,
+      "loss": 0.529,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5061638868745467,
+      "grad_norm": 0.44850371734571,
+      "learning_rate": 5.3235257344037996e-06,
+      "loss": 0.5363,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5066473289823543,
+      "grad_norm": 0.4687290830806752,
+      "learning_rate": 5.3155482016398995e-06,
+      "loss": 0.5335,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5071307710901619,
+      "grad_norm": 0.43863314490483407,
+      "learning_rate": 5.307569862309363e-06,
+      "loss": 0.5269,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5076142131979695,
+      "grad_norm": 0.44499358455418236,
+      "learning_rate": 5.29959073680547e-06,
+      "loss": 0.5269,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5080976553057771,
+      "grad_norm": 0.4443034689772289,
+      "learning_rate": 5.2916108455235084e-06,
+      "loss": 0.5247,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5085810974135847,
+      "grad_norm": 0.41196428985814043,
+      "learning_rate": 5.2836302088607235e-06,
+      "loss": 0.5257,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5090645395213923,
+      "grad_norm": 0.4196111969098687,
+      "learning_rate": 5.275648847216263e-06,
+      "loss": 0.5326,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5095479816291999,
+      "grad_norm": 0.4396027925039716,
+      "learning_rate": 5.267666780991135e-06,
+      "loss": 0.5384,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5100314237370075,
+      "grad_norm": 0.4918850008899013,
+      "learning_rate": 5.259684030588141e-06,
+      "loss": 0.5217,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5105148658448151,
+      "grad_norm": 0.4636011028874602,
+      "learning_rate": 5.251700616411836e-06,
+      "loss": 0.5292,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5109983079526227,
+      "grad_norm": 0.45080015588791633,
+      "learning_rate": 5.243716558868469e-06,
+      "loss": 0.5335,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5114817500604303,
+      "grad_norm": 0.4430606358386943,
+      "learning_rate": 5.235731878365935e-06,
+      "loss": 0.5366,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5119651921682379,
+      "grad_norm": 0.4765633831252005,
+      "learning_rate": 5.22774659531372e-06,
+      "loss": 0.5343,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5124486342760455,
+      "grad_norm": 0.47759199812886083,
+      "learning_rate": 5.219760730122854e-06,
+      "loss": 0.5318,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5129320763838531,
+      "grad_norm": 0.4277450357502153,
+      "learning_rate": 5.211774303205849e-06,
+      "loss": 0.5055,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5134155184916607,
+      "grad_norm": 0.42824551582926373,
+      "learning_rate": 5.203787334976655e-06,
+      "loss": 0.5015,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5138989605994683,
+      "grad_norm": 0.5041405711353805,
+      "learning_rate": 5.195799845850611e-06,
+      "loss": 0.525,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5143824027072758,
+      "grad_norm": 0.4306413852615145,
+      "learning_rate": 5.18781185624438e-06,
+      "loss": 0.5265,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5148658448150834,
+      "grad_norm": 0.4361432944536921,
+      "learning_rate": 5.179823386575908e-06,
+      "loss": 0.5311,
+      "step": 1065
+    },
+    {
+      "epoch": 0.515349286922891,
+      "grad_norm": 0.4240690760700838,
+      "learning_rate": 5.171834457264364e-06,
+      "loss": 0.5286,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5158327290306985,
+      "grad_norm": 0.47256746355593565,
+      "learning_rate": 5.1638450887301006e-06,
+      "loss": 0.5282,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5163161711385061,
+      "grad_norm": 0.4445988854998036,
+      "learning_rate": 5.155855301394585e-06,
+      "loss": 0.527,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5167996132463137,
+      "grad_norm": 0.46405127876443125,
+      "learning_rate": 5.147865115680357e-06,
+      "loss": 0.5289,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5172830553541213,
+      "grad_norm": 0.45801111107179227,
+      "learning_rate": 5.139874552010975e-06,
+      "loss": 0.531,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5177664974619289,
+      "grad_norm": 0.454345233725776,
+      "learning_rate": 5.131883630810966e-06,
+      "loss": 0.5428,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5182499395697365,
+      "grad_norm": 0.4649418568564353,
+      "learning_rate": 5.123892372505768e-06,
+      "loss": 0.524,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5187333816775441,
+      "grad_norm": 0.4609921972037312,
+      "learning_rate": 5.11590079752168e-06,
+      "loss": 0.5337,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5192168237853517,
+      "grad_norm": 0.4257268056499296,
+      "learning_rate": 5.107908926285813e-06,
+      "loss": 0.5247,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5197002658931593,
+      "grad_norm": 0.4277925159892485,
+      "learning_rate": 5.099916779226032e-06,
+      "loss": 0.5314,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5201837080009669,
+      "grad_norm": 0.46374173796570095,
+      "learning_rate": 5.091924376770912e-06,
+      "loss": 0.5267,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5206671501087745,
+      "grad_norm": 0.46413399868377414,
+      "learning_rate": 5.083931739349675e-06,
+      "loss": 0.5227,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5211505922165821,
+      "grad_norm": 0.42805047816232233,
+      "learning_rate": 5.075938887392149e-06,
+      "loss": 0.5148,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5216340343243897,
+      "grad_norm": 0.42348168283938675,
+      "learning_rate": 5.0679458413287055e-06,
+      "loss": 0.5168,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5221174764321972,
+      "grad_norm": 0.48349127354280697,
+      "learning_rate": 5.059952621590216e-06,
+      "loss": 0.5274,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5226009185400048,
+      "grad_norm": 0.447828940551305,
+      "learning_rate": 5.051959248607993e-06,
+      "loss": 0.5251,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5230843606478124,
+      "grad_norm": 0.45995219297674483,
+      "learning_rate": 5.043965742813744e-06,
+      "loss": 0.5246,
+      "step": 1082
+    },
+    {
+      "epoch": 0.52356780275562,
+      "grad_norm": 0.4744443087812869,
+      "learning_rate": 5.035972124639511e-06,
+      "loss": 0.5299,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5240512448634276,
+      "grad_norm": 0.4140590259366713,
+      "learning_rate": 5.02797841451763e-06,
+      "loss": 0.5273,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5245346869712352,
+      "grad_norm": 0.4421687910150297,
+      "learning_rate": 5.019984632880665e-06,
+      "loss": 0.5342,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5250181290790428,
+      "grad_norm": 0.4665180180836343,
+      "learning_rate": 5.011990800161369e-06,
+      "loss": 0.5314,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5255015711868504,
+      "grad_norm": 0.45569707259235365,
+      "learning_rate": 5.00399693679262e-06,
+      "loss": 0.5291,
+      "step": 1087
+    },
+    {
+      "epoch": 0.525985013294658,
+      "grad_norm": 0.3966652463954235,
+      "learning_rate": 4.9960030632073815e-06,
+      "loss": 0.4852,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5264684554024656,
+      "grad_norm": 0.4156779611396039,
+      "learning_rate": 4.988009199838632e-06,
+      "loss": 0.5266,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5269518975102732,
+      "grad_norm": 0.4459927015276638,
+      "learning_rate": 4.980015367119336e-06,
+      "loss": 0.5128,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5274353396180808,
+      "grad_norm": 0.4470940459251613,
+      "learning_rate": 4.9720215854823716e-06,
+      "loss": 0.5215,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5279187817258884,
+      "grad_norm": 0.4379040861596386,
+      "learning_rate": 4.96402787536049e-06,
+      "loss": 0.529,
+      "step": 1092
+    },
+    {
+      "epoch": 0.528402223833696,
+      "grad_norm": 0.42846117019918506,
+      "learning_rate": 4.956034257186258e-06,
+      "loss": 0.5196,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5288856659415035,
+      "grad_norm": 0.45571668646782787,
+      "learning_rate": 4.9480407513920086e-06,
+      "loss": 0.527,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5293691080493111,
+      "grad_norm": 0.4680208745726276,
+      "learning_rate": 4.940047378409786e-06,
+      "loss": 0.523,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5298525501571187,
+      "grad_norm": 0.44029959009092084,
+      "learning_rate": 4.932054158671295e-06,
+      "loss": 0.5244,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5303359922649262,
+      "grad_norm": 0.45471058687459337,
+      "learning_rate": 4.924061112607853e-06,
+      "loss": 0.532,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5308194343727338,
+      "grad_norm": 0.43297324457453135,
+      "learning_rate": 4.9160682606503255e-06,
+      "loss": 0.5226,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5313028764805414,
+      "grad_norm": 0.4269816974887771,
+      "learning_rate": 4.908075623229089e-06,
+      "loss": 0.5242,
+      "step": 1099
+    },
+    {
+      "epoch": 0.531786318588349,
+      "grad_norm": 0.4280455143246443,
+      "learning_rate": 4.900083220773968e-06,
+      "loss": 0.5082,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5322697606961566,
+      "grad_norm": 0.4422972364197889,
+      "learning_rate": 4.892091073714189e-06,
+      "loss": 0.5162,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5327532028039642,
+      "grad_norm": 0.44177794728378245,
+      "learning_rate": 4.88409920247832e-06,
+      "loss": 0.5287,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5332366449117718,
+      "grad_norm": 0.4384466281073893,
+      "learning_rate": 4.876107627494234e-06,
+      "loss": 0.5224,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5337200870195794,
+      "grad_norm": 0.42791066381831555,
+      "learning_rate": 4.868116369189033e-06,
+      "loss": 0.519,
+      "step": 1104
+    },
+    {
+      "epoch": 0.534203529127387,
+      "grad_norm": 0.43287603203122865,
+      "learning_rate": 4.860125447989026e-06,
+      "loss": 0.5315,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5346869712351946,
+      "grad_norm": 0.40835340581900775,
+      "learning_rate": 4.852134884319646e-06,
+      "loss": 0.5013,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5351704133430022,
+      "grad_norm": 0.4564877621647975,
+      "learning_rate": 4.844144698605418e-06,
+      "loss": 0.5163,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5356538554508098,
+      "grad_norm": 0.4350849446506223,
+      "learning_rate": 4.836154911269902e-06,
+      "loss": 0.5216,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5361372975586174,
+      "grad_norm": 0.45777989132001196,
+      "learning_rate": 4.8281655427356375e-06,
+      "loss": 0.5248,
+      "step": 1109
+    },
+    {
+      "epoch": 0.536620739666425,
+      "grad_norm": 0.463821530722432,
+      "learning_rate": 4.820176613424095e-06,
+      "loss": 0.5302,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5371041817742326,
+      "grad_norm": 0.44517887708797427,
+      "learning_rate": 4.812188143755621e-06,
+      "loss": 0.5214,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5375876238820402,
+      "grad_norm": 0.5017058167871686,
+      "learning_rate": 4.80420015414939e-06,
+      "loss": 0.5305,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5380710659898477,
+      "grad_norm": 0.4672768776486219,
+      "learning_rate": 4.796212665023345e-06,
+      "loss": 0.5237,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5385545080976553,
+      "grad_norm": 0.4419995316595335,
+      "learning_rate": 4.788225696794153e-06,
+      "loss": 0.5277,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5390379502054629,
+      "grad_norm": 0.44730511943868706,
+      "learning_rate": 4.780239269877147e-06,
+      "loss": 0.5313,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5395213923132705,
+      "grad_norm": 0.45204850110236905,
+      "learning_rate": 4.7722534046862805e-06,
+      "loss": 0.5231,
+      "step": 1116
+    },
+    {
+      "epoch": 0.540004834421078,
+      "grad_norm": 0.45779960123303604,
+      "learning_rate": 4.764268121634066e-06,
+      "loss": 0.507,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5404882765288856,
+      "grad_norm": 0.41695878051285573,
+      "learning_rate": 4.7562834411315324e-06,
+      "loss": 0.5042,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5409717186366932,
+      "grad_norm": 0.4513235919318933,
+      "learning_rate": 4.748299383588167e-06,
+      "loss": 0.5258,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5414551607445008,
+      "grad_norm": 0.467590473541893,
+      "learning_rate": 4.74031596941186e-06,
+      "loss": 0.5375,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5419386028523084,
+      "grad_norm": 0.4524946558422428,
+      "learning_rate": 4.7323332190088675e-06,
+      "loss": 0.5199,
+      "step": 1121
+    },
+    {
+      "epoch": 0.542422044960116,
+      "grad_norm": 0.42113263739650647,
+      "learning_rate": 4.7243511527837374e-06,
+      "loss": 0.5251,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5429054870679236,
+      "grad_norm": 0.4636225245244688,
+      "learning_rate": 4.716369791139279e-06,
+      "loss": 0.5308,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5433889291757312,
+      "grad_norm": 0.4398566927754983,
+      "learning_rate": 4.708389154476492e-06,
+      "loss": 0.5201,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5438723712835388,
+      "grad_norm": 0.43330211663967066,
+      "learning_rate": 4.7004092631945315e-06,
+      "loss": 0.5258,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5443558133913464,
+      "grad_norm": 0.44482177202458967,
+      "learning_rate": 4.692430137690638e-06,
+      "loss": 0.5222,
+      "step": 1126
+    },
+    {
+      "epoch": 0.544839255499154,
+      "grad_norm": 0.4318024796342877,
+      "learning_rate": 4.684451798360102e-06,
+      "loss": 0.5204,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5453226976069616,
+      "grad_norm": 0.45444414518744425,
+      "learning_rate": 4.6764742655962e-06,
+      "loss": 0.5255,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5458061397147692,
+      "grad_norm": 0.4372839038666406,
+      "learning_rate": 4.6684975597901526e-06,
+      "loss": 0.5275,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5462895818225767,
+      "grad_norm": 0.43767287064021165,
+      "learning_rate": 4.660521701331058e-06,
+      "loss": 0.5046,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5467730239303843,
+      "grad_norm": 0.42146624184063447,
+      "learning_rate": 4.652546710605857e-06,
+      "loss": 0.5284,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5472564660381919,
+      "grad_norm": 0.4528806888549099,
+      "learning_rate": 4.644572607999267e-06,
+      "loss": 0.5234,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5477399081459995,
+      "grad_norm": 0.41615032339392954,
+      "learning_rate": 4.636599413893734e-06,
+      "loss": 0.5149,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5482233502538071,
+      "grad_norm": 0.4080780400252472,
+      "learning_rate": 4.628627148669391e-06,
+      "loss": 0.5069,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5487067923616147,
+      "grad_norm": 0.4376373218589361,
+      "learning_rate": 4.620655832703984e-06,
+      "loss": 0.5232,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5491902344694223,
+      "grad_norm": 0.44817086720050736,
+      "learning_rate": 4.612685486372846e-06,
+      "loss": 0.5284,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5496736765772299,
+      "grad_norm": 0.4269717422470433,
+      "learning_rate": 4.604716130048818e-06,
+      "loss": 0.5292,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5501571186850375,
+      "grad_norm": 0.4463889311004707,
+      "learning_rate": 4.596747784102221e-06,
+      "loss": 0.5296,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5506405607928451,
+      "grad_norm": 0.4544805899306494,
+      "learning_rate": 4.588780468900787e-06,
+      "loss": 0.5342,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5511240029006527,
+      "grad_norm": 0.43223974279832084,
+      "learning_rate": 4.580814204809618e-06,
+      "loss": 0.5278,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5516074450084603,
+      "grad_norm": 0.4526101073084036,
+      "learning_rate": 4.572849012191126e-06,
+      "loss": 0.5274,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5520908871162679,
+      "grad_norm": 0.46135315991278786,
+      "learning_rate": 4.564884911404986e-06,
+      "loss": 0.5308,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5525743292240755,
+      "grad_norm": 0.451744932774668,
+      "learning_rate": 4.5569219228080805e-06,
+      "loss": 0.5228,
+      "step": 1143
+    },
+    {
+      "epoch": 0.553057771331883,
+      "grad_norm": 0.4431407162306295,
+      "learning_rate": 4.54896006675445e-06,
+      "loss": 0.5001,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5535412134396906,
+      "grad_norm": 0.4115173802208087,
+      "learning_rate": 4.540999363595242e-06,
+      "loss": 0.4963,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5540246555474981,
+      "grad_norm": 0.418962800898634,
+      "learning_rate": 4.5330398336786526e-06,
+      "loss": 0.5277,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5545080976553057,
+      "grad_norm": 0.46372386433812574,
+      "learning_rate": 4.525081497349887e-06,
+      "loss": 0.5427,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5549915397631133,
+      "grad_norm": 0.4530031866173555,
+      "learning_rate": 4.517124374951086e-06,
+      "loss": 0.5178,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5554749818709209,
+      "grad_norm": 0.440822304311767,
+      "learning_rate": 4.509168486821304e-06,
+      "loss": 0.5225,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5559584239787285,
+      "grad_norm": 0.4127961848352273,
+      "learning_rate": 4.501213853296425e-06,
+      "loss": 0.523,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5564418660865361,
+      "grad_norm": 0.461288293810827,
+      "learning_rate": 4.493260494709141e-06,
+      "loss": 0.5251,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5569253081943437,
+      "grad_norm": 0.43458067348060225,
+      "learning_rate": 4.48530843138887e-06,
+      "loss": 0.5419,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5574087503021513,
+      "grad_norm": 0.42472609884541546,
+      "learning_rate": 4.477357683661734e-06,
+      "loss": 0.5318,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5578921924099589,
+      "grad_norm": 0.4256107384490566,
+      "learning_rate": 4.469408271850479e-06,
+      "loss": 0.506,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5583756345177665,
+      "grad_norm": 0.4098807197744025,
+      "learning_rate": 4.4614602162744455e-06,
+      "loss": 0.5152,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5588590766255741,
+      "grad_norm": 0.4390148331821114,
+      "learning_rate": 4.453513537249503e-06,
+      "loss": 0.527,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5593425187333817,
+      "grad_norm": 0.4711204869094555,
+      "learning_rate": 4.445568255088003e-06,
+      "loss": 0.5247,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5598259608411893,
+      "grad_norm": 0.41653734784347574,
+      "learning_rate": 4.4376243900987296e-06,
+      "loss": 0.5232,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5603094029489969,
+      "grad_norm": 0.44411194102788987,
+      "learning_rate": 4.429681962586839e-06,
+      "loss": 0.5365,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5607928450568045,
+      "grad_norm": 0.44003715852304276,
+      "learning_rate": 4.421740992853818e-06,
+      "loss": 0.5311,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5612762871646121,
+      "grad_norm": 0.43445702749764165,
+      "learning_rate": 4.413801501197424e-06,
+      "loss": 0.5192,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5617597292724196,
+      "grad_norm": 0.4315566299087621,
+      "learning_rate": 4.405863507911638e-06,
+      "loss": 0.5007,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5622431713802272,
+      "grad_norm": 0.3786991552102392,
+      "learning_rate": 4.3979270332866105e-06,
+      "loss": 0.4867,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5627266134880348,
+      "grad_norm": 0.4244074214223604,
+      "learning_rate": 4.389992097608613e-06,
+      "loss": 0.5271,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5632100555958424,
+      "grad_norm": 0.430673286759424,
+      "learning_rate": 4.3820587211599745e-06,
+      "loss": 0.5292,
+      "step": 1165
+    },
+    {
+      "epoch": 0.56369349770365,
+      "grad_norm": 0.46109079230422667,
+      "learning_rate": 4.374126924219052e-06,
+      "loss": 0.5197,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5641769398114576,
+      "grad_norm": 0.4275407630254517,
+      "learning_rate": 4.366196727060152e-06,
+      "loss": 0.5196,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5646603819192652,
+      "grad_norm": 0.4407289737300367,
+      "learning_rate": 4.3582681499535e-06,
+      "loss": 0.5206,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5651438240270727,
+      "grad_norm": 0.4119248969357382,
+      "learning_rate": 4.3503412131651765e-06,
+      "loss": 0.5341,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5656272661348803,
+      "grad_norm": 0.4621359921349015,
+      "learning_rate": 4.342415936957073e-06,
+      "loss": 0.5225,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5661107082426879,
+      "grad_norm": 0.45885127710831636,
+      "learning_rate": 4.334492341586833e-06,
+      "loss": 0.5328,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5665941503504955,
+      "grad_norm": 0.4217251248929531,
+      "learning_rate": 4.326570447307804e-06,
+      "loss": 0.5024,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5670775924583031,
+      "grad_norm": 0.43349297844561585,
+      "learning_rate": 4.318650274368989e-06,
+      "loss": 0.5302,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5675610345661107,
+      "grad_norm": 0.44173213520395654,
+      "learning_rate": 4.310731843014985e-06,
+      "loss": 0.518,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5680444766739183,
+      "grad_norm": 0.44550331806578247,
+      "learning_rate": 4.302815173485944e-06,
+      "loss": 0.5262,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5685279187817259,
+      "grad_norm": 0.41790214183276025,
+      "learning_rate": 4.294900286017509e-06,
+      "loss": 0.5249,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5690113608895335,
+      "grad_norm": 0.4406462454965783,
+      "learning_rate": 4.286987200840772e-06,
+      "loss": 0.5399,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5694948029973411,
+      "grad_norm": 0.42899916006281247,
+      "learning_rate": 4.279075938182214e-06,
+      "loss": 0.522,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5699782451051486,
+      "grad_norm": 0.39425395091405524,
+      "learning_rate": 4.271166518263662e-06,
+      "loss": 0.4916,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5704616872129562,
+      "grad_norm": 0.4325306883401716,
+      "learning_rate": 4.263258961302232e-06,
+      "loss": 0.5297,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5709451293207638,
+      "grad_norm": 0.42524338171666753,
+      "learning_rate": 4.255353287510272e-06,
+      "loss": 0.524,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.44279478264935357,
+      "learning_rate": 4.247449517095329e-06,
+      "loss": 0.5215,
+      "step": 1182
+    },
+    {
+      "epoch": 0.571912013536379,
+      "grad_norm": 0.4197931530174695,
+      "learning_rate": 4.239547670260069e-06,
+      "loss": 0.5099,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5723954556441866,
+      "grad_norm": 0.4310563818067479,
+      "learning_rate": 4.231647767202254e-06,
+      "loss": 0.5191,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5728788977519942,
+      "grad_norm": 0.42356166524796646,
+      "learning_rate": 4.223749828114672e-06,
+      "loss": 0.528,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5733623398598018,
+      "grad_norm": 0.4095445643557658,
+      "learning_rate": 4.215853873185093e-06,
+      "loss": 0.4862,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5738457819676094,
+      "grad_norm": 0.47015761976525233,
+      "learning_rate": 4.2079599225962115e-06,
+      "loss": 0.5183,
+      "step": 1187
+    },
+    {
+      "epoch": 0.574329224075417,
+      "grad_norm": 0.4175313339147143,
+      "learning_rate": 4.2000679965256045e-06,
+      "loss": 0.5045,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5748126661832246,
+      "grad_norm": 0.43952093100735207,
+      "learning_rate": 4.192178115145668e-06,
+      "loss": 0.4942,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5752961082910322,
+      "grad_norm": 0.41782905847478197,
+      "learning_rate": 4.184290298623578e-06,
+      "loss": 0.4962,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5757795503988398,
+      "grad_norm": 0.4321674737064156,
+      "learning_rate": 4.176404567121225e-06,
+      "loss": 0.5397,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5762629925066474,
+      "grad_norm": 0.4802929727845466,
+      "learning_rate": 4.16852094079518e-06,
+      "loss": 0.5249,
+      "step": 1192
+    },
+    {
+      "epoch": 0.576746434614455,
+      "grad_norm": 0.4829071925445664,
+      "learning_rate": 4.160639439796624e-06,
+      "loss": 0.5192,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5772298767222626,
+      "grad_norm": 0.41462388567526964,
+      "learning_rate": 4.152760084271305e-06,
+      "loss": 0.5224,
+      "step": 1194
+    },
+    {
+      "epoch": 0.57771331883007,
+      "grad_norm": 0.41336785715866026,
+      "learning_rate": 4.1448828943595e-06,
+      "loss": 0.531,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5781967609378776,
+      "grad_norm": 0.4424735299028098,
+      "learning_rate": 4.1370078901959306e-06,
+      "loss": 0.5309,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5786802030456852,
+      "grad_norm": 0.43755728178912967,
+      "learning_rate": 4.129135091909752e-06,
+      "loss": 0.5314,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5791636451534928,
+      "grad_norm": 0.4234445834897067,
+      "learning_rate": 4.121264519624463e-06,
+      "loss": 0.4971,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5796470872613004,
+      "grad_norm": 0.4465261505400547,
+      "learning_rate": 4.113396193457887e-06,
+      "loss": 0.5421,
+      "step": 1199
+    },
+    {
+      "epoch": 0.580130529369108,
+      "grad_norm": 0.43864096822609294,
+      "learning_rate": 4.105530133522096e-06,
+      "loss": 0.5286,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5806139714769156,
+      "grad_norm": 0.44812939972631666,
+      "learning_rate": 4.0976663599233745e-06,
+      "loss": 0.5283,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5810974135847232,
+      "grad_norm": 0.41015688502886277,
+      "learning_rate": 4.08980489276216e-06,
+      "loss": 0.5031,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5815808556925308,
+      "grad_norm": 0.4371592707730246,
+      "learning_rate": 4.081945752133e-06,
+      "loss": 0.5247,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5820642978003384,
+      "grad_norm": 0.4545452075633195,
+      "learning_rate": 4.074088958124488e-06,
+      "loss": 0.5233,
+      "step": 1204
+    },
+    {
+      "epoch": 0.582547739908146,
+      "grad_norm": 0.4034341944937049,
+      "learning_rate": 4.066234530819222e-06,
+      "loss": 0.4997,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5830311820159536,
+      "grad_norm": 0.44158682076629885,
+      "learning_rate": 4.058382490293755e-06,
+      "loss": 0.5327,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5835146241237612,
+      "grad_norm": 0.44126859021633347,
+      "learning_rate": 4.050532856618532e-06,
+      "loss": 0.5172,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5839980662315688,
+      "grad_norm": 0.44983862546632325,
+      "learning_rate": 4.0426856498578515e-06,
+      "loss": 0.5321,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5844815083393764,
+      "grad_norm": 0.4306823960936312,
+      "learning_rate": 4.034840890069805e-06,
+      "loss": 0.5286,
+      "step": 1209
+    },
+    {
+      "epoch": 0.584964950447184,
+      "grad_norm": 0.46647403495000134,
+      "learning_rate": 4.0269985973062325e-06,
+      "loss": 0.5205,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5854483925549916,
+      "grad_norm": 0.43512362259260234,
+      "learning_rate": 4.019158791612662e-06,
+      "loss": 0.5186,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5859318346627991,
+      "grad_norm": 0.42474616162914414,
+      "learning_rate": 4.0113214930282765e-06,
+      "loss": 0.5019,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5864152767706067,
+      "grad_norm": 0.4496255266024326,
+      "learning_rate": 4.003486721585834e-06,
+      "loss": 0.5292,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5868987188784143,
+      "grad_norm": 0.4692064186214037,
+      "learning_rate": 3.995654497311649e-06,
+      "loss": 0.524,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5873821609862219,
+      "grad_norm": 0.47446797627391324,
+      "learning_rate": 3.987824840225512e-06,
+      "loss": 0.5296,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5878656030940295,
+      "grad_norm": 0.45351628397562704,
+      "learning_rate": 3.979997770340664e-06,
+      "loss": 0.5191,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5883490452018371,
+      "grad_norm": 0.4469658212338775,
+      "learning_rate": 3.972173307663721e-06,
+      "loss": 0.5259,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5888324873096447,
+      "grad_norm": 0.42106841746243373,
+      "learning_rate": 3.964351472194642e-06,
+      "loss": 0.5169,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5893159294174523,
+      "grad_norm": 0.46081869250650287,
+      "learning_rate": 3.95653228392667e-06,
+      "loss": 0.5324,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5897993715252599,
+      "grad_norm": 0.43787433991965447,
+      "learning_rate": 3.9487157628462784e-06,
+      "loss": 0.5253,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5902828136330674,
+      "grad_norm": 0.4446710684339124,
+      "learning_rate": 3.940901928933127e-06,
+      "loss": 0.5207,
+      "step": 1221
+    },
+    {
+      "epoch": 0.590766255740875,
+      "grad_norm": 0.4878921355151572,
+      "learning_rate": 3.933090802160004e-06,
+      "loss": 0.5216,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5912496978486826,
+      "grad_norm": 0.43014793362728476,
+      "learning_rate": 3.925282402492779e-06,
+      "loss": 0.5158,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5917331399564902,
+      "grad_norm": 0.4785437280134671,
+      "learning_rate": 3.917476749890351e-06,
+      "loss": 0.5337,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5922165820642978,
+      "grad_norm": 0.4374190045344671,
+      "learning_rate": 3.909673864304597e-06,
+      "loss": 0.528,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5927000241721054,
+      "grad_norm": 0.4547685840805692,
+      "learning_rate": 3.901873765680322e-06,
+      "loss": 0.5366,
+      "step": 1226
+    },
+    {
+      "epoch": 0.593183466279913,
+      "grad_norm": 0.4594395711536057,
+      "learning_rate": 3.894076473955207e-06,
+      "loss": 0.5202,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5936669083877205,
+      "grad_norm": 0.5197830445285274,
+      "learning_rate": 3.886282009059757e-06,
+      "loss": 0.5293,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5941503504955281,
+      "grad_norm": 0.438870970435199,
+      "learning_rate": 3.878490390917253e-06,
+      "loss": 0.521,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5946337926033357,
+      "grad_norm": 0.43955571970066076,
+      "learning_rate": 3.8707016394436985e-06,
+      "loss": 0.5219,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5951172347111433,
+      "grad_norm": 0.4319400896121196,
+      "learning_rate": 3.86291577454777e-06,
+      "loss": 0.5274,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5956006768189509,
+      "grad_norm": 0.42866945423374303,
+      "learning_rate": 3.855132816130767e-06,
+      "loss": 0.5287,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5960841189267585,
+      "grad_norm": 0.4905566924544953,
+      "learning_rate": 3.847352784086556e-06,
+      "loss": 0.5214,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5965675610345661,
+      "grad_norm": 0.4815238783873125,
+      "learning_rate": 3.839575698301529e-06,
+      "loss": 0.5348,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5970510031423737,
+      "grad_norm": 0.4611545386580029,
+      "learning_rate": 3.831801578654541e-06,
+      "loss": 0.521,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5975344452501813,
+      "grad_norm": 0.44138236799554953,
+      "learning_rate": 3.8240304450168716e-06,
+      "loss": 0.5012,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5980178873579889,
+      "grad_norm": 0.4729557809805851,
+      "learning_rate": 3.8162623172521615e-06,
+      "loss": 0.5239,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5985013294657965,
+      "grad_norm": 0.47751713396917145,
+      "learning_rate": 3.808497215216374e-06,
+      "loss": 0.518,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5989847715736041,
+      "grad_norm": 0.4632727328684821,
+      "learning_rate": 3.8007351587577342e-06,
+      "loss": 0.5212,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5994682136814117,
+      "grad_norm": 0.43452448731632676,
+      "learning_rate": 3.7929761677166847e-06,
+      "loss": 0.5256,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5999516557892193,
+      "grad_norm": 0.41707858133064807,
+      "learning_rate": 3.7852202619258327e-06,
+      "loss": 0.5258,
+      "step": 1241
+    },
+    {
+      "epoch": 0.6004350978970269,
+      "grad_norm": 0.42484738876263983,
+      "learning_rate": 3.777467461209895e-06,
+      "loss": 0.5226,
+      "step": 1242
+    },
+    {
+      "epoch": 0.6009185400048345,
+      "grad_norm": 0.46278022630826876,
+      "learning_rate": 3.76971778538566e-06,
+      "loss": 0.5265,
+      "step": 1243
+    },
+    {
+      "epoch": 0.6014019821126421,
+      "grad_norm": 0.4136178806145892,
+      "learning_rate": 3.76197125426192e-06,
+      "loss": 0.521,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6018854242204495,
+      "grad_norm": 0.42100500706131366,
+      "learning_rate": 3.754227887639434e-06,
+      "loss": 0.5119,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6023688663282571,
+      "grad_norm": 0.4123940096542578,
+      "learning_rate": 3.7464877053108706e-06,
+      "loss": 0.5258,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6028523084360647,
+      "grad_norm": 0.4195138029502561,
+      "learning_rate": 3.7387507270607617e-06,
+      "loss": 0.529,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6033357505438723,
+      "grad_norm": 0.45482061749805036,
+      "learning_rate": 3.7310169726654444e-06,
+      "loss": 0.528,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6038191926516799,
+      "grad_norm": 0.4371974850319641,
+      "learning_rate": 3.7232864618930217e-06,
+      "loss": 0.5182,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6043026347594875,
+      "grad_norm": 0.43099991632770085,
+      "learning_rate": 3.715559214503298e-06,
+      "loss": 0.5133,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6047860768672951,
+      "grad_norm": 0.4048318396541149,
+      "learning_rate": 3.707835250247745e-06,
+      "loss": 0.4877,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6052695189751027,
+      "grad_norm": 0.43040118158566426,
+      "learning_rate": 3.7001145888694335e-06,
+      "loss": 0.5256,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6057529610829103,
+      "grad_norm": 0.43151561170744174,
+      "learning_rate": 3.6923972501029996e-06,
+      "loss": 0.5028,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6062364031907179,
+      "grad_norm": 0.4168861988285146,
+      "learning_rate": 3.684683253674583e-06,
+      "loss": 0.5249,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6067198452985255,
+      "grad_norm": 0.4341349343804055,
+      "learning_rate": 3.676972619301776e-06,
+      "loss": 0.5119,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6072032874063331,
+      "grad_norm": 0.45716952276328937,
+      "learning_rate": 3.6692653666935875e-06,
+      "loss": 0.5262,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6076867295141407,
+      "grad_norm": 0.4304219021067875,
+      "learning_rate": 3.6615615155503703e-06,
+      "loss": 0.5168,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6081701716219483,
+      "grad_norm": 0.4098861167543977,
+      "learning_rate": 3.6538610855637953e-06,
+      "loss": 0.5193,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6086536137297559,
+      "grad_norm": 0.45344600794993284,
+      "learning_rate": 3.6461640964167755e-06,
+      "loss": 0.5213,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6091370558375635,
+      "grad_norm": 0.408785994340013,
+      "learning_rate": 3.638470567783442e-06,
+      "loss": 0.4982,
+      "step": 1260
+    },
+    {
+      "epoch": 0.609620497945371,
+      "grad_norm": 0.4481815917202834,
+      "learning_rate": 3.630780519329069e-06,
+      "loss": 0.5329,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6101039400531786,
+      "grad_norm": 0.4271346926738626,
+      "learning_rate": 3.623093970710043e-06,
+      "loss": 0.5278,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6105873821609862,
+      "grad_norm": 0.4443519643294819,
+      "learning_rate": 3.615410941573799e-06,
+      "loss": 0.5358,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6110708242687938,
+      "grad_norm": 0.41093410035963396,
+      "learning_rate": 3.607731451558783e-06,
+      "loss": 0.4973,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6115542663766014,
+      "grad_norm": 0.4526917508007144,
+      "learning_rate": 3.6000555202943872e-06,
+      "loss": 0.5223,
+      "step": 1265
+    },
+    {
+      "epoch": 0.612037708484409,
+      "grad_norm": 0.458600770079082,
+      "learning_rate": 3.59238316740091e-06,
+      "loss": 0.5206,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6125211505922166,
+      "grad_norm": 0.4390522720994981,
+      "learning_rate": 3.584714412489506e-06,
+      "loss": 0.5306,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6130045927000242,
+      "grad_norm": 0.41799452227515504,
+      "learning_rate": 3.5770492751621292e-06,
+      "loss": 0.5029,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6134880348078318,
+      "grad_norm": 0.43636835439611227,
+      "learning_rate": 3.5693877750114903e-06,
+      "loss": 0.5167,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6139714769156394,
+      "grad_norm": 0.43211619164840076,
+      "learning_rate": 3.5617299316209984e-06,
+      "loss": 0.5049,
+      "step": 1270
+    },
+    {
+      "epoch": 0.614454919023447,
+      "grad_norm": 0.42547953690955836,
+      "learning_rate": 3.5540757645647217e-06,
+      "loss": 0.4939,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6149383611312546,
+      "grad_norm": 0.4170182860561763,
+      "learning_rate": 3.546425293407324e-06,
+      "loss": 0.5199,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6154218032390621,
+      "grad_norm": 0.4222424350681242,
+      "learning_rate": 3.5387785377040316e-06,
+      "loss": 0.5132,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6159052453468697,
+      "grad_norm": 0.4552864030500758,
+      "learning_rate": 3.531135517000561e-06,
+      "loss": 0.5269,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6163886874546773,
+      "grad_norm": 0.43855551812482985,
+      "learning_rate": 3.523496250833098e-06,
+      "loss": 0.5122,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6168721295624849,
+      "grad_norm": 0.46449768544610603,
+      "learning_rate": 3.515860758728214e-06,
+      "loss": 0.5234,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6173555716702925,
+      "grad_norm": 0.46363731713711515,
+      "learning_rate": 3.5082290602028492e-06,
+      "loss": 0.5269,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6178390137781,
+      "grad_norm": 0.42830833200680596,
+      "learning_rate": 3.5006011747642366e-06,
+      "loss": 0.5177,
+      "step": 1278
+    },
+    {
+      "epoch": 0.6183224558859076,
+      "grad_norm": 0.45403297568672957,
+      "learning_rate": 3.492977121909865e-06,
+      "loss": 0.5329,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6188058979937152,
+      "grad_norm": 0.4707193517755616,
+      "learning_rate": 3.4853569211274306e-06,
+      "loss": 0.5275,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6192893401015228,
+      "grad_norm": 0.4382585499371175,
+      "learning_rate": 3.4777405918947795e-06,
+      "loss": 0.5117,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6197727822093304,
+      "grad_norm": 0.43247243116721396,
+      "learning_rate": 3.4701281536798638e-06,
+      "loss": 0.5274,
+      "step": 1282
+    },
+    {
+      "epoch": 0.620256224317138,
+      "grad_norm": 0.4457194202455219,
+      "learning_rate": 3.462519625940688e-06,
+      "loss": 0.5282,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6207396664249456,
+      "grad_norm": 0.45208181214242377,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.5224,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6212231085327532,
+      "grad_norm": 0.43419084686544124,
+      "learning_rate": 3.4473143796715537e-06,
+      "loss": 0.5221,
+      "step": 1285
+    },
+    {
+      "epoch": 0.6217065506405608,
+      "grad_norm": 0.45834613229050314,
+      "learning_rate": 3.4397177000074307e-06,
+      "loss": 0.5286,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6221899927483684,
+      "grad_norm": 0.4551713196802731,
+      "learning_rate": 3.4321250085506174e-06,
+      "loss": 0.519,
+      "step": 1287
+    },
+    {
+      "epoch": 0.622673434856176,
+      "grad_norm": 0.45716705644935435,
+      "learning_rate": 3.4245363247086477e-06,
+      "loss": 0.5291,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6231568769639836,
+      "grad_norm": 0.4214582611290155,
+      "learning_rate": 3.4169516678788096e-06,
+      "loss": 0.5084,
+      "step": 1289
+    },
+    {
+      "epoch": 0.6236403190717912,
+      "grad_norm": 0.43304150590207136,
+      "learning_rate": 3.4093710574480926e-06,
+      "loss": 0.5181,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6241237611795988,
+      "grad_norm": 0.4201723521255349,
+      "learning_rate": 3.4017945127931517e-06,
+      "loss": 0.5215,
+      "step": 1291
+    },
+    {
+      "epoch": 0.6246072032874064,
+      "grad_norm": 0.43513341637312203,
+      "learning_rate": 3.394222053280245e-06,
+      "loss": 0.5219,
+      "step": 1292
+    },
+    {
+      "epoch": 0.625090645395214,
+      "grad_norm": 0.41778354176218346,
+      "learning_rate": 3.386653698265189e-06,
+      "loss": 0.5295,
+      "step": 1293
+    },
+    {
+      "epoch": 0.6255740875030215,
+      "grad_norm": 0.4289158469228602,
+      "learning_rate": 3.3790894670933096e-06,
+      "loss": 0.4993,
+      "step": 1294
+    },
+    {
+      "epoch": 0.626057529610829,
+      "grad_norm": 0.4362452363722638,
+      "learning_rate": 3.3715293790993906e-06,
+      "loss": 0.5212,
+      "step": 1295
+    },
+    {
+      "epoch": 0.6265409717186367,
+      "grad_norm": 0.4287020140556269,
+      "learning_rate": 3.3639734536076263e-06,
+      "loss": 0.5145,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6270244138264442,
+      "grad_norm": 0.4371619386426176,
+      "learning_rate": 3.356421709931573e-06,
+      "loss": 0.5263,
+      "step": 1297
+    },
+    {
+      "epoch": 0.6275078559342518,
+      "grad_norm": 0.4558196176944881,
+      "learning_rate": 3.348874167374093e-06,
+      "loss": 0.5193,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6279912980420594,
+      "grad_norm": 0.433400379634257,
+      "learning_rate": 3.341330845227316e-06,
+      "loss": 0.5342,
+      "step": 1299
+    },
+    {
+      "epoch": 0.628474740149867,
+      "grad_norm": 0.44874534866829735,
+      "learning_rate": 3.33379176277258e-06,
+      "loss": 0.5192,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6289581822576746,
+      "grad_norm": 0.42814892447345076,
+      "learning_rate": 3.326256939280389e-06,
+      "loss": 0.5222,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6294416243654822,
+      "grad_norm": 0.4175511142456795,
+      "learning_rate": 3.3187263940103587e-06,
+      "loss": 0.5008,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6299250664732898,
+      "grad_norm": 0.43012465279292367,
+      "learning_rate": 3.3112001462111666e-06,
+      "loss": 0.5252,
+      "step": 1303
+    },
+    {
+      "epoch": 0.6304085085810974,
+      "grad_norm": 0.4343943982087317,
+      "learning_rate": 3.3036782151205134e-06,
+      "loss": 0.5269,
+      "step": 1304
+    },
+    {
+      "epoch": 0.630891950688905,
+      "grad_norm": 0.4399879585469454,
+      "learning_rate": 3.296160619965056e-06,
+      "loss": 0.5348,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6313753927967126,
+      "grad_norm": 0.41469773355380823,
+      "learning_rate": 3.2886473799603793e-06,
+      "loss": 0.5315,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6318588349045202,
+      "grad_norm": 0.4118723735223055,
+      "learning_rate": 3.2811385143109254e-06,
+      "loss": 0.5199,
+      "step": 1307
+    },
+    {
+      "epoch": 0.6323422770123278,
+      "grad_norm": 0.42911493217824326,
+      "learning_rate": 3.2736340422099633e-06,
+      "loss": 0.5287,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6328257191201354,
+      "grad_norm": 0.42229266051589764,
+      "learning_rate": 3.2661339828395263e-06,
+      "loss": 0.5034,
+      "step": 1309
+    },
+    {
+      "epoch": 0.633309161227943,
+      "grad_norm": 0.4171629681078508,
+      "learning_rate": 3.2586383553703723e-06,
+      "loss": 0.5169,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6337926033357505,
+      "grad_norm": 0.4335972937122555,
+      "learning_rate": 3.2511471789619274e-06,
+      "loss": 0.5084,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6342760454435581,
+      "grad_norm": 0.42465524204681926,
+      "learning_rate": 3.2436604727622447e-06,
+      "loss": 0.5126,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6347594875513657,
+      "grad_norm": 0.4075010185296818,
+      "learning_rate": 3.2361782559079465e-06,
+      "loss": 0.5158,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6352429296591733,
+      "grad_norm": 0.4308541178278502,
+      "learning_rate": 3.228700547524184e-06,
+      "loss": 0.5145,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6357263717669809,
+      "grad_norm": 0.428226473042128,
+      "learning_rate": 3.221227366724581e-06,
+      "loss": 0.5146,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6362098138747885,
+      "grad_norm": 0.4299253923140349,
+      "learning_rate": 3.2137587326111896e-06,
+      "loss": 0.5207,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6366932559825961,
+      "grad_norm": 0.4181473728080887,
+      "learning_rate": 3.206294664274443e-06,
+      "loss": 0.5268,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6371766980904037,
+      "grad_norm": 0.4474913351660176,
+      "learning_rate": 3.198835180793097e-06,
+      "loss": 0.5277,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6376601401982113,
+      "grad_norm": 0.432676302059824,
+      "learning_rate": 3.1913803012341987e-06,
+      "loss": 0.5195,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6381435823060189,
+      "grad_norm": 0.44131963361999216,
+      "learning_rate": 3.183930044653014e-06,
+      "loss": 0.5157,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6386270244138265,
+      "grad_norm": 0.4300971230394046,
+      "learning_rate": 3.176484430093007e-06,
+      "loss": 0.5312,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6391104665216341,
+      "grad_norm": 0.44998544162450493,
+      "learning_rate": 3.1690434765857604e-06,
+      "loss": 0.523,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6395939086294417,
+      "grad_norm": 0.4472778564601584,
+      "learning_rate": 3.1616072031509594e-06,
+      "loss": 0.5155,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6400773507372493,
+      "grad_norm": 0.4278886331070072,
+      "learning_rate": 3.154175628796311e-06,
+      "loss": 0.5214,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6405607928450568,
+      "grad_norm": 0.4354238172365612,
+      "learning_rate": 3.146748772517523e-06,
+      "loss": 0.5202,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6410442349528644,
+      "grad_norm": 0.48506314375368076,
+      "learning_rate": 3.139326653298236e-06,
+      "loss": 0.5226,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6415276770606719,
+      "grad_norm": 0.42689588677759766,
+      "learning_rate": 3.1319092901099847e-06,
+      "loss": 0.5192,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6420111191684795,
+      "grad_norm": 0.44087925863432936,
+      "learning_rate": 3.1244967019121496e-06,
+      "loss": 0.5127,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6424945612762871,
+      "grad_norm": 0.42605880529211515,
+      "learning_rate": 3.117088907651902e-06,
+      "loss": 0.5176,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6429780033840947,
+      "grad_norm": 0.42151112378627853,
+      "learning_rate": 3.109685926264161e-06,
+      "loss": 0.5165,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6434614454919023,
+      "grad_norm": 0.4268979096606091,
+      "learning_rate": 3.102287776671544e-06,
+      "loss": 0.5339,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6439448875997099,
+      "grad_norm": 0.4392831430463705,
+      "learning_rate": 3.094894477784318e-06,
+      "loss": 0.5166,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6444283297075175,
+      "grad_norm": 0.4146076328411298,
+      "learning_rate": 3.0875060485003496e-06,
+      "loss": 0.5274,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6449117718153251,
+      "grad_norm": 0.42157017978455763,
+      "learning_rate": 3.080122507705062e-06,
+      "loss": 0.5243,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6453952139231327,
+      "grad_norm": 0.41448886842899935,
+      "learning_rate": 3.0727438742713766e-06,
+      "loss": 0.4982,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6458786560309403,
+      "grad_norm": 0.4311221980804021,
+      "learning_rate": 3.0653701670596805e-06,
+      "loss": 0.5222,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6463620981387479,
+      "grad_norm": 0.42860321112689415,
+      "learning_rate": 3.0580014049177566e-06,
+      "loss": 0.5203,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6468455402465555,
+      "grad_norm": 0.42363404336164073,
+      "learning_rate": 3.0506376066807632e-06,
+      "loss": 0.5131,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6473289823543631,
+      "grad_norm": 0.4352925336415172,
+      "learning_rate": 3.0432787911711553e-06,
+      "loss": 0.5244,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6478124244621707,
+      "grad_norm": 0.4213888881156135,
+      "learning_rate": 3.0359249771986605e-06,
+      "loss": 0.5114,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6482958665699783,
+      "grad_norm": 0.44310700039649703,
+      "learning_rate": 3.028576183560221e-06,
+      "loss": 0.5345,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6487793086777859,
+      "grad_norm": 0.46384351518694394,
+      "learning_rate": 3.021232429039944e-06,
+      "loss": 0.5103,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6492627507855935,
+      "grad_norm": 0.420074285156663,
+      "learning_rate": 3.01389373240906e-06,
+      "loss": 0.5307,
+      "step": 1343
+    },
+    {
+      "epoch": 0.649746192893401,
+      "grad_norm": 0.46717936599561755,
+      "learning_rate": 3.006560112425867e-06,
+      "loss": 0.5146,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6502296350012086,
+      "grad_norm": 0.4411181920960548,
+      "learning_rate": 2.999231587835691e-06,
+      "loss": 0.5113,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6507130771090162,
+      "grad_norm": 0.4353633086273451,
+      "learning_rate": 2.9919081773708293e-06,
+      "loss": 0.5195,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6511965192168238,
+      "grad_norm": 0.4555577360086876,
+      "learning_rate": 2.9845898997505102e-06,
+      "loss": 0.5201,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6516799613246314,
+      "grad_norm": 0.43234112941806857,
+      "learning_rate": 2.9772767736808406e-06,
+      "loss": 0.5194,
+      "step": 1348
+    },
+    {
+      "epoch": 0.652163403432439,
+      "grad_norm": 0.43499867909496204,
+      "learning_rate": 2.9699688178547615e-06,
+      "loss": 0.5252,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6526468455402465,
+      "grad_norm": 0.4243771527145887,
+      "learning_rate": 2.962666050951997e-06,
+      "loss": 0.5122,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6531302876480541,
+      "grad_norm": 0.4158873835022681,
+      "learning_rate": 2.9553684916390053e-06,
+      "loss": 0.5092,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6536137297558617,
+      "grad_norm": 0.4494940146550669,
+      "learning_rate": 2.948076158568939e-06,
+      "loss": 0.5256,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6540971718636693,
+      "grad_norm": 0.39860657725442444,
+      "learning_rate": 2.940789070381587e-06,
+      "loss": 0.5001,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6545806139714769,
+      "grad_norm": 0.41241371956595596,
+      "learning_rate": 2.933507245703335e-06,
+      "loss": 0.5174,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6550640560792845,
+      "grad_norm": 0.41370768103275585,
+      "learning_rate": 2.9262307031471132e-06,
+      "loss": 0.5142,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6555474981870921,
+      "grad_norm": 0.4457235721009106,
+      "learning_rate": 2.918959461312353e-06,
+      "loss": 0.5212,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6560309402948997,
+      "grad_norm": 0.44189033261031596,
+      "learning_rate": 2.911693538784931e-06,
+      "loss": 0.5315,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6565143824027073,
+      "grad_norm": 0.417441027459776,
+      "learning_rate": 2.904432954137136e-06,
+      "loss": 0.5197,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6569978245105149,
+      "grad_norm": 0.41446617019595194,
+      "learning_rate": 2.897177725927599e-06,
+      "loss": 0.4977,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6574812666183224,
+      "grad_norm": 0.424230216094722,
+      "learning_rate": 2.889927872701278e-06,
+      "loss": 0.5319,
+      "step": 1360
+    },
+    {
+      "epoch": 0.65796470872613,
+      "grad_norm": 0.42985260891150956,
+      "learning_rate": 2.8826834129893755e-06,
+      "loss": 0.5166,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6584481508339376,
+      "grad_norm": 0.38447115699767576,
+      "learning_rate": 2.8754443653093186e-06,
+      "loss": 0.4786,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6589315929417452,
+      "grad_norm": 0.4044201835297723,
+      "learning_rate": 2.8682107481646915e-06,
+      "loss": 0.5216,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6594150350495528,
+      "grad_norm": 0.4022645350617216,
+      "learning_rate": 2.8609825800452063e-06,
+      "loss": 0.4988,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6598984771573604,
+      "grad_norm": 0.4401229027571195,
+      "learning_rate": 2.853759879426644e-06,
+      "loss": 0.5181,
+      "step": 1365
+    },
+    {
+      "epoch": 0.660381919265168,
+      "grad_norm": 0.42349548120906483,
+      "learning_rate": 2.8465426647708067e-06,
+      "loss": 0.5163,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6608653613729756,
+      "grad_norm": 0.40013997451662586,
+      "learning_rate": 2.8393309545254776e-06,
+      "loss": 0.5214,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6613488034807832,
+      "grad_norm": 0.4359244125864156,
+      "learning_rate": 2.8321247671243695e-06,
+      "loss": 0.5179,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6618322455885908,
+      "grad_norm": 0.41425895295471055,
+      "learning_rate": 2.82492412098708e-06,
+      "loss": 0.5081,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6623156876963984,
+      "grad_norm": 0.4210065663342879,
+      "learning_rate": 2.8177290345190387e-06,
+      "loss": 0.5194,
+      "step": 1370
+    },
+    {
+      "epoch": 0.662799129804206,
+      "grad_norm": 0.4028980901393777,
+      "learning_rate": 2.8105395261114666e-06,
+      "loss": 0.5234,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6632825719120136,
+      "grad_norm": 0.4325922757476261,
+      "learning_rate": 2.803355614141327e-06,
+      "loss": 0.5188,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6637660140198212,
+      "grad_norm": 0.4308186918740408,
+      "learning_rate": 2.7961773169712803e-06,
+      "loss": 0.5125,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6642494561276288,
+      "grad_norm": 0.4211885259856405,
+      "learning_rate": 2.7890046529496284e-06,
+      "loss": 0.5233,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6647328982354364,
+      "grad_norm": 0.4304676159038956,
+      "learning_rate": 2.7818376404102832e-06,
+      "loss": 0.5188,
+      "step": 1375
+    },
+    {
+      "epoch": 0.665216340343244,
+      "grad_norm": 0.4137521174014562,
+      "learning_rate": 2.774676297672701e-06,
+      "loss": 0.5248,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6656997824510514,
+      "grad_norm": 0.4389331875357886,
+      "learning_rate": 2.7675206430418542e-06,
+      "loss": 0.5265,
+      "step": 1377
+    },
+    {
+      "epoch": 0.666183224558859,
+      "grad_norm": 0.46429330512304384,
+      "learning_rate": 2.7603706948081745e-06,
+      "loss": 0.5211,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.4260734411731187,
+      "learning_rate": 2.753226471247501e-06,
+      "loss": 0.517,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6671501087744742,
+      "grad_norm": 0.4189810127916622,
+      "learning_rate": 2.7460879906210485e-06,
+      "loss": 0.5107,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6676335508822818,
+      "grad_norm": 0.41204910620329505,
+      "learning_rate": 2.7389552711753477e-06,
+      "loss": 0.5191,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6681169929900894,
+      "grad_norm": 0.4267680612975131,
+      "learning_rate": 2.731828331142207e-06,
+      "loss": 0.5128,
+      "step": 1382
+    },
+    {
+      "epoch": 0.668600435097897,
+      "grad_norm": 0.42901984315752384,
+      "learning_rate": 2.7247071887386544e-06,
+      "loss": 0.5257,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6690838772057046,
+      "grad_norm": 0.4146728225846163,
+      "learning_rate": 2.7175918621669074e-06,
+      "loss": 0.5184,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6695673193135122,
+      "grad_norm": 0.4782489091382579,
+      "learning_rate": 2.7104823696143136e-06,
+      "loss": 0.5298,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6700507614213198,
+      "grad_norm": 0.40703765978893935,
+      "learning_rate": 2.70337872925331e-06,
+      "loss": 0.5111,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6705342035291274,
+      "grad_norm": 0.4109547447766556,
+      "learning_rate": 2.6962809592413726e-06,
+      "loss": 0.5002,
+      "step": 1387
+    },
+    {
+      "epoch": 0.671017645636935,
+      "grad_norm": 0.4498968198632276,
+      "learning_rate": 2.6891890777209696e-06,
+      "loss": 0.5256,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6715010877447426,
+      "grad_norm": 0.42208190857564254,
+      "learning_rate": 2.68210310281953e-06,
+      "loss": 0.5193,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6719845298525502,
+      "grad_norm": 0.41822528698390377,
+      "learning_rate": 2.67502305264937e-06,
+      "loss": 0.5163,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6724679719603578,
+      "grad_norm": 0.4218034674050614,
+      "learning_rate": 2.667948945307674e-06,
+      "loss": 0.5174,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6729514140681654,
+      "grad_norm": 0.4206471334382422,
+      "learning_rate": 2.6608807988764252e-06,
+      "loss": 0.4936,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6734348561759729,
+      "grad_norm": 0.42181885072694014,
+      "learning_rate": 2.653818631422378e-06,
+      "loss": 0.5138,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6739182982837805,
+      "grad_norm": 0.4261589725068296,
+      "learning_rate": 2.6467624609970005e-06,
+      "loss": 0.5145,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6744017403915881,
+      "grad_norm": 0.40519700853309554,
+      "learning_rate": 2.6397123056364364e-06,
+      "loss": 0.5013,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6748851824993957,
+      "grad_norm": 0.4510436140721377,
+      "learning_rate": 2.6326681833614464e-06,
+      "loss": 0.5184,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6753686246072033,
+      "grad_norm": 0.45003681113297744,
+      "learning_rate": 2.6256301121773775e-06,
+      "loss": 0.5149,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6758520667150109,
+      "grad_norm": 0.4263810181960221,
+      "learning_rate": 2.618598110074105e-06,
+      "loss": 0.5115,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6763355088228185,
+      "grad_norm": 0.4263005297393967,
+      "learning_rate": 2.6115721950259977e-06,
+      "loss": 0.5243,
+      "step": 1399
+    },
+    {
+      "epoch": 0.676818950930626,
+      "grad_norm": 0.4563967376255983,
+      "learning_rate": 2.6045523849918553e-06,
+      "loss": 0.5314,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6773023930384336,
+      "grad_norm": 0.4478593986013541,
+      "learning_rate": 2.5975386979148792e-06,
+      "loss": 0.5179,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6777858351462412,
+      "grad_norm": 0.40707261007936574,
+      "learning_rate": 2.590531151722622e-06,
+      "loss": 0.5165,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6782692772540488,
+      "grad_norm": 0.45689999209163507,
+      "learning_rate": 2.5835297643269326e-06,
+      "loss": 0.5212,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6787527193618564,
+      "grad_norm": 0.41521250872284,
+      "learning_rate": 2.576534553623925e-06,
+      "loss": 0.5197,
+      "step": 1404
+    },
+    {
+      "epoch": 0.679236161469664,
+      "grad_norm": 0.3969654860159799,
+      "learning_rate": 2.5695455374939147e-06,
+      "loss": 0.4939,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6797196035774716,
+      "grad_norm": 0.4115250925249713,
+      "learning_rate": 2.5625627338014004e-06,
+      "loss": 0.5242,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6802030456852792,
+      "grad_norm": 0.4253454941567133,
+      "learning_rate": 2.5555861603949832e-06,
+      "loss": 0.513,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6806864877930868,
+      "grad_norm": 0.4448844424181978,
+      "learning_rate": 2.548615835107352e-06,
+      "loss": 0.5047,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6811699299008944,
+      "grad_norm": 0.41222858577096244,
+      "learning_rate": 2.5416517757552157e-06,
+      "loss": 0.5286,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6816533720087019,
+      "grad_norm": 0.42285086542458045,
+      "learning_rate": 2.534694000139273e-06,
+      "loss": 0.5169,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6821368141165095,
+      "grad_norm": 0.4122433378845125,
+      "learning_rate": 2.5277425260441616e-06,
+      "loss": 0.515,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6826202562243171,
+      "grad_norm": 0.4362061175188878,
+      "learning_rate": 2.520797371238406e-06,
+      "loss": 0.5225,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6831036983321247,
+      "grad_norm": 0.4411789430289944,
+      "learning_rate": 2.513858553474382e-06,
+      "loss": 0.5191,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6835871404399323,
+      "grad_norm": 0.4415744443134195,
+      "learning_rate": 2.506926090488269e-06,
+      "loss": 0.5306,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6840705825477399,
+      "grad_norm": 0.4477316137829116,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.5248,
+      "step": 1415
+    },
+    {
+      "epoch": 0.6845540246555475,
+      "grad_norm": 0.41099572818531255,
+      "learning_rate": 2.4930802997132213e-06,
+      "loss": 0.5218,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6850374667633551,
+      "grad_norm": 0.4009913145578469,
+      "learning_rate": 2.486167007315243e-06,
+      "loss": 0.5189,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6855209088711627,
+      "grad_norm": 0.43880257019064667,
+      "learning_rate": 2.479260140476999e-06,
+      "loss": 0.5114,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6860043509789703,
+      "grad_norm": 0.4322007294880164,
+      "learning_rate": 2.4723597168529984e-06,
+      "loss": 0.5066,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6864877930867779,
+      "grad_norm": 0.4106120224272021,
+      "learning_rate": 2.465465754081277e-06,
+      "loss": 0.4888,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6869712351945855,
+      "grad_norm": 0.422067985874925,
+      "learning_rate": 2.458578269783364e-06,
+      "loss": 0.5155,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6874546773023931,
+      "grad_norm": 0.4136266956566046,
+      "learning_rate": 2.4516972815642166e-06,
+      "loss": 0.5143,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6879381194102007,
+      "grad_norm": 0.4335536983962682,
+      "learning_rate": 2.444822807012204e-06,
+      "loss": 0.5196,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6884215615180083,
+      "grad_norm": 0.42723749184962806,
+      "learning_rate": 2.4379548636990343e-06,
+      "loss": 0.5136,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6889050036258159,
+      "grad_norm": 0.4307011628135296,
+      "learning_rate": 2.4310934691797207e-06,
+      "loss": 0.5305,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6893884457336233,
+      "grad_norm": 0.45161428649005025,
+      "learning_rate": 2.4242386409925435e-06,
+      "loss": 0.5048,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6898718878414309,
+      "grad_norm": 0.4351186095813856,
+      "learning_rate": 2.4173903966589957e-06,
+      "loss": 0.5216,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6903553299492385,
+      "grad_norm": 0.4128958039987362,
+      "learning_rate": 2.410548753683743e-06,
+      "loss": 0.5206,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6908387720570461,
+      "grad_norm": 0.4185374425485222,
+      "learning_rate": 2.4037137295545737e-06,
+      "loss": 0.5205,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6913222141648537,
+      "grad_norm": 0.4256083734187945,
+      "learning_rate": 2.396885341742361e-06,
+      "loss": 0.4804,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6918056562726613,
+      "grad_norm": 0.411514639053229,
+      "learning_rate": 2.390063607701016e-06,
+      "loss": 0.5194,
+      "step": 1431
+    },
+    {
+      "epoch": 0.6922890983804689,
+      "grad_norm": 0.43241514860902464,
+      "learning_rate": 2.3832485448674407e-06,
+      "loss": 0.53,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6927725404882765,
+      "grad_norm": 0.4291596725507727,
+      "learning_rate": 2.3764401706614832e-06,
+      "loss": 0.5144,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6932559825960841,
+      "grad_norm": 0.42041788788695633,
+      "learning_rate": 2.369638502485897e-06,
+      "loss": 0.5148,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6937394247038917,
+      "grad_norm": 0.4482987713314786,
+      "learning_rate": 2.3628435577262947e-06,
+      "loss": 0.5191,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6942228668116993,
+      "grad_norm": 0.42573448798758273,
+      "learning_rate": 2.3560553537511043e-06,
+      "loss": 0.5021,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6947063089195069,
+      "grad_norm": 0.41739963072931596,
+      "learning_rate": 2.3492739079115214e-06,
+      "loss": 0.5061,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6951897510273145,
+      "grad_norm": 0.4366261411331466,
+      "learning_rate": 2.3424992375414655e-06,
+      "loss": 0.5133,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6956731931351221,
+      "grad_norm": 0.42225675860612266,
+      "learning_rate": 2.3357313599575422e-06,
+      "loss": 0.5254,
+      "step": 1439
+    },
+    {
+      "epoch": 0.6961566352429297,
+      "grad_norm": 0.4347650420428982,
+      "learning_rate": 2.3289702924589914e-06,
+      "loss": 0.5143,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6966400773507373,
+      "grad_norm": 0.4220266027824235,
+      "learning_rate": 2.3222160523276486e-06,
+      "loss": 0.5194,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6971235194585449,
+      "grad_norm": 0.400495176856287,
+      "learning_rate": 2.3154686568278933e-06,
+      "loss": 0.5315,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6976069615663524,
+      "grad_norm": 0.4149083634198192,
+      "learning_rate": 2.3087281232066134e-06,
+      "loss": 0.5109,
+      "step": 1443
+    },
+    {
+      "epoch": 0.69809040367416,
+      "grad_norm": 0.43831779922906355,
+      "learning_rate": 2.3019944686931554e-06,
+      "loss": 0.5256,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6985738457819676,
+      "grad_norm": 0.4379300687242213,
+      "learning_rate": 2.2952677104992855e-06,
+      "loss": 0.5287,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6990572878897752,
+      "grad_norm": 0.43973213205463885,
+      "learning_rate": 2.2885478658191364e-06,
+      "loss": 0.5192,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6995407299975828,
+      "grad_norm": 0.42002084857343974,
+      "learning_rate": 2.281834951829174e-06,
+      "loss": 0.521,
+      "step": 1447
+    },
+    {
+      "epoch": 0.7000241721053904,
+      "grad_norm": 0.38595076036167364,
+      "learning_rate": 2.2751289856881487e-06,
+      "loss": 0.4869,
+      "step": 1448
+    },
+    {
+      "epoch": 0.700507614213198,
+      "grad_norm": 0.436647846778714,
+      "learning_rate": 2.268429984537048e-06,
+      "loss": 0.5216,
+      "step": 1449
+    },
+    {
+      "epoch": 0.7009910563210056,
+      "grad_norm": 0.4140253730185284,
+      "learning_rate": 2.2617379654990623e-06,
+      "loss": 0.5165,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7014744984288132,
+      "grad_norm": 0.4644944125638521,
+      "learning_rate": 2.255052945679525e-06,
+      "loss": 0.5183,
+      "step": 1451
+    },
+    {
+      "epoch": 0.7019579405366208,
+      "grad_norm": 0.41536119938345195,
+      "learning_rate": 2.248374942165894e-06,
+      "loss": 0.5231,
+      "step": 1452
+    },
+    {
+      "epoch": 0.7024413826444283,
+      "grad_norm": 0.4012349549582878,
+      "learning_rate": 2.241703972027679e-06,
+      "loss": 0.5168,
+      "step": 1453
+    },
+    {
+      "epoch": 0.7029248247522359,
+      "grad_norm": 0.4521292215779327,
+      "learning_rate": 2.23504005231642e-06,
+      "loss": 0.5158,
+      "step": 1454
+    },
+    {
+      "epoch": 0.7034082668600435,
+      "grad_norm": 0.4172271643387044,
+      "learning_rate": 2.2283832000656304e-06,
+      "loss": 0.4941,
+      "step": 1455
+    },
+    {
+      "epoch": 0.7038917089678511,
+      "grad_norm": 0.421958406666486,
+      "learning_rate": 2.221733432290762e-06,
+      "loss": 0.5209,
+      "step": 1456
+    },
+    {
+      "epoch": 0.7043751510756587,
+      "grad_norm": 0.42224698163781604,
+      "learning_rate": 2.2150907659891566e-06,
+      "loss": 0.5173,
+      "step": 1457
+    },
+    {
+      "epoch": 0.7048585931834663,
+      "grad_norm": 0.43523243642666853,
+      "learning_rate": 2.2084552181400087e-06,
+      "loss": 0.5186,
+      "step": 1458
+    },
+    {
+      "epoch": 0.7053420352912738,
+      "grad_norm": 0.4437233504227722,
+      "learning_rate": 2.201826805704308e-06,
+      "loss": 0.5125,
+      "step": 1459
+    },
+    {
+      "epoch": 0.7058254773990814,
+      "grad_norm": 0.42532048824174346,
+      "learning_rate": 2.195205545624813e-06,
+      "loss": 0.5243,
+      "step": 1460
+    },
+    {
+      "epoch": 0.706308919506889,
+      "grad_norm": 0.4322950043512432,
+      "learning_rate": 2.188591454826e-06,
+      "loss": 0.5135,
+      "step": 1461
+    },
+    {
+      "epoch": 0.7067923616146966,
+      "grad_norm": 0.4272575345234204,
+      "learning_rate": 2.181984550214015e-06,
+      "loss": 0.5116,
+      "step": 1462
+    },
+    {
+      "epoch": 0.7072758037225042,
+      "grad_norm": 0.41921770884395154,
+      "learning_rate": 2.175384848676639e-06,
+      "loss": 0.5165,
+      "step": 1463
+    },
+    {
+      "epoch": 0.7077592458303118,
+      "grad_norm": 0.43176187181049736,
+      "learning_rate": 2.168792367083243e-06,
+      "loss": 0.5138,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7082426879381194,
+      "grad_norm": 0.41695232513283254,
+      "learning_rate": 2.162207122284742e-06,
+      "loss": 0.5091,
+      "step": 1465
+    },
+    {
+      "epoch": 0.708726130045927,
+      "grad_norm": 0.41339935320490057,
+      "learning_rate": 2.155629131113549e-06,
+      "loss": 0.5158,
+      "step": 1466
+    },
+    {
+      "epoch": 0.7092095721537346,
+      "grad_norm": 0.40689486411834114,
+      "learning_rate": 2.1490584103835433e-06,
+      "loss": 0.4847,
+      "step": 1467
+    },
+    {
+      "epoch": 0.7096930142615422,
+      "grad_norm": 0.417060588337446,
+      "learning_rate": 2.142494976890011e-06,
+      "loss": 0.5241,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7101764563693498,
+      "grad_norm": 0.4289677663647557,
+      "learning_rate": 2.135938847409625e-06,
+      "loss": 0.5206,
+      "step": 1469
+    },
+    {
+      "epoch": 0.7106598984771574,
+      "grad_norm": 0.43410470718447147,
+      "learning_rate": 2.1293900387003742e-06,
+      "loss": 0.4931,
+      "step": 1470
+    },
+    {
+      "epoch": 0.711143340584965,
+      "grad_norm": 0.42958196993128944,
+      "learning_rate": 2.1228485675015455e-06,
+      "loss": 0.5204,
+      "step": 1471
+    },
+    {
+      "epoch": 0.7116267826927726,
+      "grad_norm": 0.4311771692424152,
+      "learning_rate": 2.1163144505336634e-06,
+      "loss": 0.5219,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7121102248005802,
+      "grad_norm": 0.4150104118521869,
+      "learning_rate": 2.109787704498459e-06,
+      "loss": 0.519,
+      "step": 1473
+    },
+    {
+      "epoch": 0.7125936669083878,
+      "grad_norm": 0.43013467795196153,
+      "learning_rate": 2.1032683460788223e-06,
+      "loss": 0.4979,
+      "step": 1474
+    },
+    {
+      "epoch": 0.7130771090161954,
+      "grad_norm": 0.4303795815833922,
+      "learning_rate": 2.0967563919387563e-06,
+      "loss": 0.5256,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7135605511240029,
+      "grad_norm": 0.4386538663824397,
+      "learning_rate": 2.0902518587233418e-06,
+      "loss": 0.5195,
+      "step": 1476
+    },
+    {
+      "epoch": 0.7140439932318104,
+      "grad_norm": 0.41141211228553354,
+      "learning_rate": 2.08375476305869e-06,
+      "loss": 0.5238,
+      "step": 1477
+    },
+    {
+      "epoch": 0.714527435339618,
+      "grad_norm": 0.3832973623968104,
+      "learning_rate": 2.077265121551903e-06,
+      "loss": 0.4914,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7150108774474256,
+      "grad_norm": 0.4396380345403612,
+      "learning_rate": 2.0707829507910237e-06,
+      "loss": 0.5224,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7154943195552332,
+      "grad_norm": 0.4084969868928133,
+      "learning_rate": 2.0643082673450053e-06,
+      "loss": 0.5214,
+      "step": 1480
+    },
+    {
+      "epoch": 0.7159777616630408,
+      "grad_norm": 0.41940449704789057,
+      "learning_rate": 2.05784108776366e-06,
+      "loss": 0.5098,
+      "step": 1481
+    },
+    {
+      "epoch": 0.7164612037708484,
+      "grad_norm": 0.4368606150106444,
+      "learning_rate": 2.051381428577622e-06,
+      "loss": 0.5213,
+      "step": 1482
+    },
+    {
+      "epoch": 0.716944645878656,
+      "grad_norm": 0.4475169176125263,
+      "learning_rate": 2.044929306298298e-06,
+      "loss": 0.5169,
+      "step": 1483
+    },
+    {
+      "epoch": 0.7174280879864636,
+      "grad_norm": 0.4192404761939798,
+      "learning_rate": 2.0384847374178346e-06,
+      "loss": 0.5214,
+      "step": 1484
+    },
+    {
+      "epoch": 0.7179115300942712,
+      "grad_norm": 0.4000794067095613,
+      "learning_rate": 2.0320477384090665e-06,
+      "loss": 0.5002,
+      "step": 1485
+    },
+    {
+      "epoch": 0.7183949722020788,
+      "grad_norm": 0.4083964682274076,
+      "learning_rate": 2.0256183257254837e-06,
+      "loss": 0.5057,
+      "step": 1486
+    },
+    {
+      "epoch": 0.7188784143098864,
+      "grad_norm": 0.4286205023949667,
+      "learning_rate": 2.0191965158011854e-06,
+      "loss": 0.4815,
+      "step": 1487
+    },
+    {
+      "epoch": 0.719361856417694,
+      "grad_norm": 0.40907099979637535,
+      "learning_rate": 2.012782325050831e-06,
+      "loss": 0.5283,
+      "step": 1488
+    },
+    {
+      "epoch": 0.7198452985255016,
+      "grad_norm": 0.41946463733283473,
+      "learning_rate": 2.006375769869611e-06,
+      "loss": 0.522,
+      "step": 1489
+    },
+    {
+      "epoch": 0.7203287406333092,
+      "grad_norm": 0.4222854300641897,
+      "learning_rate": 1.9999768666331974e-06,
+      "loss": 0.5132,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7208121827411168,
+      "grad_norm": 0.3830302288103666,
+      "learning_rate": 1.9935856316977044e-06,
+      "loss": 0.4938,
+      "step": 1491
+    },
+    {
+      "epoch": 0.7212956248489243,
+      "grad_norm": 0.47757660690611003,
+      "learning_rate": 1.987202081399639e-06,
+      "loss": 0.5251,
+      "step": 1492
+    },
+    {
+      "epoch": 0.7217790669567319,
+      "grad_norm": 0.3992903621119011,
+      "learning_rate": 1.9808262320558724e-06,
+      "loss": 0.506,
+      "step": 1493
+    },
+    {
+      "epoch": 0.7222625090645395,
+      "grad_norm": 0.41142424465140587,
+      "learning_rate": 1.9744580999635902e-06,
+      "loss": 0.5143,
+      "step": 1494
+    },
+    {
+      "epoch": 0.7227459511723471,
+      "grad_norm": 0.4124129943865437,
+      "learning_rate": 1.968097701400252e-06,
+      "loss": 0.5245,
+      "step": 1495
+    },
+    {
+      "epoch": 0.7232293932801547,
+      "grad_norm": 0.4312737875038871,
+      "learning_rate": 1.9617450526235464e-06,
+      "loss": 0.5178,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7237128353879623,
+      "grad_norm": 0.43509903197162936,
+      "learning_rate": 1.9554001698713572e-06,
+      "loss": 0.5131,
+      "step": 1497
+    },
+    {
+      "epoch": 0.7241962774957699,
+      "grad_norm": 0.4260008705271214,
+      "learning_rate": 1.949063069361717e-06,
+      "loss": 0.5136,
+      "step": 1498
+    },
+    {
+      "epoch": 0.7246797196035775,
+      "grad_norm": 0.42356802738060345,
+      "learning_rate": 1.9427337672927632e-06,
+      "loss": 0.5146,
+      "step": 1499
+    },
+    {
+      "epoch": 0.7251631617113851,
+      "grad_norm": 0.4027997963462275,
+      "learning_rate": 1.936412279842705e-06,
+      "loss": 0.4913,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7256466038191927,
+      "grad_norm": 0.4124397793510055,
+      "learning_rate": 1.9300986231697705e-06,
+      "loss": 0.5175,
+      "step": 1501
+    },
+    {
+      "epoch": 0.7261300459270003,
+      "grad_norm": 0.4442811918906246,
+      "learning_rate": 1.9237928134121757e-06,
+      "loss": 0.516,
+      "step": 1502
+    },
+    {
+      "epoch": 0.7266134880348079,
+      "grad_norm": 0.4393627100062481,
+      "learning_rate": 1.9174948666880805e-06,
+      "loss": 0.5155,
+      "step": 1503
+    },
+    {
+      "epoch": 0.7270969301426154,
+      "grad_norm": 0.43133527501756386,
+      "learning_rate": 1.9112047990955446e-06,
+      "loss": 0.5136,
+      "step": 1504
+    },
+    {
+      "epoch": 0.727580372250423,
+      "grad_norm": 0.45322135855021595,
+      "learning_rate": 1.9049226267124844e-06,
+      "loss": 0.5172,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7280638143582306,
+      "grad_norm": 0.41078461158260915,
+      "learning_rate": 1.8986483655966408e-06,
+      "loss": 0.5179,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7285472564660382,
+      "grad_norm": 0.4178604053793329,
+      "learning_rate": 1.8923820317855307e-06,
+      "loss": 0.5076,
+      "step": 1507
+    },
+    {
+      "epoch": 0.7290306985738458,
+      "grad_norm": 0.42623268157040256,
+      "learning_rate": 1.8861236412964106e-06,
+      "loss": 0.5172,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7295141406816533,
+      "grad_norm": 0.42835046843347674,
+      "learning_rate": 1.879873210126229e-06,
+      "loss": 0.5259,
+      "step": 1509
+    },
+    {
+      "epoch": 0.7299975827894609,
+      "grad_norm": 0.4196504177616674,
+      "learning_rate": 1.873630754251588e-06,
+      "loss": 0.5177,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7304810248972685,
+      "grad_norm": 0.4079644120305993,
+      "learning_rate": 1.8673962896287152e-06,
+      "loss": 0.5201,
+      "step": 1511
+    },
+    {
+      "epoch": 0.7309644670050761,
+      "grad_norm": 0.43892341061011425,
+      "learning_rate": 1.8611698321933991e-06,
+      "loss": 0.5186,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7314479091128837,
+      "grad_norm": 0.42683430911112086,
+      "learning_rate": 1.8549513978609707e-06,
+      "loss": 0.5111,
+      "step": 1513
+    },
+    {
+      "epoch": 0.7319313512206913,
+      "grad_norm": 0.41062878136002484,
+      "learning_rate": 1.8487410025262436e-06,
+      "loss": 0.5103,
+      "step": 1514
+    },
+    {
+      "epoch": 0.7324147933284989,
+      "grad_norm": 0.4256013874707191,
+      "learning_rate": 1.8425386620634961e-06,
+      "loss": 0.5167,
+      "step": 1515
+    },
+    {
+      "epoch": 0.7328982354363065,
+      "grad_norm": 0.4388797350675763,
+      "learning_rate": 1.8363443923264046e-06,
+      "loss": 0.5125,
+      "step": 1516
+    },
+    {
+      "epoch": 0.7333816775441141,
+      "grad_norm": 0.4394233254146738,
+      "learning_rate": 1.8301582091480264e-06,
+      "loss": 0.5217,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7338651196519217,
+      "grad_norm": 0.41564422037394944,
+      "learning_rate": 1.8239801283407393e-06,
+      "loss": 0.5164,
+      "step": 1518
+    },
+    {
+      "epoch": 0.7343485617597293,
+      "grad_norm": 0.4173422643681329,
+      "learning_rate": 1.8178101656962188e-06,
+      "loss": 0.5205,
+      "step": 1519
+    },
+    {
+      "epoch": 0.7348320038675369,
+      "grad_norm": 0.39698118648442665,
+      "learning_rate": 1.8116483369853853e-06,
+      "loss": 0.4835,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7353154459753445,
+      "grad_norm": 0.42300362992419904,
+      "learning_rate": 1.8054946579583732e-06,
+      "loss": 0.5143,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7357988880831521,
+      "grad_norm": 0.42464469919772974,
+      "learning_rate": 1.7993491443444771e-06,
+      "loss": 0.5129,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7362823301909597,
+      "grad_norm": 0.4501988280108448,
+      "learning_rate": 1.7932118118521274e-06,
+      "loss": 0.5131,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7367657722987673,
+      "grad_norm": 0.41493548901611477,
+      "learning_rate": 1.787082676168842e-06,
+      "loss": 0.5268,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7372492144065748,
+      "grad_norm": 0.4436917707906808,
+      "learning_rate": 1.7809617529611828e-06,
+      "loss": 0.5126,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7377326565143824,
+      "grad_norm": 0.39767655781448813,
+      "learning_rate": 1.7748490578747257e-06,
+      "loss": 0.4945,
+      "step": 1526
+    },
+    {
+      "epoch": 0.73821609862219,
+      "grad_norm": 0.4281607415979641,
+      "learning_rate": 1.7687446065340074e-06,
+      "loss": 0.5189,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7386995407299976,
+      "grad_norm": 0.4123906023331037,
+      "learning_rate": 1.7626484145425038e-06,
+      "loss": 0.5117,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7391829828378051,
+      "grad_norm": 0.39861909677156787,
+      "learning_rate": 1.7565604974825678e-06,
+      "loss": 0.4917,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7396664249456127,
+      "grad_norm": 0.4164290248459804,
+      "learning_rate": 1.7504808709154104e-06,
+      "loss": 0.5187,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7401498670534203,
+      "grad_norm": 0.4149617264710624,
+      "learning_rate": 1.744409550381041e-06,
+      "loss": 0.529,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7406333091612279,
+      "grad_norm": 0.402995768205116,
+      "learning_rate": 1.7383465513982517e-06,
+      "loss": 0.4906,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7411167512690355,
+      "grad_norm": 0.4357911248878148,
+      "learning_rate": 1.7322918894645525e-06,
+      "loss": 0.5209,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7416001933768431,
+      "grad_norm": 0.4310636351470309,
+      "learning_rate": 1.7262455800561456e-06,
+      "loss": 0.529,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7420836354846507,
+      "grad_norm": 0.40110062198063573,
+      "learning_rate": 1.7202076386278876e-06,
+      "loss": 0.5218,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7425670775924583,
+      "grad_norm": 0.4044655145984996,
+      "learning_rate": 1.7141780806132429e-06,
+      "loss": 0.5038,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7430505197002659,
+      "grad_norm": 0.4169687562172726,
+      "learning_rate": 1.70815692142425e-06,
+      "loss": 0.5094,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7435339618080735,
+      "grad_norm": 0.3892005945860465,
+      "learning_rate": 1.702144176451473e-06,
+      "loss": 0.4909,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7440174039158811,
+      "grad_norm": 0.4059894671987348,
+      "learning_rate": 1.696139861063974e-06,
+      "loss": 0.5231,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7445008460236887,
+      "grad_norm": 0.4235285224343199,
+      "learning_rate": 1.690143990609268e-06,
+      "loss": 0.5116,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7449842881314963,
+      "grad_norm": 0.4066059462995061,
+      "learning_rate": 1.6841565804132843e-06,
+      "loss": 0.5159,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7454677302393038,
+      "grad_norm": 0.41374792014057904,
+      "learning_rate": 1.6781776457803227e-06,
+      "loss": 0.5146,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7459511723471114,
+      "grad_norm": 0.41330516594974576,
+      "learning_rate": 1.6722072019930242e-06,
+      "loss": 0.4841,
+      "step": 1543
+    },
+    {
+      "epoch": 0.746434614454919,
+      "grad_norm": 0.4342078760633199,
+      "learning_rate": 1.6662452643123234e-06,
+      "loss": 0.5181,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7469180565627266,
+      "grad_norm": 0.4366803318877013,
+      "learning_rate": 1.660291847977415e-06,
+      "loss": 0.5056,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7474014986705342,
+      "grad_norm": 0.4107968782550443,
+      "learning_rate": 1.6543469682057105e-06,
+      "loss": 0.5102,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7478849407783418,
+      "grad_norm": 0.43703346533243426,
+      "learning_rate": 1.6484106401927991e-06,
+      "loss": 0.517,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7483683828861494,
+      "grad_norm": 0.4185149815126949,
+      "learning_rate": 1.6424828791124159e-06,
+      "loss": 0.5162,
+      "step": 1548
+    },
+    {
+      "epoch": 0.748851824993957,
+      "grad_norm": 0.3941815905233016,
+      "learning_rate": 1.6365637001163958e-06,
+      "loss": 0.4694,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7493352671017646,
+      "grad_norm": 0.4069386532862478,
+      "learning_rate": 1.6306531183346387e-06,
+      "loss": 0.5172,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7498187092095722,
+      "grad_norm": 0.44449597102378385,
+      "learning_rate": 1.624751148875065e-06,
+      "loss": 0.5227,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7503021513173798,
+      "grad_norm": 0.4200070436877298,
+      "learning_rate": 1.6188578068235855e-06,
+      "loss": 0.5227,
+      "step": 1552
+    },
+    {
+      "epoch": 0.7507855934251874,
+      "grad_norm": 0.4134676341568954,
+      "learning_rate": 1.6129731072440586e-06,
+      "loss": 0.5197,
+      "step": 1553
+    },
+    {
+      "epoch": 0.751269035532995,
+      "grad_norm": 0.4342416540931307,
+      "learning_rate": 1.6070970651782514e-06,
+      "loss": 0.5234,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7517524776408026,
+      "grad_norm": 0.4621699665968105,
+      "learning_rate": 1.6012296956457972e-06,
+      "loss": 0.5224,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7522359197486101,
+      "grad_norm": 0.39794619123328484,
+      "learning_rate": 1.5953710136441685e-06,
+      "loss": 0.5222,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7527193618564177,
+      "grad_norm": 0.39795969856270086,
+      "learning_rate": 1.5895210341486279e-06,
+      "loss": 0.4697,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7532028039642252,
+      "grad_norm": 0.4348573897259895,
+      "learning_rate": 1.583679772112196e-06,
+      "loss": 0.5256,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7536862460720328,
+      "grad_norm": 0.4108494121358044,
+      "learning_rate": 1.5778472424656083e-06,
+      "loss": 0.5185,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7541696881798404,
+      "grad_norm": 0.41224584403564757,
+      "learning_rate": 1.5720234601172767e-06,
+      "loss": 0.5203,
+      "step": 1560
+    },
+    {
+      "epoch": 0.754653130287648,
+      "grad_norm": 0.4348874788487397,
+      "learning_rate": 1.566208439953265e-06,
+      "loss": 0.5189,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7551365723954556,
+      "grad_norm": 0.42842919833727694,
+      "learning_rate": 1.5604021968372286e-06,
+      "loss": 0.5111,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7556200145032632,
+      "grad_norm": 0.43772492324957596,
+      "learning_rate": 1.5546047456103964e-06,
+      "loss": 0.5147,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7561034566110708,
+      "grad_norm": 0.41431446343362865,
+      "learning_rate": 1.548816101091517e-06,
+      "loss": 0.5149,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7565868987188784,
+      "grad_norm": 0.40777837421338714,
+      "learning_rate": 1.5430362780768343e-06,
+      "loss": 0.5117,
+      "step": 1565
+    },
+    {
+      "epoch": 0.757070340826686,
+      "grad_norm": 0.4454487846070906,
+      "learning_rate": 1.537265291340042e-06,
+      "loss": 0.5074,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7575537829344936,
+      "grad_norm": 0.46396843002779686,
+      "learning_rate": 1.531503155632249e-06,
+      "loss": 0.5223,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7580372250423012,
+      "grad_norm": 0.41741600165011983,
+      "learning_rate": 1.5257498856819353e-06,
+      "loss": 0.5158,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7585206671501088,
+      "grad_norm": 0.4059061868499258,
+      "learning_rate": 1.5200054961949233e-06,
+      "loss": 0.5049,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7590041092579164,
+      "grad_norm": 0.41330390270516437,
+      "learning_rate": 1.5142700018543382e-06,
+      "loss": 0.5305,
+      "step": 1570
+    },
+    {
+      "epoch": 0.759487551365724,
+      "grad_norm": 0.43099056056318497,
+      "learning_rate": 1.508543417320562e-06,
+      "loss": 0.5212,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7599709934735316,
+      "grad_norm": 0.39882553101049034,
+      "learning_rate": 1.5028257572312105e-06,
+      "loss": 0.4883,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7604544355813392,
+      "grad_norm": 0.4581685557000849,
+      "learning_rate": 1.4971170362010774e-06,
+      "loss": 0.5225,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7609378776891468,
+      "grad_norm": 0.4428964310587446,
+      "learning_rate": 1.4914172688221213e-06,
+      "loss": 0.5195,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7614213197969543,
+      "grad_norm": 0.4170791170307987,
+      "learning_rate": 1.485726469663401e-06,
+      "loss": 0.5294,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 0.4212168944035229,
+      "learning_rate": 1.4800446532710627e-06,
+      "loss": 0.5143,
+      "step": 1576
+    },
+    {
+      "epoch": 0.7623882040125695,
+      "grad_norm": 0.4317778496296824,
+      "learning_rate": 1.4743718341682806e-06,
+      "loss": 0.5242,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7628716461203771,
+      "grad_norm": 0.3887549768642727,
+      "learning_rate": 1.468708026855245e-06,
+      "loss": 0.4927,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7633550882281847,
+      "grad_norm": 0.41991973562573803,
+      "learning_rate": 1.463053245809099e-06,
+      "loss": 0.5248,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7638385303359922,
+      "grad_norm": 0.41267795471721197,
+      "learning_rate": 1.457407505483921e-06,
+      "loss": 0.5187,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7643219724437998,
+      "grad_norm": 0.44716407911896383,
+      "learning_rate": 1.4517708203106763e-06,
+      "loss": 0.523,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7648054145516074,
+      "grad_norm": 0.4254440302923612,
+      "learning_rate": 1.446143204697187e-06,
+      "loss": 0.5233,
+      "step": 1582
+    },
+    {
+      "epoch": 0.765288856659415,
+      "grad_norm": 0.39996785018921494,
+      "learning_rate": 1.4405246730280946e-06,
+      "loss": 0.5172,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7657722987672226,
+      "grad_norm": 0.443369622770567,
+      "learning_rate": 1.4349152396648153e-06,
+      "loss": 0.5183,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7662557408750302,
+      "grad_norm": 0.40505843584897416,
+      "learning_rate": 1.4293149189455146e-06,
+      "loss": 0.5161,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7667391829828378,
+      "grad_norm": 0.4077704595280849,
+      "learning_rate": 1.4237237251850634e-06,
+      "loss": 0.5107,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7672226250906454,
+      "grad_norm": 0.40791039312028615,
+      "learning_rate": 1.4181416726750052e-06,
+      "loss": 0.5146,
+      "step": 1587
+    },
+    {
+      "epoch": 0.767706067198453,
+      "grad_norm": 0.41705043398231784,
+      "learning_rate": 1.4125687756835132e-06,
+      "loss": 0.4812,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7681895093062606,
+      "grad_norm": 0.4235182346193989,
+      "learning_rate": 1.4070050484553644e-06,
+      "loss": 0.5129,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7686729514140682,
+      "grad_norm": 0.414137655909364,
+      "learning_rate": 1.4014505052118893e-06,
+      "loss": 0.5236,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7691563935218757,
+      "grad_norm": 0.43611300077847176,
+      "learning_rate": 1.3959051601509537e-06,
+      "loss": 0.5345,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7696398356296833,
+      "grad_norm": 0.410845648388898,
+      "learning_rate": 1.3903690274469029e-06,
+      "loss": 0.5115,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7701232777374909,
+      "grad_norm": 0.3961083948871449,
+      "learning_rate": 1.3848421212505404e-06,
+      "loss": 0.5168,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7706067198452985,
+      "grad_norm": 0.42179325369386034,
+      "learning_rate": 1.37932445568908e-06,
+      "loss": 0.5125,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7710901619531061,
+      "grad_norm": 0.4213217250215216,
+      "learning_rate": 1.3738160448661253e-06,
+      "loss": 0.5267,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7715736040609137,
+      "grad_norm": 0.4143253090473424,
+      "learning_rate": 1.3683169028616155e-06,
+      "loss": 0.5178,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7720570461687213,
+      "grad_norm": 0.4171850827541685,
+      "learning_rate": 1.3628270437317993e-06,
+      "loss": 0.5211,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7725404882765289,
+      "grad_norm": 0.39565458081679644,
+      "learning_rate": 1.3573464815092003e-06,
+      "loss": 0.5055,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7730239303843365,
+      "grad_norm": 0.4271922188091497,
+      "learning_rate": 1.3518752302025773e-06,
+      "loss": 0.5279,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7735073724921441,
+      "grad_norm": 0.4151739224827406,
+      "learning_rate": 1.3464133037968914e-06,
+      "loss": 0.5239,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7739908145999517,
+      "grad_norm": 0.3960683162461613,
+      "learning_rate": 1.3409607162532628e-06,
+      "loss": 0.4987,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7744742567077593,
+      "grad_norm": 0.43044333694614223,
+      "learning_rate": 1.3355174815089477e-06,
+      "loss": 0.5273,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7749576988155669,
+      "grad_norm": 0.4121649380386113,
+      "learning_rate": 1.3300836134772916e-06,
+      "loss": 0.5162,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7754411409233745,
+      "grad_norm": 0.4005354058641754,
+      "learning_rate": 1.3246591260477015e-06,
+      "loss": 0.5167,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7759245830311821,
+      "grad_norm": 0.3951020817933521,
+      "learning_rate": 1.3192440330856005e-06,
+      "loss": 0.5251,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7764080251389897,
+      "grad_norm": 0.42611917105831465,
+      "learning_rate": 1.3138383484324063e-06,
+      "loss": 0.5252,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7768914672467973,
+      "grad_norm": 0.40098636118444037,
+      "learning_rate": 1.308442085905482e-06,
+      "loss": 0.5101,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7773749093546047,
+      "grad_norm": 0.4404415072756006,
+      "learning_rate": 1.30305525929811e-06,
+      "loss": 0.5224,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7778583514624123,
+      "grad_norm": 0.40390400609014704,
+      "learning_rate": 1.297677882379455e-06,
+      "loss": 0.5191,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7783417935702199,
+      "grad_norm": 0.43645719023114843,
+      "learning_rate": 1.2923099688945234e-06,
+      "loss": 0.5096,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7788252356780275,
+      "grad_norm": 0.401799031041578,
+      "learning_rate": 1.2869515325641357e-06,
+      "loss": 0.4812,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7793086777858351,
+      "grad_norm": 0.40544675897829047,
+      "learning_rate": 1.281602587084887e-06,
+      "loss": 0.5211,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7797921198936427,
+      "grad_norm": 0.4166351291750946,
+      "learning_rate": 1.2762631461291148e-06,
+      "loss": 0.5294,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7802755620014503,
+      "grad_norm": 0.4334981607396633,
+      "learning_rate": 1.2709332233448573e-06,
+      "loss": 0.5096,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7807590041092579,
+      "grad_norm": 0.437984950036233,
+      "learning_rate": 1.2656128323558286e-06,
+      "loss": 0.5135,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7812424462170655,
+      "grad_norm": 0.41467240914944964,
+      "learning_rate": 1.2603019867613764e-06,
+      "loss": 0.5162,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7817258883248731,
+      "grad_norm": 0.40797210573439474,
+      "learning_rate": 1.2550007001364518e-06,
+      "loss": 0.5064,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7822093304326807,
+      "grad_norm": 0.40625079236189654,
+      "learning_rate": 1.2497089860315675e-06,
+      "loss": 0.5057,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7826927725404883,
+      "grad_norm": 0.3973135238618207,
+      "learning_rate": 1.244426857972773e-06,
+      "loss": 0.5125,
+      "step": 1619
+    },
+    {
+      "epoch": 0.7831762146482959,
+      "grad_norm": 0.41758654400468537,
+      "learning_rate": 1.239154329461615e-06,
+      "loss": 0.5146,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7836596567561035,
+      "grad_norm": 0.4546571879884002,
+      "learning_rate": 1.233891413975098e-06,
+      "loss": 0.5138,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7841430988639111,
+      "grad_norm": 0.4501304501527847,
+      "learning_rate": 1.228638124965661e-06,
+      "loss": 0.5111,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7846265409717187,
+      "grad_norm": 0.40173574952002505,
+      "learning_rate": 1.223394475861131e-06,
+      "loss": 0.5134,
+      "step": 1623
+    },
+    {
+      "epoch": 0.7851099830795262,
+      "grad_norm": 0.4105768174048188,
+      "learning_rate": 1.2181604800646996e-06,
+      "loss": 0.5092,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7855934251873338,
+      "grad_norm": 0.39390517153871624,
+      "learning_rate": 1.212936150954882e-06,
+      "loss": 0.498,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7860768672951414,
+      "grad_norm": 0.41453725871465896,
+      "learning_rate": 1.207721501885486e-06,
+      "loss": 0.5063,
+      "step": 1626
+    },
+    {
+      "epoch": 0.786560309402949,
+      "grad_norm": 0.44249465126635484,
+      "learning_rate": 1.2025165461855714e-06,
+      "loss": 0.5212,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7870437515107566,
+      "grad_norm": 0.4079816768267276,
+      "learning_rate": 1.1973212971594262e-06,
+      "loss": 0.5155,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7875271936185642,
+      "grad_norm": 0.4318458945961838,
+      "learning_rate": 1.1921357680865258e-06,
+      "loss": 0.5183,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7880106357263718,
+      "grad_norm": 0.40656198305401237,
+      "learning_rate": 1.1869599722215013e-06,
+      "loss": 0.4949,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7884940778341794,
+      "grad_norm": 0.4056814293942294,
+      "learning_rate": 1.181793922794102e-06,
+      "loss": 0.5206,
+      "step": 1631
+    },
+    {
+      "epoch": 0.788977519941987,
+      "grad_norm": 0.42895763169120843,
+      "learning_rate": 1.1766376330091684e-06,
+      "loss": 0.503,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7894609620497945,
+      "grad_norm": 0.4165970675717556,
+      "learning_rate": 1.1714911160465924e-06,
+      "loss": 0.5255,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7899444041576021,
+      "grad_norm": 0.4123917311937627,
+      "learning_rate": 1.1663543850612847e-06,
+      "loss": 0.5169,
+      "step": 1634
+    },
+    {
+      "epoch": 0.7904278462654097,
+      "grad_norm": 0.41612583641837364,
+      "learning_rate": 1.1612274531831463e-06,
+      "loss": 0.4938,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7909112883732173,
+      "grad_norm": 0.40728900719245686,
+      "learning_rate": 1.1561103335170242e-06,
+      "loss": 0.5222,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7913947304810249,
+      "grad_norm": 0.4348645075910405,
+      "learning_rate": 1.1510030391426941e-06,
+      "loss": 0.5192,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7918781725888325,
+      "grad_norm": 0.4086546804175218,
+      "learning_rate": 1.1459055831148074e-06,
+      "loss": 0.5232,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7923616146966401,
+      "grad_norm": 0.40880965205946446,
+      "learning_rate": 1.140817978462876e-06,
+      "loss": 0.5212,
+      "step": 1639
+    },
+    {
+      "epoch": 0.7928450568044476,
+      "grad_norm": 0.3893016631161895,
+      "learning_rate": 1.1357402381912224e-06,
+      "loss": 0.4873,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7933284989122552,
+      "grad_norm": 0.4215992969510908,
+      "learning_rate": 1.1306723752789672e-06,
+      "loss": 0.5211,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7938119410200628,
+      "grad_norm": 0.420615559845491,
+      "learning_rate": 1.1256144026799703e-06,
+      "loss": 0.5179,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7942953831278704,
+      "grad_norm": 0.39236133338098145,
+      "learning_rate": 1.1205663333228217e-06,
+      "loss": 0.4911,
+      "step": 1643
+    },
+    {
+      "epoch": 0.794778825235678,
+      "grad_norm": 0.4158254754636244,
+      "learning_rate": 1.1155281801107897e-06,
+      "loss": 0.5146,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7952622673434856,
+      "grad_norm": 0.4092049660763265,
+      "learning_rate": 1.1104999559218022e-06,
+      "loss": 0.5063,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7957457094512932,
+      "grad_norm": 0.43121118572534733,
+      "learning_rate": 1.1054816736084057e-06,
+      "loss": 0.5122,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7962291515591008,
+      "grad_norm": 0.40574034047521074,
+      "learning_rate": 1.1004733459977325e-06,
+      "loss": 0.5089,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7967125936669084,
+      "grad_norm": 0.4343773778355907,
+      "learning_rate": 1.0954749858914727e-06,
+      "loss": 0.5177,
+      "step": 1648
+    },
+    {
+      "epoch": 0.797196035774716,
+      "grad_norm": 0.429877165339691,
+      "learning_rate": 1.0904866060658376e-06,
+      "loss": 0.5211,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7976794778825236,
+      "grad_norm": 0.4108995062804379,
+      "learning_rate": 1.0855082192715294e-06,
+      "loss": 0.5174,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7981629199903312,
+      "grad_norm": 0.4018616150052113,
+      "learning_rate": 1.0805398382337035e-06,
+      "loss": 0.5049,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7986463620981388,
+      "grad_norm": 0.4089174910335269,
+      "learning_rate": 1.0755814756519445e-06,
+      "loss": 0.5226,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7991298042059464,
+      "grad_norm": 0.3964537076582955,
+      "learning_rate": 1.0706331442002226e-06,
+      "loss": 0.5095,
+      "step": 1653
+    },
+    {
+      "epoch": 0.799613246313754,
+      "grad_norm": 0.4267767025207229,
+      "learning_rate": 1.0656948565268782e-06,
+      "loss": 0.5168,
+      "step": 1654
+    },
+    {
+      "epoch": 0.8000966884215616,
+      "grad_norm": 0.41363796984886936,
+      "learning_rate": 1.0607666252545673e-06,
+      "loss": 0.5128,
+      "step": 1655
+    },
+    {
+      "epoch": 0.8005801305293692,
+      "grad_norm": 0.43264146945425214,
+      "learning_rate": 1.0558484629802502e-06,
+      "loss": 0.514,
+      "step": 1656
+    },
+    {
+      "epoch": 0.8010635726371766,
+      "grad_norm": 0.42544390140386235,
+      "learning_rate": 1.0509403822751425e-06,
+      "loss": 0.512,
+      "step": 1657
+    },
+    {
+      "epoch": 0.8015470147449842,
+      "grad_norm": 0.3932679351449648,
+      "learning_rate": 1.0460423956846955e-06,
+      "loss": 0.4941,
+      "step": 1658
+    },
+    {
+      "epoch": 0.8020304568527918,
+      "grad_norm": 0.4096876585407803,
+      "learning_rate": 1.041154515728559e-06,
+      "loss": 0.5088,
+      "step": 1659
+    },
+    {
+      "epoch": 0.8025138989605994,
+      "grad_norm": 0.4173497731763413,
+      "learning_rate": 1.0362767549005454e-06,
+      "loss": 0.5119,
+      "step": 1660
+    },
+    {
+      "epoch": 0.802997341068407,
+      "grad_norm": 0.41012015779324845,
+      "learning_rate": 1.0314091256686065e-06,
+      "loss": 0.5212,
+      "step": 1661
+    },
+    {
+      "epoch": 0.8034807831762146,
+      "grad_norm": 0.38447439239259856,
+      "learning_rate": 1.0265516404747943e-06,
+      "loss": 0.5052,
+      "step": 1662
+    },
+    {
+      "epoch": 0.8039642252840222,
+      "grad_norm": 0.40948392634706504,
+      "learning_rate": 1.0217043117352337e-06,
+      "loss": 0.5109,
+      "step": 1663
+    },
+    {
+      "epoch": 0.8044476673918298,
+      "grad_norm": 0.40148827230751766,
+      "learning_rate": 1.0168671518400853e-06,
+      "loss": 0.5118,
+      "step": 1664
+    },
+    {
+      "epoch": 0.8049311094996374,
+      "grad_norm": 0.3939565441232479,
+      "learning_rate": 1.0120401731535213e-06,
+      "loss": 0.4879,
+      "step": 1665
+    },
+    {
+      "epoch": 0.805414551607445,
+      "grad_norm": 0.4394864393242481,
+      "learning_rate": 1.0072233880136872e-06,
+      "loss": 0.5104,
+      "step": 1666
+    },
+    {
+      "epoch": 0.8058979937152526,
+      "grad_norm": 0.4318770671908104,
+      "learning_rate": 1.0024168087326764e-06,
+      "loss": 0.5235,
+      "step": 1667
+    },
+    {
+      "epoch": 0.8063814358230602,
+      "grad_norm": 0.4281259140520081,
+      "learning_rate": 9.976204475964907e-07,
+      "loss": 0.5149,
+      "step": 1668
+    },
+    {
+      "epoch": 0.8068648779308678,
+      "grad_norm": 0.43979946361695016,
+      "learning_rate": 9.92834316865015e-07,
+      "loss": 0.5191,
+      "step": 1669
+    },
+    {
+      "epoch": 0.8073483200386754,
+      "grad_norm": 0.4312412015437643,
+      "learning_rate": 9.88058428771987e-07,
+      "loss": 0.5188,
+      "step": 1670
+    },
+    {
+      "epoch": 0.807831762146483,
+      "grad_norm": 0.4461824252192259,
+      "learning_rate": 9.832927955249605e-07,
+      "loss": 0.518,
+      "step": 1671
+    },
+    {
+      "epoch": 0.8083152042542906,
+      "grad_norm": 0.40455517199845253,
+      "learning_rate": 9.785374293052802e-07,
+      "loss": 0.5279,
+      "step": 1672
+    },
+    {
+      "epoch": 0.8087986463620981,
+      "grad_norm": 0.40500700400967726,
+      "learning_rate": 9.737923422680424e-07,
+      "loss": 0.5267,
+      "step": 1673
+    },
+    {
+      "epoch": 0.8092820884699057,
+      "grad_norm": 0.4053422468834684,
+      "learning_rate": 9.690575465420733e-07,
+      "loss": 0.5098,
+      "step": 1674
+    },
+    {
+      "epoch": 0.8097655305777133,
+      "grad_norm": 0.41221923071964073,
+      "learning_rate": 9.643330542298929e-07,
+      "loss": 0.5171,
+      "step": 1675
+    },
+    {
+      "epoch": 0.8102489726855209,
+      "grad_norm": 0.4289210188727792,
+      "learning_rate": 9.596188774076849e-07,
+      "loss": 0.5164,
+      "step": 1676
+    },
+    {
+      "epoch": 0.8107324147933285,
+      "grad_norm": 0.4119920227929362,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.5167,
+      "step": 1677
+    },
+    {
+      "epoch": 0.8112158569011361,
+      "grad_norm": 0.43146374267443927,
+      "learning_rate": 9.50221518406047e-07,
+      "loss": 0.5198,
+      "step": 1678
+    },
+    {
+      "epoch": 0.8116992990089437,
+      "grad_norm": 0.3915995001014536,
+      "learning_rate": 9.455383602470247e-07,
+      "loss": 0.5194,
+      "step": 1679
+    },
+    {
+      "epoch": 0.8121827411167513,
+      "grad_norm": 0.42092897815810126,
+      "learning_rate": 9.408655656187282e-07,
+      "loss": 0.5154,
+      "step": 1680
+    },
+    {
+      "epoch": 0.8126661832245589,
+      "grad_norm": 0.43929014126287974,
+      "learning_rate": 9.362031464651955e-07,
+      "loss": 0.5111,
+      "step": 1681
+    },
+    {
+      "epoch": 0.8131496253323665,
+      "grad_norm": 0.419403258433708,
+      "learning_rate": 9.31551114703943e-07,
+      "loss": 0.5175,
+      "step": 1682
+    },
+    {
+      "epoch": 0.813633067440174,
+      "grad_norm": 0.4235039718034734,
+      "learning_rate": 9.269094822259439e-07,
+      "loss": 0.5219,
+      "step": 1683
+    },
+    {
+      "epoch": 0.8141165095479816,
+      "grad_norm": 0.403949404981181,
+      "learning_rate": 9.22278260895581e-07,
+      "loss": 0.5257,
+      "step": 1684
+    },
+    {
+      "epoch": 0.8145999516557892,
+      "grad_norm": 0.40201626032689436,
+      "learning_rate": 9.176574625506324e-07,
+      "loss": 0.5065,
+      "step": 1685
+    },
+    {
+      "epoch": 0.8150833937635968,
+      "grad_norm": 0.42029809516611727,
+      "learning_rate": 9.130470990022283e-07,
+      "loss": 0.5198,
+      "step": 1686
+    },
+    {
+      "epoch": 0.8155668358714044,
+      "grad_norm": 0.4443584968330059,
+      "learning_rate": 9.084471820348306e-07,
+      "loss": 0.5054,
+      "step": 1687
+    },
+    {
+      "epoch": 0.816050277979212,
+      "grad_norm": 0.4011266291605723,
+      "learning_rate": 9.038577234061979e-07,
+      "loss": 0.481,
+      "step": 1688
+    },
+    {
+      "epoch": 0.8165337200870196,
+      "grad_norm": 0.4116565403445696,
+      "learning_rate": 8.992787348473575e-07,
+      "loss": 0.512,
+      "step": 1689
+    },
+    {
+      "epoch": 0.8170171621948271,
+      "grad_norm": 0.3855753519601646,
+      "learning_rate": 8.947102280625708e-07,
+      "loss": 0.4919,
+      "step": 1690
+    },
+    {
+      "epoch": 0.8175006043026347,
+      "grad_norm": 0.3998193393341577,
+      "learning_rate": 8.901522147293107e-07,
+      "loss": 0.5063,
+      "step": 1691
+    },
+    {
+      "epoch": 0.8179840464104423,
+      "grad_norm": 0.40465428030335077,
+      "learning_rate": 8.856047064982276e-07,
+      "loss": 0.4969,
+      "step": 1692
+    },
+    {
+      "epoch": 0.8184674885182499,
+      "grad_norm": 0.3993077607842942,
+      "learning_rate": 8.810677149931168e-07,
+      "loss": 0.5123,
+      "step": 1693
+    },
+    {
+      "epoch": 0.8189509306260575,
+      "grad_norm": 0.41845032917424874,
+      "learning_rate": 8.765412518108957e-07,
+      "loss": 0.5222,
+      "step": 1694
+    },
+    {
+      "epoch": 0.8194343727338651,
+      "grad_norm": 0.4482989172909152,
+      "learning_rate": 8.720253285215685e-07,
+      "loss": 0.5245,
+      "step": 1695
+    },
+    {
+      "epoch": 0.8199178148416727,
+      "grad_norm": 0.4096945568958353,
+      "learning_rate": 8.675199566682002e-07,
+      "loss": 0.4987,
+      "step": 1696
+    },
+    {
+      "epoch": 0.8204012569494803,
+      "grad_norm": 0.42715377043083036,
+      "learning_rate": 8.630251477668828e-07,
+      "loss": 0.4956,
+      "step": 1697
+    },
+    {
+      "epoch": 0.8208846990572879,
+      "grad_norm": 0.42586545844645524,
+      "learning_rate": 8.585409133067119e-07,
+      "loss": 0.5096,
+      "step": 1698
+    },
+    {
+      "epoch": 0.8213681411650955,
+      "grad_norm": 0.43766586659276707,
+      "learning_rate": 8.540672647497483e-07,
+      "loss": 0.5136,
+      "step": 1699
+    },
+    {
+      "epoch": 0.8218515832729031,
+      "grad_norm": 0.4371618341766256,
+      "learning_rate": 8.49604213531004e-07,
+      "loss": 0.5213,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8223350253807107,
+      "grad_norm": 0.4375571316772861,
+      "learning_rate": 8.451517710583934e-07,
+      "loss": 0.5051,
+      "step": 1701
+    },
+    {
+      "epoch": 0.8228184674885183,
+      "grad_norm": 0.4132441919616583,
+      "learning_rate": 8.407099487127207e-07,
+      "loss": 0.5257,
+      "step": 1702
+    },
+    {
+      "epoch": 0.8233019095963259,
+      "grad_norm": 0.42607745465695845,
+      "learning_rate": 8.362787578476395e-07,
+      "loss": 0.5249,
+      "step": 1703
+    },
+    {
+      "epoch": 0.8237853517041335,
+      "grad_norm": 0.4075673839523143,
+      "learning_rate": 8.318582097896316e-07,
+      "loss": 0.5058,
+      "step": 1704
+    },
+    {
+      "epoch": 0.8242687938119411,
+      "grad_norm": 0.42693741052199397,
+      "learning_rate": 8.274483158379759e-07,
+      "loss": 0.5111,
+      "step": 1705
+    },
+    {
+      "epoch": 0.8247522359197486,
+      "grad_norm": 0.39832416179935565,
+      "learning_rate": 8.230490872647146e-07,
+      "loss": 0.4938,
+      "step": 1706
+    },
+    {
+      "epoch": 0.8252356780275562,
+      "grad_norm": 0.422151557962671,
+      "learning_rate": 8.18660535314631e-07,
+      "loss": 0.5183,
+      "step": 1707
+    },
+    {
+      "epoch": 0.8257191201353637,
+      "grad_norm": 0.4003210551929738,
+      "learning_rate": 8.142826712052177e-07,
+      "loss": 0.5131,
+      "step": 1708
+    },
+    {
+      "epoch": 0.8262025622431713,
+      "grad_norm": 0.41552515229148246,
+      "learning_rate": 8.099155061266495e-07,
+      "loss": 0.5104,
+      "step": 1709
+    },
+    {
+      "epoch": 0.8266860043509789,
+      "grad_norm": 0.4199192751255081,
+      "learning_rate": 8.055590512417499e-07,
+      "loss": 0.504,
+      "step": 1710
+    },
+    {
+      "epoch": 0.8271694464587865,
+      "grad_norm": 0.4183052253157522,
+      "learning_rate": 8.012133176859705e-07,
+      "loss": 0.5183,
+      "step": 1711
+    },
+    {
+      "epoch": 0.8276528885665941,
+      "grad_norm": 0.40771463289221466,
+      "learning_rate": 7.968783165673554e-07,
+      "loss": 0.5134,
+      "step": 1712
+    },
+    {
+      "epoch": 0.8281363306744017,
+      "grad_norm": 0.4201027836512912,
+      "learning_rate": 7.925540589665187e-07,
+      "loss": 0.5074,
+      "step": 1713
+    },
+    {
+      "epoch": 0.8286197727822093,
+      "grad_norm": 0.395143526726159,
+      "learning_rate": 7.882405559366091e-07,
+      "loss": 0.4907,
+      "step": 1714
+    },
+    {
+      "epoch": 0.8291032148900169,
+      "grad_norm": 0.39924930985003787,
+      "learning_rate": 7.839378185032897e-07,
+      "loss": 0.5107,
+      "step": 1715
+    },
+    {
+      "epoch": 0.8295866569978245,
+      "grad_norm": 0.4132095601626946,
+      "learning_rate": 7.796458576647015e-07,
+      "loss": 0.5185,
+      "step": 1716
+    },
+    {
+      "epoch": 0.8300700991056321,
+      "grad_norm": 0.40587201306044,
+      "learning_rate": 7.753646843914465e-07,
+      "loss": 0.5182,
+      "step": 1717
+    },
+    {
+      "epoch": 0.8305535412134397,
+      "grad_norm": 0.4094094956774689,
+      "learning_rate": 7.710943096265461e-07,
+      "loss": 0.5029,
+      "step": 1718
+    },
+    {
+      "epoch": 0.8310369833212473,
+      "grad_norm": 0.41067812349491495,
+      "learning_rate": 7.668347442854218e-07,
+      "loss": 0.5021,
+      "step": 1719
+    },
+    {
+      "epoch": 0.8315204254290549,
+      "grad_norm": 0.39956787890532264,
+      "learning_rate": 7.625859992558665e-07,
+      "loss": 0.5206,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8320038675368625,
+      "grad_norm": 0.43928086956712875,
+      "learning_rate": 7.583480853980158e-07,
+      "loss": 0.5134,
+      "step": 1721
+    },
+    {
+      "epoch": 0.8324873096446701,
+      "grad_norm": 0.4100632271699525,
+      "learning_rate": 7.541210135443188e-07,
+      "loss": 0.5184,
+      "step": 1722
+    },
+    {
+      "epoch": 0.8329707517524776,
+      "grad_norm": 0.3961555211112688,
+      "learning_rate": 7.499047944995108e-07,
+      "loss": 0.5222,
+      "step": 1723
+    },
+    {
+      "epoch": 0.8334541938602852,
+      "grad_norm": 0.41401758140390904,
+      "learning_rate": 7.45699439040588e-07,
+      "loss": 0.5149,
+      "step": 1724
+    },
+    {
+      "epoch": 0.8339376359680928,
+      "grad_norm": 0.41725576477900833,
+      "learning_rate": 7.415049579167783e-07,
+      "loss": 0.5086,
+      "step": 1725
+    },
+    {
+      "epoch": 0.8344210780759004,
+      "grad_norm": 0.40808361223845036,
+      "learning_rate": 7.37321361849514e-07,
+      "loss": 0.5171,
+      "step": 1726
+    },
+    {
+      "epoch": 0.834904520183708,
+      "grad_norm": 0.4044441513281848,
+      "learning_rate": 7.331486615324024e-07,
+      "loss": 0.4931,
+      "step": 1727
+    },
+    {
+      "epoch": 0.8353879622915156,
+      "grad_norm": 0.39255016483428246,
+      "learning_rate": 7.289868676312023e-07,
+      "loss": 0.4895,
+      "step": 1728
+    },
+    {
+      "epoch": 0.8358714043993232,
+      "grad_norm": 0.4273894357037594,
+      "learning_rate": 7.248359907837959e-07,
+      "loss": 0.5141,
+      "step": 1729
+    },
+    {
+      "epoch": 0.8363548465071308,
+      "grad_norm": 0.41270523260835523,
+      "learning_rate": 7.206960416001563e-07,
+      "loss": 0.5053,
+      "step": 1730
+    },
+    {
+      "epoch": 0.8368382886149384,
+      "grad_norm": 0.42210989792552517,
+      "learning_rate": 7.165670306623296e-07,
+      "loss": 0.515,
+      "step": 1731
+    },
+    {
+      "epoch": 0.837321730722746,
+      "grad_norm": 0.4005116526979819,
+      "learning_rate": 7.124489685243985e-07,
+      "loss": 0.5084,
+      "step": 1732
+    },
+    {
+      "epoch": 0.8378051728305536,
+      "grad_norm": 0.42730888005294004,
+      "learning_rate": 7.08341865712463e-07,
+      "loss": 0.5149,
+      "step": 1733
+    },
+    {
+      "epoch": 0.8382886149383612,
+      "grad_norm": 0.3946117211995092,
+      "learning_rate": 7.042457327246088e-07,
+      "loss": 0.5272,
+      "step": 1734
+    },
+    {
+      "epoch": 0.8387720570461688,
+      "grad_norm": 0.40058125990145727,
+      "learning_rate": 7.001605800308825e-07,
+      "loss": 0.5173,
+      "step": 1735
+    },
+    {
+      "epoch": 0.8392554991539763,
+      "grad_norm": 0.39419621537510763,
+      "learning_rate": 6.960864180732618e-07,
+      "loss": 0.5182,
+      "step": 1736
+    },
+    {
+      "epoch": 0.8397389412617839,
+      "grad_norm": 0.4302451888948554,
+      "learning_rate": 6.920232572656349e-07,
+      "loss": 0.5145,
+      "step": 1737
+    },
+    {
+      "epoch": 0.8402223833695915,
+      "grad_norm": 0.39221396906385003,
+      "learning_rate": 6.879711079937667e-07,
+      "loss": 0.5079,
+      "step": 1738
+    },
+    {
+      "epoch": 0.840705825477399,
+      "grad_norm": 0.4210023704512398,
+      "learning_rate": 6.839299806152799e-07,
+      "loss": 0.5061,
+      "step": 1739
+    },
+    {
+      "epoch": 0.8411892675852066,
+      "grad_norm": 0.4031707044630559,
+      "learning_rate": 6.79899885459619e-07,
+      "loss": 0.5174,
+      "step": 1740
+    },
+    {
+      "epoch": 0.8416727096930142,
+      "grad_norm": 0.40104705743190977,
+      "learning_rate": 6.758808328280325e-07,
+      "loss": 0.4981,
+      "step": 1741
+    },
+    {
+      "epoch": 0.8421561518008218,
+      "grad_norm": 0.4158859718137932,
+      "learning_rate": 6.718728329935448e-07,
+      "loss": 0.5216,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8426395939086294,
+      "grad_norm": 0.4140963838597211,
+      "learning_rate": 6.678758962009241e-07,
+      "loss": 0.5154,
+      "step": 1743
+    },
+    {
+      "epoch": 0.843123036016437,
+      "grad_norm": 0.41926365963573253,
+      "learning_rate": 6.638900326666653e-07,
+      "loss": 0.5181,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8436064781242446,
+      "grad_norm": 0.4007033614343704,
+      "learning_rate": 6.599152525789531e-07,
+      "loss": 0.4772,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8440899202320522,
+      "grad_norm": 0.4266694328755557,
+      "learning_rate": 6.559515660976506e-07,
+      "loss": 0.5153,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8445733623398598,
+      "grad_norm": 0.42158713984389296,
+      "learning_rate": 6.519989833542567e-07,
+      "loss": 0.5218,
+      "step": 1747
+    },
+    {
+      "epoch": 0.8450568044476674,
+      "grad_norm": 0.4190422236566301,
+      "learning_rate": 6.480575144518931e-07,
+      "loss": 0.5267,
+      "step": 1748
+    },
+    {
+      "epoch": 0.845540246555475,
+      "grad_norm": 0.42322451653416415,
+      "learning_rate": 6.441271694652701e-07,
+      "loss": 0.517,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8460236886632826,
+      "grad_norm": 0.4050974433698499,
+      "learning_rate": 6.402079584406673e-07,
+      "loss": 0.523,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8465071307710902,
+      "grad_norm": 0.39927068510798064,
+      "learning_rate": 6.36299891395904e-07,
+      "loss": 0.4943,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8469905728788978,
+      "grad_norm": 0.40520913199613756,
+      "learning_rate": 6.32402978320315e-07,
+      "loss": 0.519,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8474740149867054,
+      "grad_norm": 0.41524557234436116,
+      "learning_rate": 6.285172291747232e-07,
+      "loss": 0.5087,
+      "step": 1753
+    },
+    {
+      "epoch": 0.847957457094513,
+      "grad_norm": 0.39348055940589066,
+      "learning_rate": 6.246426538914174e-07,
+      "loss": 0.5135,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8484408992023206,
+      "grad_norm": 0.40472211918575973,
+      "learning_rate": 6.207792623741249e-07,
+      "loss": 0.5181,
+      "step": 1755
+    },
+    {
+      "epoch": 0.8489243413101281,
+      "grad_norm": 0.3797781522780497,
+      "learning_rate": 6.169270644979836e-07,
+      "loss": 0.4718,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8494077834179357,
+      "grad_norm": 0.4172564454240539,
+      "learning_rate": 6.130860701095226e-07,
+      "loss": 0.5093,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8498912255257433,
+      "grad_norm": 0.460481903524328,
+      "learning_rate": 6.092562890266341e-07,
+      "loss": 0.5245,
+      "step": 1758
+    },
+    {
+      "epoch": 0.8503746676335509,
+      "grad_norm": 0.40544203017797725,
+      "learning_rate": 6.054377310385479e-07,
+      "loss": 0.5067,
+      "step": 1759
+    },
+    {
+      "epoch": 0.8508581097413584,
+      "grad_norm": 0.4225253280006634,
+      "learning_rate": 6.016304059058031e-07,
+      "loss": 0.5169,
+      "step": 1760
+    },
+    {
+      "epoch": 0.851341551849166,
+      "grad_norm": 0.40769266639259943,
+      "learning_rate": 5.97834323360233e-07,
+      "loss": 0.5243,
+      "step": 1761
+    },
+    {
+      "epoch": 0.8518249939569736,
+      "grad_norm": 0.42284940262412657,
+      "learning_rate": 5.940494931049262e-07,
+      "loss": 0.5194,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8523084360647812,
+      "grad_norm": 0.3916025337851957,
+      "learning_rate": 5.902759248142187e-07,
+      "loss": 0.4975,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8527918781725888,
+      "grad_norm": 0.41326270414280697,
+      "learning_rate": 5.86513628133652e-07,
+      "loss": 0.5154,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8532753202803964,
+      "grad_norm": 0.40856588365868324,
+      "learning_rate": 5.827626126799613e-07,
+      "loss": 0.5154,
+      "step": 1765
+    },
+    {
+      "epoch": 0.853758762388204,
+      "grad_norm": 0.42831173680710594,
+      "learning_rate": 5.790228880410426e-07,
+      "loss": 0.5163,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8542422044960116,
+      "grad_norm": 0.4218590594382107,
+      "learning_rate": 5.75294463775935e-07,
+      "loss": 0.517,
+      "step": 1767
+    },
+    {
+      "epoch": 0.8547256466038192,
+      "grad_norm": 0.38253864809006055,
+      "learning_rate": 5.715773494147919e-07,
+      "loss": 0.4929,
+      "step": 1768
+    },
+    {
+      "epoch": 0.8552090887116268,
+      "grad_norm": 0.40270548702028475,
+      "learning_rate": 5.678715544588547e-07,
+      "loss": 0.5088,
+      "step": 1769
+    },
+    {
+      "epoch": 0.8556925308194344,
+      "grad_norm": 0.4229953125269584,
+      "learning_rate": 5.641770883804365e-07,
+      "loss": 0.5258,
+      "step": 1770
+    },
+    {
+      "epoch": 0.856175972927242,
+      "grad_norm": 0.4037677845049078,
+      "learning_rate": 5.604939606228887e-07,
+      "loss": 0.5095,
+      "step": 1771
+    },
+    {
+      "epoch": 0.8566594150350495,
+      "grad_norm": 0.39977977942883575,
+      "learning_rate": 5.568221806005847e-07,
+      "loss": 0.5128,
+      "step": 1772
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.4175904938844971,
+      "learning_rate": 5.531617576988879e-07,
+      "loss": 0.5114,
+      "step": 1773
+    },
+    {
+      "epoch": 0.8576262992506647,
+      "grad_norm": 0.41812393010867166,
+      "learning_rate": 5.495127012741352e-07,
+      "loss": 0.5188,
+      "step": 1774
+    },
+    {
+      "epoch": 0.8581097413584723,
+      "grad_norm": 0.38871202154348194,
+      "learning_rate": 5.45875020653609e-07,
+      "loss": 0.4882,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8585931834662799,
+      "grad_norm": 0.40175664384357557,
+      "learning_rate": 5.422487251355146e-07,
+      "loss": 0.5088,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8590766255740875,
+      "grad_norm": 0.4258611448475652,
+      "learning_rate": 5.386338239889549e-07,
+      "loss": 0.5136,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8595600676818951,
+      "grad_norm": 0.38040545155326977,
+      "learning_rate": 5.350303264539091e-07,
+      "loss": 0.4692,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8600435097897027,
+      "grad_norm": 0.4185751036827134,
+      "learning_rate": 5.314382417412062e-07,
+      "loss": 0.516,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8605269518975103,
+      "grad_norm": 0.4237092619379993,
+      "learning_rate": 5.278575790325052e-07,
+      "loss": 0.5146,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8610103940053179,
+      "grad_norm": 0.4173802982789206,
+      "learning_rate": 5.242883474802696e-07,
+      "loss": 0.5125,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8614938361131255,
+      "grad_norm": 0.41838440801291993,
+      "learning_rate": 5.207305562077403e-07,
+      "loss": 0.5177,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8619772782209331,
+      "grad_norm": 0.4779855097218796,
+      "learning_rate": 5.1718421430892e-07,
+      "loss": 0.5304,
+      "step": 1783
+    },
+    {
+      "epoch": 0.8624607203287407,
+      "grad_norm": 0.37738685143261025,
+      "learning_rate": 5.136493308485446e-07,
+      "loss": 0.486,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8629441624365483,
+      "grad_norm": 0.39963258309250466,
+      "learning_rate": 5.101259148620618e-07,
+      "loss": 0.4959,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8634276045443559,
+      "grad_norm": 0.39604391770722097,
+      "learning_rate": 5.066139753556049e-07,
+      "loss": 0.4993,
+      "step": 1786
+    },
+    {
+      "epoch": 0.8639110466521635,
+      "grad_norm": 0.40732958269577874,
+      "learning_rate": 5.031135213059756e-07,
+      "loss": 0.5153,
+      "step": 1787
+    },
+    {
+      "epoch": 0.864394488759971,
+      "grad_norm": 0.3897806967927546,
+      "learning_rate": 4.99624561660616e-07,
+      "loss": 0.4871,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8648779308677785,
+      "grad_norm": 0.4048723969181331,
+      "learning_rate": 4.961471053375899e-07,
+      "loss": 0.512,
+      "step": 1789
+    },
+    {
+      "epoch": 0.8653613729755861,
+      "grad_norm": 0.4203351282800037,
+      "learning_rate": 4.926811612255539e-07,
+      "loss": 0.5121,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8658448150833937,
+      "grad_norm": 0.39858565202586066,
+      "learning_rate": 4.892267381837396e-07,
+      "loss": 0.5011,
+      "step": 1791
+    },
+    {
+      "epoch": 0.8663282571912013,
+      "grad_norm": 0.4344627773200746,
+      "learning_rate": 4.857838450419339e-07,
+      "loss": 0.5103,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8668116992990089,
+      "grad_norm": 0.40293448022650774,
+      "learning_rate": 4.823524906004468e-07,
+      "loss": 0.5138,
+      "step": 1793
+    },
+    {
+      "epoch": 0.8672951414068165,
+      "grad_norm": 0.41801337173969716,
+      "learning_rate": 4.789326836300983e-07,
+      "loss": 0.5151,
+      "step": 1794
+    },
+    {
+      "epoch": 0.8677785835146241,
+      "grad_norm": 0.4058943681689954,
+      "learning_rate": 4.7552443287218866e-07,
+      "loss": 0.5098,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8682620256224317,
+      "grad_norm": 0.42652856984845416,
+      "learning_rate": 4.7212774703848273e-07,
+      "loss": 0.508,
+      "step": 1796
+    },
+    {
+      "epoch": 0.8687454677302393,
+      "grad_norm": 0.4211824745719729,
+      "learning_rate": 4.687426348111834e-07,
+      "loss": 0.5122,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8692289098380469,
+      "grad_norm": 0.4022753726796167,
+      "learning_rate": 4.65369104842911e-07,
+      "loss": 0.52,
+      "step": 1798
+    },
+    {
+      "epoch": 0.8697123519458545,
+      "grad_norm": 0.40283890754002527,
+      "learning_rate": 4.620071657566777e-07,
+      "loss": 0.5072,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8701957940536621,
+      "grad_norm": 0.4283203699114763,
+      "learning_rate": 4.586568261458729e-07,
+      "loss": 0.5096,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8706792361614697,
+      "grad_norm": 0.38328034805235095,
+      "learning_rate": 4.553180945742336e-07,
+      "loss": 0.4861,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8711626782692773,
+      "grad_norm": 0.4122501087059972,
+      "learning_rate": 4.5199097957582816e-07,
+      "loss": 0.5136,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8716461203770849,
+      "grad_norm": 0.41264272327652995,
+      "learning_rate": 4.486754896550288e-07,
+      "loss": 0.5012,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8721295624848925,
+      "grad_norm": 0.41725003600600513,
+      "learning_rate": 4.45371633286496e-07,
+      "loss": 0.5185,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8726130045927,
+      "grad_norm": 0.4078148663174146,
+      "learning_rate": 4.4207941891515335e-07,
+      "loss": 0.5135,
+      "step": 1805
+    },
+    {
+      "epoch": 0.8730964467005076,
+      "grad_norm": 0.4209684818924423,
+      "learning_rate": 4.3879885495616505e-07,
+      "loss": 0.512,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8735798888083152,
+      "grad_norm": 0.3832255061477332,
+      "learning_rate": 4.3552994979491836e-07,
+      "loss": 0.5131,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8740633309161228,
+      "grad_norm": 0.415646535369065,
+      "learning_rate": 4.322727117869951e-07,
+      "loss": 0.5156,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8745467730239304,
+      "grad_norm": 0.403529021224522,
+      "learning_rate": 4.290271492581627e-07,
+      "loss": 0.5225,
+      "step": 1809
+    },
+    {
+      "epoch": 0.875030215131738,
+      "grad_norm": 0.4248226663595473,
+      "learning_rate": 4.257932705043372e-07,
+      "loss": 0.5276,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8755136572395456,
+      "grad_norm": 0.42279657022545747,
+      "learning_rate": 4.2257108379157586e-07,
+      "loss": 0.5224,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8759970993473531,
+      "grad_norm": 0.4140176038814713,
+      "learning_rate": 4.1936059735604497e-07,
+      "loss": 0.5161,
+      "step": 1812
+    },
+    {
+      "epoch": 0.8764805414551607,
+      "grad_norm": 0.39792458444383394,
+      "learning_rate": 4.161618194040079e-07,
+      "loss": 0.5277,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8769639835629683,
+      "grad_norm": 0.39549573015495143,
+      "learning_rate": 4.129747581117993e-07,
+      "loss": 0.5053,
+      "step": 1814
+    },
+    {
+      "epoch": 0.8774474256707759,
+      "grad_norm": 0.4160802633412061,
+      "learning_rate": 4.0979942162580387e-07,
+      "loss": 0.516,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8779308677785835,
+      "grad_norm": 0.41239251805984983,
+      "learning_rate": 4.06635818062438e-07,
+      "loss": 0.5278,
+      "step": 1816
+    },
+    {
+      "epoch": 0.8784143098863911,
+      "grad_norm": 0.40871873580107365,
+      "learning_rate": 4.0348395550812713e-07,
+      "loss": 0.5294,
+      "step": 1817
+    },
+    {
+      "epoch": 0.8788977519941987,
+      "grad_norm": 0.40365670038657436,
+      "learning_rate": 4.003438420192873e-07,
+      "loss": 0.5158,
+      "step": 1818
+    },
+    {
+      "epoch": 0.8793811941020063,
+      "grad_norm": 0.4255428234546921,
+      "learning_rate": 3.9721548562229985e-07,
+      "loss": 0.5114,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8798646362098139,
+      "grad_norm": 0.41203315649756733,
+      "learning_rate": 3.9409889431349656e-07,
+      "loss": 0.5116,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8803480783176215,
+      "grad_norm": 0.4149872650348109,
+      "learning_rate": 3.9099407605913576e-07,
+      "loss": 0.5099,
+      "step": 1821
+    },
+    {
+      "epoch": 0.880831520425429,
+      "grad_norm": 0.4258100076362105,
+      "learning_rate": 3.879010387953841e-07,
+      "loss": 0.5175,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8813149625332366,
+      "grad_norm": 0.3902355927247227,
+      "learning_rate": 3.84819790428293e-07,
+      "loss": 0.498,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8817984046410442,
+      "grad_norm": 0.40842472365457144,
+      "learning_rate": 3.8175033883378233e-07,
+      "loss": 0.518,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8822818467488518,
+      "grad_norm": 0.4221970543634826,
+      "learning_rate": 3.7869269185761613e-07,
+      "loss": 0.5216,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8827652888566594,
+      "grad_norm": 0.40616883661281006,
+      "learning_rate": 3.7564685731538985e-07,
+      "loss": 0.5066,
+      "step": 1826
+    },
+    {
+      "epoch": 0.883248730964467,
+      "grad_norm": 0.4061562407072031,
+      "learning_rate": 3.7261284299249967e-07,
+      "loss": 0.517,
+      "step": 1827
+    },
+    {
+      "epoch": 0.8837321730722746,
+      "grad_norm": 0.4079225433423233,
+      "learning_rate": 3.695906566441304e-07,
+      "loss": 0.4959,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8842156151800822,
+      "grad_norm": 0.38197368709112006,
+      "learning_rate": 3.665803059952344e-07,
+      "loss": 0.4871,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8846990572878898,
+      "grad_norm": 0.411849076052872,
+      "learning_rate": 3.63581798740511e-07,
+      "loss": 0.5143,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8851824993956974,
+      "grad_norm": 0.393276210273132,
+      "learning_rate": 3.605951425443871e-07,
+      "loss": 0.4936,
+      "step": 1831
+    },
+    {
+      "epoch": 0.885665941503505,
+      "grad_norm": 0.369604359657528,
+      "learning_rate": 3.576203450409943e-07,
+      "loss": 0.4684,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8861493836113126,
+      "grad_norm": 0.43326466002005165,
+      "learning_rate": 3.5465741383415684e-07,
+      "loss": 0.5104,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8866328257191202,
+      "grad_norm": 0.41527359664646213,
+      "learning_rate": 3.5170635649736497e-07,
+      "loss": 0.519,
+      "step": 1834
+    },
+    {
+      "epoch": 0.8871162678269278,
+      "grad_norm": 0.41356740894281485,
+      "learning_rate": 3.487671805737597e-07,
+      "loss": 0.508,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8875997099347354,
+      "grad_norm": 0.4050751048123327,
+      "learning_rate": 3.4583989357611037e-07,
+      "loss": 0.5135,
+      "step": 1836
+    },
+    {
+      "epoch": 0.888083152042543,
+      "grad_norm": 0.3923610722591795,
+      "learning_rate": 3.4292450298679945e-07,
+      "loss": 0.5075,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8885665941503504,
+      "grad_norm": 0.41919225013002887,
+      "learning_rate": 3.400210162577999e-07,
+      "loss": 0.5166,
+      "step": 1838
+    },
+    {
+      "epoch": 0.889050036258158,
+      "grad_norm": 0.42118222715491443,
+      "learning_rate": 3.371294408106585e-07,
+      "loss": 0.523,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8895334783659656,
+      "grad_norm": 0.3952238335142466,
+      "learning_rate": 3.3424978403647443e-07,
+      "loss": 0.5138,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8900169204737732,
+      "grad_norm": 0.4163195177412695,
+      "learning_rate": 3.313820532958817e-07,
+      "loss": 0.5274,
+      "step": 1841
+    },
+    {
+      "epoch": 0.8905003625815808,
+      "grad_norm": 0.3930314520659748,
+      "learning_rate": 3.285262559190322e-07,
+      "loss": 0.4991,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8909838046893884,
+      "grad_norm": 0.4336804309313973,
+      "learning_rate": 3.256823992055741e-07,
+      "loss": 0.5009,
+      "step": 1843
+    },
+    {
+      "epoch": 0.891467246797196,
+      "grad_norm": 0.41714068524986875,
+      "learning_rate": 3.228504904246349e-07,
+      "loss": 0.5238,
+      "step": 1844
+    },
+    {
+      "epoch": 0.8919506889050036,
+      "grad_norm": 0.41848606366751967,
+      "learning_rate": 3.20030536814801e-07,
+      "loss": 0.5202,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8924341310128112,
+      "grad_norm": 0.422964314144621,
+      "learning_rate": 3.1722254558410047e-07,
+      "loss": 0.5104,
+      "step": 1846
+    },
+    {
+      "epoch": 0.8929175731206188,
+      "grad_norm": 0.41539348703446205,
+      "learning_rate": 3.144265239099864e-07,
+      "loss": 0.5152,
+      "step": 1847
+    },
+    {
+      "epoch": 0.8934010152284264,
+      "grad_norm": 0.3936271006898258,
+      "learning_rate": 3.1164247893931575e-07,
+      "loss": 0.5071,
+      "step": 1848
+    },
+    {
+      "epoch": 0.893884457336234,
+      "grad_norm": 0.4152031331913687,
+      "learning_rate": 3.088704177883306e-07,
+      "loss": 0.5181,
+      "step": 1849
+    },
+    {
+      "epoch": 0.8943678994440416,
+      "grad_norm": 0.4176432021270733,
+      "learning_rate": 3.06110347542643e-07,
+      "loss": 0.5235,
+      "step": 1850
+    },
+    {
+      "epoch": 0.8948513415518492,
+      "grad_norm": 0.3954219378639727,
+      "learning_rate": 3.033622752572157e-07,
+      "loss": 0.5019,
+      "step": 1851
+    },
+    {
+      "epoch": 0.8953347836596568,
+      "grad_norm": 0.45830856560980365,
+      "learning_rate": 3.0062620795634214e-07,
+      "loss": 0.5263,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8958182257674644,
+      "grad_norm": 0.4009466020951186,
+      "learning_rate": 2.9790215263363174e-07,
+      "loss": 0.5222,
+      "step": 1853
+    },
+    {
+      "epoch": 0.896301667875272,
+      "grad_norm": 0.3933495297633584,
+      "learning_rate": 2.951901162519877e-07,
+      "loss": 0.5233,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8967851099830795,
+      "grad_norm": 0.39895160904445,
+      "learning_rate": 2.9249010574359636e-07,
+      "loss": 0.5212,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8972685520908871,
+      "grad_norm": 0.42068899596041226,
+      "learning_rate": 2.898021280098995e-07,
+      "loss": 0.5168,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8977519941986947,
+      "grad_norm": 0.39971963228555085,
+      "learning_rate": 2.8712618992158656e-07,
+      "loss": 0.5084,
+      "step": 1857
+    },
+    {
+      "epoch": 0.8982354363065023,
+      "grad_norm": 0.3999616227972635,
+      "learning_rate": 2.8446229831856964e-07,
+      "loss": 0.5088,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8987188784143099,
+      "grad_norm": 0.4001447692276326,
+      "learning_rate": 2.8181046000997136e-07,
+      "loss": 0.521,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8992023205221175,
+      "grad_norm": 0.41592034251039167,
+      "learning_rate": 2.791706817741041e-07,
+      "loss": 0.5072,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8996857626299251,
+      "grad_norm": 0.4445686187455443,
+      "learning_rate": 2.765429703584538e-07,
+      "loss": 0.5148,
+      "step": 1861
+    },
+    {
+      "epoch": 0.9001692047377327,
+      "grad_norm": 0.40228802491920107,
+      "learning_rate": 2.739273324796621e-07,
+      "loss": 0.5262,
+      "step": 1862
+    },
+    {
+      "epoch": 0.9006526468455403,
+      "grad_norm": 0.40404504261863744,
+      "learning_rate": 2.7132377482351037e-07,
+      "loss": 0.5147,
+      "step": 1863
+    },
+    {
+      "epoch": 0.9011360889533478,
+      "grad_norm": 0.3986359660989621,
+      "learning_rate": 2.687323040449025e-07,
+      "loss": 0.5172,
+      "step": 1864
+    },
+    {
+      "epoch": 0.9016195310611554,
+      "grad_norm": 0.42039178580411435,
+      "learning_rate": 2.6615292676784533e-07,
+      "loss": 0.5191,
+      "step": 1865
+    },
+    {
+      "epoch": 0.902102973168963,
+      "grad_norm": 0.4168785648766661,
+      "learning_rate": 2.635856495854372e-07,
+      "loss": 0.5116,
+      "step": 1866
+    },
+    {
+      "epoch": 0.9025864152767706,
+      "grad_norm": 0.4006359687639295,
+      "learning_rate": 2.6103047905984224e-07,
+      "loss": 0.5243,
+      "step": 1867
+    },
+    {
+      "epoch": 0.9030698573845782,
+      "grad_norm": 0.4136741219117099,
+      "learning_rate": 2.584874217222855e-07,
+      "loss": 0.516,
+      "step": 1868
+    },
+    {
+      "epoch": 0.9035532994923858,
+      "grad_norm": 0.41454758895188654,
+      "learning_rate": 2.5595648407302496e-07,
+      "loss": 0.5299,
+      "step": 1869
+    },
+    {
+      "epoch": 0.9040367416001934,
+      "grad_norm": 0.43072596167116733,
+      "learning_rate": 2.53437672581342e-07,
+      "loss": 0.5192,
+      "step": 1870
+    },
+    {
+      "epoch": 0.9045201837080009,
+      "grad_norm": 0.413346134850188,
+      "learning_rate": 2.5093099368551974e-07,
+      "loss": 0.5135,
+      "step": 1871
+    },
+    {
+      "epoch": 0.9050036258158085,
+      "grad_norm": 0.44414111234791465,
+      "learning_rate": 2.484364537928341e-07,
+      "loss": 0.5248,
+      "step": 1872
+    },
+    {
+      "epoch": 0.9054870679236161,
+      "grad_norm": 0.41031454686253116,
+      "learning_rate": 2.45954059279529e-07,
+      "loss": 0.5198,
+      "step": 1873
+    },
+    {
+      "epoch": 0.9059705100314237,
+      "grad_norm": 0.3982976345229948,
+      "learning_rate": 2.4348381649080486e-07,
+      "loss": 0.5163,
+      "step": 1874
+    },
+    {
+      "epoch": 0.9064539521392313,
+      "grad_norm": 0.4007617837820295,
+      "learning_rate": 2.41025731740801e-07,
+      "loss": 0.511,
+      "step": 1875
+    },
+    {
+      "epoch": 0.9069373942470389,
+      "grad_norm": 0.40168617787804406,
+      "learning_rate": 2.3857981131258037e-07,
+      "loss": 0.5114,
+      "step": 1876
+    },
+    {
+      "epoch": 0.9074208363548465,
+      "grad_norm": 0.38110421429609603,
+      "learning_rate": 2.3614606145811347e-07,
+      "loss": 0.4992,
+      "step": 1877
+    },
+    {
+      "epoch": 0.9079042784626541,
+      "grad_norm": 0.3870732423514054,
+      "learning_rate": 2.3372448839825978e-07,
+      "loss": 0.4887,
+      "step": 1878
+    },
+    {
+      "epoch": 0.9083877205704617,
+      "grad_norm": 0.39979584331802676,
+      "learning_rate": 2.3131509832275633e-07,
+      "loss": 0.5122,
+      "step": 1879
+    },
+    {
+      "epoch": 0.9088711626782693,
+      "grad_norm": 0.3996732608438804,
+      "learning_rate": 2.2891789739019733e-07,
+      "loss": 0.5102,
+      "step": 1880
+    },
+    {
+      "epoch": 0.9093546047860769,
+      "grad_norm": 0.40968516048558534,
+      "learning_rate": 2.2653289172802295e-07,
+      "loss": 0.5049,
+      "step": 1881
+    },
+    {
+      "epoch": 0.9098380468938845,
+      "grad_norm": 0.4006751726323446,
+      "learning_rate": 2.241600874324984e-07,
+      "loss": 0.5144,
+      "step": 1882
+    },
+    {
+      "epoch": 0.9103214890016921,
+      "grad_norm": 0.4066456668668,
+      "learning_rate": 2.2179949056870432e-07,
+      "loss": 0.5184,
+      "step": 1883
+    },
+    {
+      "epoch": 0.9108049311094997,
+      "grad_norm": 0.4179374057794063,
+      "learning_rate": 2.194511071705141e-07,
+      "loss": 0.5131,
+      "step": 1884
+    },
+    {
+      "epoch": 0.9112883732173073,
+      "grad_norm": 0.419480536858942,
+      "learning_rate": 2.1711494324058724e-07,
+      "loss": 0.5147,
+      "step": 1885
+    },
+    {
+      "epoch": 0.9117718153251149,
+      "grad_norm": 0.40624640146953556,
+      "learning_rate": 2.1479100475034598e-07,
+      "loss": 0.5084,
+      "step": 1886
+    },
+    {
+      "epoch": 0.9122552574329225,
+      "grad_norm": 0.40367583928635464,
+      "learning_rate": 2.1247929763996534e-07,
+      "loss": 0.4832,
+      "step": 1887
+    },
+    {
+      "epoch": 0.91273869954073,
+      "grad_norm": 0.3989060344990105,
+      "learning_rate": 2.101798278183542e-07,
+      "loss": 0.5144,
+      "step": 1888
+    },
+    {
+      "epoch": 0.9132221416485375,
+      "grad_norm": 0.3998308893808953,
+      "learning_rate": 2.0789260116314215e-07,
+      "loss": 0.5081,
+      "step": 1889
+    },
+    {
+      "epoch": 0.9137055837563451,
+      "grad_norm": 0.4063990008087812,
+      "learning_rate": 2.0561762352066638e-07,
+      "loss": 0.5109,
+      "step": 1890
+    },
+    {
+      "epoch": 0.9141890258641527,
+      "grad_norm": 0.4167108480628528,
+      "learning_rate": 2.0335490070595208e-07,
+      "loss": 0.5186,
+      "step": 1891
+    },
+    {
+      "epoch": 0.9146724679719603,
+      "grad_norm": 0.39430080435851855,
+      "learning_rate": 2.011044385027011e-07,
+      "loss": 0.5101,
+      "step": 1892
+    },
+    {
+      "epoch": 0.9151559100797679,
+      "grad_norm": 0.42096559238441866,
+      "learning_rate": 1.988662426632765e-07,
+      "loss": 0.5078,
+      "step": 1893
+    },
+    {
+      "epoch": 0.9156393521875755,
+      "grad_norm": 0.39723951707790667,
+      "learning_rate": 1.9664031890868795e-07,
+      "loss": 0.5223,
+      "step": 1894
+    },
+    {
+      "epoch": 0.9161227942953831,
+      "grad_norm": 0.3912147208179025,
+      "learning_rate": 1.9442667292857432e-07,
+      "loss": 0.509,
+      "step": 1895
+    },
+    {
+      "epoch": 0.9166062364031907,
+      "grad_norm": 0.4054442997347736,
+      "learning_rate": 1.922253103811944e-07,
+      "loss": 0.4972,
+      "step": 1896
+    },
+    {
+      "epoch": 0.9170896785109983,
+      "grad_norm": 0.4117401816100168,
+      "learning_rate": 1.9003623689340777e-07,
+      "loss": 0.5143,
+      "step": 1897
+    },
+    {
+      "epoch": 0.9175731206188059,
+      "grad_norm": 0.40528953423093284,
+      "learning_rate": 1.8785945806066297e-07,
+      "loss": 0.5186,
+      "step": 1898
+    },
+    {
+      "epoch": 0.9180565627266135,
+      "grad_norm": 0.4027696401480633,
+      "learning_rate": 1.85694979446982e-07,
+      "loss": 0.5167,
+      "step": 1899
+    },
+    {
+      "epoch": 0.9185400048344211,
+      "grad_norm": 0.38938110778215645,
+      "learning_rate": 1.835428065849465e-07,
+      "loss": 0.5141,
+      "step": 1900
+    },
+    {
+      "epoch": 0.9190234469422287,
+      "grad_norm": 0.3958049685314876,
+      "learning_rate": 1.814029449756849e-07,
+      "loss": 0.5231,
+      "step": 1901
+    },
+    {
+      "epoch": 0.9195068890500363,
+      "grad_norm": 0.4039199277502588,
+      "learning_rate": 1.7927540008885414e-07,
+      "loss": 0.5088,
+      "step": 1902
+    },
+    {
+      "epoch": 0.9199903311578439,
+      "grad_norm": 0.40426884197944674,
+      "learning_rate": 1.7716017736263192e-07,
+      "loss": 0.5129,
+      "step": 1903
+    },
+    {
+      "epoch": 0.9204737732656514,
+      "grad_norm": 0.41358470698939953,
+      "learning_rate": 1.7505728220369667e-07,
+      "loss": 0.5203,
+      "step": 1904
+    },
+    {
+      "epoch": 0.920957215373459,
+      "grad_norm": 0.4250820090378729,
+      "learning_rate": 1.729667199872187e-07,
+      "loss": 0.5223,
+      "step": 1905
+    },
+    {
+      "epoch": 0.9214406574812666,
+      "grad_norm": 0.40899977989644076,
+      "learning_rate": 1.70888496056843e-07,
+      "loss": 0.5107,
+      "step": 1906
+    },
+    {
+      "epoch": 0.9219240995890742,
+      "grad_norm": 0.4187760713922149,
+      "learning_rate": 1.6882261572467862e-07,
+      "loss": 0.5142,
+      "step": 1907
+    },
+    {
+      "epoch": 0.9224075416968818,
+      "grad_norm": 0.39684261118945696,
+      "learning_rate": 1.6676908427128103e-07,
+      "loss": 0.4847,
+      "step": 1908
+    },
+    {
+      "epoch": 0.9228909838046894,
+      "grad_norm": 0.4124141033869449,
+      "learning_rate": 1.64727906945642e-07,
+      "loss": 0.5063,
+      "step": 1909
+    },
+    {
+      "epoch": 0.923374425912497,
+      "grad_norm": 0.4104731721152495,
+      "learning_rate": 1.6269908896517638e-07,
+      "loss": 0.5035,
+      "step": 1910
+    },
+    {
+      "epoch": 0.9238578680203046,
+      "grad_norm": 0.38208183163995635,
+      "learning_rate": 1.6068263551570596e-07,
+      "loss": 0.4855,
+      "step": 1911
+    },
+    {
+      "epoch": 0.9243413101281122,
+      "grad_norm": 0.37943822460943005,
+      "learning_rate": 1.5867855175144885e-07,
+      "loss": 0.4863,
+      "step": 1912
+    },
+    {
+      "epoch": 0.9248247522359198,
+      "grad_norm": 0.4169103989292416,
+      "learning_rate": 1.5668684279500245e-07,
+      "loss": 0.5077,
+      "step": 1913
+    },
+    {
+      "epoch": 0.9253081943437274,
+      "grad_norm": 0.41157707540822663,
+      "learning_rate": 1.5470751373733773e-07,
+      "loss": 0.5184,
+      "step": 1914
+    },
+    {
+      "epoch": 0.925791636451535,
+      "grad_norm": 0.39771451862665147,
+      "learning_rate": 1.5274056963777817e-07,
+      "loss": 0.5094,
+      "step": 1915
+    },
+    {
+      "epoch": 0.9262750785593425,
+      "grad_norm": 0.4092987974762817,
+      "learning_rate": 1.507860155239921e-07,
+      "loss": 0.5154,
+      "step": 1916
+    },
+    {
+      "epoch": 0.9267585206671501,
+      "grad_norm": 0.3854503813446518,
+      "learning_rate": 1.488438563919764e-07,
+      "loss": 0.4938,
+      "step": 1917
+    },
+    {
+      "epoch": 0.9272419627749577,
+      "grad_norm": 0.3900052964813903,
+      "learning_rate": 1.4691409720604732e-07,
+      "loss": 0.5077,
+      "step": 1918
+    },
+    {
+      "epoch": 0.9277254048827653,
+      "grad_norm": 0.40750712678387396,
+      "learning_rate": 1.449967428988247e-07,
+      "loss": 0.5145,
+      "step": 1919
+    },
+    {
+      "epoch": 0.9282088469905729,
+      "grad_norm": 0.4023813113333878,
+      "learning_rate": 1.4309179837122045e-07,
+      "loss": 0.5291,
+      "step": 1920
+    },
+    {
+      "epoch": 0.9286922890983804,
+      "grad_norm": 0.38502235475455626,
+      "learning_rate": 1.411992684924257e-07,
+      "loss": 0.5119,
+      "step": 1921
+    },
+    {
+      "epoch": 0.929175731206188,
+      "grad_norm": 0.40862887218787325,
+      "learning_rate": 1.3931915809990039e-07,
+      "loss": 0.5106,
+      "step": 1922
+    },
+    {
+      "epoch": 0.9296591733139956,
+      "grad_norm": 0.4123756674563694,
+      "learning_rate": 1.374514719993575e-07,
+      "loss": 0.5126,
+      "step": 1923
+    },
+    {
+      "epoch": 0.9301426154218032,
+      "grad_norm": 0.41456641529199556,
+      "learning_rate": 1.3559621496475438e-07,
+      "loss": 0.5145,
+      "step": 1924
+    },
+    {
+      "epoch": 0.9306260575296108,
+      "grad_norm": 0.4049152537963314,
+      "learning_rate": 1.3375339173827551e-07,
+      "loss": 0.5261,
+      "step": 1925
+    },
+    {
+      "epoch": 0.9311094996374184,
+      "grad_norm": 0.37450439680837744,
+      "learning_rate": 1.3192300703032733e-07,
+      "loss": 0.474,
+      "step": 1926
+    },
+    {
+      "epoch": 0.931592941745226,
+      "grad_norm": 0.41100475742292075,
+      "learning_rate": 1.3010506551952018e-07,
+      "loss": 0.5134,
+      "step": 1927
+    },
+    {
+      "epoch": 0.9320763838530336,
+      "grad_norm": 0.41369315234307685,
+      "learning_rate": 1.2829957185265863e-07,
+      "loss": 0.52,
+      "step": 1928
+    },
+    {
+      "epoch": 0.9325598259608412,
+      "grad_norm": 0.3885589982730842,
+      "learning_rate": 1.2650653064473106e-07,
+      "loss": 0.5031,
+      "step": 1929
+    },
+    {
+      "epoch": 0.9330432680686488,
+      "grad_norm": 0.3951920703691663,
+      "learning_rate": 1.2472594647889357e-07,
+      "loss": 0.5092,
+      "step": 1930
+    },
+    {
+      "epoch": 0.9335267101764564,
+      "grad_norm": 0.40947647060207415,
+      "learning_rate": 1.2295782390646494e-07,
+      "loss": 0.5177,
+      "step": 1931
+    },
+    {
+      "epoch": 0.934010152284264,
+      "grad_norm": 0.390574491653679,
+      "learning_rate": 1.2120216744690716e-07,
+      "loss": 0.5133,
+      "step": 1932
+    },
+    {
+      "epoch": 0.9344935943920716,
+      "grad_norm": 0.4045498383765011,
+      "learning_rate": 1.194589815878211e-07,
+      "loss": 0.5163,
+      "step": 1933
+    },
+    {
+      "epoch": 0.9349770364998792,
+      "grad_norm": 0.40440549648310886,
+      "learning_rate": 1.177282707849281e-07,
+      "loss": 0.5181,
+      "step": 1934
+    },
+    {
+      "epoch": 0.9354604786076868,
+      "grad_norm": 0.4024689599876574,
+      "learning_rate": 1.1601003946206723e-07,
+      "loss": 0.5181,
+      "step": 1935
+    },
+    {
+      "epoch": 0.9359439207154944,
+      "grad_norm": 0.3986512567562451,
+      "learning_rate": 1.1430429201117476e-07,
+      "loss": 0.5032,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9364273628233019,
+      "grad_norm": 0.39397430112101045,
+      "learning_rate": 1.1261103279227858e-07,
+      "loss": 0.5178,
+      "step": 1937
+    },
+    {
+      "epoch": 0.9369108049311095,
+      "grad_norm": 0.4291769455926264,
+      "learning_rate": 1.1093026613348601e-07,
+      "loss": 0.5196,
+      "step": 1938
+    },
+    {
+      "epoch": 0.937394247038917,
+      "grad_norm": 0.3917679927009391,
+      "learning_rate": 1.0926199633097156e-07,
+      "loss": 0.4919,
+      "step": 1939
+    },
+    {
+      "epoch": 0.9378776891467246,
+      "grad_norm": 0.42599062790587783,
+      "learning_rate": 1.0760622764896866e-07,
+      "loss": 0.5147,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9383611312545322,
+      "grad_norm": 0.4023777838627757,
+      "learning_rate": 1.0596296431975406e-07,
+      "loss": 0.5156,
+      "step": 1941
+    },
+    {
+      "epoch": 0.9388445733623398,
+      "grad_norm": 0.3966354448847634,
+      "learning_rate": 1.0433221054364174e-07,
+      "loss": 0.5065,
+      "step": 1942
+    },
+    {
+      "epoch": 0.9393280154701474,
+      "grad_norm": 0.4013413460541232,
+      "learning_rate": 1.0271397048897014e-07,
+      "loss": 0.5053,
+      "step": 1943
+    },
+    {
+      "epoch": 0.939811457577955,
+      "grad_norm": 0.37653088174864213,
+      "learning_rate": 1.0110824829209164e-07,
+      "loss": 0.4939,
+      "step": 1944
+    },
+    {
+      "epoch": 0.9402948996857626,
+      "grad_norm": 0.399035469753345,
+      "learning_rate": 9.951504805735979e-08,
+      "loss": 0.5106,
+      "step": 1945
+    },
+    {
+      "epoch": 0.9407783417935702,
+      "grad_norm": 0.3991989592342914,
+      "learning_rate": 9.793437385712479e-08,
+      "loss": 0.5153,
+      "step": 1946
+    },
+    {
+      "epoch": 0.9412617839013778,
+      "grad_norm": 0.4057585743893453,
+      "learning_rate": 9.636622973171583e-08,
+      "loss": 0.51,
+      "step": 1947
+    },
+    {
+      "epoch": 0.9417452260091854,
+      "grad_norm": 0.4054528739627977,
+      "learning_rate": 9.481061968943717e-08,
+      "loss": 0.516,
+      "step": 1948
+    },
+    {
+      "epoch": 0.942228668116993,
+      "grad_norm": 0.3921980636127477,
+      "learning_rate": 9.3267547706552e-08,
+      "loss": 0.5051,
+      "step": 1949
+    },
+    {
+      "epoch": 0.9427121102248006,
+      "grad_norm": 0.39913144156030567,
+      "learning_rate": 9.17370177272775e-08,
+      "loss": 0.5055,
+      "step": 1950
+    },
+    {
+      "epoch": 0.9431955523326082,
+      "grad_norm": 0.4004586843938766,
+      "learning_rate": 9.021903366377093e-08,
+      "loss": 0.5164,
+      "step": 1951
+    },
+    {
+      "epoch": 0.9436789944404158,
+      "grad_norm": 0.4037223050343566,
+      "learning_rate": 8.8713599396123e-08,
+      "loss": 0.5098,
+      "step": 1952
+    },
+    {
+      "epoch": 0.9441624365482234,
+      "grad_norm": 0.39850858877215634,
+      "learning_rate": 8.72207187723445e-08,
+      "loss": 0.5211,
+      "step": 1953
+    },
+    {
+      "epoch": 0.9446458786560309,
+      "grad_norm": 0.41059877409881057,
+      "learning_rate": 8.5740395608358e-08,
+      "loss": 0.5121,
+      "step": 1954
+    },
+    {
+      "epoch": 0.9451293207638385,
+      "grad_norm": 0.40573184845060545,
+      "learning_rate": 8.427263368798955e-08,
+      "loss": 0.5256,
+      "step": 1955
+    },
+    {
+      "epoch": 0.9456127628716461,
+      "grad_norm": 0.3966583772201167,
+      "learning_rate": 8.281743676295639e-08,
+      "loss": 0.5183,
+      "step": 1956
+    },
+    {
+      "epoch": 0.9460962049794537,
+      "grad_norm": 0.40701943797191764,
+      "learning_rate": 8.13748085528604e-08,
+      "loss": 0.5135,
+      "step": 1957
+    },
+    {
+      "epoch": 0.9465796470872613,
+      "grad_norm": 0.37606341196980025,
+      "learning_rate": 7.99447527451741e-08,
+      "loss": 0.4903,
+      "step": 1958
+    },
+    {
+      "epoch": 0.9470630891950689,
+      "grad_norm": 0.4114856897492863,
+      "learning_rate": 7.852727299523577e-08,
+      "loss": 0.5068,
+      "step": 1959
+    },
+    {
+      "epoch": 0.9475465313028765,
+      "grad_norm": 0.4093526523044555,
+      "learning_rate": 7.71223729262377e-08,
+      "loss": 0.5127,
+      "step": 1960
+    },
+    {
+      "epoch": 0.9480299734106841,
+      "grad_norm": 0.4130076310229578,
+      "learning_rate": 7.573005612921903e-08,
+      "loss": 0.5121,
+      "step": 1961
+    },
+    {
+      "epoch": 0.9485134155184917,
+      "grad_norm": 0.40254945616875554,
+      "learning_rate": 7.435032616305238e-08,
+      "loss": 0.5178,
+      "step": 1962
+    },
+    {
+      "epoch": 0.9489968576262993,
+      "grad_norm": 0.4108181664423654,
+      "learning_rate": 7.298318655443893e-08,
+      "loss": 0.5078,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9494802997341069,
+      "grad_norm": 0.3954161759006289,
+      "learning_rate": 7.162864079789777e-08,
+      "loss": 0.5137,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9499637418419145,
+      "grad_norm": 0.3993428213266096,
+      "learning_rate": 7.028669235575714e-08,
+      "loss": 0.496,
+      "step": 1965
+    },
+    {
+      "epoch": 0.950447183949722,
+      "grad_norm": 0.41493027982851327,
+      "learning_rate": 6.895734465814597e-08,
+      "loss": 0.5257,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9509306260575296,
+      "grad_norm": 0.38537633628397905,
+      "learning_rate": 6.764060110298287e-08,
+      "loss": 0.5208,
+      "step": 1967
+    },
+    {
+      "epoch": 0.9514140681653372,
+      "grad_norm": 0.41057398606285567,
+      "learning_rate": 6.633646505597113e-08,
+      "loss": 0.5224,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9518975102731448,
+      "grad_norm": 0.4420797620121168,
+      "learning_rate": 6.504493985058813e-08,
+      "loss": 0.5108,
+      "step": 1969
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 0.39854773939873966,
+      "learning_rate": 6.376602878807592e-08,
+      "loss": 0.5134,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9528643944887599,
+      "grad_norm": 0.4104047856111181,
+      "learning_rate": 6.249973513743345e-08,
+      "loss": 0.5079,
+      "step": 1971
+    },
+    {
+      "epoch": 0.9533478365965675,
+      "grad_norm": 0.40077931999667527,
+      "learning_rate": 6.124606213541052e-08,
+      "loss": 0.5196,
+      "step": 1972
+    },
+    {
+      "epoch": 0.9538312787043751,
+      "grad_norm": 0.43500257686302385,
+      "learning_rate": 6.000501298649653e-08,
+      "loss": 0.5197,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9543147208121827,
+      "grad_norm": 0.4186094433656202,
+      "learning_rate": 5.8776590862911764e-08,
+      "loss": 0.5135,
+      "step": 1974
+    },
+    {
+      "epoch": 0.9547981629199903,
+      "grad_norm": 0.4119358911199865,
+      "learning_rate": 5.756079890460342e-08,
+      "loss": 0.5137,
+      "step": 1975
+    },
+    {
+      "epoch": 0.9552816050277979,
+      "grad_norm": 0.39694645564275877,
+      "learning_rate": 5.635764021923229e-08,
+      "loss": 0.5121,
+      "step": 1976
+    },
+    {
+      "epoch": 0.9557650471356055,
+      "grad_norm": 0.4154887872586203,
+      "learning_rate": 5.5167117882171104e-08,
+      "loss": 0.516,
+      "step": 1977
+    },
+    {
+      "epoch": 0.9562484892434131,
+      "grad_norm": 0.7692472130509296,
+      "learning_rate": 5.3989234936489556e-08,
+      "loss": 0.5055,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9567319313512207,
+      "grad_norm": 0.4198618304821996,
+      "learning_rate": 5.2823994392951497e-08,
+      "loss": 0.5094,
+      "step": 1979
+    },
+    {
+      "epoch": 0.9572153734590283,
+      "grad_norm": 0.39385026351820934,
+      "learning_rate": 5.167139923000553e-08,
+      "loss": 0.4933,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9576988155668359,
+      "grad_norm": 0.4159053427086944,
+      "learning_rate": 5.053145239377777e-08,
+      "loss": 0.4936,
+      "step": 1981
+    },
+    {
+      "epoch": 0.9581822576746435,
+      "grad_norm": 0.3990167973444839,
+      "learning_rate": 4.940415679806465e-08,
+      "loss": 0.5124,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9586656997824511,
+      "grad_norm": 0.4012277528608715,
+      "learning_rate": 4.828951532432457e-08,
+      "loss": 0.5151,
+      "step": 1983
+    },
+    {
+      "epoch": 0.9591491418902587,
+      "grad_norm": 0.4099731484035176,
+      "learning_rate": 4.718753082167071e-08,
+      "loss": 0.5191,
+      "step": 1984
+    },
+    {
+      "epoch": 0.9596325839980663,
+      "grad_norm": 0.41474696363438857,
+      "learning_rate": 4.6098206106863774e-08,
+      "loss": 0.515,
+      "step": 1985
+    },
+    {
+      "epoch": 0.9601160261058739,
+      "grad_norm": 0.4044716506352786,
+      "learning_rate": 4.5021543964306466e-08,
+      "loss": 0.5123,
+      "step": 1986
+    },
+    {
+      "epoch": 0.9605994682136814,
+      "grad_norm": 0.40133573312591214,
+      "learning_rate": 4.395754714603351e-08,
+      "loss": 0.5133,
+      "step": 1987
+    },
+    {
+      "epoch": 0.961082910321489,
+      "grad_norm": 0.4089192998561785,
+      "learning_rate": 4.290621837170661e-08,
+      "loss": 0.5236,
+      "step": 1988
+    },
+    {
+      "epoch": 0.9615663524292966,
+      "grad_norm": 0.39452360352891674,
+      "learning_rate": 4.186756032860728e-08,
+      "loss": 0.5137,
+      "step": 1989
+    },
+    {
+      "epoch": 0.9620497945371042,
+      "grad_norm": 0.39867371724056727,
+      "learning_rate": 4.08415756716285e-08,
+      "loss": 0.5093,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9625332366449117,
+      "grad_norm": 0.357065447847406,
+      "learning_rate": 3.9828267023269696e-08,
+      "loss": 0.4505,
+      "step": 1991
+    },
+    {
+      "epoch": 0.9630166787527193,
+      "grad_norm": 0.427089982663271,
+      "learning_rate": 3.8827636973630126e-08,
+      "loss": 0.5101,
+      "step": 1992
+    },
+    {
+      "epoch": 0.9635001208605269,
+      "grad_norm": 0.4025101063369687,
+      "learning_rate": 3.783968808039995e-08,
+      "loss": 0.5245,
+      "step": 1993
+    },
+    {
+      "epoch": 0.9639835629683345,
+      "grad_norm": 0.4012223737061637,
+      "learning_rate": 3.68644228688575e-08,
+      "loss": 0.514,
+      "step": 1994
+    },
+    {
+      "epoch": 0.9644670050761421,
+      "grad_norm": 0.39715847085154765,
+      "learning_rate": 3.590184383185758e-08,
+      "loss": 0.507,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9649504471839497,
+      "grad_norm": 0.4019019064729592,
+      "learning_rate": 3.4951953429831484e-08,
+      "loss": 0.5093,
+      "step": 1996
+    },
+    {
+      "epoch": 0.9654338892917573,
+      "grad_norm": 0.3997438820964838,
+      "learning_rate": 3.401475409077426e-08,
+      "loss": 0.4987,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9659173313995649,
+      "grad_norm": 0.42247021949710184,
+      "learning_rate": 3.309024821024354e-08,
+      "loss": 0.5099,
+      "step": 1998
+    },
+    {
+      "epoch": 0.9664007735073725,
+      "grad_norm": 0.4228197536210846,
+      "learning_rate": 3.2178438151350685e-08,
+      "loss": 0.5181,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9668842156151801,
+      "grad_norm": 0.40195549014330173,
+      "learning_rate": 3.127932624475638e-08,
+      "loss": 0.5118,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9673676577229877,
+      "grad_norm": 0.40083823310970984,
+      "learning_rate": 3.039291478866169e-08,
+      "loss": 0.5265,
+      "step": 2001
+    },
+    {
+      "epoch": 0.9678510998307953,
+      "grad_norm": 0.4054162095867977,
+      "learning_rate": 2.9519206048807535e-08,
+      "loss": 0.5173,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9683345419386028,
+      "grad_norm": 0.4091589042260666,
+      "learning_rate": 2.8658202258462498e-08,
+      "loss": 0.5199,
+      "step": 2003
+    },
+    {
+      "epoch": 0.9688179840464104,
+      "grad_norm": 0.37360554914951866,
+      "learning_rate": 2.7809905618422227e-08,
+      "loss": 0.4667,
+      "step": 2004
+    },
+    {
+      "epoch": 0.969301426154218,
+      "grad_norm": 0.4264262470861418,
+      "learning_rate": 2.6974318297001144e-08,
+      "loss": 0.5208,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9697848682620256,
+      "grad_norm": 0.4133603239690626,
+      "learning_rate": 2.615144243002743e-08,
+      "loss": 0.5049,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9702683103698332,
+      "grad_norm": 0.41234915425778607,
+      "learning_rate": 2.534128012083914e-08,
+      "loss": 0.5215,
+      "step": 2007
+    },
+    {
+      "epoch": 0.9707517524776408,
+      "grad_norm": 0.39530605693418713,
+      "learning_rate": 2.4543833440275332e-08,
+      "loss": 0.5096,
+      "step": 2008
+    },
+    {
+      "epoch": 0.9712351945854484,
+      "grad_norm": 0.42034129099753553,
+      "learning_rate": 2.375910442667495e-08,
+      "loss": 0.5111,
+      "step": 2009
+    },
+    {
+      "epoch": 0.971718636693256,
+      "grad_norm": 0.4128961831040994,
+      "learning_rate": 2.298709508586794e-08,
+      "loss": 0.5136,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9722020788010636,
+      "grad_norm": 0.40946352601157776,
+      "learning_rate": 2.2227807391172474e-08,
+      "loss": 0.5239,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9726855209088712,
+      "grad_norm": 0.3999251664775986,
+      "learning_rate": 2.1481243283389408e-08,
+      "loss": 0.514,
+      "step": 2012
+    },
+    {
+      "epoch": 0.9731689630166788,
+      "grad_norm": 0.412440530125608,
+      "learning_rate": 2.074740467079672e-08,
+      "loss": 0.5174,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9736524051244864,
+      "grad_norm": 0.40416741933458183,
+      "learning_rate": 2.002629342914453e-08,
+      "loss": 0.5173,
+      "step": 2014
+    },
+    {
+      "epoch": 0.974135847232294,
+      "grad_norm": 0.4091688587167212,
+      "learning_rate": 1.9317911401651734e-08,
+      "loss": 0.5035,
+      "step": 2015
+    },
+    {
+      "epoch": 0.9746192893401016,
+      "grad_norm": 0.41181207482580323,
+      "learning_rate": 1.862226039899995e-08,
+      "loss": 0.5194,
+      "step": 2016
+    },
+    {
+      "epoch": 0.9751027314479092,
+      "grad_norm": 0.38917062513507206,
+      "learning_rate": 1.7939342199329023e-08,
+      "loss": 0.5081,
+      "step": 2017
+    },
+    {
+      "epoch": 0.9755861735557168,
+      "grad_norm": 0.4058095413062891,
+      "learning_rate": 1.7269158548232633e-08,
+      "loss": 0.514,
+      "step": 2018
+    },
+    {
+      "epoch": 0.9760696156635243,
+      "grad_norm": 0.3909310997249257,
+      "learning_rate": 1.661171115875493e-08,
+      "loss": 0.5086,
+      "step": 2019
+    },
+    {
+      "epoch": 0.9765530577713318,
+      "grad_norm": 0.3924463631554743,
+      "learning_rate": 1.5967001711383877e-08,
+      "loss": 0.5074,
+      "step": 2020
+    },
+    {
+      "epoch": 0.9770364998791394,
+      "grad_norm": 0.3897349184690982,
+      "learning_rate": 1.5335031854049055e-08,
+      "loss": 0.5164,
+      "step": 2021
+    },
+    {
+      "epoch": 0.977519941986947,
+      "grad_norm": 0.4006696563415638,
+      "learning_rate": 1.4715803202116075e-08,
+      "loss": 0.516,
+      "step": 2022
+    },
+    {
+      "epoch": 0.9780033840947546,
+      "grad_norm": 0.39127207135897235,
+      "learning_rate": 1.4109317338383832e-08,
+      "loss": 0.4864,
+      "step": 2023
+    },
+    {
+      "epoch": 0.9784868262025622,
+      "grad_norm": 0.41074499955315413,
+      "learning_rate": 1.3515575813078386e-08,
+      "loss": 0.5276,
+      "step": 2024
+    },
+    {
+      "epoch": 0.9789702683103698,
+      "grad_norm": 0.3989638057067789,
+      "learning_rate": 1.2934580143851294e-08,
+      "loss": 0.5116,
+      "step": 2025
+    },
+    {
+      "epoch": 0.9794537104181774,
+      "grad_norm": 0.4179669163774858,
+      "learning_rate": 1.2366331815774069e-08,
+      "loss": 0.5169,
+      "step": 2026
+    },
+    {
+      "epoch": 0.979937152525985,
+      "grad_norm": 0.420952308284563,
+      "learning_rate": 1.1810832281335394e-08,
+      "loss": 0.5221,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9804205946337926,
+      "grad_norm": 0.40444089801366945,
+      "learning_rate": 1.1268082960436688e-08,
+      "loss": 0.526,
+      "step": 2028
+    },
+    {
+      "epoch": 0.9809040367416002,
+      "grad_norm": 0.39774599938725236,
+      "learning_rate": 1.0738085240389883e-08,
+      "loss": 0.5158,
+      "step": 2029
+    },
+    {
+      "epoch": 0.9813874788494078,
+      "grad_norm": 0.40528876629152616,
+      "learning_rate": 1.0220840475910765e-08,
+      "loss": 0.5148,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9818709209572154,
+      "grad_norm": 0.3880426443734388,
+      "learning_rate": 9.716349989118412e-09,
+      "loss": 0.4977,
+      "step": 2031
+    },
+    {
+      "epoch": 0.982354363065023,
+      "grad_norm": 0.397796748759872,
+      "learning_rate": 9.224615069532428e-09,
+      "loss": 0.5183,
+      "step": 2032
+    },
+    {
+      "epoch": 0.9828378051728306,
+      "grad_norm": 0.41455233095701044,
+      "learning_rate": 8.745636974066274e-09,
+      "loss": 0.5151,
+      "step": 2033
+    },
+    {
+      "epoch": 0.9833212472806382,
+      "grad_norm": 0.39149878382311915,
+      "learning_rate": 8.279416927026163e-09,
+      "loss": 0.4852,
+      "step": 2034
+    },
+    {
+      "epoch": 0.9838046893884458,
+      "grad_norm": 0.428044296219464,
+      "learning_rate": 7.82595612010828e-09,
+      "loss": 0.5088,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9842881314962533,
+      "grad_norm": 0.39066146033771326,
+      "learning_rate": 7.385255712395456e-09,
+      "loss": 0.5092,
+      "step": 2036
+    },
+    {
+      "epoch": 0.9847715736040609,
+      "grad_norm": 0.3944546565780817,
+      "learning_rate": 6.9573168303532775e-09,
+      "loss": 0.5048,
+      "step": 2037
+    },
+    {
+      "epoch": 0.9852550157118685,
+      "grad_norm": 0.3875137169857006,
+      "learning_rate": 6.542140567827871e-09,
+      "loss": 0.5166,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9857384578196761,
+      "grad_norm": 0.3942008356766705,
+      "learning_rate": 6.1397279860431205e-09,
+      "loss": 0.4846,
+      "step": 2039
+    },
+    {
+      "epoch": 0.9862218999274837,
+      "grad_norm": 0.4122295335735729,
+      "learning_rate": 5.750080113598455e-09,
+      "loss": 0.5191,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9867053420352913,
+      "grad_norm": 0.4027922091033447,
+      "learning_rate": 5.373197946464403e-09,
+      "loss": 0.509,
+      "step": 2041
+    },
+    {
+      "epoch": 0.9871887841430989,
+      "grad_norm": 0.41834346896126373,
+      "learning_rate": 5.009082447983149e-09,
+      "loss": 0.52,
+      "step": 2042
+    },
+    {
+      "epoch": 0.9876722262509064,
+      "grad_norm": 0.4293813449158595,
+      "learning_rate": 4.65773454886298e-09,
+      "loss": 0.5131,
+      "step": 2043
+    },
+    {
+      "epoch": 0.988155668358714,
+      "grad_norm": 0.3928243234499177,
+      "learning_rate": 4.319155147176624e-09,
+      "loss": 0.515,
+      "step": 2044
+    },
+    {
+      "epoch": 0.9886391104665216,
+      "grad_norm": 0.4041253868270013,
+      "learning_rate": 3.9933451083612464e-09,
+      "loss": 0.5001,
+      "step": 2045
+    },
+    {
+      "epoch": 0.9891225525743292,
+      "grad_norm": 0.4010404057661429,
+      "learning_rate": 3.6803052652134572e-09,
+      "loss": 0.5077,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9896059946821368,
+      "grad_norm": 0.4298866341542551,
+      "learning_rate": 3.3800364178881996e-09,
+      "loss": 0.5112,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9900894367899444,
+      "grad_norm": 0.39490801058055036,
+      "learning_rate": 3.092539333896527e-09,
+      "loss": 0.5087,
+      "step": 2048
+    },
+    {
+      "epoch": 0.990572878897752,
+      "grad_norm": 0.4110997944280951,
+      "learning_rate": 2.817814748104497e-09,
+      "loss": 0.5044,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9910563210055596,
+      "grad_norm": 0.4386800273412446,
+      "learning_rate": 2.555863362730393e-09,
+      "loss": 0.5217,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9915397631133672,
+      "grad_norm": 0.411463570514358,
+      "learning_rate": 2.30668584734306e-09,
+      "loss": 0.5117,
+      "step": 2051
+    },
+    {
+      "epoch": 0.9920232052211748,
+      "grad_norm": 0.40419424446848334,
+      "learning_rate": 2.070282838859683e-09,
+      "loss": 0.5056,
+      "step": 2052
+    },
+    {
+      "epoch": 0.9925066473289823,
+      "grad_norm": 0.40703300041841234,
+      "learning_rate": 1.8466549415463442e-09,
+      "loss": 0.5319,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9929900894367899,
+      "grad_norm": 0.42530845899182246,
+      "learning_rate": 1.635802727013025e-09,
+      "loss": 0.5138,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9934735315445975,
+      "grad_norm": 0.4058607106461754,
+      "learning_rate": 1.4377267342158274e-09,
+      "loss": 0.4883,
+      "step": 2055
+    },
+    {
+      "epoch": 0.9939569736524051,
+      "grad_norm": 0.4062782037922318,
+      "learning_rate": 1.2524274694525329e-09,
+      "loss": 0.5225,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9944404157602127,
+      "grad_norm": 4.724922277274833,
+      "learning_rate": 1.0799054063626024e-09,
+      "loss": 0.5232,
+      "step": 2057
+    },
+    {
+      "epoch": 0.9949238578680203,
+      "grad_norm": 0.3906054598062824,
+      "learning_rate": 9.201609859271765e-10,
+      "loss": 0.5132,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9954072999758279,
+      "grad_norm": 0.3973857264750407,
+      "learning_rate": 7.731946164657445e-10,
+      "loss": 0.5083,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9958907420836355,
+      "grad_norm": 0.39787099969142303,
+      "learning_rate": 6.390066736355893e-10,
+      "loss": 0.5138,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9963741841914431,
+      "grad_norm": 0.4073985533388715,
+      "learning_rate": 5.17597500432343e-10,
+      "loss": 0.5134,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9968576262992507,
+      "grad_norm": 0.4217388952314898,
+      "learning_rate": 4.089674071872107e-10,
+      "loss": 0.5204,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9973410684070583,
+      "grad_norm": 0.3938957473860418,
+      "learning_rate": 3.131166715680811e-10,
+      "loss": 0.5134,
+      "step": 2063
+    },
+    {
+      "epoch": 0.9978245105148659,
+      "grad_norm": 0.4133454970429142,
+      "learning_rate": 2.3004553857675082e-10,
+      "loss": 0.5136,
+      "step": 2064
+    },
+    {
+      "epoch": 0.9983079526226735,
+      "grad_norm": 3.9528345451843885,
+      "learning_rate": 1.5975422055003465e-10,
+      "loss": 0.5088,
+      "step": 2065
+    },
+    {
+      "epoch": 0.9987913947304811,
+      "grad_norm": 0.4124191883874225,
+      "learning_rate": 1.022428971581002e-10,
+      "loss": 0.5106,
+      "step": 2066
+    },
+    {
+      "epoch": 0.9992748368382887,
+      "grad_norm": 0.42214447971757757,
+      "learning_rate": 5.751171540391287e-11,
+      "loss": 0.513,
+      "step": 2067
+    },
+    {
+      "epoch": 0.9997582789460963,
+      "grad_norm": 0.4373529560530007,
+      "learning_rate": 2.556078962490105e-11,
+      "loss": 0.5278,
+      "step": 2068
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4373529560530007,
+      "learning_rate": 6.390201489625547e-12,
+      "loss": 0.4723,
+      "step": 2069
+    },
+    {
+      "epoch": 1.0,
+      "step": 2069,
+      "total_flos": 2898754626256896.0,
+      "train_loss": 0.5475362745847836,
+      "train_runtime": 128435.5344,
+      "train_samples_per_second": 2.061,
+      "train_steps_per_second": 0.016
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2069,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 208,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2898754626256896.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}