diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8631 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 1227,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.007773875878433,
+      "learning_rate": 5.405405405405406e-07,
+      "loss": 1.0642,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.857206220727576,
+      "learning_rate": 1.0810810810810812e-06,
+      "loss": 1.1634,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7967680760923868,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 1.0318,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.902180242609355,
+      "learning_rate": 2.1621621621621623e-06,
+      "loss": 1.0997,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0866839964676887,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 1.0777,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6411177721266106,
+      "learning_rate": 3.2432432432432437e-06,
+      "loss": 1.1159,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8484602651995545,
+      "learning_rate": 3.7837837837837844e-06,
+      "loss": 1.1075,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3661117374745435,
+      "learning_rate": 4.324324324324325e-06,
+      "loss": 1.0968,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6897613858242715,
+      "learning_rate": 4.864864864864866e-06,
+      "loss": 1.0019,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1067843807324762,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 1.0525,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1949647364804228,
+      "learning_rate": 5.945945945945947e-06,
+      "loss": 0.9823,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1650155608059656,
+      "learning_rate": 6.486486486486487e-06,
+      "loss": 1.0187,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.337839660443822,
+      "learning_rate": 7.027027027027028e-06,
+      "loss": 0.9799,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.706949908048817,
+      "learning_rate": 7.567567567567569e-06,
+      "loss": 1.0251,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.538502149835905,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 0.9682,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.368696091479067,
+      "learning_rate": 8.64864864864865e-06,
+      "loss": 1.0244,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.065537158684757,
+      "learning_rate": 9.189189189189191e-06,
+      "loss": 1.0074,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3166065957554898,
+      "learning_rate": 9.729729729729732e-06,
+      "loss": 0.9828,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0227898100960249,
+      "learning_rate": 1.027027027027027e-05,
+      "loss": 0.9473,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1052630394089142,
+      "learning_rate": 1.0810810810810812e-05,
+      "loss": 0.8935,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0310684987266245,
+      "learning_rate": 1.1351351351351352e-05,
+      "loss": 0.9551,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.047549783181247,
+      "learning_rate": 1.1891891891891894e-05,
+      "loss": 0.9808,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8944923205462866,
+      "learning_rate": 1.2432432432432433e-05,
+      "loss": 0.9036,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8946413055843416,
+      "learning_rate": 1.2972972972972975e-05,
+      "loss": 0.9881,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9958101084749376,
+      "learning_rate": 1.3513513513513515e-05,
+      "loss": 1.0941,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.87651499567766,
+      "learning_rate": 1.4054054054054055e-05,
+      "loss": 0.9625,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.805066306166799,
+      "learning_rate": 1.4594594594594596e-05,
+      "loss": 0.9283,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7049202710021768,
+      "learning_rate": 1.5135135135135138e-05,
+      "loss": 0.927,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8128712588755304,
+      "learning_rate": 1.5675675675675676e-05,
+      "loss": 0.8556,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.694832518978497,
+      "learning_rate": 1.6216216216216218e-05,
+      "loss": 0.9383,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9257102494736904,
+      "learning_rate": 1.6756756756756757e-05,
+      "loss": 0.9333,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7510360548090936,
+      "learning_rate": 1.72972972972973e-05,
+      "loss": 0.9407,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7538350811054269,
+      "learning_rate": 1.783783783783784e-05,
+      "loss": 0.9213,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7543809555126735,
+      "learning_rate": 1.8378378378378383e-05,
+      "loss": 0.9341,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7381433651076192,
+      "learning_rate": 1.891891891891892e-05,
+      "loss": 0.8983,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7180289400292585,
+      "learning_rate": 1.9459459459459463e-05,
+      "loss": 0.8925,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7794830978461427,
+      "learning_rate": 2e-05,
+      "loss": 0.8603,
+      "step": 37
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.172653145104533,
+      "learning_rate": 1.999996515218322e-05,
+      "loss": 0.9164,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1635885899730565,
+      "learning_rate": 1.999986060897575e-05,
+      "loss": 0.8851,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6221780237162938,
+      "learning_rate": 1.999968637110621e-05,
+      "loss": 0.8973,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.905869585897509,
+      "learning_rate": 1.9999442439788964e-05,
+      "loss": 0.9,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7139073653989279,
+      "learning_rate": 1.999912881672411e-05,
+      "loss": 0.8876,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0597066496178233,
+      "learning_rate": 1.9998745504097457e-05,
+      "loss": 1.102,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7124480177534466,
+      "learning_rate": 1.9998292504580528e-05,
+      "loss": 0.901,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8247381247566109,
+      "learning_rate": 1.9997769821330534e-05,
+      "loss": 0.8346,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6360988812031477,
+      "learning_rate": 1.999717745799035e-05,
+      "loss": 0.873,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6691109246694686,
+      "learning_rate": 1.9996515418688493e-05,
+      "loss": 0.8622,
+      "step": 47
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5206145667560276,
+      "learning_rate": 1.9995783708039075e-05,
+      "loss": 0.9121,
+      "step": 48
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8929405632667002,
+      "learning_rate": 1.9994982331141813e-05,
+      "loss": 0.8964,
+      "step": 49
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7713737306605302,
+      "learning_rate": 1.9994111293581947e-05,
+      "loss": 0.8909,
+      "step": 50
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6654831932835349,
+      "learning_rate": 1.9993170601430233e-05,
+      "loss": 0.8645,
+      "step": 51
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7559249453024011,
+      "learning_rate": 1.999216026124288e-05,
+      "loss": 0.9169,
+      "step": 52
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7220497377862712,
+      "learning_rate": 1.9991080280061518e-05,
+      "loss": 0.8841,
+      "step": 53
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8444439771114949,
+      "learning_rate": 1.9989930665413148e-05,
+      "loss": 0.8727,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9566137667114166,
+      "learning_rate": 1.998871142531008e-05,
+      "loss": 0.8886,
+      "step": 55
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8257943147168417,
+      "learning_rate": 1.9987422568249886e-05,
+      "loss": 0.9217,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0434974088970088,
+      "learning_rate": 1.998606410321534e-05,
+      "loss": 0.8799,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7461768192626946,
+      "learning_rate": 1.9984636039674342e-05,
+      "loss": 0.8855,
+      "step": 58
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6188639876065991,
+      "learning_rate": 1.9983138387579877e-05,
+      "loss": 0.8638,
+      "step": 59
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8999602349060933,
+      "learning_rate": 1.9981571157369925e-05,
+      "loss": 0.8853,
+      "step": 60
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3174850456215421,
+      "learning_rate": 1.9979934359967398e-05,
+      "loss": 0.8998,
+      "step": 61
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7995372518873074,
+      "learning_rate": 1.9978228006780056e-05,
+      "loss": 0.8682,
+      "step": 62
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8252725966611132,
+      "learning_rate": 1.9976452109700436e-05,
+      "loss": 0.8696,
+      "step": 63
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0495709519973797,
+      "learning_rate": 1.997460668110577e-05,
+      "loss": 0.8594,
+      "step": 64
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9390841981330008,
+      "learning_rate": 1.997269173385788e-05,
+      "loss": 0.8611,
+      "step": 65
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9476981679588746,
+      "learning_rate": 1.9970707281303126e-05,
+      "loss": 0.9989,
+      "step": 66
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8440449864694928,
+      "learning_rate": 1.9968653337272262e-05,
+      "loss": 0.8122,
+      "step": 67
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8847901924859267,
+      "learning_rate": 1.996652991608039e-05,
+      "loss": 0.8571,
+      "step": 68
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6549842142806805,
+      "learning_rate": 1.996433703252682e-05,
+      "loss": 0.8925,
+      "step": 69
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9223720864117396,
+      "learning_rate": 1.9962074701895006e-05,
+      "loss": 0.8879,
+      "step": 70
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9764677442649146,
+      "learning_rate": 1.9959742939952393e-05,
+      "loss": 0.8117,
+      "step": 71
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9850273266670845,
+      "learning_rate": 1.9957341762950346e-05,
+      "loss": 0.8666,
+      "step": 72
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.962692728833449,
+      "learning_rate": 1.9954871187624027e-05,
+      "loss": 0.9053,
+      "step": 73
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9454258895371642,
+      "learning_rate": 1.9952331231192256e-05,
+      "loss": 0.864,
+      "step": 74
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9096512226666619,
+      "learning_rate": 1.994972191135743e-05,
+      "loss": 0.87,
+      "step": 75
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.981353470849719,
+      "learning_rate": 1.9947043246305365e-05,
+      "loss": 0.8859,
+      "step": 76
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9193231471198929,
+      "learning_rate": 1.9944295254705187e-05,
+      "loss": 0.8526,
+      "step": 77
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8905271619805811,
+      "learning_rate": 1.9941477955709195e-05,
+      "loss": 0.8716,
+      "step": 78
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9184422966842619,
+      "learning_rate": 1.993859136895274e-05,
+      "loss": 0.8354,
+      "step": 79
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3999409668377503,
+      "learning_rate": 1.9935635514554062e-05,
+      "loss": 0.8626,
+      "step": 80
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1628411466199544,
+      "learning_rate": 1.9932610413114184e-05,
+      "loss": 0.8391,
+      "step": 81
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7359320092864131,
+      "learning_rate": 1.9929516085716736e-05,
+      "loss": 0.8917,
+      "step": 82
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0705737746636836,
+      "learning_rate": 1.9926352553927833e-05,
+      "loss": 0.8575,
+      "step": 83
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9057245860836382,
+      "learning_rate": 1.992311983979591e-05,
+      "loss": 1.0095,
+      "step": 84
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.723952695240946,
+      "learning_rate": 1.9919817965851566e-05,
+      "loss": 0.891,
+      "step": 85
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8192498054074059,
+      "learning_rate": 1.991644695510743e-05,
+      "loss": 0.8735,
+      "step": 86
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6388338595986612,
+      "learning_rate": 1.9913006831057967e-05,
+      "loss": 0.8567,
+      "step": 87
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7395229777318928,
+      "learning_rate": 1.990949761767935e-05,
+      "loss": 0.833,
+      "step": 88
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6960470315375072,
+      "learning_rate": 1.9905919339429254e-05,
+      "loss": 0.8656,
+      "step": 89
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8126549022017021,
+      "learning_rate": 1.9902272021246718e-05,
+      "loss": 0.8377,
+      "step": 90
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8001333245737844,
+      "learning_rate": 1.989855568855196e-05,
+      "loss": 0.9069,
+      "step": 91
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6246257857224763,
+      "learning_rate": 1.9894770367246197e-05,
+      "loss": 0.8756,
+      "step": 92
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7135504832280406,
+      "learning_rate": 1.9890916083711463e-05,
+      "loss": 0.8565,
+      "step": 93
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0049924597661066,
+      "learning_rate": 1.9886992864810427e-05,
+      "loss": 0.8694,
+      "step": 94
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6289843920776702,
+      "learning_rate": 1.9883000737886222e-05,
+      "loss": 0.8444,
+      "step": 95
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8240416323557646,
+      "learning_rate": 1.9878939730762222e-05,
+      "loss": 0.8231,
+      "step": 96
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6314357606841658,
+      "learning_rate": 1.9874809871741877e-05,
+      "loss": 0.8336,
+      "step": 97
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8695356040869762,
+      "learning_rate": 1.9870611189608494e-05,
+      "loss": 0.8757,
+      "step": 98
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.688194653525997,
+      "learning_rate": 1.9866343713625066e-05,
+      "loss": 0.8585,
+      "step": 99
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9499108164870441,
+      "learning_rate": 1.9862007473534026e-05,
+      "loss": 0.8779,
+      "step": 100
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7550314742203836,
+      "learning_rate": 1.985760249955708e-05,
+      "loss": 0.9369,
+      "step": 101
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3902820113381522,
+      "learning_rate": 1.9853128822394976e-05,
+      "loss": 0.8713,
+      "step": 102
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6578487819760912,
+      "learning_rate": 1.9848586473227286e-05,
+      "loss": 0.8673,
+      "step": 103
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7466051990180269,
+      "learning_rate": 1.9843975483712198e-05,
+      "loss": 0.8488,
+      "step": 104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7817084597439625,
+      "learning_rate": 1.98392958859863e-05,
+      "loss": 0.8308,
+      "step": 105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7680665356429073,
+      "learning_rate": 1.983454771266434e-05,
+      "loss": 0.8361,
+      "step": 106
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7849892604067501,
+      "learning_rate": 1.982973099683902e-05,
+      "loss": 0.8729,
+      "step": 107
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7529350832962131,
+      "learning_rate": 1.982484577208074e-05,
+      "loss": 0.8519,
+      "step": 108
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8620625120138664,
+      "learning_rate": 1.9819892072437388e-05,
+      "loss": 0.7973,
+      "step": 109
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6995472310920965,
+      "learning_rate": 1.9814869932434083e-05,
+      "loss": 0.8492,
+      "step": 110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8193968845664128,
+      "learning_rate": 1.980977938707295e-05,
+      "loss": 0.8526,
+      "step": 111
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.786906492728911,
+      "learning_rate": 1.980462047183287e-05,
+      "loss": 0.9187,
+      "step": 112
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7477484388682966,
+      "learning_rate": 1.9799393222669224e-05,
+      "loss": 0.8019,
+      "step": 113
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8694975251318111,
+      "learning_rate": 1.979409767601366e-05,
+      "loss": 0.753,
+      "step": 114
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.749484757929014,
+      "learning_rate": 1.9788733868773826e-05,
+      "loss": 0.855,
+      "step": 115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6696158439558821,
+      "learning_rate": 1.9783301838333116e-05,
+      "loss": 0.8614,
+      "step": 116
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9846436399717825,
+      "learning_rate": 1.977780162255041e-05,
+      "loss": 0.8198,
+      "step": 117
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7139828628211472,
+      "learning_rate": 1.977223325975981e-05,
+      "loss": 0.8288,
+      "step": 118
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8428451010859759,
+      "learning_rate": 1.9766596788770373e-05,
+      "loss": 0.8732,
+      "step": 119
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7104517536296814,
+      "learning_rate": 1.9760892248865843e-05,
+      "loss": 0.9592,
+      "step": 120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3639065171428235,
+      "learning_rate": 1.975511967980437e-05,
+      "loss": 0.8146,
+      "step": 121
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6637197642842009,
+      "learning_rate": 1.9749279121818235e-05,
+      "loss": 0.807,
+      "step": 122
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.831864853940694,
+      "learning_rate": 1.974337061561359e-05,
+      "loss": 0.8403,
+      "step": 123
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7545282967284532,
+      "learning_rate": 1.9737394202370132e-05,
+      "loss": 0.8677,
+      "step": 124
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7474853254184157,
+      "learning_rate": 1.9731349923740857e-05,
+      "loss": 0.8173,
+      "step": 125
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7835828879337019,
+      "learning_rate": 1.9725237821851745e-05,
+      "loss": 0.8202,
+      "step": 126
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8679607102361868,
+      "learning_rate": 1.9719057939301477e-05,
+      "loss": 0.8572,
+      "step": 127
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8649305059529203,
+      "learning_rate": 1.971281031916114e-05,
+      "loss": 0.8942,
+      "step": 128
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8490273539418802,
+      "learning_rate": 1.9706495004973914e-05,
+      "loss": 0.8357,
+      "step": 129
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.183384939083542,
+      "learning_rate": 1.9700112040754784e-05,
+      "loss": 0.8165,
+      "step": 130
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7254090129598603,
+      "learning_rate": 1.9693661470990225e-05,
+      "loss": 0.7967,
+      "step": 131
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7903493887431122,
+      "learning_rate": 1.9687143340637885e-05,
+      "loss": 0.8667,
+      "step": 132
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.867657995780601,
+      "learning_rate": 1.9680557695126294e-05,
+      "loss": 0.92,
+      "step": 133
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7634965502772261,
+      "learning_rate": 1.9673904580354523e-05,
+      "loss": 0.8427,
+      "step": 134
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.056001273456717,
+      "learning_rate": 1.9667184042691877e-05,
+      "loss": 0.7941,
+      "step": 135
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9191218337807702,
+      "learning_rate": 1.9660396128977566e-05,
+      "loss": 0.8569,
+      "step": 136
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.751994877158226,
+      "learning_rate": 1.9653540886520387e-05,
+      "loss": 0.8513,
+      "step": 137
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8871085347681905,
+      "learning_rate": 1.9646618363098388e-05,
+      "loss": 0.8112,
+      "step": 138
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8656846964398582,
+      "learning_rate": 1.9639628606958535e-05,
+      "loss": 0.8135,
+      "step": 139
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3922165128687753,
+      "learning_rate": 1.963257166681637e-05,
+      "loss": 0.8452,
+      "step": 140
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7714516685024676,
+      "learning_rate": 1.9625447591855695e-05,
+      "loss": 0.9326,
+      "step": 141
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3113355565847573,
+      "learning_rate": 1.961825643172819e-05,
+      "loss": 0.8614,
+      "step": 142
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8196547585997562,
+      "learning_rate": 1.961099823655311e-05,
+      "loss": 0.8289,
+      "step": 143
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9426085355836699,
+      "learning_rate": 1.9603673056916907e-05,
+      "loss": 0.8067,
+      "step": 144
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2221433642931228,
+      "learning_rate": 1.9596280943872875e-05,
+      "loss": 0.8446,
+      "step": 145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.866819365404402,
+      "learning_rate": 1.958882194894082e-05,
+      "loss": 0.8458,
+      "step": 146
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7194063741485811,
+      "learning_rate": 1.9581296124106682e-05,
+      "loss": 0.8168,
+      "step": 147
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0846392923721617,
+      "learning_rate": 1.957370352182217e-05,
+      "loss": 0.7994,
+      "step": 148
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8933535671346733,
+      "learning_rate": 1.956604419500441e-05,
+      "loss": 0.829,
+      "step": 149
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9755100456578889,
+      "learning_rate": 1.9558318197035562e-05,
+      "loss": 0.8855,
+      "step": 150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7708397549879094,
+      "learning_rate": 1.955052558176246e-05,
+      "loss": 0.8083,
+      "step": 151
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9460543617272675,
+      "learning_rate": 1.9542666403496232e-05,
+      "loss": 0.8821,
+      "step": 152
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1174490467064162,
+      "learning_rate": 1.9534740717011917e-05,
+      "loss": 0.8064,
+      "step": 153
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0002826862266712,
+      "learning_rate": 1.9526748577548085e-05,
+      "loss": 0.8397,
+      "step": 154
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8630790310082779,
+      "learning_rate": 1.9518690040806465e-05,
+      "loss": 0.8199,
+      "step": 155
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9216650450071842,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 0.8125,
+      "step": 156
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7621019446797,
+      "learning_rate": 1.9502374000610152e-05,
+      "loss": 0.7973,
+      "step": 157
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0963348788120029,
+      "learning_rate": 1.9494116610871133e-05,
+      "loss": 0.7595,
+      "step": 158
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0722046322852692,
+      "learning_rate": 1.9485793051284884e-05,
+      "loss": 0.863,
+      "step": 159
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9038147563488605,
+      "learning_rate": 1.9477403379862976e-05,
+      "loss": 0.8286,
+      "step": 160
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1491249222228266,
+      "learning_rate": 1.9468947655077763e-05,
+      "loss": 0.8453,
+      "step": 161
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0097893592901905,
+      "learning_rate": 1.946042593586195e-05,
+      "loss": 0.8337,
+      "step": 162
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1126492132392058,
+      "learning_rate": 1.94518382816082e-05,
+      "loss": 0.8449,
+      "step": 163
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9148283461966971,
+      "learning_rate": 1.944318475216871e-05,
+      "loss": 0.8283,
+      "step": 164
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9186582857491254,
+      "learning_rate": 1.9434465407854803e-05,
+      "loss": 0.8535,
+      "step": 165
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.856650925062199,
+      "learning_rate": 1.9425680309436506e-05,
+      "loss": 0.829,
+      "step": 166
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0221504910055796,
+      "learning_rate": 1.941682951814212e-05,
+      "loss": 0.7882,
+      "step": 167
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8386380980059264,
+      "learning_rate": 1.9407913095657785e-05,
+      "loss": 0.7817,
+      "step": 168
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0113267911380253,
+      "learning_rate": 1.939893110412708e-05,
+      "loss": 0.862,
+      "step": 169
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.074343720189095,
+      "learning_rate": 1.938988360615057e-05,
+      "loss": 0.8605,
+      "step": 170
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.192518302320637,
+      "learning_rate": 1.9380770664785352e-05,
+      "loss": 0.8146,
+      "step": 171
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.385276160608002,
+      "learning_rate": 1.9371592343544655e-05,
+      "loss": 0.8383,
+      "step": 172
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1099211868000982,
+      "learning_rate": 1.9362348706397374e-05,
+      "loss": 0.8066,
+      "step": 173
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8907583522755066,
+      "learning_rate": 1.935303981776762e-05,
+      "loss": 0.8368,
+      "step": 174
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.73097937043873,
+      "learning_rate": 1.934366574253428e-05,
+      "loss": 0.8317,
+      "step": 175
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.03459393756036,
+      "learning_rate": 1.9334226546030567e-05,
+      "loss": 0.8386,
+      "step": 176
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7115534813257642,
+      "learning_rate": 1.932472229404356e-05,
+      "loss": 0.8082,
+      "step": 177
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9738920097757501,
+      "learning_rate": 1.9315153052813743e-05,
+      "loss": 1.0401,
+      "step": 178
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6695616131500474,
+      "learning_rate": 1.9305518889034552e-05,
+      "loss": 0.7911,
+      "step": 179
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8882400162791445,
+      "learning_rate": 1.9295819869851904e-05,
+      "loss": 0.797,
+      "step": 180
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7158354503340638,
+      "learning_rate": 1.928605606286372e-05,
+      "loss": 0.8094,
+      "step": 181
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9729306959715439,
+      "learning_rate": 1.927622753611948e-05,
+      "loss": 0.8346,
+      "step": 182
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3671394256058627,
+      "learning_rate": 1.9266334358119718e-05,
+      "loss": 0.8378,
+      "step": 183
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0109742540266808,
+      "learning_rate": 1.9256376597815565e-05,
+      "loss": 0.8641,
+      "step": 184
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8290426252052272,
+      "learning_rate": 1.9246354324608264e-05,
+      "loss": 0.78,
+      "step": 185
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9368719853478257,
+      "learning_rate": 1.9236267608348682e-05,
+      "loss": 0.8456,
+      "step": 186
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6616351032800075,
+      "learning_rate": 1.922611651933683e-05,
+      "loss": 0.8226,
+      "step": 187
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8469865376322364,
+      "learning_rate": 1.9215901128321364e-05,
+      "loss": 0.8711,
+      "step": 188
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7056169332256019,
+      "learning_rate": 1.9205621506499097e-05,
+      "loss": 0.8005,
+      "step": 189
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8084263213554583,
+      "learning_rate": 1.919527772551451e-05,
+      "loss": 0.7888,
+      "step": 190
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6750836711884648,
+      "learning_rate": 1.9184869857459233e-05,
+      "loss": 0.83,
+      "step": 191
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7046946668075518,
+      "learning_rate": 1.9174397974871563e-05,
+      "loss": 0.8652,
+      "step": 192
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6444973595395219,
+      "learning_rate": 1.9163862150735958e-05,
+      "loss": 0.7793,
+      "step": 193
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8679453081284159,
+      "learning_rate": 1.9153262458482496e-05,
+      "loss": 0.8068,
+      "step": 194
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8909839733358594,
+      "learning_rate": 1.914259897198642e-05,
+      "loss": 0.853,
+      "step": 195
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7205031552535607,
+      "learning_rate": 1.9131871765567568e-05,
+      "loss": 0.8152,
+      "step": 196
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7454290278397927,
+      "learning_rate": 1.912108091398988e-05,
+      "loss": 1.0208,
+      "step": 197
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8167232377232934,
+      "learning_rate": 1.9110226492460886e-05,
+      "loss": 0.7924,
+      "step": 198
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8665081008668859,
+      "learning_rate": 1.909930857663116e-05,
+      "loss": 0.787,
+      "step": 199
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8201490370356377,
+      "learning_rate": 1.9088327242593813e-05,
+      "loss": 0.7933,
+      "step": 200
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8590510153306822,
+      "learning_rate": 1.9077282566883947e-05,
+      "loss": 0.853,
+      "step": 201
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7065318942585537,
+      "learning_rate": 1.906617462647813e-05,
+      "loss": 0.7735,
+      "step": 202
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8005160540460853,
+      "learning_rate": 1.9055003498793848e-05,
+      "loss": 0.817,
+      "step": 203
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8469602646196659,
+      "learning_rate": 1.9043769261688997e-05,
+      "loss": 0.8161,
+      "step": 204
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9870256381329746,
+      "learning_rate": 1.903247199346129e-05,
+      "loss": 0.8276,
+      "step": 205
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8636796902689928,
+      "learning_rate": 1.9021111772847765e-05,
+      "loss": 0.7976,
+      "step": 206
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7652419141149845,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 0.781,
+      "step": 207
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7408397012725758,
+      "learning_rate": 1.8998202791604554e-05,
+      "loss": 0.7995,
+      "step": 208
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.748126568200572,
+      "learning_rate": 1.8986654190640462e-05,
+      "loss": 0.825,
+      "step": 209
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8031654515366501,
+      "learning_rate": 1.8975042956620632e-05,
+      "loss": 0.813,
+      "step": 210
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6509807697864332,
+      "learning_rate": 1.896336917047029e-05,
+      "loss": 0.7945,
+      "step": 211
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7321849618567235,
+      "learning_rate": 1.8951632913550625e-05,
+      "loss": 0.8017,
+      "step": 212
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8238238876070657,
+      "learning_rate": 1.8939834267658228e-05,
+      "loss": 0.8199,
+      "step": 213
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7264966898825242,
+      "learning_rate": 1.8927973315024508e-05,
+      "loss": 0.8456,
+      "step": 214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7216343694661315,
+      "learning_rate": 1.8916050138315127e-05,
+      "loss": 0.824,
+      "step": 215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9128153988173803,
+      "learning_rate": 1.8904064820629418e-05,
+      "loss": 0.7902,
+      "step": 216
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.696451881050563,
+      "learning_rate": 1.8892017445499812e-05,
+      "loss": 0.8144,
+      "step": 217
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7797719086177407,
+      "learning_rate": 1.8879908096891256e-05,
+      "loss": 0.7954,
+      "step": 218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8202053059181028,
+      "learning_rate": 1.886773685920062e-05,
+      "loss": 0.9373,
+      "step": 219
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2329671655561862,
+      "learning_rate": 1.8855503817256118e-05,
+      "loss": 0.8346,
+      "step": 220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7906067245726275,
+      "learning_rate": 1.884320905631671e-05,
+      "loss": 0.8026,
+      "step": 221
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7928517645859342,
+      "learning_rate": 1.8830852662071507e-05,
+      "loss": 0.8416,
+      "step": 222
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6942970529326871,
+      "learning_rate": 1.8818434720639188e-05,
+      "loss": 0.7786,
+      "step": 223
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9905384736223096,
+      "learning_rate": 1.880595531856738e-05,
+      "loss": 0.8406,
+      "step": 224
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7683385859011829,
+      "learning_rate": 1.879341454283207e-05,
+      "loss": 0.8048,
+      "step": 225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0170090831809877,
+      "learning_rate": 1.878081248083698e-05,
+      "loss": 0.8263,
+      "step": 226
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6355229648322088,
+      "learning_rate": 1.876814922041299e-05,
+      "loss": 0.8063,
+      "step": 227
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1541092215034703,
+      "learning_rate": 1.8755424849817486e-05,
+      "loss": 0.8202,
+      "step": 228
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1151352544088526,
+      "learning_rate": 1.8742639457733785e-05,
+      "loss": 0.8644,
+      "step": 229
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8729194226767725,
+      "learning_rate": 1.872979313327048e-05,
+      "loss": 0.7613,
+      "step": 230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7254563499363096,
+      "learning_rate": 1.871688596596084e-05,
+      "loss": 0.8498,
+      "step": 231
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7211424264159457,
+      "learning_rate": 1.8703918045762197e-05,
+      "loss": 0.7991,
+      "step": 232
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8504123540124784,
+      "learning_rate": 1.8690889463055285e-05,
+      "loss": 0.8309,
+      "step": 233
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7918790539782127,
+      "learning_rate": 1.8677800308643637e-05,
+      "loss": 0.7899,
+      "step": 234
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9516806365444205,
+      "learning_rate": 1.8664650673752945e-05,
+      "loss": 0.817,
+      "step": 235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9522413376277401,
+      "learning_rate": 1.8651440650030423e-05,
+      "loss": 0.8127,
+      "step": 236
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9057407910576015,
+      "learning_rate": 1.8638170329544164e-05,
+      "loss": 0.8388,
+      "step": 237
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8148299014130318,
+      "learning_rate": 1.8624839804782512e-05,
+      "loss": 1.0315,
+      "step": 238
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7894999475546289,
+      "learning_rate": 1.8611449168653404e-05,
+      "loss": 0.7736,
+      "step": 239
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8058300175691486,
+      "learning_rate": 1.8597998514483724e-05,
+      "loss": 0.7612,
+      "step": 240
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9336623161773804,
+      "learning_rate": 1.8584487936018663e-05,
+      "loss": 0.8398,
+      "step": 241
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0080372531775952,
+      "learning_rate": 1.857091752742105e-05,
+      "loss": 0.8085,
+      "step": 242
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0292964356875516,
+      "learning_rate": 1.8557287383270703e-05,
+      "loss": 0.8024,
+      "step": 243
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.239119494122877,
+      "learning_rate": 1.8543597598563786e-05,
+      "loss": 0.8134,
+      "step": 244
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.888774986361828,
+      "learning_rate": 1.8529848268712114e-05,
+      "loss": 0.824,
+      "step": 245
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.382607081499277,
+      "learning_rate": 1.8516039489542513e-05,
+      "loss": 0.8181,
+      "step": 246
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9080045065631199,
+      "learning_rate": 1.8502171357296144e-05,
+      "loss": 0.8623,
+      "step": 247
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7996615599558734,
+      "learning_rate": 1.848824396862783e-05,
+      "loss": 0.7958,
+      "step": 248
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8023744918333462,
+      "learning_rate": 1.84742574206054e-05,
+      "loss": 0.7735,
+      "step": 249
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.847552175548068,
+      "learning_rate": 1.846021181070898e-05,
+      "loss": 0.834,
+      "step": 250
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8022786221871363,
+      "learning_rate": 1.8446107236830336e-05,
+      "loss": 0.8121,
+      "step": 251
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.222839904080142,
+      "learning_rate": 1.8431943797272187e-05,
+      "loss": 0.8103,
+      "step": 252
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7106852200395943,
+      "learning_rate": 1.841772159074753e-05,
+      "loss": 0.7986,
+      "step": 253
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1559562451730465,
+      "learning_rate": 1.840344071637893e-05,
+      "loss": 1.0136,
+      "step": 254
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6948918577000602,
+      "learning_rate": 1.8389101273697846e-05,
+      "loss": 0.8127,
+      "step": 255
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.701022368043413,
+      "learning_rate": 1.837470336264393e-05,
+      "loss": 0.8328,
+      "step": 256
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8141149948859727,
+      "learning_rate": 1.8360247083564343e-05,
+      "loss": 0.792,
+      "step": 257
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7752765290977001,
+      "learning_rate": 1.834573253721303e-05,
+      "loss": 0.8307,
+      "step": 258
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.787903085548626,
+      "learning_rate": 1.833115982475004e-05,
+      "loss": 0.791,
+      "step": 259
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.824693684709889,
+      "learning_rate": 1.8316529047740826e-05,
+      "loss": 0.8211,
+      "step": 260
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.67521565625097,
+      "learning_rate": 1.8301840308155507e-05,
+      "loss": 0.799,
+      "step": 261
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7484949547737446,
+      "learning_rate": 1.828709370836819e-05,
+      "loss": 0.8248,
+      "step": 262
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.758343471272287,
+      "learning_rate": 1.827228935115623e-05,
+      "loss": 0.8483,
+      "step": 263
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6651885420127152,
+      "learning_rate": 1.8257427339699534e-05,
+      "loss": 0.7807,
+      "step": 264
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7485735280879325,
+      "learning_rate": 1.8242507777579833e-05,
+      "loss": 0.8371,
+      "step": 265
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6342491128069362,
+      "learning_rate": 1.8227530768779968e-05,
+      "loss": 0.764,
+      "step": 266
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8420473618103914,
+      "learning_rate": 1.8212496417683135e-05,
+      "loss": 0.8199,
+      "step": 267
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7930370391924098,
+      "learning_rate": 1.8197404829072214e-05,
+      "loss": 0.8011,
+      "step": 268
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6427130503442278,
+      "learning_rate": 1.8182256108128973e-05,
+      "loss": 0.8026,
+      "step": 269
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7509947254180463,
+      "learning_rate": 1.8167050360433394e-05,
+      "loss": 0.785,
+      "step": 270
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.200799118365208,
+      "learning_rate": 1.815178769196289e-05,
+      "loss": 0.8073,
+      "step": 271
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.710860040908139,
+      "learning_rate": 1.81364682090916e-05,
+      "loss": 0.8243,
+      "step": 272
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6014451142481027,
+      "learning_rate": 1.8121092018589635e-05,
+      "loss": 0.813,
+      "step": 273
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8351346551277475,
+      "learning_rate": 1.810565922762232e-05,
+      "loss": 0.9587,
+      "step": 274
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6061529178110853,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.7788,
+      "step": 275
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6204620893939564,
+      "learning_rate": 1.8074624274924645e-05,
+      "loss": 0.81,
+      "step": 276
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5511626063200689,
+      "learning_rate": 1.805902232949435e-05,
+      "loss": 0.8005,
+      "step": 277
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5875562247342742,
+      "learning_rate": 1.8043364216197342e-05,
+      "loss": 0.799,
+      "step": 278
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5995434221560016,
+      "learning_rate": 1.802765004416383e-05,
+      "loss": 0.8124,
+      "step": 279
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5752729074278959,
+      "learning_rate": 1.801187992291473e-05,
+      "loss": 0.7968,
+      "step": 280
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5982940262865365,
+      "learning_rate": 1.7996053962360908e-05,
+      "loss": 0.8256,
+      "step": 281
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6024116558250863,
+      "learning_rate": 1.7980172272802398e-05,
+      "loss": 0.7673,
+      "step": 282
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6125341886072998,
+      "learning_rate": 1.7964234964927634e-05,
+      "loss": 0.8479,
+      "step": 283
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6436031238325665,
+      "learning_rate": 1.79482421498127e-05,
+      "loss": 0.7594,
+      "step": 284
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5634456574130875,
+      "learning_rate": 1.7932193938920526e-05,
+      "loss": 0.8126,
+      "step": 285
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6175642207699951,
+      "learning_rate": 1.7916090444100145e-05,
+      "loss": 0.798,
+      "step": 286
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.598131491091419,
+      "learning_rate": 1.789993177758588e-05,
+      "loss": 0.8002,
+      "step": 287
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.027174071576615,
+      "learning_rate": 1.788371805199658e-05,
+      "loss": 0.8131,
+      "step": 288
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6693182596046154,
+      "learning_rate": 1.7867449380334834e-05,
+      "loss": 0.8359,
+      "step": 289
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6492272635056814,
+      "learning_rate": 1.785112587598618e-05,
+      "loss": 0.8005,
+      "step": 290
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5947319464240111,
+      "learning_rate": 1.7834747652718313e-05,
+      "loss": 0.8174,
+      "step": 291
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7492961675394055,
+      "learning_rate": 1.78183148246803e-05,
+      "loss": 0.7894,
+      "step": 292
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9454785190416699,
+      "learning_rate": 1.7801827506401775e-05,
+      "loss": 0.7723,
+      "step": 293
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6602657627865639,
+      "learning_rate": 1.778528581279215e-05,
+      "loss": 0.7719,
+      "step": 294
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7824848359553706,
+      "learning_rate": 1.776868985913981e-05,
+      "loss": 0.9638,
+      "step": 295
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6336227436814612,
+      "learning_rate": 1.77520397611113e-05,
+      "loss": 0.7848,
+      "step": 296
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8823460968034843,
+      "learning_rate": 1.773533563475053e-05,
+      "loss": 0.8113,
+      "step": 297
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6479456921759581,
+      "learning_rate": 1.7718577596477976e-05,
+      "loss": 0.8053,
+      "step": 298
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5872332028859741,
+      "learning_rate": 1.7701765763089846e-05,
+      "loss": 0.7737,
+      "step": 299
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.680444374255502,
+      "learning_rate": 1.7684900251757268e-05,
+      "loss": 0.8477,
+      "step": 300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6475683440405768,
+      "learning_rate": 1.7667981180025498e-05,
+      "loss": 0.7973,
+      "step": 301
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4697042879712467,
+      "learning_rate": 1.7651008665813083e-05,
+      "loss": 0.7773,
+      "step": 302
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7001840076718483,
+      "learning_rate": 1.763398282741103e-05,
+      "loss": 0.7749,
+      "step": 303
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7258290132575279,
+      "learning_rate": 1.7616903783482002e-05,
+      "loss": 0.8152,
+      "step": 304
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7788016916834134,
+      "learning_rate": 1.7599771653059473e-05,
+      "loss": 0.8132,
+      "step": 305
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5618779844658957,
+      "learning_rate": 1.758258655554692e-05,
+      "loss": 0.7918,
+      "step": 306
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6559239803235809,
+      "learning_rate": 1.7565348610716963e-05,
+      "loss": 0.7686,
+      "step": 307
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6781266166025134,
+      "learning_rate": 1.7548057938710547e-05,
+      "loss": 0.8068,
+      "step": 308
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5935684937621687,
+      "learning_rate": 1.7530714660036112e-05,
+      "loss": 0.7825,
+      "step": 309
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.682828943161483,
+      "learning_rate": 1.7513318895568734e-05,
+      "loss": 0.825,
+      "step": 310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6320604129930825,
+      "learning_rate": 1.74958707665493e-05,
+      "loss": 0.7679,
+      "step": 311
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.721558504672073,
+      "learning_rate": 1.7478370394583647e-05,
+      "loss": 0.771,
+      "step": 312
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.607874320861536,
+      "learning_rate": 1.7460817901641728e-05,
+      "loss": 0.8117,
+      "step": 313
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9192188757451555,
+      "learning_rate": 1.7443213410056752e-05,
+      "loss": 0.8111,
+      "step": 314
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6544124565730924,
+      "learning_rate": 1.742555704252434e-05,
+      "loss": 0.7753,
+      "step": 315
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7442583155065514,
+      "learning_rate": 1.7407848922101663e-05,
+      "loss": 0.8154,
+      "step": 316
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6277567883653218,
+      "learning_rate": 1.7390089172206594e-05,
+      "loss": 0.7741,
+      "step": 317
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6968007424097566,
+      "learning_rate": 1.737227791661683e-05,
+      "loss": 0.8003,
+      "step": 318
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5515127854870717,
+      "learning_rate": 1.735441527946904e-05,
+      "loss": 0.7669,
+      "step": 319
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8310780975785218,
+      "learning_rate": 1.7336501385258017e-05,
+      "loss": 0.7632,
+      "step": 320
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6718262602850404,
+      "learning_rate": 1.7318536358835773e-05,
+      "loss": 0.7886,
+      "step": 321
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.574834965912156,
+      "learning_rate": 1.73005203254107e-05,
+      "loss": 0.8237,
+      "step": 322
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6604177413546964,
+      "learning_rate": 1.728245341054669e-05,
+      "loss": 0.827,
+      "step": 323
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.646120544293474,
+      "learning_rate": 1.7264335740162244e-05,
+      "loss": 0.7479,
+      "step": 324
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8154193183764218,
+      "learning_rate": 1.7246167440529612e-05,
+      "loss": 0.8172,
+      "step": 325
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5650331513721365,
+      "learning_rate": 1.7227948638273918e-05,
+      "loss": 0.798,
+      "step": 326
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7069044165134257,
+      "learning_rate": 1.720967946037225e-05,
+      "loss": 0.7905,
+      "step": 327
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5888297213745091,
+      "learning_rate": 1.7191360034152806e-05,
+      "loss": 0.7622,
+      "step": 328
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0064568101221139,
+      "learning_rate": 1.7172990487293984e-05,
+      "loss": 0.8184,
+      "step": 329
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5416331805709381,
+      "learning_rate": 1.7154570947823506e-05,
+      "loss": 0.797,
+      "step": 330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6489804026798072,
+      "learning_rate": 1.7136101544117526e-05,
+      "loss": 0.8228,
+      "step": 331
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6841785411242994,
+      "learning_rate": 1.7117582404899714e-05,
+      "loss": 1.0055,
+      "step": 332
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6598129618848795,
+      "learning_rate": 1.7099013659240385e-05,
+      "loss": 0.7431,
+      "step": 333
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6134501920745908,
+      "learning_rate": 1.7080395436555593e-05,
+      "loss": 0.8113,
+      "step": 334
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6387013226081621,
+      "learning_rate": 1.7061727866606215e-05,
+      "loss": 0.8031,
+      "step": 335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6891990965964899,
+      "learning_rate": 1.704301107949707e-05,
+      "loss": 0.7809,
+      "step": 336
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5806379954761808,
+      "learning_rate": 1.7024245205675986e-05,
+      "loss": 0.76,
+      "step": 337
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7934708813107365,
+      "learning_rate": 1.700543037593291e-05,
+      "loss": 0.7723,
+      "step": 338
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5913134705320557,
+      "learning_rate": 1.6986566721398992e-05,
+      "loss": 0.7884,
+      "step": 339
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7517340344465387,
+      "learning_rate": 1.6967654373545664e-05,
+      "loss": 0.8036,
+      "step": 340
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5841435397955429,
+      "learning_rate": 1.694869346418374e-05,
+      "loss": 0.7377,
+      "step": 341
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7216162404341944,
+      "learning_rate": 1.692968412546247e-05,
+      "loss": 0.7713,
+      "step": 342
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7004655133084476,
+      "learning_rate": 1.691062648986865e-05,
+      "loss": 0.8185,
+      "step": 343
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6079511632931519,
+      "learning_rate": 1.6891520690225672e-05,
+      "loss": 0.8027,
+      "step": 344
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7131962046389391,
+      "learning_rate": 1.687236685969263e-05,
+      "loss": 0.7729,
+      "step": 345
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6882375893712311,
+      "learning_rate": 1.685316513176335e-05,
+      "loss": 0.7978,
+      "step": 346
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.170656858443233,
+      "learning_rate": 1.6833915640265485e-05,
+      "loss": 0.7715,
+      "step": 347
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7034714294473754,
+      "learning_rate": 1.6814618519359603e-05,
+      "loss": 0.8107,
+      "step": 348
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6666661465486498,
+      "learning_rate": 1.6795273903538198e-05,
+      "loss": 0.7615,
+      "step": 349
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6724702358096429,
+      "learning_rate": 1.6775881927624796e-05,
+      "loss": 0.9828,
+      "step": 350
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6840050925526813,
+      "learning_rate": 1.675644272677301e-05,
+      "loss": 0.7559,
+      "step": 351
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6987734228979375,
+      "learning_rate": 1.6736956436465573e-05,
+      "loss": 0.8257,
+      "step": 352
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6201167808484361,
+      "learning_rate": 1.6717423192513424e-05,
+      "loss": 0.7701,
+      "step": 353
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6777000192216286,
+      "learning_rate": 1.6697843131054746e-05,
+      "loss": 0.7614,
+      "step": 354
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6938342275640759,
+      "learning_rate": 1.6678216388554012e-05,
+      "loss": 0.7893,
+      "step": 355
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8430582698793689,
+      "learning_rate": 1.6658543101801055e-05,
+      "loss": 0.7717,
+      "step": 356
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6142869595841061,
+      "learning_rate": 1.6638823407910085e-05,
+      "loss": 0.8026,
+      "step": 357
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7731234446741879,
+      "learning_rate": 1.6619057444318766e-05,
+      "loss": 0.7607,
+      "step": 358
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8541820745078345,
+      "learning_rate": 1.659924534878723e-05,
+      "loss": 0.8514,
+      "step": 359
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5836013868608032,
+      "learning_rate": 1.657938725939713e-05,
+      "loss": 0.7515,
+      "step": 360
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.95967586694985,
+      "learning_rate": 1.655948331455068e-05,
+      "loss": 0.8016,
+      "step": 361
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6765831351393802,
+      "learning_rate": 1.6539533652969683e-05,
+      "loss": 0.7482,
+      "step": 362
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2245156483586421,
+      "learning_rate": 1.651953841369457e-05,
+      "loss": 0.8454,
+      "step": 363
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6468321257286775,
+      "learning_rate": 1.6499497736083434e-05,
+      "loss": 0.7723,
+      "step": 364
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7371716430396992,
+      "learning_rate": 1.6479411759811038e-05,
+      "loss": 0.7534,
+      "step": 365
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5826978668941484,
+      "learning_rate": 1.6459280624867876e-05,
+      "loss": 0.7778,
+      "step": 366
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6652383099837994,
+      "learning_rate": 1.6439104471559157e-05,
+      "loss": 0.8147,
+      "step": 367
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8110860288034121,
+      "learning_rate": 1.6418883440503868e-05,
+      "loss": 0.8337,
+      "step": 368
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3285184703679314,
+      "learning_rate": 1.6398617672633765e-05,
+      "loss": 0.7602,
+      "step": 369
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9601858847911988,
+      "learning_rate": 1.6378307309192397e-05,
+      "loss": 0.7718,
+      "step": 370
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8593581183557336,
+      "learning_rate": 1.6357952491734136e-05,
+      "loss": 0.7633,
+      "step": 371
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.273586647821306,
+      "learning_rate": 1.6337553362123165e-05,
+      "loss": 0.7777,
+      "step": 372
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7356442561970055,
+      "learning_rate": 1.631711006253251e-05,
+      "loss": 1.0121,
+      "step": 373
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7619075391796934,
+      "learning_rate": 1.6296622735443047e-05,
+      "loss": 0.7332,
+      "step": 374
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.689803480402404,
+      "learning_rate": 1.6276091523642504e-05,
+      "loss": 0.7401,
+      "step": 375
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9674138943431938,
+      "learning_rate": 1.625551657022445e-05,
+      "loss": 0.8228,
+      "step": 376
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6272691112825868,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.7824,
+      "step": 377
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0106871584729111,
+      "learning_rate": 1.6214236012433464e-05,
+      "loss": 0.7246,
+      "step": 378
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.721794803030869,
+      "learning_rate": 1.619353069576799e-05,
+      "loss": 0.7673,
+      "step": 379
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7071425264905243,
+      "learning_rate": 1.617278221289793e-05,
+      "loss": 0.7719,
+      "step": 380
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7176794995754178,
+      "learning_rate": 1.6151990708431157e-05,
+      "loss": 0.754,
+      "step": 381
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8044642741307192,
+      "learning_rate": 1.6131156327275372e-05,
+      "loss": 0.7958,
+      "step": 382
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9693128170679643,
+      "learning_rate": 1.6110279214637115e-05,
+      "loss": 0.7376,
+      "step": 383
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8108707867726541,
+      "learning_rate": 1.6089359516020752e-05,
+      "loss": 0.853,
+      "step": 384
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7594586595638613,
+      "learning_rate": 1.606839737722744e-05,
+      "loss": 0.8074,
+      "step": 385
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7891417506979658,
+      "learning_rate": 1.6047392944354136e-05,
+      "loss": 0.7877,
+      "step": 386
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8708640144666252,
+      "learning_rate": 1.6026346363792565e-05,
+      "loss": 0.7448,
+      "step": 387
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8172621350650159,
+      "learning_rate": 1.6005257782228208e-05,
+      "loss": 0.7788,
+      "step": 388
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0623371975328977,
+      "learning_rate": 1.5984127346639264e-05,
+      "loss": 0.7857,
+      "step": 389
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7999896496543831,
+      "learning_rate": 1.5962955204295646e-05,
+      "loss": 0.7933,
+      "step": 390
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7660295242393487,
+      "learning_rate": 1.594174150275794e-05,
+      "loss": 0.9643,
+      "step": 391
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.717361455049709,
+      "learning_rate": 1.5920486389876383e-05,
+      "loss": 0.7441,
+      "step": 392
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.122329218085076,
+      "learning_rate": 1.5899190013789833e-05,
+      "loss": 0.805,
+      "step": 393
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7638540625510556,
+      "learning_rate": 1.5877852522924733e-05,
+      "loss": 0.7914,
+      "step": 394
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9066916181175102,
+      "learning_rate": 1.5856474065994074e-05,
+      "loss": 0.8047,
+      "step": 395
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9040774827898911,
+      "learning_rate": 1.5835054791996366e-05,
+      "loss": 0.7699,
+      "step": 396
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8473895372185793,
+      "learning_rate": 1.58135948502146e-05,
+      "loss": 0.7305,
+      "step": 397
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6874563590929637,
+      "learning_rate": 1.5792094390215198e-05,
+      "loss": 0.8251,
+      "step": 398
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7612657677801873,
+      "learning_rate": 1.5770553561846975e-05,
+      "loss": 0.8009,
+      "step": 399
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7916540486012381,
+      "learning_rate": 1.57489725152401e-05,
+      "loss": 0.7765,
+      "step": 400
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8048032657229598,
+      "learning_rate": 1.5727351400805054e-05,
+      "loss": 0.7453,
+      "step": 401
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7133504846273055,
+      "learning_rate": 1.5705690369231552e-05,
+      "loss": 0.7977,
+      "step": 402
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3745615367033641,
+      "learning_rate": 1.568398957148753e-05,
+      "loss": 0.8125,
+      "step": 403
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8005271999154986,
+      "learning_rate": 1.5662249158818072e-05,
+      "loss": 0.7513,
+      "step": 404
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9831332729083802,
+      "learning_rate": 1.5640469282744364e-05,
+      "loss": 0.7647,
+      "step": 405
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8314471189910663,
+      "learning_rate": 1.561865009506263e-05,
+      "loss": 0.7846,
+      "step": 406
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7626155036186595,
+      "learning_rate": 1.5596791747843083e-05,
+      "loss": 0.7917,
+      "step": 407
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0599715605661562,
+      "learning_rate": 1.5574894393428856e-05,
+      "loss": 0.8398,
+      "step": 408
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7910640451426868,
+      "learning_rate": 1.555295818443495e-05,
+      "loss": 0.766,
+      "step": 409
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7871938684887542,
+      "learning_rate": 1.553098327374716e-05,
+      "loss": 0.7889,
+      "step": 410
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9698395589398837,
+      "learning_rate": 1.5508969814521026e-05,
+      "loss": 0.8024,
+      "step": 411
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8033914445898368,
+      "learning_rate": 1.5486917960180742e-05,
+      "loss": 0.7744,
+      "step": 412
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6628497468510182,
+      "learning_rate": 1.54648278644181e-05,
+      "loss": 0.7413,
+      "step": 413
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9217242400807502,
+      "learning_rate": 1.5442699681191426e-05,
+      "loss": 0.7864,
+      "step": 414
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.811233947502682,
+      "learning_rate": 1.5420533564724495e-05,
+      "loss": 0.7445,
+      "step": 415
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.765673941073733,
+      "learning_rate": 1.539832966950546e-05,
+      "loss": 0.7919,
+      "step": 416
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6759133259142388,
+      "learning_rate": 1.5376088150285777e-05,
+      "loss": 0.7902,
+      "step": 417
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.057668541779849,
+      "learning_rate": 1.5353809162079117e-05,
+      "loss": 0.7824,
+      "step": 418
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7244298543014138,
+      "learning_rate": 1.533149286016031e-05,
+      "loss": 0.8022,
+      "step": 419
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.687577311662202,
+      "learning_rate": 1.5309139400064224e-05,
+      "loss": 0.7739,
+      "step": 420
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8019917992897705,
+      "learning_rate": 1.528674893758473e-05,
+      "loss": 0.7898,
+      "step": 421
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6706760093465938,
+      "learning_rate": 1.526432162877356e-05,
+      "loss": 0.7756,
+      "step": 422
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7536796353801893,
+      "learning_rate": 1.5241857629939272e-05,
+      "loss": 0.7542,
+      "step": 423
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6737498261058815,
+      "learning_rate": 1.521935709764613e-05,
+      "loss": 0.7884,
+      "step": 424
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7544587375551581,
+      "learning_rate": 1.5196820188713011e-05,
+      "loss": 0.7778,
+      "step": 425
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5924981010277505,
+      "learning_rate": 1.5174247060212341e-05,
+      "loss": 0.747,
+      "step": 426
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7700661534076043,
+      "learning_rate": 1.515163786946896e-05,
+      "loss": 0.9493,
+      "step": 427
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6824224717873821,
+      "learning_rate": 1.5128992774059063e-05,
+      "loss": 0.8195,
+      "step": 428
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6704951692102668,
+      "learning_rate": 1.510631193180907e-05,
+      "loss": 0.7717,
+      "step": 429
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7279577213556054,
+      "learning_rate": 1.5083595500794554e-05,
+      "loss": 0.7688,
+      "step": 430
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.623387227842246,
+      "learning_rate": 1.5060843639339116e-05,
+      "loss": 0.9065,
+      "step": 431
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0131778798993434,
+      "learning_rate": 1.5038056506013297e-05,
+      "loss": 0.8092,
+      "step": 432
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8858346719669161,
+      "learning_rate": 1.5015234259633467e-05,
+      "loss": 0.83,
+      "step": 433
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8600915968010218,
+      "learning_rate": 1.4992377059260718e-05,
+      "loss": 0.7814,
+      "step": 434
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8548804532754122,
+      "learning_rate": 1.4969485064199756e-05,
+      "loss": 0.8208,
+      "step": 435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8060958322598917,
+      "learning_rate": 1.4946558433997792e-05,
+      "loss": 0.7775,
+      "step": 436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7743650711435145,
+      "learning_rate": 1.4923597328443423e-05,
+      "loss": 0.8338,
+      "step": 437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6800111582472981,
+      "learning_rate": 1.4900601907565535e-05,
+      "loss": 0.7804,
+      "step": 438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7380426668823814,
+      "learning_rate": 1.4877572331632169e-05,
+      "loss": 0.8113,
+      "step": 439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7615992533005829,
+      "learning_rate": 1.485450876114941e-05,
+      "loss": 0.8581,
+      "step": 440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7271495550193666,
+      "learning_rate": 1.4831411356860276e-05,
+      "loss": 0.7777,
+      "step": 441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6984635872415319,
+      "learning_rate": 1.4808280279743594e-05,
+      "loss": 0.8291,
+      "step": 442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5937003199888604,
+      "learning_rate": 1.4785115691012866e-05,
+      "loss": 0.791,
+      "step": 443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8721590355867014,
+      "learning_rate": 1.4761917752115162e-05,
+      "loss": 0.8295,
+      "step": 444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8671497167343813,
+      "learning_rate": 1.4738686624729987e-05,
+      "loss": 0.7737,
+      "step": 445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.583210769614291,
+      "learning_rate": 1.4715422470768157e-05,
+      "loss": 0.7826,
+      "step": 446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6142041222097552,
+      "learning_rate": 1.4692125452370664e-05,
+      "loss": 0.7694,
+      "step": 447
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7380633060687146,
+      "learning_rate": 1.4668795731907555e-05,
+      "loss": 1.0192,
+      "step": 448
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6343702209093677,
+      "learning_rate": 1.4645433471976793e-05,
+      "loss": 0.8006,
+      "step": 449
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7267722282305372,
+      "learning_rate": 1.4622038835403135e-05,
+      "loss": 0.7953,
+      "step": 450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8726432038004055,
+      "learning_rate": 1.4598611985236975e-05,
+      "loss": 0.7691,
+      "step": 451
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6259543834747051,
+      "learning_rate": 1.4575153084753233e-05,
+      "loss": 0.7872,
+      "step": 452
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7088977273999924,
+      "learning_rate": 1.4551662297450201e-05,
+      "loss": 0.8336,
+      "step": 453
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6610143682471902,
+      "learning_rate": 1.4528139787048414e-05,
+      "loss": 0.7588,
+      "step": 454
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6661031193290639,
+      "learning_rate": 1.4504585717489494e-05,
+      "loss": 0.7398,
+      "step": 455
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5883158111701801,
+      "learning_rate": 1.448100025293502e-05,
+      "loss": 0.8063,
+      "step": 456
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8273380607905003,
+      "learning_rate": 1.4457383557765385e-05,
+      "loss": 0.8606,
+      "step": 457
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6159772814282898,
+      "learning_rate": 1.4433735796578638e-05,
+      "loss": 0.7829,
+      "step": 458
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1182211217316638,
+      "learning_rate": 1.4410057134189355e-05,
+      "loss": 0.7953,
+      "step": 459
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5245656338795248,
+      "learning_rate": 1.438634773562747e-05,
+      "loss": 0.8118,
+      "step": 460
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6610468730028848,
+      "learning_rate": 1.4362607766137138e-05,
+      "loss": 0.8152,
+      "step": 461
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7570767637495006,
+      "learning_rate": 1.4338837391175582e-05,
+      "loss": 0.8235,
+      "step": 462
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6384420925819746,
+      "learning_rate": 1.4315036776411937e-05,
+      "loss": 0.7506,
+      "step": 463
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5294577398246358,
+      "learning_rate": 1.429120608772609e-05,
+      "loss": 0.7773,
+      "step": 464
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8877518187360096,
+      "learning_rate": 1.4267345491207543e-05,
+      "loss": 0.8051,
+      "step": 465
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6745225822924297,
+      "learning_rate": 1.4243455153154231e-05,
+      "loss": 0.7941,
+      "step": 466
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8544464238299039,
+      "learning_rate": 1.4219535240071378e-05,
+      "loss": 0.786,
+      "step": 467
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8478958549978071,
+      "learning_rate": 1.4195585918670338e-05,
+      "loss": 0.7775,
+      "step": 468
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6784458346444917,
+      "learning_rate": 1.4171607355867419e-05,
+      "loss": 0.837,
+      "step": 469
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6825583550768195,
+      "learning_rate": 1.4147599718782737e-05,
+      "loss": 0.8254,
+      "step": 470
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7259821187015854,
+      "learning_rate": 1.4123563174739036e-05,
+      "loss": 0.7751,
+      "step": 471
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6211191130354058,
+      "learning_rate": 1.4099497891260538e-05,
+      "loss": 0.727,
+      "step": 472
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8129040891480699,
+      "learning_rate": 1.4075404036071755e-05,
+      "loss": 0.8183,
+      "step": 473
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7299185810826833,
+      "learning_rate": 1.4051281777096339e-05,
+      "loss": 0.8363,
+      "step": 474
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6253752795602247,
+      "learning_rate": 1.4027131282455908e-05,
+      "loss": 0.7734,
+      "step": 475
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9252242765967938,
+      "learning_rate": 1.4002952720468856e-05,
+      "loss": 0.7957,
+      "step": 476
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.662524953453812,
+      "learning_rate": 1.397874625964921e-05,
+      "loss": 0.7653,
+      "step": 477
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7535299004096885,
+      "learning_rate": 1.3954512068705425e-05,
+      "loss": 0.8456,
+      "step": 478
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6625288597074174,
+      "learning_rate": 1.3930250316539237e-05,
+      "loss": 0.7957,
+      "step": 479
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6510839029205088,
+      "learning_rate": 1.3905961172244462e-05,
+      "loss": 0.753,
+      "step": 480
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7405268689198073,
+      "learning_rate": 1.3881644805105827e-05,
+      "loss": 0.7561,
+      "step": 481
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.726402466151544,
+      "learning_rate": 1.3857301384597796e-05,
+      "loss": 0.8411,
+      "step": 482
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7600168007900396,
+      "learning_rate": 1.3832931080383383e-05,
+      "loss": 0.8339,
+      "step": 483
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7657880258814093,
+      "learning_rate": 1.3808534062312964e-05,
+      "loss": 0.7519,
+      "step": 484
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6719216747458312,
+      "learning_rate": 1.3784110500423104e-05,
+      "loss": 0.9929,
+      "step": 485
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6006960125249745,
+      "learning_rate": 1.3759660564935363e-05,
+      "loss": 0.7533,
+      "step": 486
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8370329436473629,
+      "learning_rate": 1.3735184426255117e-05,
+      "loss": 0.7906,
+      "step": 487
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6279212014409818,
+      "learning_rate": 1.3710682254970366e-05,
+      "loss": 0.7704,
+      "step": 488
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6695443565778791,
+      "learning_rate": 1.368615422185054e-05,
+      "loss": 0.7526,
+      "step": 489
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.619318072343979,
+      "learning_rate": 1.3661600497845329e-05,
+      "loss": 0.7453,
+      "step": 490
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7213284059307696,
+      "learning_rate": 1.363702125408346e-05,
+      "loss": 0.8114,
+      "step": 491
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6365642929512615,
+      "learning_rate": 1.3612416661871532e-05,
+      "loss": 0.7801,
+      "step": 492
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.748406196060676,
+      "learning_rate": 1.358778689269281e-05,
+      "loss": 0.7641,
+      "step": 493
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6118564266770413,
+      "learning_rate": 1.3563132118206031e-05,
+      "loss": 0.7923,
+      "step": 494
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7079948170238669,
+      "learning_rate": 1.3538452510244207e-05,
+      "loss": 0.7572,
+      "step": 495
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7263943095850882,
+      "learning_rate": 1.3513748240813429e-05,
+      "loss": 0.7857,
+      "step": 496
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9045416758056893,
+      "learning_rate": 1.348901948209167e-05,
+      "loss": 0.7604,
+      "step": 497
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.197927298334419,
+      "learning_rate": 1.3464266406427575e-05,
+      "loss": 0.774,
+      "step": 498
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6302600390222566,
+      "learning_rate": 1.3439489186339283e-05,
+      "loss": 0.7664,
+      "step": 499
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5685784965088436,
+      "learning_rate": 1.3414687994513191e-05,
+      "loss": 0.7517,
+      "step": 500
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6484485498313265,
+      "learning_rate": 1.3389863003802779e-05,
+      "loss": 0.7686,
+      "step": 501
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.635757434663971,
+      "learning_rate": 1.3365014387227393e-05,
+      "loss": 0.7878,
+      "step": 502
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6090083522142717,
+      "learning_rate": 1.3340142317971042e-05,
+      "loss": 0.9632,
+      "step": 503
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0997994523842576,
+      "learning_rate": 1.3315246969381186e-05,
+      "loss": 0.752,
+      "step": 504
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7003987844733678,
+      "learning_rate": 1.3290328514967535e-05,
+      "loss": 0.798,
+      "step": 505
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6033469948526544,
+      "learning_rate": 1.3265387128400833e-05,
+      "loss": 0.7423,
+      "step": 506
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8132942516843006,
+      "learning_rate": 1.324042298351166e-05,
+      "loss": 0.8234,
+      "step": 507
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6803995667856012,
+      "learning_rate": 1.3215436254289202e-05,
+      "loss": 0.796,
+      "step": 508
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5649689062062885,
+      "learning_rate": 1.3190427114880052e-05,
+      "loss": 0.7409,
+      "step": 509
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6054551315226072,
+      "learning_rate": 1.316539573958699e-05,
+      "loss": 0.7736,
+      "step": 510
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5970968927673034,
+      "learning_rate": 1.3140342302867774e-05,
+      "loss": 0.7458,
+      "step": 511
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6861191830879908,
+      "learning_rate": 1.3115266979333917e-05,
+      "loss": 0.8015,
+      "step": 512
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8063436215664004,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.7721,
+      "step": 513
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5756481877151032,
+      "learning_rate": 1.306505137102983e-05,
+      "loss": 0.7295,
+      "step": 514
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7701346628575035,
+      "learning_rate": 1.3039911436240462e-05,
+      "loss": 0.8079,
+      "step": 515
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5578532495208303,
+      "learning_rate": 1.3014750314595744e-05,
+      "loss": 0.7713,
+      "step": 516
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7919768635466691,
+      "learning_rate": 1.2989568181457704e-05,
+      "loss": 0.7674,
+      "step": 517
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5530019686414475,
+      "learning_rate": 1.2964365212334817e-05,
+      "loss": 0.7586,
+      "step": 518
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6161711854921614,
+      "learning_rate": 1.293914158288077e-05,
+      "loss": 0.7712,
+      "step": 519
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7078077760162985,
+      "learning_rate": 1.2913897468893249e-05,
+      "loss": 0.7668,
+      "step": 520
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6784435776831489,
+      "learning_rate": 1.2888633046312705e-05,
+      "loss": 0.8143,
+      "step": 521
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.729755938131584,
+      "learning_rate": 1.2863348491221129e-05,
+      "loss": 0.7524,
+      "step": 522
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5796683639729194,
+      "learning_rate": 1.2838043979840833e-05,
+      "loss": 0.7474,
+      "step": 523
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1662401145107417,
+      "learning_rate": 1.2812719688533214e-05,
+      "loss": 0.7946,
+      "step": 524
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7361012529206089,
+      "learning_rate": 1.2787375793797518e-05,
+      "loss": 0.8032,
+      "step": 525
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6767434257869064,
+      "learning_rate": 1.2762012472269634e-05,
+      "loss": 0.862,
+      "step": 526
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6774891987026581,
+      "learning_rate": 1.2736629900720832e-05,
+      "loss": 0.7587,
+      "step": 527
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.625145304806134,
+      "learning_rate": 1.2711228256056553e-05,
+      "loss": 0.7706,
+      "step": 528
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7262979186322793,
+      "learning_rate": 1.268580771531517e-05,
+      "loss": 0.8015,
+      "step": 529
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6486795629074259,
+      "learning_rate": 1.2660368455666752e-05,
+      "loss": 0.7772,
+      "step": 530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7703310637121329,
+      "learning_rate": 1.2634910654411831e-05,
+      "loss": 0.6797,
+      "step": 531
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7365619049807075,
+      "learning_rate": 1.2609434488980168e-05,
+      "loss": 0.756,
+      "step": 532
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8986641524841823,
+      "learning_rate": 1.258394013692951e-05,
+      "loss": 0.7917,
+      "step": 533
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6215363105380581,
+      "learning_rate": 1.2558427775944357e-05,
+      "loss": 0.7671,
+      "step": 534
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7519323999865509,
+      "learning_rate": 1.2532897583834728e-05,
+      "loss": 0.7496,
+      "step": 535
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7027261832645963,
+      "learning_rate": 1.2507349738534913e-05,
+      "loss": 0.7737,
+      "step": 536
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.757713853581307,
+      "learning_rate": 1.248178441810224e-05,
+      "loss": 0.7895,
+      "step": 537
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7094518582215956,
+      "learning_rate": 1.2456201800715828e-05,
+      "loss": 0.8381,
+      "step": 538
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5929702976214644,
+      "learning_rate": 1.2430602064675348e-05,
+      "loss": 0.7391,
+      "step": 539
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7364577108815441,
+      "learning_rate": 1.2404985388399787e-05,
+      "loss": 0.7374,
+      "step": 540
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7278821156159085,
+      "learning_rate": 1.2379351950426188e-05,
+      "loss": 0.7763,
+      "step": 541
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6410150722164706,
+      "learning_rate": 1.2353701929408425e-05,
+      "loss": 0.7989,
+      "step": 542
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7601186779984663,
+      "learning_rate": 1.2328035504115945e-05,
+      "loss": 0.7505,
+      "step": 543
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9768689122063738,
+      "learning_rate": 1.2302352853432521e-05,
+      "loss": 0.9428,
+      "step": 544
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6453971664409188,
+      "learning_rate": 1.227665415635502e-05,
+      "loss": 0.7859,
+      "step": 545
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.753768183204473,
+      "learning_rate": 1.225093959199213e-05,
+      "loss": 0.8573,
+      "step": 546
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5593815359874007,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.7327,
+      "step": 547
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6375353315153037,
+      "learning_rate": 1.2199463578396688e-05,
+      "loss": 0.7986,
+      "step": 548
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6747963317047685,
+      "learning_rate": 1.2173702487929469e-05,
+      "loss": 0.7532,
+      "step": 549
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6315999963965212,
+      "learning_rate": 1.2147926247705043e-05,
+      "loss": 0.7809,
+      "step": 550
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.667585194132604,
+      "learning_rate": 1.2122135037372553e-05,
+      "loss": 0.7909,
+      "step": 551
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6094128778248644,
+      "learning_rate": 1.2096329036685469e-05,
+      "loss": 0.7614,
+      "step": 552
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6454600039671129,
+      "learning_rate": 1.2070508425500344e-05,
+      "loss": 0.748,
+      "step": 553
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5676522472691308,
+      "learning_rate": 1.204467338377557e-05,
+      "loss": 0.7535,
+      "step": 554
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.807446168381916,
+      "learning_rate": 1.2018824091570103e-05,
+      "loss": 0.7849,
+      "step": 555
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5757875842995284,
+      "learning_rate": 1.1992960729042227e-05,
+      "loss": 0.7352,
+      "step": 556
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7799528644632541,
+      "learning_rate": 1.1967083476448282e-05,
+      "loss": 0.7777,
+      "step": 557
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6054440446129793,
+      "learning_rate": 1.1941192514141421e-05,
+      "loss": 0.7503,
+      "step": 558
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7362160680846206,
+      "learning_rate": 1.1915288022570346e-05,
+      "loss": 0.8426,
+      "step": 559
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7366519420445967,
+      "learning_rate": 1.1889370182278053e-05,
+      "loss": 0.7661,
+      "step": 560
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6816080959735485,
+      "learning_rate": 1.1863439173900573e-05,
+      "loss": 0.939,
+      "step": 561
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5993717118510358,
+      "learning_rate": 1.1837495178165706e-05,
+      "loss": 0.7459,
+      "step": 562
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6106534974460044,
+      "learning_rate": 1.1811538375891778e-05,
+      "loss": 0.7906,
+      "step": 563
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6049210913585981,
+      "learning_rate": 1.1785568947986368e-05,
+      "loss": 0.7359,
+      "step": 564
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7219161372401054,
+      "learning_rate": 1.175958707544505e-05,
+      "loss": 0.7759,
+      "step": 565
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6192707974481667,
+      "learning_rate": 1.1733592939350125e-05,
+      "loss": 0.7626,
+      "step": 566
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6887399718261602,
+      "learning_rate": 1.1707586720869375e-05,
+      "loss": 0.7545,
+      "step": 567
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6572149287532413,
+      "learning_rate": 1.1681568601254792e-05,
+      "loss": 0.7683,
+      "step": 568
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.866750787510793,
+      "learning_rate": 1.16555387618413e-05,
+      "loss": 0.7666,
+      "step": 569
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6347821196111028,
+      "learning_rate": 1.1629497384045522e-05,
+      "loss": 0.7328,
+      "step": 570
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7587886889757064,
+      "learning_rate": 1.1603444649364485e-05,
+      "loss": 0.7615,
+      "step": 571
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7818702478299131,
+      "learning_rate": 1.1577380739374376e-05,
+      "loss": 0.8278,
+      "step": 572
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9502351308120257,
+      "learning_rate": 1.1551305835729267e-05,
+      "loss": 0.732,
+      "step": 573
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7621435065374188,
+      "learning_rate": 1.1525220120159852e-05,
+      "loss": 0.7447,
+      "step": 574
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.832381031268132,
+      "learning_rate": 1.1499123774472176e-05,
+      "loss": 0.7863,
+      "step": 575
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.558546425583418,
+      "learning_rate": 1.1473016980546377e-05,
+      "loss": 0.7416,
+      "step": 576
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6318644327080798,
+      "learning_rate": 1.1446899920335407e-05,
+      "loss": 0.7982,
+      "step": 577
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6292613602674206,
+      "learning_rate": 1.1420772775863771e-05,
+      "loss": 0.7448,
+      "step": 578
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7788993541840272,
+      "learning_rate": 1.1394635729226255e-05,
+      "loss": 0.7954,
+      "step": 579
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5791071446526155,
+      "learning_rate": 1.136848896258667e-05,
+      "loss": 0.8304,
+      "step": 580
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7307853591665858,
+      "learning_rate": 1.1342332658176556e-05,
+      "loss": 0.7437,
+      "step": 581
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5630076894851149,
+      "learning_rate": 1.1316166998293937e-05,
+      "loss": 0.7502,
+      "step": 582
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7001343715681112,
+      "learning_rate": 1.1289992165302036e-05,
+      "loss": 0.7847,
+      "step": 583
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6073073869753334,
+      "learning_rate": 1.1263808341628009e-05,
+      "loss": 0.7592,
+      "step": 584
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6605909483255391,
+      "learning_rate": 1.1237615709761673e-05,
+      "loss": 0.7229,
+      "step": 585
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7465119507323211,
+      "learning_rate": 1.1211414452254239e-05,
+      "loss": 0.753,
+      "step": 586
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6833588979790411,
+      "learning_rate": 1.118520475171703e-05,
+      "loss": 0.7678,
+      "step": 587
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7488758227701683,
+      "learning_rate": 1.115898679082021e-05,
+      "loss": 0.8195,
+      "step": 588
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8939690206912403,
+      "learning_rate": 1.1132760752291524e-05,
+      "loss": 0.7306,
+      "step": 589
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8436465030431038,
+      "learning_rate": 1.1106526818915008e-05,
+      "loss": 0.7453,
+      "step": 590
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7290339965362036,
+      "learning_rate": 1.1080285173529724e-05,
+      "loss": 0.7364,
+      "step": 591
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1864151890054417,
+      "learning_rate": 1.1054035999028478e-05,
+      "loss": 0.8039,
+      "step": 592
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9056334790916831,
+      "learning_rate": 1.102777947835656e-05,
+      "loss": 0.8287,
+      "step": 593
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6862915212725087,
+      "learning_rate": 1.1001515794510449e-05,
+      "loss": 0.7359,
+      "step": 594
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7125474709428455,
+      "learning_rate": 1.0975245130536555e-05,
+      "loss": 0.7756,
+      "step": 595
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7430767394465932,
+      "learning_rate": 1.0948967669529935e-05,
+      "loss": 0.762,
+      "step": 596
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.79278995217457,
+      "learning_rate": 1.092268359463302e-05,
+      "loss": 0.7776,
+      "step": 597
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6689877452440369,
+      "learning_rate": 1.0896393089034336e-05,
+      "loss": 0.7191,
+      "step": 598
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6327631426910685,
+      "learning_rate": 1.0870096335967226e-05,
+      "loss": 0.7264,
+      "step": 599
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8459814834711223,
+      "learning_rate": 1.0843793518708576e-05,
+      "loss": 0.7806,
+      "step": 600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6038286187184484,
+      "learning_rate": 1.0817484820577537e-05,
+      "loss": 0.7439,
+      "step": 601
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.632129818207202,
+      "learning_rate": 1.0791170424934248e-05,
+      "loss": 0.9901,
+      "step": 602
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6832925105538883,
+      "learning_rate": 1.0764850515178555e-05,
+      "loss": 0.72,
+      "step": 603
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9139026161184021,
+      "learning_rate": 1.073852527474874e-05,
+      "loss": 0.7864,
+      "step": 604
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.680882816031925,
+      "learning_rate": 1.0712194887120232e-05,
+      "loss": 0.7534,
+      "step": 605
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6104239907066029,
+      "learning_rate": 1.0685859535804336e-05,
+      "loss": 0.7487,
+      "step": 606
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8198419274415747,
+      "learning_rate": 1.0659519404346955e-05,
+      "loss": 0.7146,
+      "step": 607
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7127564686318812,
+      "learning_rate": 1.0633174676327298e-05,
+      "loss": 0.7651,
+      "step": 608
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.717421066140589,
+      "learning_rate": 1.060682553535662e-05,
+      "loss": 0.8024,
+      "step": 609
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6664300090928785,
+      "learning_rate": 1.0580472165076928e-05,
+      "loss": 0.7597,
+      "step": 610
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.43475749051984,
+      "learning_rate": 1.05541147491597e-05,
+      "loss": 0.7469,
+      "step": 611
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6675555155203455,
+      "learning_rate": 1.0527753471304625e-05,
+      "loss": 0.7278,
+      "step": 612
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.044738042966426,
+      "learning_rate": 1.0501388515238293e-05,
+      "loss": 0.7722,
+      "step": 613
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5983948860655425,
+      "learning_rate": 1.047502006471294e-05,
+      "loss": 0.7387,
+      "step": 614
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7176520425443301,
+      "learning_rate": 1.044864830350515e-05,
+      "loss": 0.736,
+      "step": 615
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7134332353570483,
+      "learning_rate": 1.0422273415414586e-05,
+      "loss": 0.7325,
+      "step": 616
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6806840654700838,
+      "learning_rate": 1.0395895584262696e-05,
+      "loss": 0.7541,
+      "step": 617
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7101909418100236,
+      "learning_rate": 1.0369514993891451e-05,
+      "loss": 0.7902,
+      "step": 618
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6396121691194859,
+      "learning_rate": 1.0343131828162045e-05,
+      "loss": 0.7869,
+      "step": 619
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5810017948697849,
+      "learning_rate": 1.0316746270953622e-05,
+      "loss": 0.7214,
+      "step": 620
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7770222028817707,
+      "learning_rate": 1.0290358506161994e-05,
+      "loss": 0.7332,
+      "step": 621
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5764422047833285,
+      "learning_rate": 1.0263968717698365e-05,
+      "loss": 0.7628,
+      "step": 622
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6783932106974461,
+      "learning_rate": 1.0237577089488032e-05,
+      "loss": 0.7571,
+      "step": 623
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5878404969986016,
+      "learning_rate": 1.0211183805469122e-05,
+      "loss": 0.7527,
+      "step": 624
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8064089439725569,
+      "learning_rate": 1.01847890495913e-05,
+      "loss": 0.7478,
+      "step": 625
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5639333999999218,
+      "learning_rate": 1.0158393005814492e-05,
+      "loss": 0.7394,
+      "step": 626
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7676623505207083,
+      "learning_rate": 1.013199585810759e-05,
+      "loss": 0.7937,
+      "step": 627
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6789316679421448,
+      "learning_rate": 1.0105597790447195e-05,
+      "loss": 0.7393,
+      "step": 628
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6024218898837607,
+      "learning_rate": 1.0079198986816307e-05,
+      "loss": 0.7721,
+      "step": 629
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6332792961437728,
+      "learning_rate": 1.0052799631203066e-05,
+      "loss": 0.7376,
+      "step": 630
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6981839307417284,
+      "learning_rate": 1.0026399907599447e-05,
+      "loss": 0.8223,
+      "step": 631
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6952917363972095,
+      "learning_rate": 1e-05,
+      "loss": 0.7374,
+      "step": 632
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5831442608276801,
+      "learning_rate": 9.973600092400556e-06,
+      "loss": 0.7398,
+      "step": 633
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7474790100525177,
+      "learning_rate": 9.94720036879694e-06,
+      "loss": 0.7313,
+      "step": 634
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5847803828896533,
+      "learning_rate": 9.920801013183696e-06,
+      "loss": 0.7634,
+      "step": 635
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7027299278806401,
+      "learning_rate": 9.894402209552809e-06,
+      "loss": 0.8274,
+      "step": 636
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8754295375528324,
+      "learning_rate": 9.868004141892412e-06,
+      "loss": 0.7275,
+      "step": 637
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6841209239344636,
+      "learning_rate": 9.841606994185513e-06,
+      "loss": 0.9869,
+      "step": 638
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6091003039095542,
+      "learning_rate": 9.815210950408703e-06,
+      "loss": 0.7214,
+      "step": 639
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9379841308063579,
+      "learning_rate": 9.788816194530882e-06,
+      "loss": 0.7459,
+      "step": 640
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9123368597008271,
+      "learning_rate": 9.762422910511971e-06,
+      "loss": 0.7414,
+      "step": 641
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6470923418614004,
+      "learning_rate": 9.73603128230164e-06,
+      "loss": 0.7917,
+      "step": 642
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6067981884715613,
+      "learning_rate": 9.70964149383801e-06,
+      "loss": 0.7184,
+      "step": 643
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8042488566224163,
+      "learning_rate": 9.683253729046383e-06,
+      "loss": 0.8172,
+      "step": 644
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6902931017992471,
+      "learning_rate": 9.65686817183796e-06,
+      "loss": 0.7167,
+      "step": 645
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7350859939010641,
+      "learning_rate": 9.630485006108554e-06,
+      "loss": 0.7833,
+      "step": 646
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6587300157834868,
+      "learning_rate": 9.604104415737309e-06,
+      "loss": 0.7369,
+      "step": 647
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7757511692185833,
+      "learning_rate": 9.57772658458542e-06,
+      "loss": 0.7493,
+      "step": 648
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7152975559911465,
+      "learning_rate": 9.551351696494854e-06,
+      "loss": 0.7312,
+      "step": 649
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8865465586642246,
+      "learning_rate": 9.524979935287064e-06,
+      "loss": 0.7722,
+      "step": 650
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7969256669562351,
+      "learning_rate": 9.498611484761707e-06,
+      "loss": 0.8119,
+      "step": 651
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5969766219754291,
+      "learning_rate": 9.472246528695377e-06,
+      "loss": 0.7287,
+      "step": 652
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6327397463367398,
+      "learning_rate": 9.445885250840301e-06,
+      "loss": 0.7266,
+      "step": 653
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.77166765941748,
+      "learning_rate": 9.419527834923074e-06,
+      "loss": 0.8123,
+      "step": 654
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7601406074366925,
+      "learning_rate": 9.393174464643381e-06,
+      "loss": 0.7899,
+      "step": 655
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4841115869777695,
+      "learning_rate": 9.366825323672702e-06,
+      "loss": 0.7396,
+      "step": 656
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6661612812638977,
+      "learning_rate": 9.340480595653047e-06,
+      "loss": 0.9072,
+      "step": 657
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6649209906395811,
+      "learning_rate": 9.314140464195663e-06,
+      "loss": 0.7246,
+      "step": 658
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7646718360075878,
+      "learning_rate": 9.28780511287977e-06,
+      "loss": 0.7777,
+      "step": 659
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6018234510855817,
+      "learning_rate": 9.261474725251261e-06,
+      "loss": 0.7433,
+      "step": 660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8131652106780938,
+      "learning_rate": 9.235149484821447e-06,
+      "loss": 0.7393,
+      "step": 661
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6099167637240838,
+      "learning_rate": 9.208829575065754e-06,
+      "loss": 0.7395,
+      "step": 662
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7922642527327044,
+      "learning_rate": 9.182515179422466e-06,
+      "loss": 0.7954,
+      "step": 663
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6077828792351242,
+      "learning_rate": 9.156206481291427e-06,
+      "loss": 0.7591,
+      "step": 664
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6045109071136379,
+      "learning_rate": 9.129903664032777e-06,
+      "loss": 0.7235,
+      "step": 665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8113614972147105,
+      "learning_rate": 9.103606910965666e-06,
+      "loss": 0.7358,
+      "step": 666
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6895359020914006,
+      "learning_rate": 9.07731640536698e-06,
+      "loss": 0.7502,
+      "step": 667
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7651234987155068,
+      "learning_rate": 9.051032330470068e-06,
+      "loss": 0.762,
+      "step": 668
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7561436083484064,
+      "learning_rate": 9.024754869463448e-06,
+      "loss": 0.7437,
+      "step": 669
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.746819203502454,
+      "learning_rate": 8.998484205489554e-06,
+      "loss": 0.716,
+      "step": 670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7379548034214043,
+      "learning_rate": 8.972220521643445e-06,
+      "loss": 0.7435,
+      "step": 671
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7461596788361676,
+      "learning_rate": 8.945964000971525e-06,
+      "loss": 0.7836,
+      "step": 672
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5699720170758436,
+      "learning_rate": 8.91971482647028e-06,
+      "loss": 0.7515,
+      "step": 673
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7942387718053385,
+      "learning_rate": 8.893473181084993e-06,
+      "loss": 0.7493,
+      "step": 674
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7535270900000731,
+      "learning_rate": 8.867239247708479e-06,
+      "loss": 0.7331,
+      "step": 675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.013295231681451,
+      "learning_rate": 8.841013209179793e-06,
+      "loss": 0.7631,
+      "step": 676
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6302406635853969,
+      "learning_rate": 8.814795248282974e-06,
+      "loss": 0.7505,
+      "step": 677
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6992643271314674,
+      "learning_rate": 8.788585547745764e-06,
+      "loss": 0.7096,
+      "step": 678
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.779820463069534,
+      "learning_rate": 8.76238429023833e-06,
+      "loss": 0.8355,
+      "step": 679
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9610632189305904,
+      "learning_rate": 8.736191658371998e-06,
+      "loss": 0.7635,
+      "step": 680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6433197366035323,
+      "learning_rate": 8.71000783469797e-06,
+      "loss": 0.76,
+      "step": 681
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7329384052385725,
+      "learning_rate": 8.683833001706068e-06,
+      "loss": 0.7432,
+      "step": 682
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7208321664125152,
+      "learning_rate": 8.657667341823449e-06,
+      "loss": 0.7722,
+      "step": 683
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0847185924939973,
+      "learning_rate": 8.631511037413334e-06,
+      "loss": 0.7303,
+      "step": 684
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6733512955650026,
+      "learning_rate": 8.605364270773747e-06,
+      "loss": 0.7257,
+      "step": 685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9057810252169715,
+      "learning_rate": 8.579227224136236e-06,
+      "loss": 0.7867,
+      "step": 686
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7610450213707276,
+      "learning_rate": 8.553100079664598e-06,
+      "loss": 0.7328,
+      "step": 687
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.032793421897824,
+      "learning_rate": 8.526983019453624e-06,
+      "loss": 0.7249,
+      "step": 688
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7460521148100402,
+      "learning_rate": 8.500876225527822e-06,
+      "loss": 0.8095,
+      "step": 689
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6217376656004977,
+      "learning_rate": 8.47477987984015e-06,
+      "loss": 0.7291,
+      "step": 690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7513944819984957,
+      "learning_rate": 8.448694164270733e-06,
+      "loss": 0.7784,
+      "step": 691
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.611381649493882,
+      "learning_rate": 8.422619260625626e-06,
+      "loss": 0.7257,
+      "step": 692
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8258348891260682,
+      "learning_rate": 8.396555350635516e-06,
+      "loss": 0.7283,
+      "step": 693
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8033990109637202,
+      "learning_rate": 8.370502615954481e-06,
+      "loss": 0.7287,
+      "step": 694
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6844104502620764,
+      "learning_rate": 8.3444612381587e-06,
+      "loss": 0.7541,
+      "step": 695
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7174758443711007,
+      "learning_rate": 8.31843139874521e-06,
+      "loss": 0.7548,
+      "step": 696
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7705682367042754,
+      "learning_rate": 8.292413279130625e-06,
+      "loss": 0.7236,
+      "step": 697
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7325425528188919,
+      "learning_rate": 8.266407060649877e-06,
+      "loss": 0.9631,
+      "step": 698
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7438168089314038,
+      "learning_rate": 8.240412924554955e-06,
+      "loss": 0.7418,
+      "step": 699
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8752209712433744,
+      "learning_rate": 8.214431052013636e-06,
+      "loss": 0.7553,
+      "step": 700
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7270570440689362,
+      "learning_rate": 8.188461624108226e-06,
+      "loss": 0.7538,
+      "step": 701
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9311871390100658,
+      "learning_rate": 8.162504821834296e-06,
+      "loss": 0.7256,
+      "step": 702
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8460910752011436,
+      "learning_rate": 8.13656082609943e-06,
+      "loss": 0.7963,
+      "step": 703
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0462104978397226,
+      "learning_rate": 8.110629817721949e-06,
+      "loss": 0.7499,
+      "step": 704
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7608057738874535,
+      "learning_rate": 8.084711977429655e-06,
+      "loss": 0.7796,
+      "step": 705
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7392097145934214,
+      "learning_rate": 8.05880748585858e-06,
+      "loss": 0.7155,
+      "step": 706
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8770640216100294,
+      "learning_rate": 8.03291652355172e-06,
+      "loss": 0.777,
+      "step": 707
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7660150026026007,
+      "learning_rate": 8.007039270957778e-06,
+      "loss": 0.7176,
+      "step": 708
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6498430157447325,
+      "learning_rate": 7.9811759084299e-06,
+      "loss": 0.7328,
+      "step": 709
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8793572934679837,
+      "learning_rate": 7.955326616224432e-06,
+      "loss": 0.7703,
+      "step": 710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7540539172877317,
+      "learning_rate": 7.929491574499659e-06,
+      "loss": 0.71,
+      "step": 711
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7967733518029273,
+      "learning_rate": 7.903670963314536e-06,
+      "loss": 0.765,
+      "step": 712
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.249519160918274,
+      "learning_rate": 7.87786496262745e-06,
+      "loss": 0.7677,
+      "step": 713
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1256189932027016,
+      "learning_rate": 7.852073752294958e-06,
+      "loss": 0.7837,
+      "step": 714
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.692158189126143,
+      "learning_rate": 7.826297512070533e-06,
+      "loss": 0.8175,
+      "step": 715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9232586068259186,
+      "learning_rate": 7.800536421603317e-06,
+      "loss": 0.7576,
+      "step": 716
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6114502739855013,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.6974,
+      "step": 717
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6949965072591656,
+      "learning_rate": 7.749060408007874e-06,
+      "loss": 0.7691,
+      "step": 718
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7832277026283018,
+      "learning_rate": 7.723345843644986e-06,
+      "loss": 0.7716,
+      "step": 719
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8178339858930677,
+      "learning_rate": 7.697647146567484e-06,
+      "loss": 0.6813,
+      "step": 720
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6531440704268704,
+      "learning_rate": 7.67196449588406e-06,
+      "loss": 0.7349,
+      "step": 721
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.884513817289984,
+      "learning_rate": 7.646298070591578e-06,
+      "loss": 0.8092,
+      "step": 722
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8286456645325372,
+      "learning_rate": 7.620648049573815e-06,
+      "loss": 0.8276,
+      "step": 723
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8033417929386947,
+      "learning_rate": 7.5950146116002184e-06,
+      "loss": 0.7114,
+      "step": 724
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7603889015161699,
+      "learning_rate": 7.569397935324653e-06,
+      "loss": 0.7725,
+      "step": 725
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9531949474176249,
+      "learning_rate": 7.543798199284175e-06,
+      "loss": 0.7977,
+      "step": 726
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7572273857708323,
+      "learning_rate": 7.518215581897763e-06,
+      "loss": 0.7478,
+      "step": 727
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7636076557592235,
+      "learning_rate": 7.4926502614650865e-06,
+      "loss": 0.7808,
+      "step": 728
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6236060413891016,
+      "learning_rate": 7.4671024161652726e-06,
+      "loss": 0.6999,
+      "step": 729
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7978570504725623,
+      "learning_rate": 7.441572224055644e-06,
+      "loss": 0.7888,
+      "step": 730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7447179930016969,
+      "learning_rate": 7.416059863070492e-06,
+      "loss": 0.7442,
+      "step": 731
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8746426035477285,
+      "learning_rate": 7.390565511019834e-06,
+      "loss": 0.7899,
+      "step": 732
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6582861675742879,
+      "learning_rate": 7.36508934558817e-06,
+      "loss": 0.7106,
+      "step": 733
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9468354795541432,
+      "learning_rate": 7.33963154433325e-06,
+      "loss": 0.907,
+      "step": 734
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6533282892663995,
+      "learning_rate": 7.314192284684831e-06,
+      "loss": 0.7443,
+      "step": 735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.812137238901455,
+      "learning_rate": 7.28877174394345e-06,
+      "loss": 0.8026,
+      "step": 736
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6965759263548852,
+      "learning_rate": 7.263370099279173e-06,
+      "loss": 0.7171,
+      "step": 737
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6849201712838094,
+      "learning_rate": 7.23798752773037e-06,
+      "loss": 0.7229,
+      "step": 738
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9549845175911698,
+      "learning_rate": 7.212624206202483e-06,
+      "loss": 0.7945,
+      "step": 739
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.052252022528913,
+      "learning_rate": 7.1872803114667914e-06,
+      "loss": 0.7815,
+      "step": 740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.689623175356214,
+      "learning_rate": 7.1619560201591705e-06,
+      "loss": 0.7487,
+      "step": 741
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6810171365981369,
+      "learning_rate": 7.136651508778876e-06,
+      "loss": 0.7013,
+      "step": 742
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.713982942562677,
+      "learning_rate": 7.1113669536872996e-06,
+      "loss": 0.7662,
+      "step": 743
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.663579294068721,
+      "learning_rate": 7.086102531106755e-06,
+      "loss": 0.7475,
+      "step": 744
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6762728077074722,
+      "learning_rate": 7.060858417119233e-06,
+      "loss": 0.7439,
+      "step": 745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7151994893732979,
+      "learning_rate": 7.035634787665186e-06,
+      "loss": 0.7905,
+      "step": 746
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6432153143969728,
+      "learning_rate": 7.010431818542298e-06,
+      "loss": 0.7137,
+      "step": 747
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.783952230139737,
+      "learning_rate": 6.9852496854042585e-06,
+      "loss": 0.7605,
+      "step": 748
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6970304699431817,
+      "learning_rate": 6.96008856375954e-06,
+      "loss": 0.812,
+      "step": 749
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7276725497686215,
+      "learning_rate": 6.934948628970173e-06,
+      "loss": 0.7191,
+      "step": 750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6468020212495768,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.7424,
+      "step": 751
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6587163369250661,
+      "learning_rate": 6.884733020666086e-06,
+      "loss": 0.7355,
+      "step": 752
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8841419605755626,
+      "learning_rate": 6.859657697132229e-06,
+      "loss": 0.786,
+      "step": 753
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7041092142120744,
+      "learning_rate": 6.834604260413014e-06,
+      "loss": 0.725,
+      "step": 754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7296836400825225,
+      "learning_rate": 6.809572885119952e-06,
+      "loss": 0.9621,
+      "step": 755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8242952620435268,
+      "learning_rate": 6.7845637457108016e-06,
+      "loss": 0.6999,
+      "step": 756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9707777520062301,
+      "learning_rate": 6.759577016488343e-06,
+      "loss": 0.814,
+      "step": 757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.270220679989626,
+      "learning_rate": 6.734612871599169e-06,
+      "loss": 0.7782,
+      "step": 758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6658043806109968,
+      "learning_rate": 6.7096714850324694e-06,
+      "loss": 0.7381,
+      "step": 759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8644119452238986,
+      "learning_rate": 6.684753030618819e-06,
+      "loss": 0.7478,
+      "step": 760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2948095422306167,
+      "learning_rate": 6.659857682028963e-06,
+      "loss": 0.7368,
+      "step": 761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1208385200341908,
+      "learning_rate": 6.634985612772611e-06,
+      "loss": 0.7548,
+      "step": 762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7141702238053835,
+      "learning_rate": 6.610136996197221e-06,
+      "loss": 0.7438,
+      "step": 763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9001678672249185,
+      "learning_rate": 6.58531200548681e-06,
+      "loss": 0.7737,
+      "step": 764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7649309531201068,
+      "learning_rate": 6.560510813660719e-06,
+      "loss": 0.7079,
+      "step": 765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9068990863331087,
+      "learning_rate": 6.535733593572424e-06,
+      "loss": 0.7892,
+      "step": 766
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9367103844970543,
+      "learning_rate": 6.510980517908334e-06,
+      "loss": 0.7234,
+      "step": 767
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0317875199677935,
+      "learning_rate": 6.486251759186573e-06,
+      "loss": 0.7615,
+      "step": 768
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6815856392580176,
+      "learning_rate": 6.461547489755795e-06,
+      "loss": 0.7084,
+      "step": 769
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7061867223907953,
+      "learning_rate": 6.436867881793971e-06,
+      "loss": 0.7634,
+      "step": 770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6717403404039805,
+      "learning_rate": 6.412213107307192e-06,
+      "loss": 0.6966,
+      "step": 771
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7919695914395061,
+      "learning_rate": 6.387583338128471e-06,
+      "loss": 0.7697,
+      "step": 772
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7231903954361062,
+      "learning_rate": 6.362978745916544e-06,
+      "loss": 0.7438,
+      "step": 773
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6545041338050531,
+      "learning_rate": 6.338399502154675e-06,
+      "loss": 0.7314,
+      "step": 774
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6643492287086143,
+      "learning_rate": 6.313845778149461e-06,
+      "loss": 0.748,
+      "step": 775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8015468390919295,
+      "learning_rate": 6.289317745029637e-06,
+      "loss": 0.7195,
+      "step": 776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.668469797541779,
+      "learning_rate": 6.264815573744884e-06,
+      "loss": 0.7327,
+      "step": 777
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8820280981950381,
+      "learning_rate": 6.240339435064638e-06,
+      "loss": 0.7084,
+      "step": 778
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6786492124509902,
+      "learning_rate": 6.215889499576898e-06,
+      "loss": 0.7114,
+      "step": 779
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8680560261324012,
+      "learning_rate": 6.191465937687038e-06,
+      "loss": 0.7845,
+      "step": 780
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6170035551506227,
+      "learning_rate": 6.167068919616619e-06,
+      "loss": 0.7499,
+      "step": 781
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8194525039442561,
+      "learning_rate": 6.142698615402205e-06,
+      "loss": 0.7585,
+      "step": 782
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7064689414825139,
+      "learning_rate": 6.118355194894174e-06,
+      "loss": 0.731,
+      "step": 783
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.692755583964466,
+      "learning_rate": 6.0940388277555415e-06,
+      "loss": 0.713,
+      "step": 784
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8651835657218316,
+      "learning_rate": 6.069749683460765e-06,
+      "loss": 0.778,
+      "step": 785
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6549382183961346,
+      "learning_rate": 6.0454879312945755e-06,
+      "loss": 0.7363,
+      "step": 786
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8996504748541897,
+      "learning_rate": 6.021253740350793e-06,
+      "loss": 0.7342,
+      "step": 787
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7643783690799606,
+      "learning_rate": 5.997047279531145e-06,
+      "loss": 0.7025,
+      "step": 788
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8632984944046991,
+      "learning_rate": 5.972868717544096e-06,
+      "loss": 0.8023,
+      "step": 789
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9276634114107253,
+      "learning_rate": 5.948718222903662e-06,
+      "loss": 0.7763,
+      "step": 790
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7407164110206144,
+      "learning_rate": 5.924595963928248e-06,
+      "loss": 0.7412,
+      "step": 791
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8088761713215605,
+      "learning_rate": 5.900502108739466e-06,
+      "loss": 0.9133,
+      "step": 792
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6816915897259173,
+      "learning_rate": 5.876436825260967e-06,
+      "loss": 0.6923,
+      "step": 793
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.591642337733151,
+      "learning_rate": 5.852400281217268e-06,
+      "loss": 0.7565,
+      "step": 794
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8024314952886368,
+      "learning_rate": 5.8283926441325835e-06,
+      "loss": 0.7298,
+      "step": 795
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9382419836146352,
+      "learning_rate": 5.804414081329668e-06,
+      "loss": 0.6999,
+      "step": 796
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.680841808842482,
+      "learning_rate": 5.780464759928623e-06,
+      "loss": 0.7279,
+      "step": 797
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7684976044026549,
+      "learning_rate": 5.756544846845775e-06,
+      "loss": 0.7786,
+      "step": 798
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7690624235498952,
+      "learning_rate": 5.732654508792461e-06,
+      "loss": 0.7534,
+      "step": 799
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7425699870803733,
+      "learning_rate": 5.708793912273911e-06,
+      "loss": 0.74,
+      "step": 800
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7893551808929594,
+      "learning_rate": 5.684963223588064e-06,
+      "loss": 0.685,
+      "step": 801
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4637784929343993,
+      "learning_rate": 5.66116260882442e-06,
+      "loss": 0.7708,
+      "step": 802
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7276852426254024,
+      "learning_rate": 5.637392233862862e-06,
+      "loss": 0.7412,
+      "step": 803
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5287150373476497,
+      "learning_rate": 5.613652264372533e-06,
+      "loss": 0.7386,
+      "step": 804
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6065891000237046,
+      "learning_rate": 5.589942865810646e-06,
+      "loss": 0.712,
+      "step": 805
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1058478466835417,
+      "learning_rate": 5.566264203421361e-06,
+      "loss": 0.7799,
+      "step": 806
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3668334229972512,
+      "learning_rate": 5.542616442234618e-06,
+      "loss": 0.7261,
+      "step": 807
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2831787059191615,
+      "learning_rate": 5.518999747064981e-06,
+      "loss": 0.7821,
+      "step": 808
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6078737115015747,
+      "learning_rate": 5.495414282510511e-06,
+      "loss": 0.6895,
+      "step": 809
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8005793259292987,
+      "learning_rate": 5.471860212951588e-06,
+      "loss": 0.9211,
+      "step": 810
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.644624372544686,
+      "learning_rate": 5.448337702549802e-06,
+      "loss": 0.7303,
+      "step": 811
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0124010167103101,
+      "learning_rate": 5.42484691524677e-06,
+      "loss": 0.7735,
+      "step": 812
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6335069815575214,
+      "learning_rate": 5.401388014763027e-06,
+      "loss": 0.718,
+      "step": 813
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9782246683477449,
+      "learning_rate": 5.3779611645968696e-06,
+      "loss": 0.7376,
+      "step": 814
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9113122076157943,
+      "learning_rate": 5.354566528023207e-06,
+      "loss": 0.748,
+      "step": 815
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9178078253954092,
+      "learning_rate": 5.331204268092449e-06,
+      "loss": 0.7817,
+      "step": 816
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.771316941152822,
+      "learning_rate": 5.307874547629339e-06,
+      "loss": 0.7568,
+      "step": 817
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7144878185317688,
+      "learning_rate": 5.284577529231847e-06,
+      "loss": 0.6787,
+      "step": 818
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8437400447156539,
+      "learning_rate": 5.2613133752700145e-06,
+      "loss": 0.7867,
+      "step": 819
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7590761541659021,
+      "learning_rate": 5.238082247884838e-06,
+      "loss": 0.7236,
+      "step": 820
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9607153002369102,
+      "learning_rate": 5.214884308987136e-06,
+      "loss": 0.7167,
+      "step": 821
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7381069644761629,
+      "learning_rate": 5.191719720256407e-06,
+      "loss": 0.7198,
+      "step": 822
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9034890699074482,
+      "learning_rate": 5.168588643139724e-06,
+      "loss": 0.7599,
+      "step": 823
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0434979041972645,
+      "learning_rate": 5.145491238850592e-06,
+      "loss": 0.6964,
+      "step": 824
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.806284185598044,
+      "learning_rate": 5.122427668367838e-06,
+      "loss": 0.7849,
+      "step": 825
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.608901620488363,
+      "learning_rate": 5.099398092434469e-06,
+      "loss": 0.6992,
+      "step": 826
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0429962730190467,
+      "learning_rate": 5.076402671556578e-06,
+      "loss": 0.7261,
+      "step": 827
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0305767486388024,
+      "learning_rate": 5.053441566002214e-06,
+      "loss": 0.7976,
+      "step": 828
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7956942218050364,
+      "learning_rate": 5.030514935800247e-06,
+      "loss": 0.7378,
+      "step": 829
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7103122177212345,
+      "learning_rate": 5.007622940739288e-06,
+      "loss": 0.7176,
+      "step": 830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8103283266637169,
+      "learning_rate": 4.9847657403665365e-06,
+      "loss": 0.764,
+      "step": 831
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8027601536595025,
+      "learning_rate": 4.961943493986709e-06,
+      "loss": 0.7054,
+      "step": 832
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1786583669685364,
+      "learning_rate": 4.939156360660888e-06,
+      "loss": 0.9676,
+      "step": 833
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.334812106267259,
+      "learning_rate": 4.916404499205452e-06,
+      "loss": 0.7576,
+      "step": 834
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6814625920448021,
+      "learning_rate": 4.893688068190933e-06,
+      "loss": 0.7181,
+      "step": 835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0082682757354204,
+      "learning_rate": 4.87100722594094e-06,
+      "loss": 0.7322,
+      "step": 836
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6751452168755783,
+      "learning_rate": 4.848362130531039e-06,
+      "loss": 0.7472,
+      "step": 837
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8599728785384757,
+      "learning_rate": 4.825752939787659e-06,
+      "loss": 0.7523,
+      "step": 838
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.710395793267347,
+      "learning_rate": 4.803179811286989e-06,
+      "loss": 0.7013,
+      "step": 839
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7168749866419042,
+      "learning_rate": 4.780642902353872e-06,
+      "loss": 0.7412,
+      "step": 840
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8292201780869123,
+      "learning_rate": 4.75814237006073e-06,
+      "loss": 0.7176,
+      "step": 841
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8640117437538617,
+      "learning_rate": 4.7356783712264405e-06,
+      "loss": 0.7913,
+      "step": 842
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7450987365631756,
+      "learning_rate": 4.713251062415276e-06,
+      "loss": 0.7122,
+      "step": 843
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.317700902025769,
+      "learning_rate": 4.690860599935776e-06,
+      "loss": 0.7723,
+      "step": 844
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8197212777562881,
+      "learning_rate": 4.668507139839692e-06,
+      "loss": 0.7147,
+      "step": 845
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7464544759431686,
+      "learning_rate": 4.646190837920884e-06,
+      "loss": 0.7693,
+      "step": 846
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8653745139861386,
+      "learning_rate": 4.623911849714226e-06,
+      "loss": 0.7085,
+      "step": 847
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7841214059855006,
+      "learning_rate": 4.601670330494543e-06,
+      "loss": 0.7782,
+      "step": 848
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7209304829785745,
+      "learning_rate": 4.579466435275506e-06,
+      "loss": 0.7605,
+      "step": 849
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.067265636450835,
+      "learning_rate": 4.557300318808578e-06,
+      "loss": 0.7119,
+      "step": 850
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7131002265672343,
+      "learning_rate": 4.535172135581904e-06,
+      "loss": 0.9065,
+      "step": 851
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4257557054264252,
+      "learning_rate": 4.5130820398192645e-06,
+      "loss": 0.7249,
+      "step": 852
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8717966818561004,
+      "learning_rate": 4.491030185478976e-06,
+      "loss": 0.7965,
+      "step": 853
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7010372229617998,
+      "learning_rate": 4.469016726252838e-06,
+      "loss": 0.6888,
+      "step": 854
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7485367435150484,
+      "learning_rate": 4.447041815565053e-06,
+      "loss": 0.7653,
+      "step": 855
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8001119586291509,
+      "learning_rate": 4.425105606571145e-06,
+      "loss": 0.7334,
+      "step": 856
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7640389420778844,
+      "learning_rate": 4.403208252156921e-06,
+      "loss": 0.7187,
+      "step": 857
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8971742940194984,
+      "learning_rate": 4.381349904937371e-06,
+      "loss": 0.7767,
+      "step": 858
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1740820679055584,
+      "learning_rate": 4.359530717255639e-06,
+      "loss": 0.7249,
+      "step": 859
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6986366047436583,
+      "learning_rate": 4.33775084118193e-06,
+      "loss": 0.7179,
+      "step": 860
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8415150140135839,
+      "learning_rate": 4.316010428512472e-06,
+      "loss": 0.7352,
+      "step": 861
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.693989704657536,
+      "learning_rate": 4.294309630768452e-06,
+      "loss": 0.7198,
+      "step": 862
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.785712387628987,
+      "learning_rate": 4.272648599194948e-06,
+      "loss": 0.7342,
+      "step": 863
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7227446133003971,
+      "learning_rate": 4.2510274847599e-06,
+      "loss": 0.6988,
+      "step": 864
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.781512469207994,
+      "learning_rate": 4.229446438153027e-06,
+      "loss": 0.7605,
+      "step": 865
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9599716399355726,
+      "learning_rate": 4.2079056097848085e-06,
+      "loss": 0.7463,
+      "step": 866
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.121288987875495,
+      "learning_rate": 4.186405149785403e-06,
+      "loss": 0.7657,
+      "step": 867
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7695666862397598,
+      "learning_rate": 4.164945208003635e-06,
+      "loss": 0.7855,
+      "step": 868
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8280999391739019,
+      "learning_rate": 4.1435259340059306e-06,
+      "loss": 0.7638,
+      "step": 869
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7407140523322566,
+      "learning_rate": 4.12214747707527e-06,
+      "loss": 0.7551,
+      "step": 870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.706352331169021,
+      "learning_rate": 4.10080998621017e-06,
+      "loss": 0.7319,
+      "step": 871
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9699304617717999,
+      "learning_rate": 4.079513610123619e-06,
+      "loss": 0.7009,
+      "step": 872
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0350619563183492,
+      "learning_rate": 4.058258497242065e-06,
+      "loss": 0.6928,
+      "step": 873
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8198269776716316,
+      "learning_rate": 4.037044795704356e-06,
+      "loss": 0.7867,
+      "step": 874
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7914715535490132,
+      "learning_rate": 4.015872653360739e-06,
+      "loss": 0.7299,
+      "step": 875
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.815228058365996,
+      "learning_rate": 3.994742217771795e-06,
+      "loss": 0.7472,
+      "step": 876
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7304783307425929,
+      "learning_rate": 3.973653636207437e-06,
+      "loss": 0.7024,
+      "step": 877
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8037609450293971,
+      "learning_rate": 3.952607055645866e-06,
+      "loss": 0.7832,
+      "step": 878
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8091738584690644,
+      "learning_rate": 3.931602622772561e-06,
+      "loss": 0.8074,
+      "step": 879
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7545926922064994,
+      "learning_rate": 3.910640483979251e-06,
+      "loss": 0.7248,
+      "step": 880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9106780317338347,
+      "learning_rate": 3.889720785362884e-06,
+      "loss": 0.6981,
+      "step": 881
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.776234342011967,
+      "learning_rate": 3.86884367272463e-06,
+      "loss": 0.7394,
+      "step": 882
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7981574271569339,
+      "learning_rate": 3.848009291568845e-06,
+      "loss": 0.7459,
+      "step": 883
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9674876495340984,
+      "learning_rate": 3.827217787102072e-06,
+      "loss": 0.7604,
+      "step": 884
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7418699923992804,
+      "learning_rate": 3.8064693042320133e-06,
+      "loss": 0.6968,
+      "step": 885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7559758820093558,
+      "learning_rate": 3.7857639875665377e-06,
+      "loss": 0.7049,
+      "step": 886
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7182905108237919,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 0.9489,
+      "step": 887
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7967620746811465,
+      "learning_rate": 3.7444834297755504e-06,
+      "loss": 0.7458,
+      "step": 888
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.88108583702487,
+      "learning_rate": 3.7239084763575016e-06,
+      "loss": 0.7047,
+      "step": 889
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8852617935762098,
+      "learning_rate": 3.703377264556952e-06,
+      "loss": 0.7265,
+      "step": 890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.000729797943329,
+      "learning_rate": 3.6828899374674933e-06,
+      "loss": 0.71,
+      "step": 891
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7820465299003037,
+      "learning_rate": 3.6624466378768387e-06,
+      "loss": 0.7219,
+      "step": 892
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6854222880485018,
+      "learning_rate": 3.642047508265866e-06,
+      "loss": 0.7178,
+      "step": 893
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7766746381371006,
+      "learning_rate": 3.6216926908076043e-06,
+      "loss": 0.7315,
+      "step": 894
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0254055400633215,
+      "learning_rate": 3.601382327366236e-06,
+      "loss": 0.7462,
+      "step": 895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8356546224811126,
+      "learning_rate": 3.581116559496134e-06,
+      "loss": 0.7376,
+      "step": 896
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8806230598218974,
+      "learning_rate": 3.560895528440844e-06,
+      "loss": 0.7779,
+      "step": 897
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8008457494674663,
+      "learning_rate": 3.540719375132129e-06,
+      "loss": 0.7177,
+      "step": 898
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8139819422320145,
+      "learning_rate": 3.5205882401889624e-06,
+      "loss": 0.721,
+      "step": 899
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9471787776980874,
+      "learning_rate": 3.500502263916571e-06,
+      "loss": 0.7912,
+      "step": 900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7932989697628156,
+      "learning_rate": 3.4804615863054314e-06,
+      "loss": 0.7595,
+      "step": 901
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0649060804145318,
+      "learning_rate": 3.460466347030319e-06,
+      "loss": 0.7509,
+      "step": 902
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7142009512556069,
+      "learning_rate": 3.4405166854493266e-06,
+      "loss": 0.7318,
+      "step": 903
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7594076544359142,
+      "learning_rate": 3.4206127406028744e-06,
+      "loss": 0.7226,
+      "step": 904
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1560899315781274,
+      "learning_rate": 3.4007546512127764e-06,
+      "loss": 0.7673,
+      "step": 905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7235419177395497,
+      "learning_rate": 3.380942555681238e-06,
+      "loss": 0.7045,
+      "step": 906
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8626786351866071,
+      "learning_rate": 3.361176592089919e-06,
+      "loss": 0.7688,
+      "step": 907
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7493530543703399,
+      "learning_rate": 3.3414568981989492e-06,
+      "loss": 0.8024,
+      "step": 908
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0893622118531872,
+      "learning_rate": 3.3217836114459887e-06,
+      "loss": 0.76,
+      "step": 909
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.504320455660752,
+      "learning_rate": 3.302156868945259e-06,
+      "loss": 0.7068,
+      "step": 910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8774510119923523,
+      "learning_rate": 3.282576807486575e-06,
+      "loss": 0.7597,
+      "step": 911
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7146814184813233,
+      "learning_rate": 3.2630435635344283e-06,
+      "loss": 0.7331,
+      "step": 912
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8867282718460139,
+      "learning_rate": 3.2435572732269905e-06,
+      "loss": 0.7242,
+      "step": 913
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8530695993724191,
+      "learning_rate": 3.224118072375204e-06,
+      "loss": 0.726,
+      "step": 914
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8456246882463645,
+      "learning_rate": 3.2047260964618034e-06,
+      "loss": 0.6793,
+      "step": 915
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7033499769143633,
+      "learning_rate": 3.1853814806404003e-06,
+      "loss": 0.7412,
+      "step": 916
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0685630322408217,
+      "learning_rate": 3.1660843597345137e-06,
+      "loss": 0.7407,
+      "step": 917
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7093894273031172,
+      "learning_rate": 3.1468348682366567e-06,
+      "loss": 0.6954,
+      "step": 918
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9152791614368945,
+      "learning_rate": 3.1276331403073733e-06,
+      "loss": 0.7963,
+      "step": 919
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8560196065937299,
+      "learning_rate": 3.108479309774326e-06,
+      "loss": 0.7747,
+      "step": 920
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.750932964422409,
+      "learning_rate": 3.089373510131354e-06,
+      "loss": 0.7199,
+      "step": 921
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8897407163907054,
+      "learning_rate": 3.0703158745375316e-06,
+      "loss": 0.7136,
+      "step": 922
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7280486281611915,
+      "learning_rate": 3.051306535816264e-06,
+      "loss": 0.7178,
+      "step": 923
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7494491235280746,
+      "learning_rate": 3.032345626454335e-06,
+      "loss": 0.6934,
+      "step": 924
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9776284514532279,
+      "learning_rate": 3.013433278601011e-06,
+      "loss": 0.7572,
+      "step": 925
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7431082971092643,
+      "learning_rate": 2.9945696240670905e-06,
+      "loss": 0.6983,
+      "step": 926
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7295956931710292,
+      "learning_rate": 2.975754794324015e-06,
+      "loss": 0.736,
+      "step": 927
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.702292032302322,
+      "learning_rate": 2.9569889205029324e-06,
+      "loss": 0.7183,
+      "step": 928
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0103699872285812,
+      "learning_rate": 2.9382721333937847e-06,
+      "loss": 0.825,
+      "step": 929
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1101207666144408,
+      "learning_rate": 2.9196045634444116e-06,
+      "loss": 0.747,
+      "step": 930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8981156874327574,
+      "learning_rate": 2.9009863407596173e-06,
+      "loss": 0.6751,
+      "step": 931
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.82507627883392,
+      "learning_rate": 2.8824175951002918e-06,
+      "loss": 0.7284,
+      "step": 932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7764399943617596,
+      "learning_rate": 2.8638984558824777e-06,
+      "loss": 0.7154,
+      "step": 933
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.921676258416402,
+      "learning_rate": 2.8454290521764927e-06,
+      "loss": 0.8008,
+      "step": 934
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7943312275700011,
+      "learning_rate": 2.827009512706018e-06,
+      "loss": 0.678,
+      "step": 935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6681039830338549,
+      "learning_rate": 2.8086399658471965e-06,
+      "loss": 0.7341,
+      "step": 936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7032793048479545,
+      "learning_rate": 2.7903205396277546e-06,
+      "loss": 0.7128,
+      "step": 937
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2849422924054812,
+      "learning_rate": 2.7720513617260857e-06,
+      "loss": 0.798,
+      "step": 938
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8490817516763229,
+      "learning_rate": 2.7538325594703907e-06,
+      "loss": 0.7627,
+      "step": 939
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8197548167445705,
+      "learning_rate": 2.7356642598377604e-06,
+      "loss": 0.6781,
+      "step": 940
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9894861587012567,
+      "learning_rate": 2.7175465894533114e-06,
+      "loss": 0.7838,
+      "step": 941
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9384298357928221,
+      "learning_rate": 2.6994796745893e-06,
+      "loss": 0.7127,
+      "step": 942
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.983967841932367,
+      "learning_rate": 2.681463641164228e-06,
+      "loss": 0.7648,
+      "step": 943
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8913625927868462,
+      "learning_rate": 2.663498614741987e-06,
+      "loss": 0.6791,
+      "step": 944
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7360833304237321,
+      "learning_rate": 2.6455847205309615e-06,
+      "loss": 0.9459,
+      "step": 945
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9125594128565054,
+      "learning_rate": 2.6277220833831774e-06,
+      "loss": 0.7253,
+      "step": 946
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.424205737475593,
+      "learning_rate": 2.6099108277934105e-06,
+      "loss": 0.7665,
+      "step": 947
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6239181323551143,
+      "learning_rate": 2.5921510778983405e-06,
+      "loss": 0.6995,
+      "step": 948
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8191129976309806,
+      "learning_rate": 2.5744429574756612e-06,
+      "loss": 0.6895,
+      "step": 949
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6939973362903256,
+      "learning_rate": 2.556786589943252e-06,
+      "loss": 0.7319,
+      "step": 950
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7494463577364916,
+      "learning_rate": 2.539182098358276e-06,
+      "loss": 0.7496,
+      "step": 951
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6750590373790963,
+      "learning_rate": 2.5216296054163547e-06,
+      "loss": 0.7173,
+      "step": 952
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.839435548248877,
+      "learning_rate": 2.5041292334507028e-06,
+      "loss": 0.7078,
+      "step": 953
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.666327413083182,
+      "learning_rate": 2.4866811044312667e-06,
+      "loss": 0.7453,
+      "step": 954
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7430582725949687,
+      "learning_rate": 2.469285339963892e-06,
+      "loss": 0.7663,
+      "step": 955
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.784394608312579,
+      "learning_rate": 2.451942061289455e-06,
+      "loss": 0.7272,
+      "step": 956
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8721460252377428,
+      "learning_rate": 2.4346513892830427e-06,
+      "loss": 0.7016,
+      "step": 957
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7416628921072561,
+      "learning_rate": 2.4174134444530817e-06,
+      "loss": 0.6846,
+      "step": 958
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7453880317519894,
+      "learning_rate": 2.400228346940525e-06,
+      "loss": 0.7394,
+      "step": 959
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6651273689288474,
+      "learning_rate": 2.3830962165179995e-06,
+      "loss": 0.7053,
+      "step": 960
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7626610515540132,
+      "learning_rate": 2.3660171725889703e-06,
+      "loss": 0.7718,
+      "step": 961
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9005992075695879,
+      "learning_rate": 2.34899133418692e-06,
+      "loss": 0.7152,
+      "step": 962
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7626192939730301,
+      "learning_rate": 2.332018819974502e-06,
+      "loss": 0.935,
+      "step": 963
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8831317906248485,
+      "learning_rate": 2.3150997482427364e-06,
+      "loss": 0.7764,
+      "step": 964
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.855960088112948,
+      "learning_rate": 2.2982342369101597e-06,
+      "loss": 0.7113,
+      "step": 965
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7890122170141386,
+      "learning_rate": 2.2814224035220268e-06,
+      "loss": 0.7326,
+      "step": 966
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1040415770195264,
+      "learning_rate": 2.2646643652494693e-06,
+      "loss": 0.6817,
+      "step": 967
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8304476366154898,
+      "learning_rate": 2.2479602388887013e-06,
+      "loss": 0.7993,
+      "step": 968
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.016984492808275,
+      "learning_rate": 2.231310140860192e-06,
+      "loss": 0.6779,
+      "step": 969
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6542955078304242,
+      "learning_rate": 2.214714187207848e-06,
+      "loss": 0.7488,
+      "step": 970
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7481185633593638,
+      "learning_rate": 2.1981724935982264e-06,
+      "loss": 0.6866,
+      "step": 971
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7830187553144453,
+      "learning_rate": 2.1816851753197023e-06,
+      "loss": 0.7721,
+      "step": 972
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8752788909536485,
+      "learning_rate": 2.165252347281692e-06,
+      "loss": 0.7795,
+      "step": 973
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6299356236543497,
+      "learning_rate": 2.1488741240138245e-06,
+      "loss": 0.6954,
+      "step": 974
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7990380688458015,
+      "learning_rate": 2.132550619665168e-06,
+      "loss": 0.7333,
+      "step": 975
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2145641813362529,
+      "learning_rate": 2.1162819480034226e-06,
+      "loss": 0.7001,
+      "step": 976
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8298203255675277,
+      "learning_rate": 2.100068222414121e-06,
+      "loss": 0.7891,
+      "step": 977
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8008073678517103,
+      "learning_rate": 2.0839095558998566e-06,
+      "loss": 0.6915,
+      "step": 978
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5658161387982137,
+      "learning_rate": 2.0678060610794748e-06,
+      "loss": 0.755,
+      "step": 979
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9126071191291331,
+      "learning_rate": 2.051757850187306e-06,
+      "loss": 0.6855,
+      "step": 980
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3996962841854064,
+      "learning_rate": 2.035765035072369e-06,
+      "loss": 0.7582,
+      "step": 981
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.728188863866462,
+      "learning_rate": 2.019827727197605e-06,
+      "loss": 0.7182,
+      "step": 982
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0013721752483946,
+      "learning_rate": 2.0039460376390927e-06,
+      "loss": 0.7657,
+      "step": 983
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7105445186726845,
+      "learning_rate": 1.9881200770852695e-06,
+      "loss": 0.7218,
+      "step": 984
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0901322332170915,
+      "learning_rate": 1.9723499558361746e-06,
+      "loss": 0.7132,
+      "step": 985
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7205358168272719,
+      "learning_rate": 1.9566357838026594e-06,
+      "loss": 0.7913,
+      "step": 986
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7718716619554407,
+      "learning_rate": 1.9409776705056514e-06,
+      "loss": 0.7391,
+      "step": 987
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6804984682449529,
+      "learning_rate": 1.925375725075357e-06,
+      "loss": 0.74,
+      "step": 988
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0969108480232745,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.7127,
+      "step": 989
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8439129649895618,
+      "learning_rate": 1.8943407723776796e-06,
+      "loss": 0.7256,
+      "step": 990
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8885118400371784,
+      "learning_rate": 1.8789079814103673e-06,
+      "loss": 0.7073,
+      "step": 991
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6723163236059383,
+      "learning_rate": 1.8635317909083983e-06,
+      "loss": 0.7152,
+      "step": 992
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8588567326201004,
+      "learning_rate": 1.8482123080371095e-06,
+      "loss": 0.7577,
+      "step": 993
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.607474502977786,
+      "learning_rate": 1.8329496395666079e-06,
+      "loss": 0.7154,
+      "step": 994
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8958812172692612,
+      "learning_rate": 1.8177438918710267e-06,
+      "loss": 0.6998,
+      "step": 995
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7033894194441586,
+      "learning_rate": 1.80259517092779e-06,
+      "loss": 0.7552,
+      "step": 996
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8086032172961205,
+      "learning_rate": 1.7875035823168641e-06,
+      "loss": 0.7389,
+      "step": 997
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1253010208488041,
+      "learning_rate": 1.7724692312200376e-06,
+      "loss": 0.7464,
+      "step": 998
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6895135060688806,
+      "learning_rate": 1.7574922224201674e-06,
+      "loss": 0.7047,
+      "step": 999
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8829017369933378,
+      "learning_rate": 1.7425726603004678e-06,
+      "loss": 0.7656,
+      "step": 1000
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6860873603849518,
+      "learning_rate": 1.7277106488437745e-06,
+      "loss": 0.7239,
+      "step": 1001
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.635453834244164,
+      "learning_rate": 1.712906291631814e-06,
+      "loss": 0.7234,
+      "step": 1002
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1079926756794962,
+      "learning_rate": 1.6981596918444953e-06,
+      "loss": 0.7014,
+      "step": 1003
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7656594239859689,
+      "learning_rate": 1.6834709522591753e-06,
+      "loss": 0.9494,
+      "step": 1004
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7270266942882763,
+      "learning_rate": 1.6688401752499606e-06,
+      "loss": 0.7243,
+      "step": 1005
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7651986913982075,
+      "learning_rate": 1.6542674627869738e-06,
+      "loss": 0.7675,
+      "step": 1006
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9828422571147613,
+      "learning_rate": 1.6397529164356606e-06,
+      "loss": 0.6741,
+      "step": 1007
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7603156978757374,
+      "learning_rate": 1.625296637356072e-06,
+      "loss": 0.7573,
+      "step": 1008
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8472696700308414,
+      "learning_rate": 1.610898726302157e-06,
+      "loss": 0.7494,
+      "step": 1009
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6937112879304376,
+      "learning_rate": 1.596559283621074e-06,
+      "loss": 0.7134,
+      "step": 1010
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8748304283264113,
+      "learning_rate": 1.5822784092524734e-06,
+      "loss": 0.7262,
+      "step": 1011
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9748756629520862,
+      "learning_rate": 1.5680562027278156e-06,
+      "loss": 0.6961,
+      "step": 1012
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.643035441216947,
+      "learning_rate": 1.5538927631696687e-06,
+      "loss": 0.7522,
+      "step": 1013
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8381094528692606,
+      "learning_rate": 1.539788189291025e-06,
+      "loss": 0.7783,
+      "step": 1014
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6401184552869392,
+      "learning_rate": 1.5257425793946011e-06,
+      "loss": 0.7114,
+      "step": 1015
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.950561773518692,
+      "learning_rate": 1.511756031372168e-06,
+      "loss": 0.6774,
+      "step": 1016
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.239209946840932,
+      "learning_rate": 1.4978286427038602e-06,
+      "loss": 0.761,
+      "step": 1017
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6529592571644756,
+      "learning_rate": 1.4839605104574905e-06,
+      "loss": 0.7235,
+      "step": 1018
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9081095984829524,
+      "learning_rate": 1.470151731287891e-06,
+      "loss": 0.7803,
+      "step": 1019
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6648644445461772,
+      "learning_rate": 1.4564024014362165e-06,
+      "loss": 0.6986,
+      "step": 1020
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7844686740734905,
+      "learning_rate": 1.4427126167292994e-06,
+      "loss": 0.8966,
+      "step": 1021
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6351089237736126,
+      "learning_rate": 1.4290824725789542e-06,
+      "loss": 0.7154,
+      "step": 1022
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8221325820670042,
+      "learning_rate": 1.4155120639813392e-06,
+      "loss": 0.765,
+      "step": 1023
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6290893105949764,
+      "learning_rate": 1.4020014855162755e-06,
+      "loss": 0.717,
+      "step": 1024
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0955841139907345,
+      "learning_rate": 1.3885508313465957e-06,
+      "loss": 0.7065,
+      "step": 1025
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7737692058517823,
+      "learning_rate": 1.3751601952174887e-06,
+      "loss": 0.7318,
+      "step": 1026
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5286587140700232,
+      "learning_rate": 1.3618296704558364e-06,
+      "loss": 0.7747,
+      "step": 1027
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6888559096241698,
+      "learning_rate": 1.3485593499695816e-06,
+      "loss": 0.7084,
+      "step": 1028
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7988970224208605,
+      "learning_rate": 1.3353493262470564e-06,
+      "loss": 0.7637,
+      "step": 1029
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7511463302857055,
+      "learning_rate": 1.322199691356366e-06,
+      "loss": 0.7007,
+      "step": 1030
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8028398716836663,
+      "learning_rate": 1.3091105369447166e-06,
+      "loss": 0.7588,
+      "step": 1031
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7725771841392625,
+      "learning_rate": 1.2960819542378055e-06,
+      "loss": 0.8151,
+      "step": 1032
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6433863531637293,
+      "learning_rate": 1.2831140340391602e-06,
+      "loss": 0.6788,
+      "step": 1033
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.952916748637187,
+      "learning_rate": 1.270206866729523e-06,
+      "loss": 0.6699,
+      "step": 1034
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7951687209949158,
+      "learning_rate": 1.2573605422662193e-06,
+      "loss": 0.7637,
+      "step": 1035
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5286975245649691,
+      "learning_rate": 1.2445751501825142e-06,
+      "loss": 0.7302,
+      "step": 1036
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8506490788525439,
+      "learning_rate": 1.2318507795870138e-06,
+      "loss": 0.713,
+      "step": 1037
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6961051414975093,
+      "learning_rate": 1.2191875191630209e-06,
+      "loss": 0.7382,
+      "step": 1038
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8640017439719226,
+      "learning_rate": 1.2065854571679348e-06,
+      "loss": 0.7212,
+      "step": 1039
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.78539551449576,
+      "learning_rate": 1.19404468143262e-06,
+      "loss": 0.7239,
+      "step": 1040
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.752929393999625,
+      "learning_rate": 1.1815652793608123e-06,
+      "loss": 0.8931,
+      "step": 1041
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.673731559915361,
+      "learning_rate": 1.1691473379284945e-06,
+      "loss": 0.7193,
+      "step": 1042
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8567518134991969,
+      "learning_rate": 1.1567909436832936e-06,
+      "loss": 0.7068,
+      "step": 1043
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6509979353905454,
+      "learning_rate": 1.144496182743885e-06,
+      "loss": 0.7164,
+      "step": 1044
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7062037758911758,
+      "learning_rate": 1.132263140799381e-06,
+      "loss": 0.6837,
+      "step": 1045
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6579375636488486,
+      "learning_rate": 1.1200919031087466e-06,
+      "loss": 0.7659,
+      "step": 1046
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7422948450413535,
+      "learning_rate": 1.1079825545001887e-06,
+      "loss": 0.7113,
+      "step": 1047
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.833636441395792,
+      "learning_rate": 1.095935179370584e-06,
+      "loss": 0.7443,
+      "step": 1048
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8317534249824864,
+      "learning_rate": 1.0839498616848753e-06,
+      "loss": 0.6745,
+      "step": 1049
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0356523854290443,
+      "learning_rate": 1.0720266849754934e-06,
+      "loss": 0.7434,
+      "step": 1050
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1041023488041275,
+      "learning_rate": 1.0601657323417746e-06,
+      "loss": 0.7677,
+      "step": 1051
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6980084297806788,
+      "learning_rate": 1.0483670864493777e-06,
+      "loss": 0.7097,
+      "step": 1052
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.791631898771703,
+      "learning_rate": 1.0366308295297156e-06,
+      "loss": 0.7907,
+      "step": 1053
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6585144445269271,
+      "learning_rate": 1.0249570433793709e-06,
+      "loss": 0.7192,
+      "step": 1054
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7646207527618,
+      "learning_rate": 1.0133458093595383e-06,
+      "loss": 0.7745,
+      "step": 1055
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9764143752326625,
+      "learning_rate": 1.0017972083954508e-06,
+      "loss": 0.7666,
+      "step": 1056
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6801100288614615,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 0.6842,
+      "step": 1057
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6746840419061803,
+      "learning_rate": 9.788882271522394e-07,
+      "loss": 0.7339,
+      "step": 1058
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.907253416305685,
+      "learning_rate": 9.675280065387117e-07,
+      "loss": 0.7206,
+      "step": 1059
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.726209400044091,
+      "learning_rate": 9.56230738311007e-07,
+      "loss": 0.7547,
+      "step": 1060
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.769371891517745,
+      "learning_rate": 9.449965012061524e-07,
+      "loss": 0.7047,
+      "step": 1061
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.711127798472907,
+      "learning_rate": 9.33825373521875e-07,
+      "loss": 0.9369,
+      "step": 1062
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6701289608294342,
+      "learning_rate": 9.227174331160538e-07,
+      "loss": 0.6892,
+      "step": 1063
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.982788269405761,
+      "learning_rate": 9.116727574061879e-07,
+      "loss": 0.7798,
+      "step": 1064
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0280878200899732,
+      "learning_rate": 9.006914233688402e-07,
+      "loss": 0.7755,
+      "step": 1065
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6477681881569981,
+      "learning_rate": 8.897735075391156e-07,
+      "loss": 0.6918,
+      "step": 1066
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9025417635880701,
+      "learning_rate": 8.789190860101226e-07,
+      "loss": 0.7137,
+      "step": 1067
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.84702821356326,
+      "learning_rate": 8.68128234432436e-07,
+      "loss": 0.741,
+      "step": 1068
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0768068055276054,
+      "learning_rate": 8.574010280135824e-07,
+      "loss": 0.7568,
+      "step": 1069
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7162813325180957,
+      "learning_rate": 8.467375415175027e-07,
+      "loss": 0.6719,
+      "step": 1070
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7297868706019205,
+      "learning_rate": 8.361378492640471e-07,
+      "loss": 0.7366,
+      "step": 1071
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7050541308972996,
+      "learning_rate": 8.256020251284381e-07,
+      "loss": 0.7088,
+      "step": 1072
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9629620458225608,
+      "learning_rate": 8.151301425407699e-07,
+      "loss": 0.751,
+      "step": 1073
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6481695628169805,
+      "learning_rate": 8.047222744854943e-07,
+      "loss": 0.6811,
+      "step": 1074
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9408626040318457,
+      "learning_rate": 7.943784935009036e-07,
+      "loss": 0.7025,
+      "step": 1075
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6622493928442372,
+      "learning_rate": 7.840988716786391e-07,
+      "loss": 0.7017,
+      "step": 1076
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.322304060015091,
+      "learning_rate": 7.738834806631712e-07,
+      "loss": 0.7616,
+      "step": 1077
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7318777956353805,
+      "learning_rate": 7.637323916513195e-07,
+      "loss": 0.7139,
+      "step": 1078
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7361656060791522,
+      "learning_rate": 7.536456753917377e-07,
+      "loss": 0.746,
+      "step": 1079
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6130095676524424,
+      "learning_rate": 7.43623402184438e-07,
+      "loss": 0.7064,
+      "step": 1080
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8844999296767325,
+      "learning_rate": 7.336656418802857e-07,
+      "loss": 0.7856,
+      "step": 1081
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7614884174231038,
+      "learning_rate": 7.237724638805221e-07,
+      "loss": 0.7159,
+      "step": 1082
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8425515248166154,
+      "learning_rate": 7.13943937136281e-07,
+      "loss": 0.7488,
+      "step": 1083
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8727864878761608,
+      "learning_rate": 7.041801301480977e-07,
+      "loss": 0.6786,
+      "step": 1084
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7682495122053381,
+      "learning_rate": 6.944811109654481e-07,
+      "loss": 0.766,
+      "step": 1085
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7003901282372423,
+      "learning_rate": 6.848469471862573e-07,
+      "loss": 0.7252,
+      "step": 1086
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9855614805548879,
+      "learning_rate": 6.752777059564431e-07,
+      "loss": 0.735,
+      "step": 1087
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6982929246787412,
+      "learning_rate": 6.657734539694361e-07,
+      "loss": 0.6727,
+      "step": 1088
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0695218076662443,
+      "learning_rate": 6.563342574657228e-07,
+      "loss": 0.7589,
+      "step": 1089
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6969108206103031,
+      "learning_rate": 6.469601822323845e-07,
+      "loss": 0.737,
+      "step": 1090
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8542613682679879,
+      "learning_rate": 6.37651293602628e-07,
+      "loss": 0.7434,
+      "step": 1091
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.070070205427742,
+      "learning_rate": 6.284076564553465e-07,
+      "loss": 0.7772,
+      "step": 1092
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9709894613108512,
+      "learning_rate": 6.192293352146506e-07,
+      "loss": 0.6913,
+      "step": 1093
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8578412309136388,
+      "learning_rate": 6.101163938494359e-07,
+      "loss": 0.7945,
+      "step": 1094
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6815858254814903,
+      "learning_rate": 6.010688958729205e-07,
+      "loss": 0.7105,
+      "step": 1095
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8785810603135195,
+      "learning_rate": 5.920869043422172e-07,
+      "loss": 0.7508,
+      "step": 1096
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7696049459298279,
+      "learning_rate": 5.831704818578842e-07,
+      "loss": 0.6919,
+      "step": 1097
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8964464972262199,
+      "learning_rate": 5.743196905634951e-07,
+      "loss": 0.9438,
+      "step": 1098
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.257968792198762,
+      "learning_rate": 5.655345921451982e-07,
+      "loss": 0.7589,
+      "step": 1099
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7524332457902064,
+      "learning_rate": 5.568152478312927e-07,
+      "loss": 0.7029,
+      "step": 1100
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8609259200911433,
+      "learning_rate": 5.481617183918053e-07,
+      "loss": 0.7577,
+      "step": 1101
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.662361357241114,
+      "learning_rate": 5.395740641380532e-07,
+      "loss": 0.6769,
+      "step": 1102
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8160108402179471,
+      "learning_rate": 5.310523449222393e-07,
+      "loss": 0.7656,
+      "step": 1103
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6219049100157225,
+      "learning_rate": 5.225966201370236e-07,
+      "loss": 0.7224,
+      "step": 1104
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9431382265590498,
+      "learning_rate": 5.142069487151191e-07,
+      "loss": 0.7428,
+      "step": 1105
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.679885043251871,
+      "learning_rate": 5.058833891288683e-07,
+      "loss": 0.7351,
+      "step": 1106
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6851273951939502,
+      "learning_rate": 4.976259993898503e-07,
+      "loss": 0.7212,
+      "step": 1107
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8333080609379073,
+      "learning_rate": 4.894348370484648e-07,
+      "loss": 0.7135,
+      "step": 1108
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6753269464550602,
+      "learning_rate": 4.813099591935356e-07,
+      "loss": 0.7087,
+      "step": 1109
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8678235883683769,
+      "learning_rate": 4.732514224519158e-07,
+      "loss": 0.7692,
+      "step": 1110
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0993533773356294,
+      "learning_rate": 4.6525928298808645e-07,
+      "loss": 0.6869,
+      "step": 1111
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.632999790409631,
+      "learning_rate": 4.573335965037706e-07,
+      "loss": 0.7027,
+      "step": 1112
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8421959960330501,
+      "learning_rate": 4.49474418237541e-07,
+      "loss": 0.7078,
+      "step": 1113
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8323121149381778,
+      "learning_rate": 4.4168180296443875e-07,
+      "loss": 0.7364,
+      "step": 1114
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9064372845700429,
+      "learning_rate": 4.3395580499559276e-07,
+      "loss": 0.7084,
+      "step": 1115
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6938693276351212,
+      "learning_rate": 4.2629647817783116e-07,
+      "loss": 0.7007,
+      "step": 1116
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7598920775831073,
+      "learning_rate": 4.187038758933204e-07,
+      "loss": 0.9032,
+      "step": 1117
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8054024887878547,
+      "learning_rate": 4.1117805105918116e-07,
+      "loss": 0.7208,
+      "step": 1118
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.276577144975448,
+      "learning_rate": 4.037190561271276e-07,
+      "loss": 0.7599,
+      "step": 1119
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7106066143968043,
+      "learning_rate": 3.963269430830974e-07,
+      "loss": 0.6726,
+      "step": 1120
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.099538823481326,
+      "learning_rate": 3.89001763446889e-07,
+      "loss": 0.7393,
+      "step": 1121
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.716614482858073,
+      "learning_rate": 3.817435682718096e-07,
+      "loss": 0.7264,
+      "step": 1122
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8363541607159143,
+      "learning_rate": 3.745524081443075e-07,
+      "loss": 0.75,
+      "step": 1123
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7997627886267095,
+      "learning_rate": 3.674283331836304e-07,
+      "loss": 0.7214,
+      "step": 1124
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7228247423346588,
+      "learning_rate": 3.603713930414676e-07,
+      "loss": 0.7166,
+      "step": 1125
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8443933354700255,
+      "learning_rate": 3.5338163690161296e-07,
+      "loss": 0.7773,
+      "step": 1126
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6871583588489021,
+      "learning_rate": 3.4645911347961357e-07,
+      "loss": 0.7156,
+      "step": 1127
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8222690563327123,
+      "learning_rate": 3.3960387102243696e-07,
+      "loss": 0.7435,
+      "step": 1128
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7681369216045033,
+      "learning_rate": 3.328159573081258e-07,
+      "loss": 0.695,
+      "step": 1129
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7424532145299381,
+      "learning_rate": 3.2609541964547796e-07,
+      "loss": 0.7477,
+      "step": 1130
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7128253901529428,
+      "learning_rate": 3.194423048737072e-07,
+      "loss": 0.7363,
+      "step": 1131
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8406814187767365,
+      "learning_rate": 3.128566593621152e-07,
+      "loss": 0.791,
+      "step": 1132
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7273991567586923,
+      "learning_rate": 3.0633852900977755e-07,
+      "loss": 0.6994,
+      "step": 1133
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8998608168646478,
+      "learning_rate": 2.998879592452164e-07,
+      "loss": 0.7172,
+      "step": 1134
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8385764363694106,
+      "learning_rate": 2.9350499502608707e-07,
+      "loss": 0.7284,
+      "step": 1135
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9559379501876079,
+      "learning_rate": 2.871896808388608e-07,
+      "loss": 0.7699,
+      "step": 1136
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7705087189624642,
+      "learning_rate": 2.809420606985236e-07,
+      "loss": 0.6917,
+      "step": 1137
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8239505703146734,
+      "learning_rate": 2.747621781482568e-07,
+      "loss": 0.6501,
+      "step": 1138
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7486058376689182,
+      "learning_rate": 2.6865007625914483e-07,
+      "loss": 0.7359,
+      "step": 1139
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8979667407670275,
+      "learning_rate": 2.6260579762986905e-07,
+      "loss": 0.9232,
+      "step": 1140
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9371446392497035,
+      "learning_rate": 2.5662938438641207e-07,
+      "loss": 0.7713,
+      "step": 1141
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7616238328265476,
+      "learning_rate": 2.507208781817638e-07,
+      "loss": 0.6842,
+      "step": 1142
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6819481273668921,
+      "learning_rate": 2.44880320195634e-07,
+      "loss": 0.75,
+      "step": 1143
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7699040513552948,
+      "learning_rate": 2.3910775113416e-07,
+      "loss": 0.7496,
+      "step": 1144
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9065920454593933,
+      "learning_rate": 2.3340321122962916e-07,
+      "loss": 0.7367,
+      "step": 1145
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7961648860410553,
+      "learning_rate": 2.2776674024019286e-07,
+      "loss": 0.7206,
+      "step": 1146
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7016935733992847,
+      "learning_rate": 2.2219837744959284e-07,
+      "loss": 0.67,
+      "step": 1147
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9469793103604919,
+      "learning_rate": 2.1669816166688505e-07,
+      "loss": 0.758,
+      "step": 1148
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0206292357739108,
+      "learning_rate": 2.1126613122617546e-07,
+      "loss": 0.7796,
+      "step": 1149
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7388478114444481,
+      "learning_rate": 2.0590232398634114e-07,
+      "loss": 0.7012,
+      "step": 1150
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8222559422294684,
+      "learning_rate": 2.006067773307785e-07,
+      "loss": 0.7412,
+      "step": 1151
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.733500839416342,
+      "learning_rate": 1.9537952816713334e-07,
+      "loss": 0.7012,
+      "step": 1152
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8210352066340982,
+      "learning_rate": 1.902206129270523e-07,
+      "loss": 0.7705,
+      "step": 1153
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7123479259731154,
+      "learning_rate": 1.8513006756591957e-07,
+      "loss": 0.7304,
+      "step": 1154
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9055100556729727,
+      "learning_rate": 1.8010792756261497e-07,
+      "loss": 0.6939,
+      "step": 1155
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7301970875861544,
+      "learning_rate": 1.7515422791926195e-07,
+      "loss": 0.7142,
+      "step": 1156
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6781467989617473,
+      "learning_rate": 1.7026900316098217e-07,
+      "loss": 0.7129,
+      "step": 1157
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8578816193531952,
+      "learning_rate": 1.654522873356612e-07,
+      "loss": 0.9109,
+      "step": 1158
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7593956003492683,
+      "learning_rate": 1.6070411401370335e-07,
+      "loss": 0.6851,
+      "step": 1159
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0367887458918381,
+      "learning_rate": 1.560245162878049e-07,
+      "loss": 0.7317,
+      "step": 1160
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7032316518127807,
+      "learning_rate": 1.514135267727179e-07,
+      "loss": 0.7256,
+      "step": 1161
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.797512801384326,
+      "learning_rate": 1.4687117760502579e-07,
+      "loss": 0.7658,
+      "step": 1162
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8057064503623421,
+      "learning_rate": 1.42397500442919e-07,
+      "loss": 0.7652,
+      "step": 1163
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.646588198072519,
+      "learning_rate": 1.3799252646597428e-07,
+      "loss": 0.6841,
+      "step": 1164
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.853108664923995,
+      "learning_rate": 1.3365628637493688e-07,
+      "loss": 0.7425,
+      "step": 1165
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6575289244067559,
+      "learning_rate": 1.293888103915064e-07,
+      "loss": 0.7014,
+      "step": 1166
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.73937209230011,
+      "learning_rate": 1.2519012825812804e-07,
+      "loss": 0.734,
+      "step": 1167
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7366781028694698,
+      "learning_rate": 1.2106026923778046e-07,
+      "loss": 0.7143,
+      "step": 1168
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6711335117945008,
+      "learning_rate": 1.1699926211378054e-07,
+      "loss": 0.719,
+      "step": 1169
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0845267490415766,
+      "learning_rate": 1.130071351895734e-07,
+      "loss": 0.7425,
+      "step": 1170
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8068923255863443,
+      "learning_rate": 1.0908391628854042e-07,
+      "loss": 0.7069,
+      "step": 1171
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9886725445435967,
+      "learning_rate": 1.0522963275380494e-07,
+      "loss": 0.738,
+      "step": 1172
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7498440146370289,
+      "learning_rate": 1.0144431144804012e-07,
+      "loss": 0.7252,
+      "step": 1173
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.980294671084566,
+      "learning_rate": 9.772797875328255e-08,
+      "loss": 0.7069,
+      "step": 1174
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6681685473132211,
+      "learning_rate": 9.408066057074782e-08,
+      "loss": 0.8005,
+      "step": 1175
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7598366984979643,
+      "learning_rate": 9.0502382320653e-08,
+      "loss": 0.7208,
+      "step": 1176
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5744970801800884,
+      "learning_rate": 8.699316894203225e-08,
+      "loss": 0.7143,
+      "step": 1177
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9275017727370837,
+      "learning_rate": 8.355304489257254e-08,
+      "loss": 0.6994,
+      "step": 1178
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3216571691703607,
+      "learning_rate": 8.018203414843607e-08,
+      "loss": 0.7598,
+      "step": 1179
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6602690816736759,
+      "learning_rate": 7.688016020409473e-08,
+      "loss": 0.6863,
+      "step": 1180
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9465582392399315,
+      "learning_rate": 7.364744607216923e-08,
+      "loss": 0.7915,
+      "step": 1181
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8732116496542586,
+      "learning_rate": 7.048391428326585e-08,
+      "loss": 0.7092,
+      "step": 1182
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.000737701837792,
+      "learning_rate": 6.738958688581876e-08,
+      "loss": 0.7395,
+      "step": 1183
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.643466672368607,
+      "learning_rate": 6.43644854459391e-08,
+      "loss": 0.6816,
+      "step": 1184
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7952327474147997,
+      "learning_rate": 6.140863104726391e-08,
+      "loss": 0.804,
+      "step": 1185
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9607149170543946,
+      "learning_rate": 5.852204429080632e-08,
+      "loss": 0.7501,
+      "step": 1186
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8194884689476464,
+      "learning_rate": 5.5704745294815624e-08,
+      "loss": 0.6771,
+      "step": 1187
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9804124042216501,
+      "learning_rate": 5.295675369463737e-08,
+      "loss": 0.7476,
+      "step": 1188
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8941695128804578,
+      "learning_rate": 5.0278088642572445e-08,
+      "loss": 0.7002,
+      "step": 1189
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8044582651339719,
+      "learning_rate": 4.766876880774596e-08,
+      "loss": 0.7489,
+      "step": 1190
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8034323450731775,
+      "learning_rate": 4.512881237597744e-08,
+      "loss": 0.7218,
+      "step": 1191
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8477241770408097,
+      "learning_rate": 4.2658237049655325e-08,
+      "loss": 0.7015,
+      "step": 1192
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.65613044233294,
+      "learning_rate": 4.025706004760932e-08,
+      "loss": 0.712,
+      "step": 1193
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8284595522666801,
+      "learning_rate": 3.792529810499712e-08,
+      "loss": 0.9191,
+      "step": 1194
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7063334248674779,
+      "learning_rate": 3.56629674731801e-08,
+      "loss": 0.7508,
+      "step": 1195
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9495036030482247,
+      "learning_rate": 3.347008391961337e-08,
+      "loss": 0.7062,
+      "step": 1196
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7313543917895858,
+      "learning_rate": 3.134666272774034e-08,
+      "loss": 0.6781,
+      "step": 1197
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8791512225401776,
+      "learning_rate": 2.929271869687833e-08,
+      "loss": 0.7665,
+      "step": 1198
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7260293031012599,
+      "learning_rate": 2.7308266142119788e-08,
+      "loss": 0.7099,
+      "step": 1199
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7435078884942009,
+      "learning_rate": 2.5393318894233464e-08,
+      "loss": 0.7418,
+      "step": 1200
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.698336149609688,
+      "learning_rate": 2.354789029956561e-08,
+      "loss": 0.6931,
+      "step": 1201
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8261380521282469,
+      "learning_rate": 2.177199321994672e-08,
+      "loss": 0.7229,
+      "step": 1202
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6470651924870178,
+      "learning_rate": 2.0065640032604916e-08,
+      "loss": 0.7092,
+      "step": 1203
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0241406739812993,
+      "learning_rate": 1.842884263007716e-08,
+      "loss": 0.7348,
+      "step": 1204
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6657706082092261,
+      "learning_rate": 1.6861612420124852e-08,
+      "loss": 0.6641,
+      "step": 1205
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8118792710308739,
+      "learning_rate": 1.5363960325660565e-08,
+      "loss": 0.7915,
+      "step": 1206
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9046882809460657,
+      "learning_rate": 1.3935896784663671e-08,
+      "loss": 0.7722,
+      "step": 1207
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6539860253908187,
+      "learning_rate": 1.2577431750114833e-08,
+      "loss": 0.7126,
+      "step": 1208
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8546723828376718,
+      "learning_rate": 1.1288574689920506e-08,
+      "loss": 0.7571,
+      "step": 1209
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7816828303599752,
+      "learning_rate": 1.0069334586854106e-08,
+      "loss": 0.6894,
+      "step": 1210
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6950558128729017,
+      "learning_rate": 8.91971993848384e-09,
+      "loss": 0.7493,
+      "step": 1211
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8489320083711304,
+      "learning_rate": 7.83973875712385e-09,
+      "loss": 0.7446,
+      "step": 1212
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9769970323043154,
+      "learning_rate": 6.82939856977094e-09,
+      "loss": 0.7143,
+      "step": 1213
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6415405798229384,
+      "learning_rate": 5.888706418054613e-09,
+      "loss": 0.6698,
+      "step": 1214
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7454040190377804,
+      "learning_rate": 5.017668858189328e-09,
+      "loss": 0.9449,
+      "step": 1215
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7955489211350182,
+      "learning_rate": 4.216291960925656e-09,
+      "loss": 0.769,
+      "step": 1216
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9604809532238177,
+      "learning_rate": 3.4845813115114147e-09,
+      "loss": 0.7191,
+      "step": 1217
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9153875061323662,
+      "learning_rate": 2.8225420096494873e-09,
+      "loss": 0.7195,
+      "step": 1218
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5979009577886261,
+      "learning_rate": 2.2301786694656214e-09,
+      "loss": 0.6817,
+      "step": 1219
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8611622232350314,
+      "learning_rate": 1.7074954194729044e-09,
+      "loss": 0.7554,
+      "step": 1220
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0816072326431696,
+      "learning_rate": 1.2544959025462266e-09,
+      "loss": 0.7264,
+      "step": 1221
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2280979077744345,
+      "learning_rate": 8.711832758934169e-10,
+      "loss": 0.7011,
+      "step": 1222
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2838320846621525,
+      "learning_rate": 5.575602110363675e-10,
+      "loss": 0.6929,
+      "step": 1223
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8666666408129854,
+      "learning_rate": 3.136288937910514e-10,
+      "loss": 0.7584,
+      "step": 1224
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6861736263257758,
+      "learning_rate": 1.3939102425197803e-10,
+      "loss": 0.7071,
+      "step": 1225
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9190970558941943,
+      "learning_rate": 3.4847816781091725e-11,
+      "loss": 0.7167,
+      "step": 1226
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7104657256719872,
+      "learning_rate": 0.0,
+      "loss": 0.6672,
+      "step": 1227
+    },
+    {
+      "epoch": 1.0,
+      "step": 1227,
+      "total_flos": 0.0,
+      "train_loss": 0.013711861507039586,
+      "train_runtime": 5478.5145,
+      "train_samples_per_second": 460.564,
+      "train_steps_per_second": 0.224
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1227,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 10,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}