diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12985 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9996395746981438,
+  "eval_steps": 500,
+  "global_step": 1849,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005406379527842854,
+      "grad_norm": 5.835691452026367,
+      "learning_rate": 5.405405405405406e-08,
+      "loss": 0.8753,
+      "step": 1
+    },
+    {
+      "epoch": 0.001081275905568571,
+      "grad_norm": 5.94018030166626,
+      "learning_rate": 1.0810810810810812e-07,
+      "loss": 0.892,
+      "step": 2
+    },
+    {
+      "epoch": 0.0016219138583528565,
+      "grad_norm": 6.048556327819824,
+      "learning_rate": 1.6216216216216218e-07,
+      "loss": 0.8786,
+      "step": 3
+    },
+    {
+      "epoch": 0.002162551811137142,
+      "grad_norm": 5.985991477966309,
+      "learning_rate": 2.1621621621621625e-07,
+      "loss": 0.8707,
+      "step": 4
+    },
+    {
+      "epoch": 0.0027031897639214274,
+      "grad_norm": 5.788306713104248,
+      "learning_rate": 2.702702702702703e-07,
+      "loss": 0.8625,
+      "step": 5
+    },
+    {
+      "epoch": 0.003243827716705713,
+      "grad_norm": 6.294447422027588,
+      "learning_rate": 3.2432432432432436e-07,
+      "loss": 0.9267,
+      "step": 6
+    },
+    {
+      "epoch": 0.003784465669489998,
+      "grad_norm": 6.075927734375,
+      "learning_rate": 3.7837837837837843e-07,
+      "loss": 0.9002,
+      "step": 7
+    },
+    {
+      "epoch": 0.004325103622274284,
+      "grad_norm": 6.027771472930908,
+      "learning_rate": 4.324324324324325e-07,
+      "loss": 0.8858,
+      "step": 8
+    },
+    {
+      "epoch": 0.004865741575058569,
+      "grad_norm": 5.661100387573242,
+      "learning_rate": 4.864864864864865e-07,
+      "loss": 0.8382,
+      "step": 9
+    },
+    {
+      "epoch": 0.005406379527842855,
+      "grad_norm": 5.359391689300537,
+      "learning_rate": 5.405405405405406e-07,
+      "loss": 0.8345,
+      "step": 10
+    },
+    {
+      "epoch": 0.00594701748062714,
+      "grad_norm": 5.4135613441467285,
+      "learning_rate": 5.945945945945947e-07,
+      "loss": 0.8613,
+      "step": 11
+    },
+    {
+      "epoch": 0.006487655433411426,
+      "grad_norm": 5.188356876373291,
+      "learning_rate": 6.486486486486487e-07,
+      "loss": 0.8316,
+      "step": 12
+    },
+    {
+      "epoch": 0.0070282933861957105,
+      "grad_norm": 5.179812908172607,
+      "learning_rate": 7.027027027027028e-07,
+      "loss": 0.8574,
+      "step": 13
+    },
+    {
+      "epoch": 0.007568931338979996,
+      "grad_norm": 4.420322418212891,
+      "learning_rate": 7.567567567567569e-07,
+      "loss": 0.8381,
+      "step": 14
+    },
+    {
+      "epoch": 0.008109569291764282,
+      "grad_norm": 4.343968868255615,
+      "learning_rate": 8.108108108108109e-07,
+      "loss": 0.8244,
+      "step": 15
+    },
+    {
+      "epoch": 0.008650207244548567,
+      "grad_norm": 4.318638801574707,
+      "learning_rate": 8.64864864864865e-07,
+      "loss": 0.8635,
+      "step": 16
+    },
+    {
+      "epoch": 0.009190845197332853,
+      "grad_norm": 4.086209774017334,
+      "learning_rate": 9.189189189189191e-07,
+      "loss": 0.8179,
+      "step": 17
+    },
+    {
+      "epoch": 0.009731483150117138,
+      "grad_norm": 4.0021467208862305,
+      "learning_rate": 9.72972972972973e-07,
+      "loss": 0.7978,
+      "step": 18
+    },
+    {
+      "epoch": 0.010272121102901424,
+      "grad_norm": 2.3173251152038574,
+      "learning_rate": 1.027027027027027e-06,
+      "loss": 0.7756,
+      "step": 19
+    },
+    {
+      "epoch": 0.01081275905568571,
+      "grad_norm": 2.2132866382598877,
+      "learning_rate": 1.0810810810810812e-06,
+      "loss": 0.761,
+      "step": 20
+    },
+    {
+      "epoch": 0.011353397008469995,
+      "grad_norm": 2.2799596786499023,
+      "learning_rate": 1.1351351351351352e-06,
+      "loss": 0.7687,
+      "step": 21
+    },
+    {
+      "epoch": 0.01189403496125428,
+      "grad_norm": 2.0921261310577393,
+      "learning_rate": 1.1891891891891893e-06,
+      "loss": 0.7825,
+      "step": 22
+    },
+    {
+      "epoch": 0.012434672914038566,
+      "grad_norm": 1.862625002861023,
+      "learning_rate": 1.2432432432432434e-06,
+      "loss": 0.762,
+      "step": 23
+    },
+    {
+      "epoch": 0.012975310866822852,
+      "grad_norm": 1.887032389640808,
+      "learning_rate": 1.2972972972972974e-06,
+      "loss": 0.7788,
+      "step": 24
+    },
+    {
+      "epoch": 0.013515948819607137,
+      "grad_norm": 1.483746886253357,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 0.7347,
+      "step": 25
+    },
+    {
+      "epoch": 0.014056586772391421,
+      "grad_norm": 2.0825283527374268,
+      "learning_rate": 1.4054054054054056e-06,
+      "loss": 0.733,
+      "step": 26
+    },
+    {
+      "epoch": 0.014597224725175707,
+      "grad_norm": 2.5244522094726562,
+      "learning_rate": 1.4594594594594596e-06,
+      "loss": 0.738,
+      "step": 27
+    },
+    {
+      "epoch": 0.015137862677959992,
+      "grad_norm": 2.7919321060180664,
+      "learning_rate": 1.5135135135135137e-06,
+      "loss": 0.7233,
+      "step": 28
+    },
+    {
+      "epoch": 0.015678500630744278,
+      "grad_norm": 2.6368916034698486,
+      "learning_rate": 1.5675675675675678e-06,
+      "loss": 0.7231,
+      "step": 29
+    },
+    {
+      "epoch": 0.016219138583528563,
+      "grad_norm": 2.4845619201660156,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 0.7093,
+      "step": 30
+    },
+    {
+      "epoch": 0.01675977653631285,
+      "grad_norm": 2.3388469219207764,
+      "learning_rate": 1.675675675675676e-06,
+      "loss": 0.7384,
+      "step": 31
+    },
+    {
+      "epoch": 0.017300414489097134,
+      "grad_norm": 2.0697882175445557,
+      "learning_rate": 1.72972972972973e-06,
+      "loss": 0.724,
+      "step": 32
+    },
+    {
+      "epoch": 0.01784105244188142,
+      "grad_norm": 1.7929456233978271,
+      "learning_rate": 1.783783783783784e-06,
+      "loss": 0.7247,
+      "step": 33
+    },
+    {
+      "epoch": 0.018381690394665706,
+      "grad_norm": 1.1819241046905518,
+      "learning_rate": 1.8378378378378381e-06,
+      "loss": 0.6814,
+      "step": 34
+    },
+    {
+      "epoch": 0.01892232834744999,
+      "grad_norm": 0.9916073679924011,
+      "learning_rate": 1.8918918918918922e-06,
+      "loss": 0.7039,
+      "step": 35
+    },
+    {
+      "epoch": 0.019462966300234277,
+      "grad_norm": 1.0272109508514404,
+      "learning_rate": 1.945945945945946e-06,
+      "loss": 0.6968,
+      "step": 36
+    },
+    {
+      "epoch": 0.020003604253018562,
+      "grad_norm": 1.2608418464660645,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.7107,
+      "step": 37
+    },
+    {
+      "epoch": 0.020544242205802848,
+      "grad_norm": 1.2174832820892334,
+      "learning_rate": 2.054054054054054e-06,
+      "loss": 0.6771,
+      "step": 38
+    },
+    {
+      "epoch": 0.021084880158587133,
+      "grad_norm": 1.0341556072235107,
+      "learning_rate": 2.1081081081081085e-06,
+      "loss": 0.6796,
+      "step": 39
+    },
+    {
+      "epoch": 0.02162551811137142,
+      "grad_norm": 0.9752543568611145,
+      "learning_rate": 2.1621621621621623e-06,
+      "loss": 0.6607,
+      "step": 40
+    },
+    {
+      "epoch": 0.022166156064155704,
+      "grad_norm": 0.8276258707046509,
+      "learning_rate": 2.2162162162162166e-06,
+      "loss": 0.6314,
+      "step": 41
+    },
+    {
+      "epoch": 0.02270679401693999,
+      "grad_norm": 0.9401139616966248,
+      "learning_rate": 2.2702702702702705e-06,
+      "loss": 0.6805,
+      "step": 42
+    },
+    {
+      "epoch": 0.023247431969724276,
+      "grad_norm": 0.7445069551467896,
+      "learning_rate": 2.3243243243243247e-06,
+      "loss": 0.6743,
+      "step": 43
+    },
+    {
+      "epoch": 0.02378806992250856,
+      "grad_norm": 0.6744734644889832,
+      "learning_rate": 2.3783783783783786e-06,
+      "loss": 0.6249,
+      "step": 44
+    },
+    {
+      "epoch": 0.024328707875292847,
+      "grad_norm": 0.8370453715324402,
+      "learning_rate": 2.432432432432433e-06,
+      "loss": 0.6669,
+      "step": 45
+    },
+    {
+      "epoch": 0.024869345828077132,
+      "grad_norm": 0.7774664759635925,
+      "learning_rate": 2.4864864864864867e-06,
+      "loss": 0.6698,
+      "step": 46
+    },
+    {
+      "epoch": 0.025409983780861418,
+      "grad_norm": 0.7434698343276978,
+      "learning_rate": 2.540540540540541e-06,
+      "loss": 0.6347,
+      "step": 47
+    },
+    {
+      "epoch": 0.025950621733645703,
+      "grad_norm": 0.7597132325172424,
+      "learning_rate": 2.594594594594595e-06,
+      "loss": 0.6217,
+      "step": 48
+    },
+    {
+      "epoch": 0.02649125968642999,
+      "grad_norm": 0.7307024598121643,
+      "learning_rate": 2.648648648648649e-06,
+      "loss": 0.6406,
+      "step": 49
+    },
+    {
+      "epoch": 0.027031897639214274,
+      "grad_norm": 0.6316570043563843,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 0.5985,
+      "step": 50
+    },
+    {
+      "epoch": 0.02757253559199856,
+      "grad_norm": 0.58868408203125,
+      "learning_rate": 2.7567567567567573e-06,
+      "loss": 0.6157,
+      "step": 51
+    },
+    {
+      "epoch": 0.028113173544782842,
+      "grad_norm": 0.6674655675888062,
+      "learning_rate": 2.810810810810811e-06,
+      "loss": 0.6382,
+      "step": 52
+    },
+    {
+      "epoch": 0.028653811497567128,
+      "grad_norm": 0.6769363880157471,
+      "learning_rate": 2.8648648648648654e-06,
+      "loss": 0.6284,
+      "step": 53
+    },
+    {
+      "epoch": 0.029194449450351413,
+      "grad_norm": 0.5872988104820251,
+      "learning_rate": 2.9189189189189193e-06,
+      "loss": 0.6095,
+      "step": 54
+    },
+    {
+      "epoch": 0.0297350874031357,
+      "grad_norm": 0.5742107033729553,
+      "learning_rate": 2.9729729729729736e-06,
+      "loss": 0.6609,
+      "step": 55
+    },
+    {
+      "epoch": 0.030275725355919984,
+      "grad_norm": 0.5116291046142578,
+      "learning_rate": 3.0270270270270274e-06,
+      "loss": 0.5959,
+      "step": 56
+    },
+    {
+      "epoch": 0.03081636330870427,
+      "grad_norm": 0.4740088880062103,
+      "learning_rate": 3.0810810810810817e-06,
+      "loss": 0.5925,
+      "step": 57
+    },
+    {
+      "epoch": 0.031357001261488555,
+      "grad_norm": 0.536938488483429,
+      "learning_rate": 3.1351351351351356e-06,
+      "loss": 0.6281,
+      "step": 58
+    },
+    {
+      "epoch": 0.03189763921427284,
+      "grad_norm": 0.5394325852394104,
+      "learning_rate": 3.1891891891891894e-06,
+      "loss": 0.604,
+      "step": 59
+    },
+    {
+      "epoch": 0.03243827716705713,
+      "grad_norm": 0.5232502222061157,
+      "learning_rate": 3.2432432432432437e-06,
+      "loss": 0.5988,
+      "step": 60
+    },
+    {
+      "epoch": 0.03297891511984141,
+      "grad_norm": 0.5132797360420227,
+      "learning_rate": 3.2972972972972976e-06,
+      "loss": 0.6177,
+      "step": 61
+    },
+    {
+      "epoch": 0.0335195530726257,
+      "grad_norm": 0.46831756830215454,
+      "learning_rate": 3.351351351351352e-06,
+      "loss": 0.6274,
+      "step": 62
+    },
+    {
+      "epoch": 0.03406019102540998,
+      "grad_norm": 0.4727836549282074,
+      "learning_rate": 3.4054054054054057e-06,
+      "loss": 0.6218,
+      "step": 63
+    },
+    {
+      "epoch": 0.03460082897819427,
+      "grad_norm": 0.4939032196998596,
+      "learning_rate": 3.45945945945946e-06,
+      "loss": 0.5846,
+      "step": 64
+    },
+    {
+      "epoch": 0.035141466930978554,
+      "grad_norm": 0.46765533089637756,
+      "learning_rate": 3.513513513513514e-06,
+      "loss": 0.6012,
+      "step": 65
+    },
+    {
+      "epoch": 0.03568210488376284,
+      "grad_norm": 0.41743600368499756,
+      "learning_rate": 3.567567567567568e-06,
+      "loss": 0.6112,
+      "step": 66
+    },
+    {
+      "epoch": 0.036222742836547125,
+      "grad_norm": 0.46013137698173523,
+      "learning_rate": 3.621621621621622e-06,
+      "loss": 0.5885,
+      "step": 67
+    },
+    {
+      "epoch": 0.03676338078933141,
+      "grad_norm": 0.4405370354652405,
+      "learning_rate": 3.6756756756756763e-06,
+      "loss": 0.6377,
+      "step": 68
+    },
+    {
+      "epoch": 0.0373040187421157,
+      "grad_norm": 0.41007307171821594,
+      "learning_rate": 3.72972972972973e-06,
+      "loss": 0.6025,
+      "step": 69
+    },
+    {
+      "epoch": 0.03784465669489998,
+      "grad_norm": 0.3900786340236664,
+      "learning_rate": 3.7837837837837844e-06,
+      "loss": 0.5646,
+      "step": 70
+    },
+    {
+      "epoch": 0.03838529464768427,
+      "grad_norm": 0.45452040433883667,
+      "learning_rate": 3.837837837837838e-06,
+      "loss": 0.5868,
+      "step": 71
+    },
+    {
+      "epoch": 0.03892593260046855,
+      "grad_norm": 0.4392114579677582,
+      "learning_rate": 3.891891891891892e-06,
+      "loss": 0.5681,
+      "step": 72
+    },
+    {
+      "epoch": 0.03946657055325284,
+      "grad_norm": 0.4254467189311981,
+      "learning_rate": 3.945945945945947e-06,
+      "loss": 0.6258,
+      "step": 73
+    },
+    {
+      "epoch": 0.040007208506037124,
+      "grad_norm": 0.44197121262550354,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.6088,
+      "step": 74
+    },
+    {
+      "epoch": 0.04054784645882141,
+      "grad_norm": 0.4894886910915375,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 0.596,
+      "step": 75
+    },
+    {
+      "epoch": 0.041088484411605695,
+      "grad_norm": 0.4831041395664215,
+      "learning_rate": 4.108108108108108e-06,
+      "loss": 0.5854,
+      "step": 76
+    },
+    {
+      "epoch": 0.04162912236438998,
+      "grad_norm": 0.4371893107891083,
+      "learning_rate": 4.162162162162163e-06,
+      "loss": 0.5775,
+      "step": 77
+    },
+    {
+      "epoch": 0.04216976031717427,
+      "grad_norm": 0.4326671361923218,
+      "learning_rate": 4.216216216216217e-06,
+      "loss": 0.5871,
+      "step": 78
+    },
+    {
+      "epoch": 0.04271039826995855,
+      "grad_norm": 0.4272334575653076,
+      "learning_rate": 4.270270270270271e-06,
+      "loss": 0.5637,
+      "step": 79
+    },
+    {
+      "epoch": 0.04325103622274284,
+      "grad_norm": 0.4943785071372986,
+      "learning_rate": 4.324324324324325e-06,
+      "loss": 0.5609,
+      "step": 80
+    },
+    {
+      "epoch": 0.04379167417552712,
+      "grad_norm": 0.3759538531303406,
+      "learning_rate": 4.378378378378379e-06,
+      "loss": 0.5881,
+      "step": 81
+    },
+    {
+      "epoch": 0.04433231212831141,
+      "grad_norm": 0.4269590377807617,
+      "learning_rate": 4.432432432432433e-06,
+      "loss": 0.5822,
+      "step": 82
+    },
+    {
+      "epoch": 0.044872950081095694,
+      "grad_norm": 0.4101639986038208,
+      "learning_rate": 4.486486486486487e-06,
+      "loss": 0.5925,
+      "step": 83
+    },
+    {
+      "epoch": 0.04541358803387998,
+      "grad_norm": 0.40165165066719055,
+      "learning_rate": 4.540540540540541e-06,
+      "loss": 0.5304,
+      "step": 84
+    },
+    {
+      "epoch": 0.045954225986664266,
+      "grad_norm": 0.3949192762374878,
+      "learning_rate": 4.594594594594596e-06,
+      "loss": 0.5503,
+      "step": 85
+    },
+    {
+      "epoch": 0.04649486393944855,
+      "grad_norm": 0.4039553999900818,
+      "learning_rate": 4.6486486486486495e-06,
+      "loss": 0.5634,
+      "step": 86
+    },
+    {
+      "epoch": 0.04703550189223284,
+      "grad_norm": 0.4268937110900879,
+      "learning_rate": 4.702702702702703e-06,
+      "loss": 0.5854,
+      "step": 87
+    },
+    {
+      "epoch": 0.04757613984501712,
+      "grad_norm": 0.420392245054245,
+      "learning_rate": 4.756756756756757e-06,
+      "loss": 0.5554,
+      "step": 88
+    },
+    {
+      "epoch": 0.04811677779780141,
+      "grad_norm": 0.38791415095329285,
+      "learning_rate": 4.810810810810811e-06,
+      "loss": 0.5662,
+      "step": 89
+    },
+    {
+      "epoch": 0.04865741575058569,
+      "grad_norm": 0.38142862915992737,
+      "learning_rate": 4.864864864864866e-06,
+      "loss": 0.5621,
+      "step": 90
+    },
+    {
+      "epoch": 0.04919805370336998,
+      "grad_norm": 0.4095216691493988,
+      "learning_rate": 4.91891891891892e-06,
+      "loss": 0.5567,
+      "step": 91
+    },
+    {
+      "epoch": 0.049738691656154264,
+      "grad_norm": 0.40666601061820984,
+      "learning_rate": 4.9729729729729735e-06,
+      "loss": 0.6048,
+      "step": 92
+    },
+    {
+      "epoch": 0.05027932960893855,
+      "grad_norm": 0.37900641560554504,
+      "learning_rate": 5.027027027027027e-06,
+      "loss": 0.57,
+      "step": 93
+    },
+    {
+      "epoch": 0.050819967561722836,
+      "grad_norm": 0.4188796579837799,
+      "learning_rate": 5.081081081081082e-06,
+      "loss": 0.5547,
+      "step": 94
+    },
+    {
+      "epoch": 0.05136060551450712,
+      "grad_norm": 0.3917129933834076,
+      "learning_rate": 5.135135135135135e-06,
+      "loss": 0.5994,
+      "step": 95
+    },
+    {
+      "epoch": 0.05190124346729141,
+      "grad_norm": 0.3745051920413971,
+      "learning_rate": 5.18918918918919e-06,
+      "loss": 0.5645,
+      "step": 96
+    },
+    {
+      "epoch": 0.05244188142007569,
+      "grad_norm": 0.4055512249469757,
+      "learning_rate": 5.243243243243244e-06,
+      "loss": 0.5516,
+      "step": 97
+    },
+    {
+      "epoch": 0.05298251937285998,
+      "grad_norm": 0.40357547998428345,
+      "learning_rate": 5.297297297297298e-06,
+      "loss": 0.5414,
+      "step": 98
+    },
+    {
+      "epoch": 0.05352315732564426,
+      "grad_norm": 0.37698817253112793,
+      "learning_rate": 5.351351351351351e-06,
+      "loss": 0.5494,
+      "step": 99
+    },
+    {
+      "epoch": 0.05406379527842855,
+      "grad_norm": 0.45397427678108215,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 0.5574,
+      "step": 100
+    },
+    {
+      "epoch": 0.054604433231212834,
+      "grad_norm": 0.4567568600177765,
+      "learning_rate": 5.45945945945946e-06,
+      "loss": 0.578,
+      "step": 101
+    },
+    {
+      "epoch": 0.05514507118399712,
+      "grad_norm": 0.4371408224105835,
+      "learning_rate": 5.513513513513515e-06,
+      "loss": 0.5614,
+      "step": 102
+    },
+    {
+      "epoch": 0.0556857091367814,
+      "grad_norm": 0.4275542199611664,
+      "learning_rate": 5.567567567567568e-06,
+      "loss": 0.5884,
+      "step": 103
+    },
+    {
+      "epoch": 0.056226347089565684,
+      "grad_norm": 0.4597940742969513,
+      "learning_rate": 5.621621621621622e-06,
+      "loss": 0.5431,
+      "step": 104
+    },
+    {
+      "epoch": 0.05676698504234997,
+      "grad_norm": 0.47299715876579285,
+      "learning_rate": 5.675675675675676e-06,
+      "loss": 0.5648,
+      "step": 105
+    },
+    {
+      "epoch": 0.057307622995134255,
+      "grad_norm": 0.435273140668869,
+      "learning_rate": 5.729729729729731e-06,
+      "loss": 0.5281,
+      "step": 106
+    },
+    {
+      "epoch": 0.05784826094791854,
+      "grad_norm": 0.5185096859931946,
+      "learning_rate": 5.783783783783784e-06,
+      "loss": 0.5975,
+      "step": 107
+    },
+    {
+      "epoch": 0.058388898900702826,
+      "grad_norm": 0.4125944674015045,
+      "learning_rate": 5.837837837837839e-06,
+      "loss": 0.5627,
+      "step": 108
+    },
+    {
+      "epoch": 0.05892953685348711,
+      "grad_norm": 0.4223628640174866,
+      "learning_rate": 5.8918918918918924e-06,
+      "loss": 0.5734,
+      "step": 109
+    },
+    {
+      "epoch": 0.0594701748062714,
+      "grad_norm": 0.4168817698955536,
+      "learning_rate": 5.945945945945947e-06,
+      "loss": 0.5575,
+      "step": 110
+    },
+    {
+      "epoch": 0.06001081275905568,
+      "grad_norm": 0.4130726754665375,
+      "learning_rate": 6e-06,
+      "loss": 0.5319,
+      "step": 111
+    },
+    {
+      "epoch": 0.06055145071183997,
+      "grad_norm": 0.4145527482032776,
+      "learning_rate": 6.054054054054055e-06,
+      "loss": 0.5544,
+      "step": 112
+    },
+    {
+      "epoch": 0.061092088664624254,
+      "grad_norm": 0.47130388021469116,
+      "learning_rate": 6.108108108108109e-06,
+      "loss": 0.5454,
+      "step": 113
+    },
+    {
+      "epoch": 0.06163272661740854,
+      "grad_norm": 0.4709829092025757,
+      "learning_rate": 6.162162162162163e-06,
+      "loss": 0.5477,
+      "step": 114
+    },
+    {
+      "epoch": 0.062173364570192825,
+      "grad_norm": 0.4437786340713501,
+      "learning_rate": 6.2162162162162164e-06,
+      "loss": 0.5337,
+      "step": 115
+    },
+    {
+      "epoch": 0.06271400252297711,
+      "grad_norm": 0.44709983468055725,
+      "learning_rate": 6.270270270270271e-06,
+      "loss": 0.5584,
+      "step": 116
+    },
+    {
+      "epoch": 0.0632546404757614,
+      "grad_norm": 0.438896119594574,
+      "learning_rate": 6.324324324324325e-06,
+      "loss": 0.5374,
+      "step": 117
+    },
+    {
+      "epoch": 0.06379527842854568,
+      "grad_norm": 0.4727858901023865,
+      "learning_rate": 6.378378378378379e-06,
+      "loss": 0.5907,
+      "step": 118
+    },
+    {
+      "epoch": 0.06433591638132997,
+      "grad_norm": 0.5302643179893494,
+      "learning_rate": 6.432432432432433e-06,
+      "loss": 0.5583,
+      "step": 119
+    },
+    {
+      "epoch": 0.06487655433411425,
+      "grad_norm": 0.49883052706718445,
+      "learning_rate": 6.486486486486487e-06,
+      "loss": 0.5728,
+      "step": 120
+    },
+    {
+      "epoch": 0.06541719228689855,
+      "grad_norm": 0.45870596170425415,
+      "learning_rate": 6.540540540540541e-06,
+      "loss": 0.5372,
+      "step": 121
+    },
+    {
+      "epoch": 0.06595783023968282,
+      "grad_norm": 0.5096645951271057,
+      "learning_rate": 6.594594594594595e-06,
+      "loss": 0.5361,
+      "step": 122
+    },
+    {
+      "epoch": 0.06649846819246712,
+      "grad_norm": 0.5785489678382874,
+      "learning_rate": 6.648648648648649e-06,
+      "loss": 0.5449,
+      "step": 123
+    },
+    {
+      "epoch": 0.0670391061452514,
+      "grad_norm": 0.48732200264930725,
+      "learning_rate": 6.702702702702704e-06,
+      "loss": 0.5647,
+      "step": 124
+    },
+    {
+      "epoch": 0.06757974409803569,
+      "grad_norm": 0.5112847089767456,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 0.5576,
+      "step": 125
+    },
+    {
+      "epoch": 0.06812038205081997,
+      "grad_norm": 0.5451067686080933,
+      "learning_rate": 6.810810810810811e-06,
+      "loss": 0.5321,
+      "step": 126
+    },
+    {
+      "epoch": 0.06866102000360426,
+      "grad_norm": 0.4540032744407654,
+      "learning_rate": 6.864864864864865e-06,
+      "loss": 0.5252,
+      "step": 127
+    },
+    {
+      "epoch": 0.06920165795638854,
+      "grad_norm": 0.48304489254951477,
+      "learning_rate": 6.91891891891892e-06,
+      "loss": 0.5905,
+      "step": 128
+    },
+    {
+      "epoch": 0.06974229590917283,
+      "grad_norm": 0.44558167457580566,
+      "learning_rate": 6.972972972972973e-06,
+      "loss": 0.569,
+      "step": 129
+    },
+    {
+      "epoch": 0.07028293386195711,
+      "grad_norm": 0.4131074547767639,
+      "learning_rate": 7.027027027027028e-06,
+      "loss": 0.5453,
+      "step": 130
+    },
+    {
+      "epoch": 0.0708235718147414,
+      "grad_norm": 0.3999936878681183,
+      "learning_rate": 7.0810810810810815e-06,
+      "loss": 0.5302,
+      "step": 131
+    },
+    {
+      "epoch": 0.07136420976752568,
+      "grad_norm": 0.43754464387893677,
+      "learning_rate": 7.135135135135136e-06,
+      "loss": 0.5523,
+      "step": 132
+    },
+    {
+      "epoch": 0.07190484772030997,
+      "grad_norm": 0.4095590114593506,
+      "learning_rate": 7.189189189189189e-06,
+      "loss": 0.5603,
+      "step": 133
+    },
+    {
+      "epoch": 0.07244548567309425,
+      "grad_norm": 0.4144512712955475,
+      "learning_rate": 7.243243243243244e-06,
+      "loss": 0.5747,
+      "step": 134
+    },
+    {
+      "epoch": 0.07298612362587854,
+      "grad_norm": 0.41802504658699036,
+      "learning_rate": 7.297297297297298e-06,
+      "loss": 0.5991,
+      "step": 135
+    },
+    {
+      "epoch": 0.07352676157866282,
+      "grad_norm": 0.42772114276885986,
+      "learning_rate": 7.3513513513513525e-06,
+      "loss": 0.5364,
+      "step": 136
+    },
+    {
+      "epoch": 0.07406739953144711,
+      "grad_norm": 0.4586148262023926,
+      "learning_rate": 7.4054054054054055e-06,
+      "loss": 0.5732,
+      "step": 137
+    },
+    {
+      "epoch": 0.0746080374842314,
+      "grad_norm": 0.4603786766529083,
+      "learning_rate": 7.45945945945946e-06,
+      "loss": 0.5481,
+      "step": 138
+    },
+    {
+      "epoch": 0.07514867543701567,
+      "grad_norm": 0.4603879153728485,
+      "learning_rate": 7.513513513513514e-06,
+      "loss": 0.5536,
+      "step": 139
+    },
+    {
+      "epoch": 0.07568931338979996,
+      "grad_norm": 0.46576988697052,
+      "learning_rate": 7.567567567567569e-06,
+      "loss": 0.5667,
+      "step": 140
+    },
+    {
+      "epoch": 0.07622995134258424,
+      "grad_norm": 0.4343358874320984,
+      "learning_rate": 7.621621621621622e-06,
+      "loss": 0.5316,
+      "step": 141
+    },
+    {
+      "epoch": 0.07677058929536854,
+      "grad_norm": 0.5072187185287476,
+      "learning_rate": 7.675675675675676e-06,
+      "loss": 0.5577,
+      "step": 142
+    },
+    {
+      "epoch": 0.07731122724815281,
+      "grad_norm": 0.3946337401866913,
+      "learning_rate": 7.72972972972973e-06,
+      "loss": 0.5349,
+      "step": 143
+    },
+    {
+      "epoch": 0.0778518652009371,
+      "grad_norm": 0.4240807294845581,
+      "learning_rate": 7.783783783783784e-06,
+      "loss": 0.5337,
+      "step": 144
+    },
+    {
+      "epoch": 0.07839250315372139,
+      "grad_norm": 0.4701116681098938,
+      "learning_rate": 7.837837837837838e-06,
+      "loss": 0.5159,
+      "step": 145
+    },
+    {
+      "epoch": 0.07893314110650568,
+      "grad_norm": 0.5360553860664368,
+      "learning_rate": 7.891891891891894e-06,
+      "loss": 0.5657,
+      "step": 146
+    },
+    {
+      "epoch": 0.07947377905928996,
+      "grad_norm": 0.46461114287376404,
+      "learning_rate": 7.945945945945946e-06,
+      "loss": 0.5631,
+      "step": 147
+    },
+    {
+      "epoch": 0.08001441701207425,
+      "grad_norm": 0.5180739760398865,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.5461,
+      "step": 148
+    },
+    {
+      "epoch": 0.08055505496485853,
+      "grad_norm": 0.4858033359050751,
+      "learning_rate": 8.054054054054055e-06,
+      "loss": 0.5493,
+      "step": 149
+    },
+    {
+      "epoch": 0.08109569291764282,
+      "grad_norm": 0.4371548593044281,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 0.5284,
+      "step": 150
+    },
+    {
+      "epoch": 0.0816363308704271,
+      "grad_norm": 0.46319711208343506,
+      "learning_rate": 8.162162162162163e-06,
+      "loss": 0.5401,
+      "step": 151
+    },
+    {
+      "epoch": 0.08217696882321139,
+      "grad_norm": 0.563747763633728,
+      "learning_rate": 8.216216216216217e-06,
+      "loss": 0.5532,
+      "step": 152
+    },
+    {
+      "epoch": 0.08271760677599567,
+      "grad_norm": 0.45033544301986694,
+      "learning_rate": 8.27027027027027e-06,
+      "loss": 0.5396,
+      "step": 153
+    },
+    {
+      "epoch": 0.08325824472877996,
+      "grad_norm": 0.5136594772338867,
+      "learning_rate": 8.324324324324326e-06,
+      "loss": 0.5382,
+      "step": 154
+    },
+    {
+      "epoch": 0.08379888268156424,
+      "grad_norm": 0.5209782123565674,
+      "learning_rate": 8.378378378378378e-06,
+      "loss": 0.5493,
+      "step": 155
+    },
+    {
+      "epoch": 0.08433952063434853,
+      "grad_norm": 0.5323936343193054,
+      "learning_rate": 8.432432432432434e-06,
+      "loss": 0.539,
+      "step": 156
+    },
+    {
+      "epoch": 0.08488015858713281,
+      "grad_norm": 0.4908245801925659,
+      "learning_rate": 8.486486486486488e-06,
+      "loss": 0.5233,
+      "step": 157
+    },
+    {
+      "epoch": 0.0854207965399171,
+      "grad_norm": 0.5659683346748352,
+      "learning_rate": 8.540540540540542e-06,
+      "loss": 0.5293,
+      "step": 158
+    },
+    {
+      "epoch": 0.08596143449270138,
+      "grad_norm": 0.5987280011177063,
+      "learning_rate": 8.594594594594595e-06,
+      "loss": 0.5394,
+      "step": 159
+    },
+    {
+      "epoch": 0.08650207244548568,
+      "grad_norm": 0.48582640290260315,
+      "learning_rate": 8.64864864864865e-06,
+      "loss": 0.5356,
+      "step": 160
+    },
+    {
+      "epoch": 0.08704271039826995,
+      "grad_norm": 0.5271715521812439,
+      "learning_rate": 8.702702702702703e-06,
+      "loss": 0.493,
+      "step": 161
+    },
+    {
+      "epoch": 0.08758334835105425,
+      "grad_norm": 0.668293833732605,
+      "learning_rate": 8.756756756756759e-06,
+      "loss": 0.4996,
+      "step": 162
+    },
+    {
+      "epoch": 0.08812398630383853,
+      "grad_norm": 0.4799709916114807,
+      "learning_rate": 8.810810810810811e-06,
+      "loss": 0.5161,
+      "step": 163
+    },
+    {
+      "epoch": 0.08866462425662282,
+      "grad_norm": 0.43464991450309753,
+      "learning_rate": 8.864864864864866e-06,
+      "loss": 0.5431,
+      "step": 164
+    },
+    {
+      "epoch": 0.0892052622094071,
+      "grad_norm": 0.5938690900802612,
+      "learning_rate": 8.91891891891892e-06,
+      "loss": 0.5677,
+      "step": 165
+    },
+    {
+      "epoch": 0.08974590016219139,
+      "grad_norm": 0.575090765953064,
+      "learning_rate": 8.972972972972974e-06,
+      "loss": 0.5506,
+      "step": 166
+    },
+    {
+      "epoch": 0.09028653811497567,
+      "grad_norm": 0.44394662976264954,
+      "learning_rate": 9.027027027027028e-06,
+      "loss": 0.5415,
+      "step": 167
+    },
+    {
+      "epoch": 0.09082717606775996,
+      "grad_norm": 0.45902568101882935,
+      "learning_rate": 9.081081081081082e-06,
+      "loss": 0.501,
+      "step": 168
+    },
+    {
+      "epoch": 0.09136781402054424,
+      "grad_norm": 0.6057916283607483,
+      "learning_rate": 9.135135135135136e-06,
+      "loss": 0.5383,
+      "step": 169
+    },
+    {
+      "epoch": 0.09190845197332853,
+      "grad_norm": 0.4782038927078247,
+      "learning_rate": 9.189189189189191e-06,
+      "loss": 0.5522,
+      "step": 170
+    },
+    {
+      "epoch": 0.09244908992611281,
+      "grad_norm": 0.5645440816879272,
+      "learning_rate": 9.243243243243243e-06,
+      "loss": 0.5395,
+      "step": 171
+    },
+    {
+      "epoch": 0.0929897278788971,
+      "grad_norm": 0.5265817046165466,
+      "learning_rate": 9.297297297297299e-06,
+      "loss": 0.5352,
+      "step": 172
+    },
+    {
+      "epoch": 0.09353036583168138,
+      "grad_norm": 0.504550039768219,
+      "learning_rate": 9.351351351351353e-06,
+      "loss": 0.5538,
+      "step": 173
+    },
+    {
+      "epoch": 0.09407100378446567,
+      "grad_norm": 0.5427402257919312,
+      "learning_rate": 9.405405405405407e-06,
+      "loss": 0.5307,
+      "step": 174
+    },
+    {
+      "epoch": 0.09461164173724995,
+      "grad_norm": 0.5136035084724426,
+      "learning_rate": 9.45945945945946e-06,
+      "loss": 0.5412,
+      "step": 175
+    },
+    {
+      "epoch": 0.09515227969003424,
+      "grad_norm": 0.5547382831573486,
+      "learning_rate": 9.513513513513514e-06,
+      "loss": 0.5452,
+      "step": 176
+    },
+    {
+      "epoch": 0.09569291764281852,
+      "grad_norm": 0.4894575774669647,
+      "learning_rate": 9.567567567567568e-06,
+      "loss": 0.5645,
+      "step": 177
+    },
+    {
+      "epoch": 0.09623355559560282,
+      "grad_norm": 0.521987795829773,
+      "learning_rate": 9.621621621621622e-06,
+      "loss": 0.5371,
+      "step": 178
+    },
+    {
+      "epoch": 0.0967741935483871,
+      "grad_norm": 0.48432838916778564,
+      "learning_rate": 9.675675675675676e-06,
+      "loss": 0.4921,
+      "step": 179
+    },
+    {
+      "epoch": 0.09731483150117139,
+      "grad_norm": 0.5000583529472351,
+      "learning_rate": 9.729729729729732e-06,
+      "loss": 0.5463,
+      "step": 180
+    },
+    {
+      "epoch": 0.09785546945395567,
+      "grad_norm": 0.5252205729484558,
+      "learning_rate": 9.783783783783785e-06,
+      "loss": 0.5571,
+      "step": 181
+    },
+    {
+      "epoch": 0.09839610740673996,
+      "grad_norm": 0.49048709869384766,
+      "learning_rate": 9.83783783783784e-06,
+      "loss": 0.5456,
+      "step": 182
+    },
+    {
+      "epoch": 0.09893674535952424,
+      "grad_norm": 0.5005491375923157,
+      "learning_rate": 9.891891891891893e-06,
+      "loss": 0.5263,
+      "step": 183
+    },
+    {
+      "epoch": 0.09947738331230853,
+      "grad_norm": 0.47705668210983276,
+      "learning_rate": 9.945945945945947e-06,
+      "loss": 0.5358,
+      "step": 184
+    },
+    {
+      "epoch": 0.10001802126509281,
+      "grad_norm": 0.4576117694377899,
+      "learning_rate": 1e-05,
+      "loss": 0.5025,
+      "step": 185
+    },
+    {
+      "epoch": 0.1005586592178771,
+      "grad_norm": 0.5019325613975525,
+      "learning_rate": 9.999991088865861e-06,
+      "loss": 0.5123,
+      "step": 186
+    },
+    {
+      "epoch": 0.10109929717066138,
+      "grad_norm": 0.47290462255477905,
+      "learning_rate": 9.999964355495207e-06,
+      "loss": 0.5141,
+      "step": 187
+    },
+    {
+      "epoch": 0.10163993512344567,
+      "grad_norm": 0.6163212656974792,
+      "learning_rate": 9.999919799983327e-06,
+      "loss": 0.5468,
+      "step": 188
+    },
+    {
+      "epoch": 0.10218057307622995,
+      "grad_norm": 0.5143344402313232,
+      "learning_rate": 9.99985742248904e-06,
+      "loss": 0.5433,
+      "step": 189
+    },
+    {
+      "epoch": 0.10272121102901424,
+      "grad_norm": 0.5387307405471802,
+      "learning_rate": 9.999777223234682e-06,
+      "loss": 0.5117,
+      "step": 190
+    },
+    {
+      "epoch": 0.10326184898179852,
+      "grad_norm": 0.6247243881225586,
+      "learning_rate": 9.999679202506126e-06,
+      "loss": 0.5298,
+      "step": 191
+    },
+    {
+      "epoch": 0.10380248693458281,
+      "grad_norm": 0.4776528477668762,
+      "learning_rate": 9.999563360652757e-06,
+      "loss": 0.5227,
+      "step": 192
+    },
+    {
+      "epoch": 0.10434312488736709,
+      "grad_norm": 0.5106874108314514,
+      "learning_rate": 9.999429698087491e-06,
+      "loss": 0.5197,
+      "step": 193
+    },
+    {
+      "epoch": 0.10488376284015138,
+      "grad_norm": 0.5041946768760681,
+      "learning_rate": 9.99927821528676e-06,
+      "loss": 0.5162,
+      "step": 194
+    },
+    {
+      "epoch": 0.10542440079293566,
+      "grad_norm": 0.5362434387207031,
+      "learning_rate": 9.999108912790521e-06,
+      "loss": 0.5103,
+      "step": 195
+    },
+    {
+      "epoch": 0.10596503874571996,
+      "grad_norm": 0.4504891335964203,
+      "learning_rate": 9.99892179120224e-06,
+      "loss": 0.5253,
+      "step": 196
+    },
+    {
+      "epoch": 0.10650567669850423,
+      "grad_norm": 0.5226731896400452,
+      "learning_rate": 9.99871685118891e-06,
+      "loss": 0.5392,
+      "step": 197
+    },
+    {
+      "epoch": 0.10704631465128853,
+      "grad_norm": 0.46790215373039246,
+      "learning_rate": 9.998494093481022e-06,
+      "loss": 0.4986,
+      "step": 198
+    },
+    {
+      "epoch": 0.1075869526040728,
+      "grad_norm": 0.4980153441429138,
+      "learning_rate": 9.998253518872592e-06,
+      "loss": 0.5568,
+      "step": 199
+    },
+    {
+      "epoch": 0.1081275905568571,
+      "grad_norm": 0.5192104578018188,
+      "learning_rate": 9.997995128221131e-06,
+      "loss": 0.564,
+      "step": 200
+    },
+    {
+      "epoch": 0.10866822850964138,
+      "grad_norm": 0.48999977111816406,
+      "learning_rate": 9.997718922447669e-06,
+      "loss": 0.5315,
+      "step": 201
+    },
+    {
+      "epoch": 0.10920886646242567,
+      "grad_norm": 0.5758963823318481,
+      "learning_rate": 9.99742490253672e-06,
+      "loss": 0.5428,
+      "step": 202
+    },
+    {
+      "epoch": 0.10974950441520995,
+      "grad_norm": 0.4924217164516449,
+      "learning_rate": 9.99711306953631e-06,
+      "loss": 0.5164,
+      "step": 203
+    },
+    {
+      "epoch": 0.11029014236799424,
+      "grad_norm": 0.5533744692802429,
+      "learning_rate": 9.99678342455795e-06,
+      "loss": 0.5286,
+      "step": 204
+    },
+    {
+      "epoch": 0.11083078032077852,
+      "grad_norm": 0.6308779120445251,
+      "learning_rate": 9.996435968776646e-06,
+      "loss": 0.5488,
+      "step": 205
+    },
+    {
+      "epoch": 0.1113714182735628,
+      "grad_norm": 0.49895963072776794,
+      "learning_rate": 9.996070703430888e-06,
+      "loss": 0.5228,
+      "step": 206
+    },
+    {
+      "epoch": 0.11191205622634709,
+      "grad_norm": 0.5760313868522644,
+      "learning_rate": 9.995687629822647e-06,
+      "loss": 0.5259,
+      "step": 207
+    },
+    {
+      "epoch": 0.11245269417913137,
+      "grad_norm": 0.5599102973937988,
+      "learning_rate": 9.99528674931737e-06,
+      "loss": 0.5191,
+      "step": 208
+    },
+    {
+      "epoch": 0.11299333213191566,
+      "grad_norm": 0.7119779586791992,
+      "learning_rate": 9.99486806334398e-06,
+      "loss": 0.5246,
+      "step": 209
+    },
+    {
+      "epoch": 0.11353397008469994,
+      "grad_norm": 0.5256053805351257,
+      "learning_rate": 9.994431573394861e-06,
+      "loss": 0.52,
+      "step": 210
+    },
+    {
+      "epoch": 0.11407460803748423,
+      "grad_norm": 0.6746430397033691,
+      "learning_rate": 9.993977281025862e-06,
+      "loss": 0.5519,
+      "step": 211
+    },
+    {
+      "epoch": 0.11461524599026851,
+      "grad_norm": 0.5388162732124329,
+      "learning_rate": 9.993505187856289e-06,
+      "loss": 0.5353,
+      "step": 212
+    },
+    {
+      "epoch": 0.1151558839430528,
+      "grad_norm": 0.6086297631263733,
+      "learning_rate": 9.993015295568893e-06,
+      "loss": 0.509,
+      "step": 213
+    },
+    {
+      "epoch": 0.11569652189583708,
+      "grad_norm": 0.5672162771224976,
+      "learning_rate": 9.992507605909873e-06,
+      "loss": 0.5215,
+      "step": 214
+    },
+    {
+      "epoch": 0.11623715984862137,
+      "grad_norm": 0.6110180616378784,
+      "learning_rate": 9.991982120688865e-06,
+      "loss": 0.5482,
+      "step": 215
+    },
+    {
+      "epoch": 0.11677779780140565,
+      "grad_norm": 0.5936868786811829,
+      "learning_rate": 9.99143884177894e-06,
+      "loss": 0.5016,
+      "step": 216
+    },
+    {
+      "epoch": 0.11731843575418995,
+      "grad_norm": 0.5300704836845398,
+      "learning_rate": 9.990877771116588e-06,
+      "loss": 0.5477,
+      "step": 217
+    },
+    {
+      "epoch": 0.11785907370697422,
+      "grad_norm": 0.5491863489151001,
+      "learning_rate": 9.99029891070172e-06,
+      "loss": 0.5311,
+      "step": 218
+    },
+    {
+      "epoch": 0.11839971165975852,
+      "grad_norm": 0.590416669845581,
+      "learning_rate": 9.989702262597656e-06,
+      "loss": 0.5245,
+      "step": 219
+    },
+    {
+      "epoch": 0.1189403496125428,
+      "grad_norm": 0.4550315737724304,
+      "learning_rate": 9.989087828931121e-06,
+      "loss": 0.513,
+      "step": 220
+    },
+    {
+      "epoch": 0.11948098756532709,
+      "grad_norm": 0.6941486597061157,
+      "learning_rate": 9.988455611892237e-06,
+      "loss": 0.5153,
+      "step": 221
+    },
+    {
+      "epoch": 0.12002162551811137,
+      "grad_norm": 0.5239282846450806,
+      "learning_rate": 9.987805613734508e-06,
+      "loss": 0.5329,
+      "step": 222
+    },
+    {
+      "epoch": 0.12056226347089566,
+      "grad_norm": 0.5700839161872864,
+      "learning_rate": 9.987137836774827e-06,
+      "loss": 0.5224,
+      "step": 223
+    },
+    {
+      "epoch": 0.12110290142367994,
+      "grad_norm": 0.6924121379852295,
+      "learning_rate": 9.986452283393452e-06,
+      "loss": 0.5256,
+      "step": 224
+    },
+    {
+      "epoch": 0.12164353937646423,
+      "grad_norm": 0.5878159999847412,
+      "learning_rate": 9.985748956034007e-06,
+      "loss": 0.5569,
+      "step": 225
+    },
+    {
+      "epoch": 0.12218417732924851,
+      "grad_norm": 0.5620668530464172,
+      "learning_rate": 9.985027857203469e-06,
+      "loss": 0.5307,
+      "step": 226
+    },
+    {
+      "epoch": 0.1227248152820328,
+      "grad_norm": 0.6018951535224915,
+      "learning_rate": 9.984288989472162e-06,
+      "loss": 0.517,
+      "step": 227
+    },
+    {
+      "epoch": 0.12326545323481708,
+      "grad_norm": 0.43050438165664673,
+      "learning_rate": 9.983532355473744e-06,
+      "loss": 0.5052,
+      "step": 228
+    },
+    {
+      "epoch": 0.12380609118760137,
+      "grad_norm": 0.5617673993110657,
+      "learning_rate": 9.982757957905204e-06,
+      "loss": 0.5355,
+      "step": 229
+    },
+    {
+      "epoch": 0.12434672914038565,
+      "grad_norm": 0.5786687135696411,
+      "learning_rate": 9.981965799526846e-06,
+      "loss": 0.523,
+      "step": 230
+    },
+    {
+      "epoch": 0.12488736709316994,
+      "grad_norm": 0.42592427134513855,
+      "learning_rate": 9.981155883162281e-06,
+      "loss": 0.5169,
+      "step": 231
+    },
+    {
+      "epoch": 0.12542800504595422,
+      "grad_norm": 0.4949047267436981,
+      "learning_rate": 9.980328211698418e-06,
+      "loss": 0.5368,
+      "step": 232
+    },
+    {
+      "epoch": 0.1259686429987385,
+      "grad_norm": 0.49621617794036865,
+      "learning_rate": 9.979482788085455e-06,
+      "loss": 0.5401,
+      "step": 233
+    },
+    {
+      "epoch": 0.1265092809515228,
+      "grad_norm": 0.5175781846046448,
+      "learning_rate": 9.978619615336866e-06,
+      "loss": 0.5167,
+      "step": 234
+    },
+    {
+      "epoch": 0.12704991890430709,
+      "grad_norm": 0.5909814834594727,
+      "learning_rate": 9.977738696529387e-06,
+      "loss": 0.5132,
+      "step": 235
+    },
+    {
+      "epoch": 0.12759055685709136,
+      "grad_norm": 0.43574607372283936,
+      "learning_rate": 9.976840034803014e-06,
+      "loss": 0.5355,
+      "step": 236
+    },
+    {
+      "epoch": 0.12813119480987564,
+      "grad_norm": 0.4887847602367401,
+      "learning_rate": 9.975923633360985e-06,
+      "loss": 0.5362,
+      "step": 237
+    },
+    {
+      "epoch": 0.12867183276265995,
+      "grad_norm": 0.5564954280853271,
+      "learning_rate": 9.974989495469771e-06,
+      "loss": 0.5436,
+      "step": 238
+    },
+    {
+      "epoch": 0.12921247071544423,
+      "grad_norm": 0.4811561405658722,
+      "learning_rate": 9.974037624459063e-06,
+      "loss": 0.5005,
+      "step": 239
+    },
+    {
+      "epoch": 0.1297531086682285,
+      "grad_norm": 0.5640829205513,
+      "learning_rate": 9.973068023721761e-06,
+      "loss": 0.4908,
+      "step": 240
+    },
+    {
+      "epoch": 0.13029374662101278,
+      "grad_norm": 0.558860719203949,
+      "learning_rate": 9.972080696713962e-06,
+      "loss": 0.5369,
+      "step": 241
+    },
+    {
+      "epoch": 0.1308343845737971,
+      "grad_norm": 0.5138370394706726,
+      "learning_rate": 9.971075646954946e-06,
+      "loss": 0.545,
+      "step": 242
+    },
+    {
+      "epoch": 0.13137502252658137,
+      "grad_norm": 0.558138906955719,
+      "learning_rate": 9.970052878027169e-06,
+      "loss": 0.5029,
+      "step": 243
+    },
+    {
+      "epoch": 0.13191566047936565,
+      "grad_norm": 0.5365882515907288,
+      "learning_rate": 9.969012393576241e-06,
+      "loss": 0.5164,
+      "step": 244
+    },
+    {
+      "epoch": 0.13245629843214993,
+      "grad_norm": 0.6326988935470581,
+      "learning_rate": 9.967954197310922e-06,
+      "loss": 0.5089,
+      "step": 245
+    },
+    {
+      "epoch": 0.13299693638493423,
+      "grad_norm": 0.5007409453392029,
+      "learning_rate": 9.966878293003102e-06,
+      "loss": 0.4909,
+      "step": 246
+    },
+    {
+      "epoch": 0.1335375743377185,
+      "grad_norm": 0.5736434459686279,
+      "learning_rate": 9.965784684487794e-06,
+      "loss": 0.5446,
+      "step": 247
+    },
+    {
+      "epoch": 0.1340782122905028,
+      "grad_norm": 0.4436694383621216,
+      "learning_rate": 9.964673375663114e-06,
+      "loss": 0.5096,
+      "step": 248
+    },
+    {
+      "epoch": 0.13461885024328707,
+      "grad_norm": 0.4315429627895355,
+      "learning_rate": 9.96354437049027e-06,
+      "loss": 0.506,
+      "step": 249
+    },
+    {
+      "epoch": 0.13515948819607138,
+      "grad_norm": 0.5133793354034424,
+      "learning_rate": 9.962397672993552e-06,
+      "loss": 0.5142,
+      "step": 250
+    },
+    {
+      "epoch": 0.13570012614885565,
+      "grad_norm": 0.4259757697582245,
+      "learning_rate": 9.961233287260305e-06,
+      "loss": 0.5126,
+      "step": 251
+    },
+    {
+      "epoch": 0.13624076410163993,
+      "grad_norm": 0.497167706489563,
+      "learning_rate": 9.96005121744093e-06,
+      "loss": 0.5345,
+      "step": 252
+    },
+    {
+      "epoch": 0.1367814020544242,
+      "grad_norm": 0.5054571628570557,
+      "learning_rate": 9.958851467748863e-06,
+      "loss": 0.5147,
+      "step": 253
+    },
+    {
+      "epoch": 0.13732204000720852,
+      "grad_norm": 0.48127052187919617,
+      "learning_rate": 9.957634042460551e-06,
+      "loss": 0.4974,
+      "step": 254
+    },
+    {
+      "epoch": 0.1378626779599928,
+      "grad_norm": 0.5029913187026978,
+      "learning_rate": 9.956398945915455e-06,
+      "loss": 0.5036,
+      "step": 255
+    },
+    {
+      "epoch": 0.13840331591277708,
+      "grad_norm": 0.4597805142402649,
+      "learning_rate": 9.955146182516015e-06,
+      "loss": 0.4872,
+      "step": 256
+    },
+    {
+      "epoch": 0.13894395386556135,
+      "grad_norm": 0.45765337347984314,
+      "learning_rate": 9.95387575672765e-06,
+      "loss": 0.5019,
+      "step": 257
+    },
+    {
+      "epoch": 0.13948459181834566,
+      "grad_norm": 0.5424937009811401,
+      "learning_rate": 9.952587673078738e-06,
+      "loss": 0.5358,
+      "step": 258
+    },
+    {
+      "epoch": 0.14002522977112994,
+      "grad_norm": 0.6009652018547058,
+      "learning_rate": 9.951281936160587e-06,
+      "loss": 0.533,
+      "step": 259
+    },
+    {
+      "epoch": 0.14056586772391422,
+      "grad_norm": 0.5036184191703796,
+      "learning_rate": 9.949958550627436e-06,
+      "loss": 0.4852,
+      "step": 260
+    },
+    {
+      "epoch": 0.1411065056766985,
+      "grad_norm": 0.6518417000770569,
+      "learning_rate": 9.948617521196438e-06,
+      "loss": 0.538,
+      "step": 261
+    },
+    {
+      "epoch": 0.1416471436294828,
+      "grad_norm": 0.5519372224807739,
+      "learning_rate": 9.947258852647623e-06,
+      "loss": 0.5216,
+      "step": 262
+    },
+    {
+      "epoch": 0.14218778158226708,
+      "grad_norm": 0.6252304911613464,
+      "learning_rate": 9.945882549823906e-06,
+      "loss": 0.5188,
+      "step": 263
+    },
+    {
+      "epoch": 0.14272841953505136,
+      "grad_norm": 0.5159186124801636,
+      "learning_rate": 9.944488617631053e-06,
+      "loss": 0.5178,
+      "step": 264
+    },
+    {
+      "epoch": 0.14326905748783564,
+      "grad_norm": 0.5060036182403564,
+      "learning_rate": 9.943077061037672e-06,
+      "loss": 0.5083,
+      "step": 265
+    },
+    {
+      "epoch": 0.14380969544061994,
+      "grad_norm": 0.579529345035553,
+      "learning_rate": 9.94164788507519e-06,
+      "loss": 0.4991,
+      "step": 266
+    },
+    {
+      "epoch": 0.14435033339340422,
+      "grad_norm": 0.5514207482337952,
+      "learning_rate": 9.940201094837838e-06,
+      "loss": 0.5154,
+      "step": 267
+    },
+    {
+      "epoch": 0.1448909713461885,
+      "grad_norm": 0.5269582867622375,
+      "learning_rate": 9.938736695482636e-06,
+      "loss": 0.4937,
+      "step": 268
+    },
+    {
+      "epoch": 0.14543160929897278,
+      "grad_norm": 0.5171031951904297,
+      "learning_rate": 9.937254692229363e-06,
+      "loss": 0.5024,
+      "step": 269
+    },
+    {
+      "epoch": 0.1459722472517571,
+      "grad_norm": 0.5391225814819336,
+      "learning_rate": 9.935755090360554e-06,
+      "loss": 0.4961,
+      "step": 270
+    },
+    {
+      "epoch": 0.14651288520454137,
+      "grad_norm": 0.47390222549438477,
+      "learning_rate": 9.93423789522147e-06,
+      "loss": 0.4855,
+      "step": 271
+    },
+    {
+      "epoch": 0.14705352315732564,
+      "grad_norm": 0.4980723261833191,
+      "learning_rate": 9.932703112220084e-06,
+      "loss": 0.5481,
+      "step": 272
+    },
+    {
+      "epoch": 0.14759416111010992,
+      "grad_norm": 0.5040086507797241,
+      "learning_rate": 9.931150746827055e-06,
+      "loss": 0.52,
+      "step": 273
+    },
+    {
+      "epoch": 0.14813479906289423,
+      "grad_norm": 0.5215423107147217,
+      "learning_rate": 9.929580804575718e-06,
+      "loss": 0.5113,
+      "step": 274
+    },
+    {
+      "epoch": 0.1486754370156785,
+      "grad_norm": 0.44872352480888367,
+      "learning_rate": 9.927993291062064e-06,
+      "loss": 0.4919,
+      "step": 275
+    },
+    {
+      "epoch": 0.1492160749684628,
+      "grad_norm": 0.5068681240081787,
+      "learning_rate": 9.926388211944707e-06,
+      "loss": 0.5243,
+      "step": 276
+    },
+    {
+      "epoch": 0.14975671292124706,
+      "grad_norm": 0.5543198585510254,
+      "learning_rate": 9.924765572944879e-06,
+      "loss": 0.5083,
+      "step": 277
+    },
+    {
+      "epoch": 0.15029735087403134,
+      "grad_norm": 0.4373819828033447,
+      "learning_rate": 9.9231253798464e-06,
+      "loss": 0.5,
+      "step": 278
+    },
+    {
+      "epoch": 0.15083798882681565,
+      "grad_norm": 0.4874158501625061,
+      "learning_rate": 9.921467638495666e-06,
+      "loss": 0.5336,
+      "step": 279
+    },
+    {
+      "epoch": 0.15137862677959993,
+      "grad_norm": 0.4793112874031067,
+      "learning_rate": 9.919792354801614e-06,
+      "loss": 0.5135,
+      "step": 280
+    },
+    {
+      "epoch": 0.1519192647323842,
+      "grad_norm": 0.4696117341518402,
+      "learning_rate": 9.91809953473572e-06,
+      "loss": 0.5018,
+      "step": 281
+    },
+    {
+      "epoch": 0.15245990268516849,
+      "grad_norm": 0.5002115964889526,
+      "learning_rate": 9.916389184331957e-06,
+      "loss": 0.5332,
+      "step": 282
+    },
+    {
+      "epoch": 0.1530005406379528,
+      "grad_norm": 0.4873870015144348,
+      "learning_rate": 9.914661309686796e-06,
+      "loss": 0.5238,
+      "step": 283
+    },
+    {
+      "epoch": 0.15354117859073707,
+      "grad_norm": 0.5112310647964478,
+      "learning_rate": 9.912915916959162e-06,
+      "loss": 0.5145,
+      "step": 284
+    },
+    {
+      "epoch": 0.15408181654352135,
+      "grad_norm": 0.5636197328567505,
+      "learning_rate": 9.911153012370427e-06,
+      "loss": 0.5095,
+      "step": 285
+    },
+    {
+      "epoch": 0.15462245449630563,
+      "grad_norm": 0.4921044409275055,
+      "learning_rate": 9.909372602204385e-06,
+      "loss": 0.5083,
+      "step": 286
+    },
+    {
+      "epoch": 0.15516309244908993,
+      "grad_norm": 0.5235900282859802,
+      "learning_rate": 9.907574692807223e-06,
+      "loss": 0.5126,
+      "step": 287
+    },
+    {
+      "epoch": 0.1557037304018742,
+      "grad_norm": 0.5021162033081055,
+      "learning_rate": 9.905759290587506e-06,
+      "loss": 0.5067,
+      "step": 288
+    },
+    {
+      "epoch": 0.1562443683546585,
+      "grad_norm": 0.5399244427680969,
+      "learning_rate": 9.903926402016153e-06,
+      "loss": 0.5082,
+      "step": 289
+    },
+    {
+      "epoch": 0.15678500630744277,
+      "grad_norm": 0.48548075556755066,
+      "learning_rate": 9.902076033626409e-06,
+      "loss": 0.5091,
+      "step": 290
+    },
+    {
+      "epoch": 0.15732564426022708,
+      "grad_norm": 0.5076701045036316,
+      "learning_rate": 9.900208192013825e-06,
+      "loss": 0.5002,
+      "step": 291
+    },
+    {
+      "epoch": 0.15786628221301136,
+      "grad_norm": 0.5822921991348267,
+      "learning_rate": 9.898322883836239e-06,
+      "loss": 0.5074,
+      "step": 292
+    },
+    {
+      "epoch": 0.15840692016579563,
+      "grad_norm": 0.4363133907318115,
+      "learning_rate": 9.896420115813741e-06,
+      "loss": 0.5065,
+      "step": 293
+    },
+    {
+      "epoch": 0.1589475581185799,
+      "grad_norm": 0.47977545857429504,
+      "learning_rate": 9.894499894728665e-06,
+      "loss": 0.477,
+      "step": 294
+    },
+    {
+      "epoch": 0.15948819607136422,
+      "grad_norm": 0.5231459736824036,
+      "learning_rate": 9.892562227425541e-06,
+      "loss": 0.5164,
+      "step": 295
+    },
+    {
+      "epoch": 0.1600288340241485,
+      "grad_norm": 0.4468756318092346,
+      "learning_rate": 9.890607120811104e-06,
+      "loss": 0.4966,
+      "step": 296
+    },
+    {
+      "epoch": 0.16056947197693278,
+      "grad_norm": 0.5080546736717224,
+      "learning_rate": 9.888634581854235e-06,
+      "loss": 0.4717,
+      "step": 297
+    },
+    {
+      "epoch": 0.16111010992971705,
+      "grad_norm": 0.6117534637451172,
+      "learning_rate": 9.88664461758596e-06,
+      "loss": 0.5435,
+      "step": 298
+    },
+    {
+      "epoch": 0.16165074788250136,
+      "grad_norm": 0.44875776767730713,
+      "learning_rate": 9.884637235099414e-06,
+      "loss": 0.5162,
+      "step": 299
+    },
+    {
+      "epoch": 0.16219138583528564,
+      "grad_norm": 0.5946277379989624,
+      "learning_rate": 9.882612441549817e-06,
+      "loss": 0.5086,
+      "step": 300
+    },
+    {
+      "epoch": 0.16273202378806992,
+      "grad_norm": 0.6129415035247803,
+      "learning_rate": 9.880570244154455e-06,
+      "loss": 0.5252,
+      "step": 301
+    },
+    {
+      "epoch": 0.1632726617408542,
+      "grad_norm": 0.5008973479270935,
+      "learning_rate": 9.878510650192644e-06,
+      "loss": 0.5181,
+      "step": 302
+    },
+    {
+      "epoch": 0.1638132996936385,
+      "grad_norm": 0.6963375210762024,
+      "learning_rate": 9.876433667005711e-06,
+      "loss": 0.5238,
+      "step": 303
+    },
+    {
+      "epoch": 0.16435393764642278,
+      "grad_norm": 0.4780685305595398,
+      "learning_rate": 9.874339301996968e-06,
+      "loss": 0.5107,
+      "step": 304
+    },
+    {
+      "epoch": 0.16489457559920706,
+      "grad_norm": 0.47462719678878784,
+      "learning_rate": 9.87222756263168e-06,
+      "loss": 0.5003,
+      "step": 305
+    },
+    {
+      "epoch": 0.16543521355199134,
+      "grad_norm": 0.6904825568199158,
+      "learning_rate": 9.870098456437045e-06,
+      "loss": 0.5256,
+      "step": 306
+    },
+    {
+      "epoch": 0.16597585150477565,
+      "grad_norm": 0.4498540759086609,
+      "learning_rate": 9.867951991002162e-06,
+      "loss": 0.5088,
+      "step": 307
+    },
+    {
+      "epoch": 0.16651648945755992,
+      "grad_norm": 0.5996077656745911,
+      "learning_rate": 9.865788173978011e-06,
+      "loss": 0.5045,
+      "step": 308
+    },
+    {
+      "epoch": 0.1670571274103442,
+      "grad_norm": 0.5990262627601624,
+      "learning_rate": 9.863607013077414e-06,
+      "loss": 0.499,
+      "step": 309
+    },
+    {
+      "epoch": 0.16759776536312848,
+      "grad_norm": 0.456752210855484,
+      "learning_rate": 9.86140851607502e-06,
+      "loss": 0.5118,
+      "step": 310
+    },
+    {
+      "epoch": 0.1681384033159128,
+      "grad_norm": 0.5640468001365662,
+      "learning_rate": 9.85919269080727e-06,
+      "loss": 0.5099,
+      "step": 311
+    },
+    {
+      "epoch": 0.16867904126869707,
+      "grad_norm": 0.4840059280395508,
+      "learning_rate": 9.856959545172369e-06,
+      "loss": 0.5047,
+      "step": 312
+    },
+    {
+      "epoch": 0.16921967922148134,
+      "grad_norm": 0.4675785005092621,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 0.5176,
+      "step": 313
+    },
+    {
+      "epoch": 0.16976031717426562,
+      "grad_norm": 0.48285573720932007,
+      "learning_rate": 9.852441324702599e-06,
+      "loss": 0.515,
+      "step": 314
+    },
+    {
+      "epoch": 0.17030095512704993,
+      "grad_norm": 0.5241273045539856,
+      "learning_rate": 9.850156265972722e-06,
+      "loss": 0.5132,
+      "step": 315
+    },
+    {
+      "epoch": 0.1708415930798342,
+      "grad_norm": 0.5342344641685486,
+      "learning_rate": 9.847853919085608e-06,
+      "loss": 0.5348,
+      "step": 316
+    },
+    {
+      "epoch": 0.1713822310326185,
+      "grad_norm": 0.5915160179138184,
+      "learning_rate": 9.845534292247872e-06,
+      "loss": 0.4903,
+      "step": 317
+    },
+    {
+      "epoch": 0.17192286898540277,
+      "grad_norm": 0.5160459280014038,
+      "learning_rate": 9.843197393727713e-06,
+      "loss": 0.5048,
+      "step": 318
+    },
+    {
+      "epoch": 0.17246350693818707,
+      "grad_norm": 0.4898647665977478,
+      "learning_rate": 9.8408432318549e-06,
+      "loss": 0.5076,
+      "step": 319
+    },
+    {
+      "epoch": 0.17300414489097135,
+      "grad_norm": 0.4921559989452362,
+      "learning_rate": 9.838471815020731e-06,
+      "loss": 0.5231,
+      "step": 320
+    },
+    {
+      "epoch": 0.17354478284375563,
+      "grad_norm": 0.5118461847305298,
+      "learning_rate": 9.836083151678014e-06,
+      "loss": 0.5057,
+      "step": 321
+    },
+    {
+      "epoch": 0.1740854207965399,
+      "grad_norm": 0.48118114471435547,
+      "learning_rate": 9.833677250341027e-06,
+      "loss": 0.5581,
+      "step": 322
+    },
+    {
+      "epoch": 0.17462605874932421,
+      "grad_norm": 0.4235428273677826,
+      "learning_rate": 9.831254119585497e-06,
+      "loss": 0.4938,
+      "step": 323
+    },
+    {
+      "epoch": 0.1751666967021085,
+      "grad_norm": 0.49432626366615295,
+      "learning_rate": 9.828813768048555e-06,
+      "loss": 0.503,
+      "step": 324
+    },
+    {
+      "epoch": 0.17570733465489277,
+      "grad_norm": 0.4676252007484436,
+      "learning_rate": 9.826356204428726e-06,
+      "loss": 0.5017,
+      "step": 325
+    },
+    {
+      "epoch": 0.17624797260767705,
+      "grad_norm": 0.48674294352531433,
+      "learning_rate": 9.823881437485882e-06,
+      "loss": 0.5122,
+      "step": 326
+    },
+    {
+      "epoch": 0.17678861056046136,
+      "grad_norm": 0.5041030645370483,
+      "learning_rate": 9.821389476041212e-06,
+      "loss": 0.5014,
+      "step": 327
+    },
+    {
+      "epoch": 0.17732924851324564,
+      "grad_norm": 0.5799689888954163,
+      "learning_rate": 9.8188803289772e-06,
+      "loss": 0.5245,
+      "step": 328
+    },
+    {
+      "epoch": 0.17786988646602991,
+      "grad_norm": 0.5120707750320435,
+      "learning_rate": 9.816354005237583e-06,
+      "loss": 0.4898,
+      "step": 329
+    },
+    {
+      "epoch": 0.1784105244188142,
+      "grad_norm": 0.5724909901618958,
+      "learning_rate": 9.813810513827324e-06,
+      "loss": 0.5398,
+      "step": 330
+    },
+    {
+      "epoch": 0.1789511623715985,
+      "grad_norm": 0.6064598560333252,
+      "learning_rate": 9.811249863812581e-06,
+      "loss": 0.4939,
+      "step": 331
+    },
+    {
+      "epoch": 0.17949180032438278,
+      "grad_norm": 0.5473314523696899,
+      "learning_rate": 9.808672064320672e-06,
+      "loss": 0.5113,
+      "step": 332
+    },
+    {
+      "epoch": 0.18003243827716706,
+      "grad_norm": 0.45414504408836365,
+      "learning_rate": 9.806077124540045e-06,
+      "loss": 0.4815,
+      "step": 333
+    },
+    {
+      "epoch": 0.18057307622995133,
+      "grad_norm": 0.48928454518318176,
+      "learning_rate": 9.803465053720242e-06,
+      "loss": 0.502,
+      "step": 334
+    },
+    {
+      "epoch": 0.18111371418273564,
+      "grad_norm": 0.4803617000579834,
+      "learning_rate": 9.800835861171869e-06,
+      "loss": 0.4851,
+      "step": 335
+    },
+    {
+      "epoch": 0.18165435213551992,
+      "grad_norm": 0.4396837055683136,
+      "learning_rate": 9.798189556266559e-06,
+      "loss": 0.4956,
+      "step": 336
+    },
+    {
+      "epoch": 0.1821949900883042,
+      "grad_norm": 0.5100455284118652,
+      "learning_rate": 9.795526148436945e-06,
+      "loss": 0.5282,
+      "step": 337
+    },
+    {
+      "epoch": 0.18273562804108848,
+      "grad_norm": 0.426038920879364,
+      "learning_rate": 9.792845647176621e-06,
+      "loss": 0.4697,
+      "step": 338
+    },
+    {
+      "epoch": 0.18327626599387278,
+      "grad_norm": 0.42569661140441895,
+      "learning_rate": 9.790148062040108e-06,
+      "loss": 0.4948,
+      "step": 339
+    },
+    {
+      "epoch": 0.18381690394665706,
+      "grad_norm": 0.5453207492828369,
+      "learning_rate": 9.787433402642823e-06,
+      "loss": 0.5067,
+      "step": 340
+    },
+    {
+      "epoch": 0.18435754189944134,
+      "grad_norm": 0.5373483300209045,
+      "learning_rate": 9.784701678661045e-06,
+      "loss": 0.5227,
+      "step": 341
+    },
+    {
+      "epoch": 0.18489817985222562,
+      "grad_norm": 0.5656136274337769,
+      "learning_rate": 9.781952899831876e-06,
+      "loss": 0.5163,
+      "step": 342
+    },
+    {
+      "epoch": 0.1854388178050099,
+      "grad_norm": 0.5008209347724915,
+      "learning_rate": 9.779187075953215e-06,
+      "loss": 0.5129,
+      "step": 343
+    },
+    {
+      "epoch": 0.1859794557577942,
+      "grad_norm": 0.583916962146759,
+      "learning_rate": 9.776404216883709e-06,
+      "loss": 0.503,
+      "step": 344
+    },
+    {
+      "epoch": 0.18652009371057848,
+      "grad_norm": 0.5473840832710266,
+      "learning_rate": 9.77360433254273e-06,
+      "loss": 0.5407,
+      "step": 345
+    },
+    {
+      "epoch": 0.18706073166336276,
+      "grad_norm": 0.5509775280952454,
+      "learning_rate": 9.770787432910336e-06,
+      "loss": 0.4781,
+      "step": 346
+    },
+    {
+      "epoch": 0.18760136961614704,
+      "grad_norm": 0.6003471612930298,
+      "learning_rate": 9.767953528027238e-06,
+      "loss": 0.5375,
+      "step": 347
+    },
+    {
+      "epoch": 0.18814200756893135,
+      "grad_norm": 0.43554067611694336,
+      "learning_rate": 9.765102627994757e-06,
+      "loss": 0.4962,
+      "step": 348
+    },
+    {
+      "epoch": 0.18868264552171563,
+      "grad_norm": 0.6562443375587463,
+      "learning_rate": 9.762234742974793e-06,
+      "loss": 0.4952,
+      "step": 349
+    },
+    {
+      "epoch": 0.1892232834744999,
+      "grad_norm": 0.5603171586990356,
+      "learning_rate": 9.759349883189788e-06,
+      "loss": 0.4783,
+      "step": 350
+    },
+    {
+      "epoch": 0.18976392142728418,
+      "grad_norm": 0.5634536147117615,
+      "learning_rate": 9.756448058922697e-06,
+      "loss": 0.4884,
+      "step": 351
+    },
+    {
+      "epoch": 0.1903045593800685,
+      "grad_norm": 0.4945487082004547,
+      "learning_rate": 9.753529280516931e-06,
+      "loss": 0.5012,
+      "step": 352
+    },
+    {
+      "epoch": 0.19084519733285277,
+      "grad_norm": 0.588554322719574,
+      "learning_rate": 9.750593558376347e-06,
+      "loss": 0.5098,
+      "step": 353
+    },
+    {
+      "epoch": 0.19138583528563705,
+      "grad_norm": 0.4570363163948059,
+      "learning_rate": 9.747640902965185e-06,
+      "loss": 0.4675,
+      "step": 354
+    },
+    {
+      "epoch": 0.19192647323842132,
+      "grad_norm": 0.6398181915283203,
+      "learning_rate": 9.74467132480805e-06,
+      "loss": 0.5402,
+      "step": 355
+    },
+    {
+      "epoch": 0.19246711119120563,
+      "grad_norm": 0.4467182159423828,
+      "learning_rate": 9.741684834489866e-06,
+      "loss": 0.4916,
+      "step": 356
+    },
+    {
+      "epoch": 0.1930077491439899,
+      "grad_norm": 0.5539352297782898,
+      "learning_rate": 9.738681442655842e-06,
+      "loss": 0.4996,
+      "step": 357
+    },
+    {
+      "epoch": 0.1935483870967742,
+      "grad_norm": 0.5078352689743042,
+      "learning_rate": 9.735661160011424e-06,
+      "loss": 0.5055,
+      "step": 358
+    },
+    {
+      "epoch": 0.19408902504955847,
+      "grad_norm": 0.4970700442790985,
+      "learning_rate": 9.732623997322274e-06,
+      "loss": 0.5247,
+      "step": 359
+    },
+    {
+      "epoch": 0.19462966300234277,
+      "grad_norm": 0.4787460267543793,
+      "learning_rate": 9.729569965414214e-06,
+      "loss": 0.5114,
+      "step": 360
+    },
+    {
+      "epoch": 0.19517030095512705,
+      "grad_norm": 0.5646668672561646,
+      "learning_rate": 9.726499075173201e-06,
+      "loss": 0.4905,
+      "step": 361
+    },
+    {
+      "epoch": 0.19571093890791133,
+      "grad_norm": 0.5013266801834106,
+      "learning_rate": 9.723411337545283e-06,
+      "loss": 0.5112,
+      "step": 362
+    },
+    {
+      "epoch": 0.1962515768606956,
+      "grad_norm": 0.5415170192718506,
+      "learning_rate": 9.720306763536553e-06,
+      "loss": 0.4905,
+      "step": 363
+    },
+    {
+      "epoch": 0.19679221481347992,
+      "grad_norm": 0.443282812833786,
+      "learning_rate": 9.717185364213127e-06,
+      "loss": 0.5082,
+      "step": 364
+    },
+    {
+      "epoch": 0.1973328527662642,
+      "grad_norm": 0.6220279932022095,
+      "learning_rate": 9.714047150701082e-06,
+      "loss": 0.5117,
+      "step": 365
+    },
+    {
+      "epoch": 0.19787349071904847,
+      "grad_norm": 0.46816128492355347,
+      "learning_rate": 9.710892134186438e-06,
+      "loss": 0.5067,
+      "step": 366
+    },
+    {
+      "epoch": 0.19841412867183275,
+      "grad_norm": 0.578245222568512,
+      "learning_rate": 9.707720325915105e-06,
+      "loss": 0.518,
+      "step": 367
+    },
+    {
+      "epoch": 0.19895476662461706,
+      "grad_norm": 0.5336887240409851,
+      "learning_rate": 9.704531737192847e-06,
+      "loss": 0.513,
+      "step": 368
+    },
+    {
+      "epoch": 0.19949540457740134,
+      "grad_norm": 0.5477902889251709,
+      "learning_rate": 9.701326379385238e-06,
+      "loss": 0.4937,
+      "step": 369
+    },
+    {
+      "epoch": 0.20003604253018561,
+      "grad_norm": 0.486288845539093,
+      "learning_rate": 9.698104263917632e-06,
+      "loss": 0.4928,
+      "step": 370
+    },
+    {
+      "epoch": 0.2005766804829699,
+      "grad_norm": 0.4892939031124115,
+      "learning_rate": 9.694865402275105e-06,
+      "loss": 0.5004,
+      "step": 371
+    },
+    {
+      "epoch": 0.2011173184357542,
+      "grad_norm": 0.4445298910140991,
+      "learning_rate": 9.691609806002433e-06,
+      "loss": 0.4874,
+      "step": 372
+    },
+    {
+      "epoch": 0.20165795638853848,
+      "grad_norm": 0.4402690827846527,
+      "learning_rate": 9.688337486704038e-06,
+      "loss": 0.4714,
+      "step": 373
+    },
+    {
+      "epoch": 0.20219859434132276,
+      "grad_norm": 0.5016452670097351,
+      "learning_rate": 9.68504845604395e-06,
+      "loss": 0.5061,
+      "step": 374
+    },
+    {
+      "epoch": 0.20273923229410704,
+      "grad_norm": 0.4491643011569977,
+      "learning_rate": 9.681742725745762e-06,
+      "loss": 0.5089,
+      "step": 375
+    },
+    {
+      "epoch": 0.20327987024689134,
+      "grad_norm": 0.45842084288597107,
+      "learning_rate": 9.678420307592602e-06,
+      "loss": 0.4886,
+      "step": 376
+    },
+    {
+      "epoch": 0.20382050819967562,
+      "grad_norm": 0.4361850619316101,
+      "learning_rate": 9.675081213427076e-06,
+      "loss": 0.5183,
+      "step": 377
+    },
+    {
+      "epoch": 0.2043611461524599,
+      "grad_norm": 0.5042484998703003,
+      "learning_rate": 9.671725455151226e-06,
+      "loss": 0.4974,
+      "step": 378
+    },
+    {
+      "epoch": 0.20490178410524418,
+      "grad_norm": 0.47856807708740234,
+      "learning_rate": 9.668353044726498e-06,
+      "loss": 0.4957,
+      "step": 379
+    },
+    {
+      "epoch": 0.20544242205802848,
+      "grad_norm": 0.5412836074829102,
+      "learning_rate": 9.664963994173695e-06,
+      "loss": 0.5324,
+      "step": 380
+    },
+    {
+      "epoch": 0.20598306001081276,
+      "grad_norm": 0.5093710422515869,
+      "learning_rate": 9.66155831557293e-06,
+      "loss": 0.5335,
+      "step": 381
+    },
+    {
+      "epoch": 0.20652369796359704,
+      "grad_norm": 0.4847930073738098,
+      "learning_rate": 9.658136021063585e-06,
+      "loss": 0.4609,
+      "step": 382
+    },
+    {
+      "epoch": 0.20706433591638132,
+      "grad_norm": 0.48776888847351074,
+      "learning_rate": 9.65469712284427e-06,
+      "loss": 0.5006,
+      "step": 383
+    },
+    {
+      "epoch": 0.20760497386916563,
+      "grad_norm": 0.5098462104797363,
+      "learning_rate": 9.651241633172782e-06,
+      "loss": 0.5021,
+      "step": 384
+    },
+    {
+      "epoch": 0.2081456118219499,
+      "grad_norm": 0.4696345031261444,
+      "learning_rate": 9.647769564366048e-06,
+      "loss": 0.4987,
+      "step": 385
+    },
+    {
+      "epoch": 0.20868624977473418,
+      "grad_norm": 0.5066925883293152,
+      "learning_rate": 9.644280928800101e-06,
+      "loss": 0.5182,
+      "step": 386
+    },
+    {
+      "epoch": 0.20922688772751846,
+      "grad_norm": 0.5005125403404236,
+      "learning_rate": 9.640775738910019e-06,
+      "loss": 0.5005,
+      "step": 387
+    },
+    {
+      "epoch": 0.20976752568030277,
+      "grad_norm": 0.43679603934288025,
+      "learning_rate": 9.63725400718989e-06,
+      "loss": 0.4928,
+      "step": 388
+    },
+    {
+      "epoch": 0.21030816363308705,
+      "grad_norm": 0.5588410496711731,
+      "learning_rate": 9.633715746192762e-06,
+      "loss": 0.5374,
+      "step": 389
+    },
+    {
+      "epoch": 0.21084880158587133,
+      "grad_norm": 0.4636860489845276,
+      "learning_rate": 9.630160968530601e-06,
+      "loss": 0.5011,
+      "step": 390
+    },
+    {
+      "epoch": 0.2113894395386556,
+      "grad_norm": 0.5035771131515503,
+      "learning_rate": 9.626589686874252e-06,
+      "loss": 0.5152,
+      "step": 391
+    },
+    {
+      "epoch": 0.2119300774914399,
+      "grad_norm": 0.5176652669906616,
+      "learning_rate": 9.62300191395338e-06,
+      "loss": 0.4931,
+      "step": 392
+    },
+    {
+      "epoch": 0.2124707154442242,
+      "grad_norm": 0.4670732319355011,
+      "learning_rate": 9.619397662556434e-06,
+      "loss": 0.5099,
+      "step": 393
+    },
+    {
+      "epoch": 0.21301135339700847,
+      "grad_norm": 0.5705310702323914,
+      "learning_rate": 9.615776945530603e-06,
+      "loss": 0.5244,
+      "step": 394
+    },
+    {
+      "epoch": 0.21355199134979275,
+      "grad_norm": 0.48803743720054626,
+      "learning_rate": 9.612139775781766e-06,
+      "loss": 0.4953,
+      "step": 395
+    },
+    {
+      "epoch": 0.21409262930257705,
+      "grad_norm": 0.6569446921348572,
+      "learning_rate": 9.608486166274444e-06,
+      "loss": 0.5086,
+      "step": 396
+    },
+    {
+      "epoch": 0.21463326725536133,
+      "grad_norm": 0.5075775980949402,
+      "learning_rate": 9.60481613003176e-06,
+      "loss": 0.5471,
+      "step": 397
+    },
+    {
+      "epoch": 0.2151739052081456,
+      "grad_norm": 0.5804044604301453,
+      "learning_rate": 9.601129680135386e-06,
+      "loss": 0.5075,
+      "step": 398
+    },
+    {
+      "epoch": 0.2157145431609299,
+      "grad_norm": 0.4748011827468872,
+      "learning_rate": 9.597426829725504e-06,
+      "loss": 0.4741,
+      "step": 399
+    },
+    {
+      "epoch": 0.2162551811137142,
+      "grad_norm": 0.5516924858093262,
+      "learning_rate": 9.593707592000751e-06,
+      "loss": 0.5141,
+      "step": 400
+    },
+    {
+      "epoch": 0.21679581906649847,
+      "grad_norm": 0.49007630348205566,
+      "learning_rate": 9.58997198021818e-06,
+      "loss": 0.4965,
+      "step": 401
+    },
+    {
+      "epoch": 0.21733645701928275,
+      "grad_norm": 0.49923837184906006,
+      "learning_rate": 9.586220007693205e-06,
+      "loss": 0.4953,
+      "step": 402
+    },
+    {
+      "epoch": 0.21787709497206703,
+      "grad_norm": 0.4540408253669739,
+      "learning_rate": 9.582451687799557e-06,
+      "loss": 0.4577,
+      "step": 403
+    },
+    {
+      "epoch": 0.21841773292485134,
+      "grad_norm": 0.4953221380710602,
+      "learning_rate": 9.578667033969238e-06,
+      "loss": 0.5014,
+      "step": 404
+    },
+    {
+      "epoch": 0.21895837087763562,
+      "grad_norm": 0.5561023354530334,
+      "learning_rate": 9.574866059692471e-06,
+      "loss": 0.5158,
+      "step": 405
+    },
+    {
+      "epoch": 0.2194990088304199,
+      "grad_norm": 0.42774730920791626,
+      "learning_rate": 9.571048778517655e-06,
+      "loss": 0.4998,
+      "step": 406
+    },
+    {
+      "epoch": 0.22003964678320417,
+      "grad_norm": 0.5168089866638184,
+      "learning_rate": 9.567215204051307e-06,
+      "loss": 0.5156,
+      "step": 407
+    },
+    {
+      "epoch": 0.22058028473598848,
+      "grad_norm": 0.45053431391716003,
+      "learning_rate": 9.563365349958032e-06,
+      "loss": 0.5138,
+      "step": 408
+    },
+    {
+      "epoch": 0.22112092268877276,
+      "grad_norm": 0.4745253026485443,
+      "learning_rate": 9.55949922996045e-06,
+      "loss": 0.5122,
+      "step": 409
+    },
+    {
+      "epoch": 0.22166156064155704,
+      "grad_norm": 0.49608004093170166,
+      "learning_rate": 9.555616857839171e-06,
+      "loss": 0.4882,
+      "step": 410
+    },
+    {
+      "epoch": 0.22220219859434132,
+      "grad_norm": 0.46152451634407043,
+      "learning_rate": 9.551718247432732e-06,
+      "loss": 0.4709,
+      "step": 411
+    },
+    {
+      "epoch": 0.2227428365471256,
+      "grad_norm": 0.4751157760620117,
+      "learning_rate": 9.547803412637542e-06,
+      "loss": 0.4729,
+      "step": 412
+    },
+    {
+      "epoch": 0.2232834744999099,
+      "grad_norm": 0.5150904655456543,
+      "learning_rate": 9.543872367407854e-06,
+      "loss": 0.5082,
+      "step": 413
+    },
+    {
+      "epoch": 0.22382411245269418,
+      "grad_norm": 0.46366235613822937,
+      "learning_rate": 9.539925125755695e-06,
+      "loss": 0.4977,
+      "step": 414
+    },
+    {
+      "epoch": 0.22436475040547846,
+      "grad_norm": 0.5216416120529175,
+      "learning_rate": 9.535961701750825e-06,
+      "loss": 0.513,
+      "step": 415
+    },
+    {
+      "epoch": 0.22490538835826274,
+      "grad_norm": 0.465658962726593,
+      "learning_rate": 9.531982109520686e-06,
+      "loss": 0.4844,
+      "step": 416
+    },
+    {
+      "epoch": 0.22544602631104704,
+      "grad_norm": 0.48046615719795227,
+      "learning_rate": 9.527986363250348e-06,
+      "loss": 0.4781,
+      "step": 417
+    },
+    {
+      "epoch": 0.22598666426383132,
+      "grad_norm": 0.47210782766342163,
+      "learning_rate": 9.523974477182465e-06,
+      "loss": 0.4951,
+      "step": 418
+    },
+    {
+      "epoch": 0.2265273022166156,
+      "grad_norm": 0.4686312973499298,
+      "learning_rate": 9.519946465617217e-06,
+      "loss": 0.5024,
+      "step": 419
+    },
+    {
+      "epoch": 0.22706794016939988,
+      "grad_norm": 0.46087512373924255,
+      "learning_rate": 9.515902342912268e-06,
+      "loss": 0.4894,
+      "step": 420
+    },
+    {
+      "epoch": 0.22760857812218419,
+      "grad_norm": 0.5393832921981812,
+      "learning_rate": 9.511842123482703e-06,
+      "loss": 0.4985,
+      "step": 421
+    },
+    {
+      "epoch": 0.22814921607496846,
+      "grad_norm": 0.47389426827430725,
+      "learning_rate": 9.507765821800988e-06,
+      "loss": 0.5029,
+      "step": 422
+    },
+    {
+      "epoch": 0.22868985402775274,
+      "grad_norm": 0.5276787281036377,
+      "learning_rate": 9.503673452396909e-06,
+      "loss": 0.5029,
+      "step": 423
+    },
+    {
+      "epoch": 0.22923049198053702,
+      "grad_norm": 0.5731882452964783,
+      "learning_rate": 9.499565029857529e-06,
+      "loss": 0.5116,
+      "step": 424
+    },
+    {
+      "epoch": 0.22977112993332133,
+      "grad_norm": 0.48996415734291077,
+      "learning_rate": 9.49544056882713e-06,
+      "loss": 0.5103,
+      "step": 425
+    },
+    {
+      "epoch": 0.2303117678861056,
+      "grad_norm": 0.5771133303642273,
+      "learning_rate": 9.491300084007162e-06,
+      "loss": 0.4913,
+      "step": 426
+    },
+    {
+      "epoch": 0.23085240583888988,
+      "grad_norm": 0.42285990715026855,
+      "learning_rate": 9.48714359015619e-06,
+      "loss": 0.4914,
+      "step": 427
+    },
+    {
+      "epoch": 0.23139304379167416,
+      "grad_norm": 0.595855712890625,
+      "learning_rate": 9.482971102089845e-06,
+      "loss": 0.518,
+      "step": 428
+    },
+    {
+      "epoch": 0.23193368174445847,
+      "grad_norm": 0.41206789016723633,
+      "learning_rate": 9.478782634680765e-06,
+      "loss": 0.4696,
+      "step": 429
+    },
+    {
+      "epoch": 0.23247431969724275,
+      "grad_norm": 0.57419753074646,
+      "learning_rate": 9.47457820285855e-06,
+      "loss": 0.4933,
+      "step": 430
+    },
+    {
+      "epoch": 0.23301495765002703,
+      "grad_norm": 0.4686283469200134,
+      "learning_rate": 9.470357821609703e-06,
+      "loss": 0.4914,
+      "step": 431
+    },
+    {
+      "epoch": 0.2335555956028113,
+      "grad_norm": 0.4878191351890564,
+      "learning_rate": 9.466121505977577e-06,
+      "loss": 0.4819,
+      "step": 432
+    },
+    {
+      "epoch": 0.2340962335555956,
+      "grad_norm": 0.5020343661308289,
+      "learning_rate": 9.461869271062322e-06,
+      "loss": 0.508,
+      "step": 433
+    },
+    {
+      "epoch": 0.2346368715083799,
+      "grad_norm": 0.48259827494621277,
+      "learning_rate": 9.457601132020832e-06,
+      "loss": 0.4999,
+      "step": 434
+    },
+    {
+      "epoch": 0.23517750946116417,
+      "grad_norm": 0.439022034406662,
+      "learning_rate": 9.453317104066693e-06,
+      "loss": 0.4849,
+      "step": 435
+    },
+    {
+      "epoch": 0.23571814741394845,
+      "grad_norm": 0.43228328227996826,
+      "learning_rate": 9.44901720247012e-06,
+      "loss": 0.4841,
+      "step": 436
+    },
+    {
+      "epoch": 0.23625878536673275,
+      "grad_norm": 0.43123963475227356,
+      "learning_rate": 9.444701442557917e-06,
+      "loss": 0.5022,
+      "step": 437
+    },
+    {
+      "epoch": 0.23679942331951703,
+      "grad_norm": 0.4314519166946411,
+      "learning_rate": 9.440369839713407e-06,
+      "loss": 0.479,
+      "step": 438
+    },
+    {
+      "epoch": 0.2373400612723013,
+      "grad_norm": 0.4455517828464508,
+      "learning_rate": 9.436022409376391e-06,
+      "loss": 0.5356,
+      "step": 439
+    },
+    {
+      "epoch": 0.2378806992250856,
+      "grad_norm": 0.4760914742946625,
+      "learning_rate": 9.431659167043079e-06,
+      "loss": 0.5081,
+      "step": 440
+    },
+    {
+      "epoch": 0.2384213371778699,
+      "grad_norm": 0.4999728798866272,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 0.518,
+      "step": 441
+    },
+    {
+      "epoch": 0.23896197513065418,
+      "grad_norm": 0.397093802690506,
+      "learning_rate": 9.422885308654183e-06,
+      "loss": 0.5257,
+      "step": 442
+    },
+    {
+      "epoch": 0.23950261308343845,
+      "grad_norm": 0.4785946011543274,
+      "learning_rate": 9.418474723872609e-06,
+      "loss": 0.4708,
+      "step": 443
+    },
+    {
+      "epoch": 0.24004325103622273,
+      "grad_norm": 0.42722031474113464,
+      "learning_rate": 9.414048389642652e-06,
+      "loss": 0.508,
+      "step": 444
+    },
+    {
+      "epoch": 0.24058388898900704,
+      "grad_norm": 0.4705936908721924,
+      "learning_rate": 9.409606321741776e-06,
+      "loss": 0.4882,
+      "step": 445
+    },
+    {
+      "epoch": 0.24112452694179132,
+      "grad_norm": 0.5673285722732544,
+      "learning_rate": 9.405148536003527e-06,
+      "loss": 0.4983,
+      "step": 446
+    },
+    {
+      "epoch": 0.2416651648945756,
+      "grad_norm": 0.48440027236938477,
+      "learning_rate": 9.400675048317473e-06,
+      "loss": 0.5226,
+      "step": 447
+    },
+    {
+      "epoch": 0.24220580284735987,
+      "grad_norm": 0.5096259117126465,
+      "learning_rate": 9.396185874629158e-06,
+      "loss": 0.4979,
+      "step": 448
+    },
+    {
+      "epoch": 0.24274644080014418,
+      "grad_norm": 0.5506010055541992,
+      "learning_rate": 9.391681030940031e-06,
+      "loss": 0.4923,
+      "step": 449
+    },
+    {
+      "epoch": 0.24328707875292846,
+      "grad_norm": 0.5160098075866699,
+      "learning_rate": 9.387160533307398e-06,
+      "loss": 0.5082,
+      "step": 450
+    },
+    {
+      "epoch": 0.24382771670571274,
+      "grad_norm": 0.483447790145874,
+      "learning_rate": 9.382624397844363e-06,
+      "loss": 0.5007,
+      "step": 451
+    },
+    {
+      "epoch": 0.24436835465849702,
+      "grad_norm": 0.4803040325641632,
+      "learning_rate": 9.378072640719773e-06,
+      "loss": 0.4877,
+      "step": 452
+    },
+    {
+      "epoch": 0.24490899261128132,
+      "grad_norm": 0.5114092230796814,
+      "learning_rate": 9.373505278158152e-06,
+      "loss": 0.5008,
+      "step": 453
+    },
+    {
+      "epoch": 0.2454496305640656,
+      "grad_norm": 0.45896992087364197,
+      "learning_rate": 9.368922326439655e-06,
+      "loss": 0.4837,
+      "step": 454
+    },
+    {
+      "epoch": 0.24599026851684988,
+      "grad_norm": 0.5090027451515198,
+      "learning_rate": 9.364323801900002e-06,
+      "loss": 0.478,
+      "step": 455
+    },
+    {
+      "epoch": 0.24653090646963416,
+      "grad_norm": 0.441892147064209,
+      "learning_rate": 9.359709720930417e-06,
+      "loss": 0.511,
+      "step": 456
+    },
+    {
+      "epoch": 0.24707154442241847,
+      "grad_norm": 0.45809268951416016,
+      "learning_rate": 9.355080099977579e-06,
+      "loss": 0.5109,
+      "step": 457
+    },
+    {
+      "epoch": 0.24761218237520274,
+      "grad_norm": 0.45157548785209656,
+      "learning_rate": 9.350434955543557e-06,
+      "loss": 0.4929,
+      "step": 458
+    },
+    {
+      "epoch": 0.24815282032798702,
+      "grad_norm": 0.4648664891719818,
+      "learning_rate": 9.345774304185756e-06,
+      "loss": 0.4809,
+      "step": 459
+    },
+    {
+      "epoch": 0.2486934582807713,
+      "grad_norm": 0.5375930666923523,
+      "learning_rate": 9.341098162516848e-06,
+      "loss": 0.4998,
+      "step": 460
+    },
+    {
+      "epoch": 0.2492340962335556,
+      "grad_norm": 0.42750316858291626,
+      "learning_rate": 9.336406547204726e-06,
+      "loss": 0.4819,
+      "step": 461
+    },
+    {
+      "epoch": 0.2497747341863399,
+      "grad_norm": 0.5194140076637268,
+      "learning_rate": 9.331699474972434e-06,
+      "loss": 0.4908,
+      "step": 462
+    },
+    {
+      "epoch": 0.2503153721391242,
+      "grad_norm": 0.5086356401443481,
+      "learning_rate": 9.326976962598113e-06,
+      "loss": 0.4798,
+      "step": 463
+    },
+    {
+      "epoch": 0.25085601009190844,
+      "grad_norm": 0.4391303062438965,
+      "learning_rate": 9.322239026914938e-06,
+      "loss": 0.4989,
+      "step": 464
+    },
+    {
+      "epoch": 0.25139664804469275,
+      "grad_norm": 0.5237901210784912,
+      "learning_rate": 9.317485684811065e-06,
+      "loss": 0.5179,
+      "step": 465
+    },
+    {
+      "epoch": 0.251937285997477,
+      "grad_norm": 0.38452693819999695,
+      "learning_rate": 9.31271695322956e-06,
+      "loss": 0.4581,
+      "step": 466
+    },
+    {
+      "epoch": 0.2524779239502613,
+      "grad_norm": 0.4975660443305969,
+      "learning_rate": 9.307932849168341e-06,
+      "loss": 0.4567,
+      "step": 467
+    },
+    {
+      "epoch": 0.2530185619030456,
+      "grad_norm": 0.4287153482437134,
+      "learning_rate": 9.303133389680134e-06,
+      "loss": 0.5002,
+      "step": 468
+    },
+    {
+      "epoch": 0.25355919985582986,
+      "grad_norm": 0.5507779717445374,
+      "learning_rate": 9.298318591872381e-06,
+      "loss": 0.4925,
+      "step": 469
+    },
+    {
+      "epoch": 0.25409983780861417,
+      "grad_norm": 0.4791695475578308,
+      "learning_rate": 9.293488472907213e-06,
+      "loss": 0.5269,
+      "step": 470
+    },
+    {
+      "epoch": 0.2546404757613985,
+      "grad_norm": 0.44029316306114197,
+      "learning_rate": 9.288643050001362e-06,
+      "loss": 0.4994,
+      "step": 471
+    },
+    {
+      "epoch": 0.25518111371418273,
+      "grad_norm": 0.40624865889549255,
+      "learning_rate": 9.283782340426112e-06,
+      "loss": 0.4808,
+      "step": 472
+    },
+    {
+      "epoch": 0.25572175166696703,
+      "grad_norm": 0.44208696484565735,
+      "learning_rate": 9.278906361507238e-06,
+      "loss": 0.4944,
+      "step": 473
+    },
+    {
+      "epoch": 0.2562623896197513,
+      "grad_norm": 0.45168185234069824,
+      "learning_rate": 9.274015130624943e-06,
+      "loss": 0.4632,
+      "step": 474
+    },
+    {
+      "epoch": 0.2568030275725356,
+      "grad_norm": 0.4674747586250305,
+      "learning_rate": 9.26910866521379e-06,
+      "loss": 0.5001,
+      "step": 475
+    },
+    {
+      "epoch": 0.2573436655253199,
+      "grad_norm": 0.4525250792503357,
+      "learning_rate": 9.264186982762649e-06,
+      "loss": 0.4734,
+      "step": 476
+    },
+    {
+      "epoch": 0.25788430347810415,
+      "grad_norm": 0.48848944902420044,
+      "learning_rate": 9.25925010081463e-06,
+      "loss": 0.5054,
+      "step": 477
+    },
+    {
+      "epoch": 0.25842494143088846,
+      "grad_norm": 0.4272785484790802,
+      "learning_rate": 9.254298036967015e-06,
+      "loss": 0.4886,
+      "step": 478
+    },
+    {
+      "epoch": 0.25896557938367276,
+      "grad_norm": 0.6393560171127319,
+      "learning_rate": 9.249330808871213e-06,
+      "loss": 0.4915,
+      "step": 479
+    },
+    {
+      "epoch": 0.259506217336457,
+      "grad_norm": 0.4338541328907013,
+      "learning_rate": 9.244348434232676e-06,
+      "loss": 0.4974,
+      "step": 480
+    },
+    {
+      "epoch": 0.2600468552892413,
+      "grad_norm": 0.5929692387580872,
+      "learning_rate": 9.239350930810843e-06,
+      "loss": 0.4978,
+      "step": 481
+    },
+    {
+      "epoch": 0.26058749324202557,
+      "grad_norm": 0.523623526096344,
+      "learning_rate": 9.23433831641909e-06,
+      "loss": 0.5338,
+      "step": 482
+    },
+    {
+      "epoch": 0.2611281311948099,
+      "grad_norm": 0.45850881934165955,
+      "learning_rate": 9.229310608924643e-06,
+      "loss": 0.4928,
+      "step": 483
+    },
+    {
+      "epoch": 0.2616687691475942,
+      "grad_norm": 0.566050112247467,
+      "learning_rate": 9.224267826248536e-06,
+      "loss": 0.4998,
+      "step": 484
+    },
+    {
+      "epoch": 0.26220940710037843,
+      "grad_norm": 0.5277413129806519,
+      "learning_rate": 9.219209986365533e-06,
+      "loss": 0.4935,
+      "step": 485
+    },
+    {
+      "epoch": 0.26275004505316274,
+      "grad_norm": 0.5275083780288696,
+      "learning_rate": 9.21413710730407e-06,
+      "loss": 0.5195,
+      "step": 486
+    },
+    {
+      "epoch": 0.263290683005947,
+      "grad_norm": 0.4609350860118866,
+      "learning_rate": 9.20904920714619e-06,
+      "loss": 0.5127,
+      "step": 487
+    },
+    {
+      "epoch": 0.2638313209587313,
+      "grad_norm": 0.5389688611030579,
+      "learning_rate": 9.203946304027476e-06,
+      "loss": 0.501,
+      "step": 488
+    },
+    {
+      "epoch": 0.2643719589115156,
+      "grad_norm": 0.41184860467910767,
+      "learning_rate": 9.198828416136991e-06,
+      "loss": 0.4591,
+      "step": 489
+    },
+    {
+      "epoch": 0.26491259686429985,
+      "grad_norm": 0.5245027542114258,
+      "learning_rate": 9.193695561717207e-06,
+      "loss": 0.4657,
+      "step": 490
+    },
+    {
+      "epoch": 0.26545323481708416,
+      "grad_norm": 0.4098646938800812,
+      "learning_rate": 9.188547759063948e-06,
+      "loss": 0.5049,
+      "step": 491
+    },
+    {
+      "epoch": 0.26599387276986847,
+      "grad_norm": 0.5717266798019409,
+      "learning_rate": 9.183385026526317e-06,
+      "loss": 0.5137,
+      "step": 492
+    },
+    {
+      "epoch": 0.2665345107226527,
+      "grad_norm": 0.48556917905807495,
+      "learning_rate": 9.178207382506634e-06,
+      "loss": 0.4805,
+      "step": 493
+    },
+    {
+      "epoch": 0.267075148675437,
+      "grad_norm": 0.4661046862602234,
+      "learning_rate": 9.173014845460375e-06,
+      "loss": 0.4607,
+      "step": 494
+    },
+    {
+      "epoch": 0.2676157866282213,
+      "grad_norm": 0.4387494921684265,
+      "learning_rate": 9.167807433896091e-06,
+      "loss": 0.4992,
+      "step": 495
+    },
+    {
+      "epoch": 0.2681564245810056,
+      "grad_norm": 0.5127182006835938,
+      "learning_rate": 9.162585166375367e-06,
+      "loss": 0.5167,
+      "step": 496
+    },
+    {
+      "epoch": 0.2686970625337899,
+      "grad_norm": 0.42990031838417053,
+      "learning_rate": 9.157348061512728e-06,
+      "loss": 0.4858,
+      "step": 497
+    },
+    {
+      "epoch": 0.26923770048657414,
+      "grad_norm": 0.4679833650588989,
+      "learning_rate": 9.152096137975593e-06,
+      "loss": 0.5,
+      "step": 498
+    },
+    {
+      "epoch": 0.26977833843935844,
+      "grad_norm": 0.4706438183784485,
+      "learning_rate": 9.146829414484198e-06,
+      "loss": 0.4923,
+      "step": 499
+    },
+    {
+      "epoch": 0.27031897639214275,
+      "grad_norm": 0.4618656635284424,
+      "learning_rate": 9.14154790981154e-06,
+      "loss": 0.5015,
+      "step": 500
+    },
+    {
+      "epoch": 0.270859614344927,
+      "grad_norm": 0.473890095949173,
+      "learning_rate": 9.136251642783294e-06,
+      "loss": 0.5052,
+      "step": 501
+    },
+    {
+      "epoch": 0.2714002522977113,
+      "grad_norm": 0.4302096962928772,
+      "learning_rate": 9.130940632277757e-06,
+      "loss": 0.5216,
+      "step": 502
+    },
+    {
+      "epoch": 0.27194089025049556,
+      "grad_norm": 0.45936453342437744,
+      "learning_rate": 9.125614897225785e-06,
+      "loss": 0.5082,
+      "step": 503
+    },
+    {
+      "epoch": 0.27248152820327987,
+      "grad_norm": 0.42327070236206055,
+      "learning_rate": 9.120274456610708e-06,
+      "loss": 0.4788,
+      "step": 504
+    },
+    {
+      "epoch": 0.27302216615606417,
+      "grad_norm": 0.4344567656517029,
+      "learning_rate": 9.114919329468283e-06,
+      "loss": 0.4613,
+      "step": 505
+    },
+    {
+      "epoch": 0.2735628041088484,
+      "grad_norm": 0.44039013981819153,
+      "learning_rate": 9.10954953488661e-06,
+      "loss": 0.4888,
+      "step": 506
+    },
+    {
+      "epoch": 0.27410344206163273,
+      "grad_norm": 0.4592479169368744,
+      "learning_rate": 9.104165092006075e-06,
+      "loss": 0.4997,
+      "step": 507
+    },
+    {
+      "epoch": 0.27464408001441704,
+      "grad_norm": 0.4962833523750305,
+      "learning_rate": 9.098766020019273e-06,
+      "loss": 0.4599,
+      "step": 508
+    },
+    {
+      "epoch": 0.2751847179672013,
+      "grad_norm": 0.4496838450431824,
+      "learning_rate": 9.09335233817095e-06,
+      "loss": 0.4969,
+      "step": 509
+    },
+    {
+      "epoch": 0.2757253559199856,
+      "grad_norm": 0.44309988617897034,
+      "learning_rate": 9.08792406575792e-06,
+      "loss": 0.5136,
+      "step": 510
+    },
+    {
+      "epoch": 0.27626599387276984,
+      "grad_norm": 0.4315456748008728,
+      "learning_rate": 9.082481222129008e-06,
+      "loss": 0.4811,
+      "step": 511
+    },
+    {
+      "epoch": 0.27680663182555415,
+      "grad_norm": 0.45502516627311707,
+      "learning_rate": 9.07702382668498e-06,
+      "loss": 0.4764,
+      "step": 512
+    },
+    {
+      "epoch": 0.27734726977833846,
+      "grad_norm": 0.41209426522254944,
+      "learning_rate": 9.071551898878471e-06,
+      "loss": 0.468,
+      "step": 513
+    },
+    {
+      "epoch": 0.2778879077311227,
+      "grad_norm": 0.4675307273864746,
+      "learning_rate": 9.066065458213908e-06,
+      "loss": 0.4903,
+      "step": 514
+    },
+    {
+      "epoch": 0.278428545683907,
+      "grad_norm": 0.46809569001197815,
+      "learning_rate": 9.06056452424746e-06,
+      "loss": 0.5279,
+      "step": 515
+    },
+    {
+      "epoch": 0.2789691836366913,
+      "grad_norm": 0.5060775876045227,
+      "learning_rate": 9.055049116586951e-06,
+      "loss": 0.4866,
+      "step": 516
+    },
+    {
+      "epoch": 0.27950982158947557,
+      "grad_norm": 0.5153331756591797,
+      "learning_rate": 9.049519254891793e-06,
+      "loss": 0.5124,
+      "step": 517
+    },
+    {
+      "epoch": 0.2800504595422599,
+      "grad_norm": 0.4824049472808838,
+      "learning_rate": 9.04397495887292e-06,
+      "loss": 0.4711,
+      "step": 518
+    },
+    {
+      "epoch": 0.28059109749504413,
+      "grad_norm": 0.4388289153575897,
+      "learning_rate": 9.038416248292725e-06,
+      "loss": 0.4769,
+      "step": 519
+    },
+    {
+      "epoch": 0.28113173544782843,
+      "grad_norm": 0.5226829051971436,
+      "learning_rate": 9.03284314296497e-06,
+      "loss": 0.5135,
+      "step": 520
+    },
+    {
+      "epoch": 0.28167237340061274,
+      "grad_norm": 0.436776340007782,
+      "learning_rate": 9.02725566275473e-06,
+      "loss": 0.5041,
+      "step": 521
+    },
+    {
+      "epoch": 0.282213011353397,
+      "grad_norm": 0.5271431803703308,
+      "learning_rate": 9.021653827578322e-06,
+      "loss": 0.4836,
+      "step": 522
+    },
+    {
+      "epoch": 0.2827536493061813,
+      "grad_norm": 0.5045011043548584,
+      "learning_rate": 9.016037657403225e-06,
+      "loss": 0.487,
+      "step": 523
+    },
+    {
+      "epoch": 0.2832942872589656,
+      "grad_norm": 0.4657753109931946,
+      "learning_rate": 9.01040717224802e-06,
+      "loss": 0.4895,
+      "step": 524
+    },
+    {
+      "epoch": 0.28383492521174986,
+      "grad_norm": 0.4497890770435333,
+      "learning_rate": 9.004762392182307e-06,
+      "loss": 0.4913,
+      "step": 525
+    },
+    {
+      "epoch": 0.28437556316453416,
+      "grad_norm": 0.5092249512672424,
+      "learning_rate": 8.999103337326646e-06,
+      "loss": 0.5007,
+      "step": 526
+    },
+    {
+      "epoch": 0.2849162011173184,
+      "grad_norm": 0.48387986421585083,
+      "learning_rate": 8.993430027852476e-06,
+      "loss": 0.5121,
+      "step": 527
+    },
+    {
+      "epoch": 0.2854568390701027,
+      "grad_norm": 0.4973430037498474,
+      "learning_rate": 8.987742483982044e-06,
+      "loss": 0.4928,
+      "step": 528
+    },
+    {
+      "epoch": 0.285997477022887,
+      "grad_norm": 0.44857463240623474,
+      "learning_rate": 8.982040725988337e-06,
+      "loss": 0.5141,
+      "step": 529
+    },
+    {
+      "epoch": 0.2865381149756713,
+      "grad_norm": 0.48959702253341675,
+      "learning_rate": 8.976324774195005e-06,
+      "loss": 0.5099,
+      "step": 530
+    },
+    {
+      "epoch": 0.2870787529284556,
+      "grad_norm": 0.4419333040714264,
+      "learning_rate": 8.970594648976299e-06,
+      "loss": 0.4859,
+      "step": 531
+    },
+    {
+      "epoch": 0.2876193908812399,
+      "grad_norm": 0.43844538927078247,
+      "learning_rate": 8.964850370756978e-06,
+      "loss": 0.4931,
+      "step": 532
+    },
+    {
+      "epoch": 0.28816002883402414,
+      "grad_norm": 0.5034130811691284,
+      "learning_rate": 8.95909196001226e-06,
+      "loss": 0.4897,
+      "step": 533
+    },
+    {
+      "epoch": 0.28870066678680845,
+      "grad_norm": 0.4126366972923279,
+      "learning_rate": 8.953319437267731e-06,
+      "loss": 0.5,
+      "step": 534
+    },
+    {
+      "epoch": 0.2892413047395927,
+      "grad_norm": 0.4978686571121216,
+      "learning_rate": 8.947532823099284e-06,
+      "loss": 0.4674,
+      "step": 535
+    },
+    {
+      "epoch": 0.289781942692377,
+      "grad_norm": 0.49937352538108826,
+      "learning_rate": 8.941732138133032e-06,
+      "loss": 0.4687,
+      "step": 536
+    },
+    {
+      "epoch": 0.2903225806451613,
+      "grad_norm": 0.5455271601676941,
+      "learning_rate": 8.935917403045251e-06,
+      "loss": 0.4956,
+      "step": 537
+    },
+    {
+      "epoch": 0.29086321859794556,
+      "grad_norm": 0.46256354451179504,
+      "learning_rate": 8.930088638562296e-06,
+      "loss": 0.4607,
+      "step": 538
+    },
+    {
+      "epoch": 0.29140385655072987,
+      "grad_norm": 0.5940908193588257,
+      "learning_rate": 8.924245865460523e-06,
+      "loss": 0.4889,
+      "step": 539
+    },
+    {
+      "epoch": 0.2919444945035142,
+      "grad_norm": 0.4837222099304199,
+      "learning_rate": 8.918389104566232e-06,
+      "loss": 0.5081,
+      "step": 540
+    },
+    {
+      "epoch": 0.2924851324562984,
+      "grad_norm": 0.5419607162475586,
+      "learning_rate": 8.912518376755572e-06,
+      "loss": 0.4951,
+      "step": 541
+    },
+    {
+      "epoch": 0.29302577040908273,
+      "grad_norm": 0.47228190302848816,
+      "learning_rate": 8.906633702954482e-06,
+      "loss": 0.4769,
+      "step": 542
+    },
+    {
+      "epoch": 0.293566408361867,
+      "grad_norm": 0.4912036061286926,
+      "learning_rate": 8.900735104138605e-06,
+      "loss": 0.4911,
+      "step": 543
+    },
+    {
+      "epoch": 0.2941070463146513,
+      "grad_norm": 0.5826781392097473,
+      "learning_rate": 8.894822601333228e-06,
+      "loss": 0.4808,
+      "step": 544
+    },
+    {
+      "epoch": 0.2946476842674356,
+      "grad_norm": 0.5110523104667664,
+      "learning_rate": 8.888896215613192e-06,
+      "loss": 0.4882,
+      "step": 545
+    },
+    {
+      "epoch": 0.29518832222021985,
+      "grad_norm": 0.5374977588653564,
+      "learning_rate": 8.882955968102822e-06,
+      "loss": 0.488,
+      "step": 546
+    },
+    {
+      "epoch": 0.29572896017300415,
+      "grad_norm": 0.49603360891342163,
+      "learning_rate": 8.877001879975857e-06,
+      "loss": 0.4979,
+      "step": 547
+    },
+    {
+      "epoch": 0.29626959812578846,
+      "grad_norm": 0.5069612264633179,
+      "learning_rate": 8.87103397245537e-06,
+      "loss": 0.498,
+      "step": 548
+    },
+    {
+      "epoch": 0.2968102360785727,
+      "grad_norm": 0.5196537375450134,
+      "learning_rate": 8.865052266813686e-06,
+      "loss": 0.4882,
+      "step": 549
+    },
+    {
+      "epoch": 0.297350874031357,
+      "grad_norm": 0.4779670536518097,
+      "learning_rate": 8.85905678437232e-06,
+      "loss": 0.483,
+      "step": 550
+    },
+    {
+      "epoch": 0.29789151198414127,
+      "grad_norm": 0.5233485102653503,
+      "learning_rate": 8.853047546501893e-06,
+      "loss": 0.4952,
+      "step": 551
+    },
+    {
+      "epoch": 0.2984321499369256,
+      "grad_norm": 0.4517951011657715,
+      "learning_rate": 8.847024574622051e-06,
+      "loss": 0.5205,
+      "step": 552
+    },
+    {
+      "epoch": 0.2989727878897099,
+      "grad_norm": 0.5014324188232422,
+      "learning_rate": 8.840987890201404e-06,
+      "loss": 0.501,
+      "step": 553
+    },
+    {
+      "epoch": 0.29951342584249413,
+      "grad_norm": 0.45734500885009766,
+      "learning_rate": 8.834937514757428e-06,
+      "loss": 0.473,
+      "step": 554
+    },
+    {
+      "epoch": 0.30005406379527844,
+      "grad_norm": 0.4451104998588562,
+      "learning_rate": 8.828873469856408e-06,
+      "loss": 0.5119,
+      "step": 555
+    },
+    {
+      "epoch": 0.3005947017480627,
+      "grad_norm": 0.4849705398082733,
+      "learning_rate": 8.822795777113352e-06,
+      "loss": 0.499,
+      "step": 556
+    },
+    {
+      "epoch": 0.301135339700847,
+      "grad_norm": 0.5042188167572021,
+      "learning_rate": 8.816704458191913e-06,
+      "loss": 0.512,
+      "step": 557
+    },
+    {
+      "epoch": 0.3016759776536313,
+      "grad_norm": 0.4437592923641205,
+      "learning_rate": 8.810599534804315e-06,
+      "loss": 0.5103,
+      "step": 558
+    },
+    {
+      "epoch": 0.30221661560641555,
+      "grad_norm": 0.46505382657051086,
+      "learning_rate": 8.804481028711274e-06,
+      "loss": 0.4759,
+      "step": 559
+    },
+    {
+      "epoch": 0.30275725355919986,
+      "grad_norm": 0.5045806169509888,
+      "learning_rate": 8.798348961721925e-06,
+      "loss": 0.4896,
+      "step": 560
+    },
+    {
+      "epoch": 0.30329789151198416,
+      "grad_norm": 0.4548962116241455,
+      "learning_rate": 8.792203355693731e-06,
+      "loss": 0.4548,
+      "step": 561
+    },
+    {
+      "epoch": 0.3038385294647684,
+      "grad_norm": 0.4087465703487396,
+      "learning_rate": 8.786044232532423e-06,
+      "loss": 0.4975,
+      "step": 562
+    },
+    {
+      "epoch": 0.3043791674175527,
+      "grad_norm": 0.5508294105529785,
+      "learning_rate": 8.77987161419191e-06,
+      "loss": 0.4907,
+      "step": 563
+    },
+    {
+      "epoch": 0.30491980537033697,
+      "grad_norm": 0.53659987449646,
+      "learning_rate": 8.773685522674205e-06,
+      "loss": 0.494,
+      "step": 564
+    },
+    {
+      "epoch": 0.3054604433231213,
+      "grad_norm": 0.5059062242507935,
+      "learning_rate": 8.767485980029342e-06,
+      "loss": 0.4904,
+      "step": 565
+    },
+    {
+      "epoch": 0.3060010812759056,
+      "grad_norm": 0.462995320558548,
+      "learning_rate": 8.761273008355306e-06,
+      "loss": 0.483,
+      "step": 566
+    },
+    {
+      "epoch": 0.30654171922868984,
+      "grad_norm": 0.6344414353370667,
+      "learning_rate": 8.755046629797944e-06,
+      "loss": 0.5041,
+      "step": 567
+    },
+    {
+      "epoch": 0.30708235718147414,
+      "grad_norm": 0.455564022064209,
+      "learning_rate": 8.748806866550895e-06,
+      "loss": 0.4515,
+      "step": 568
+    },
+    {
+      "epoch": 0.30762299513425845,
+      "grad_norm": 0.5680974721908569,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 0.5019,
+      "step": 569
+    },
+    {
+      "epoch": 0.3081636330870427,
+      "grad_norm": 0.500893771648407,
+      "learning_rate": 8.736287275000755e-06,
+      "loss": 0.5346,
+      "step": 570
+    },
+    {
+      "epoch": 0.308704271039827,
+      "grad_norm": 0.43272846937179565,
+      "learning_rate": 8.730007491323167e-06,
+      "loss": 0.4932,
+      "step": 571
+    },
+    {
+      "epoch": 0.30924490899261126,
+      "grad_norm": 0.4665529727935791,
+      "learning_rate": 8.723714412206741e-06,
+      "loss": 0.5207,
+      "step": 572
+    },
+    {
+      "epoch": 0.30978554694539556,
+      "grad_norm": 0.5202204585075378,
+      "learning_rate": 8.717408060082865e-06,
+      "loss": 0.5138,
+      "step": 573
+    },
+    {
+      "epoch": 0.31032618489817987,
+      "grad_norm": 0.4143356680870056,
+      "learning_rate": 8.711088457430239e-06,
+      "loss": 0.4977,
+      "step": 574
+    },
+    {
+      "epoch": 0.3108668228509641,
+      "grad_norm": 0.4690699875354767,
+      "learning_rate": 8.704755626774796e-06,
+      "loss": 0.4854,
+      "step": 575
+    },
+    {
+      "epoch": 0.3114074608037484,
+      "grad_norm": 0.5123147964477539,
+      "learning_rate": 8.698409590689616e-06,
+      "loss": 0.493,
+      "step": 576
+    },
+    {
+      "epoch": 0.31194809875653273,
+      "grad_norm": 0.38649624586105347,
+      "learning_rate": 8.692050371794849e-06,
+      "loss": 0.4737,
+      "step": 577
+    },
+    {
+      "epoch": 0.312488736709317,
+      "grad_norm": 0.4339773952960968,
+      "learning_rate": 8.685677992757637e-06,
+      "loss": 0.485,
+      "step": 578
+    },
+    {
+      "epoch": 0.3130293746621013,
+      "grad_norm": 0.5341055989265442,
+      "learning_rate": 8.67929247629203e-06,
+      "loss": 0.4933,
+      "step": 579
+    },
+    {
+      "epoch": 0.31357001261488554,
+      "grad_norm": 0.38680076599121094,
+      "learning_rate": 8.672893845158908e-06,
+      "loss": 0.5145,
+      "step": 580
+    },
+    {
+      "epoch": 0.31411065056766985,
+      "grad_norm": 0.5113509893417358,
+      "learning_rate": 8.66648212216589e-06,
+      "loss": 0.4792,
+      "step": 581
+    },
+    {
+      "epoch": 0.31465128852045415,
+      "grad_norm": 0.41235169768333435,
+      "learning_rate": 8.660057330167267e-06,
+      "loss": 0.4734,
+      "step": 582
+    },
+    {
+      "epoch": 0.3151919264732384,
+      "grad_norm": 0.4803949296474457,
+      "learning_rate": 8.653619492063916e-06,
+      "loss": 0.4898,
+      "step": 583
+    },
+    {
+      "epoch": 0.3157325644260227,
+      "grad_norm": 0.4755302965641022,
+      "learning_rate": 8.647168630803208e-06,
+      "loss": 0.4968,
+      "step": 584
+    },
+    {
+      "epoch": 0.316273202378807,
+      "grad_norm": 0.4384644329547882,
+      "learning_rate": 8.640704769378943e-06,
+      "loss": 0.4922,
+      "step": 585
+    },
+    {
+      "epoch": 0.31681384033159127,
+      "grad_norm": 0.5590741038322449,
+      "learning_rate": 8.634227930831252e-06,
+      "loss": 0.4828,
+      "step": 586
+    },
+    {
+      "epoch": 0.3173544782843756,
+      "grad_norm": 0.45347392559051514,
+      "learning_rate": 8.627738138246529e-06,
+      "loss": 0.468,
+      "step": 587
+    },
+    {
+      "epoch": 0.3178951162371598,
+      "grad_norm": 0.4747249186038971,
+      "learning_rate": 8.621235414757337e-06,
+      "loss": 0.4905,
+      "step": 588
+    },
+    {
+      "epoch": 0.31843575418994413,
+      "grad_norm": 0.4371531903743744,
+      "learning_rate": 8.61471978354233e-06,
+      "loss": 0.4806,
+      "step": 589
+    },
+    {
+      "epoch": 0.31897639214272844,
+      "grad_norm": 0.5220611095428467,
+      "learning_rate": 8.608191267826179e-06,
+      "loss": 0.4872,
+      "step": 590
+    },
+    {
+      "epoch": 0.3195170300955127,
+      "grad_norm": 0.46987447142601013,
+      "learning_rate": 8.60164989087947e-06,
+      "loss": 0.523,
+      "step": 591
+    },
+    {
+      "epoch": 0.320057668048297,
+      "grad_norm": 0.5529326796531677,
+      "learning_rate": 8.595095676018645e-06,
+      "loss": 0.5054,
+      "step": 592
+    },
+    {
+      "epoch": 0.3205983060010813,
+      "grad_norm": 0.4231288731098175,
+      "learning_rate": 8.588528646605893e-06,
+      "loss": 0.4985,
+      "step": 593
+    },
+    {
+      "epoch": 0.32113894395386555,
+      "grad_norm": 0.5058132410049438,
+      "learning_rate": 8.581948826049086e-06,
+      "loss": 0.4989,
+      "step": 594
+    },
+    {
+      "epoch": 0.32167958190664986,
+      "grad_norm": 0.43005961179733276,
+      "learning_rate": 8.575356237801695e-06,
+      "loss": 0.4898,
+      "step": 595
+    },
+    {
+      "epoch": 0.3222202198594341,
+      "grad_norm": 0.4978388547897339,
+      "learning_rate": 8.56875090536269e-06,
+      "loss": 0.4666,
+      "step": 596
+    },
+    {
+      "epoch": 0.3227608578122184,
+      "grad_norm": 0.5346764326095581,
+      "learning_rate": 8.562132852276474e-06,
+      "loss": 0.4828,
+      "step": 597
+    },
+    {
+      "epoch": 0.3233014957650027,
+      "grad_norm": 0.5223882794380188,
+      "learning_rate": 8.555502102132792e-06,
+      "loss": 0.4619,
+      "step": 598
+    },
+    {
+      "epoch": 0.323842133717787,
+      "grad_norm": 0.5179974436759949,
+      "learning_rate": 8.548858678566643e-06,
+      "loss": 0.46,
+      "step": 599
+    },
+    {
+      "epoch": 0.3243827716705713,
+      "grad_norm": 0.44731220602989197,
+      "learning_rate": 8.542202605258204e-06,
+      "loss": 0.4712,
+      "step": 600
+    },
+    {
+      "epoch": 0.3249234096233556,
+      "grad_norm": 0.5143573880195618,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.4841,
+      "step": 601
+    },
+    {
+      "epoch": 0.32546404757613984,
+      "grad_norm": 0.40094849467277527,
+      "learning_rate": 8.528852604360518e-06,
+      "loss": 0.4825,
+      "step": 602
+    },
+    {
+      "epoch": 0.32600468552892414,
+      "grad_norm": 0.4948832094669342,
+      "learning_rate": 8.52215872435673e-06,
+      "loss": 0.5066,
+      "step": 603
+    },
+    {
+      "epoch": 0.3265453234817084,
+      "grad_norm": 0.43757307529449463,
+      "learning_rate": 8.515452289781403e-06,
+      "loss": 0.4831,
+      "step": 604
+    },
+    {
+      "epoch": 0.3270859614344927,
+      "grad_norm": 0.4551491141319275,
+      "learning_rate": 8.50873332453931e-06,
+      "loss": 0.4628,
+      "step": 605
+    },
+    {
+      "epoch": 0.327626599387277,
+      "grad_norm": 0.45467811822891235,
+      "learning_rate": 8.50200185257989e-06,
+      "loss": 0.4813,
+      "step": 606
+    },
+    {
+      "epoch": 0.32816723734006126,
+      "grad_norm": 0.4058467745780945,
+      "learning_rate": 8.495257897897166e-06,
+      "loss": 0.4756,
+      "step": 607
+    },
+    {
+      "epoch": 0.32870787529284556,
+      "grad_norm": 0.4405519664287567,
+      "learning_rate": 8.48850148452965e-06,
+      "loss": 0.4964,
+      "step": 608
+    },
+    {
+      "epoch": 0.32924851324562987,
+      "grad_norm": 0.39016538858413696,
+      "learning_rate": 8.481732636560266e-06,
+      "loss": 0.4736,
+      "step": 609
+    },
+    {
+      "epoch": 0.3297891511984141,
+      "grad_norm": 0.4444190561771393,
+      "learning_rate": 8.474951378116253e-06,
+      "loss": 0.4904,
+      "step": 610
+    },
+    {
+      "epoch": 0.3303297891511984,
+      "grad_norm": 0.45072081685066223,
+      "learning_rate": 8.468157733369102e-06,
+      "loss": 0.4906,
+      "step": 611
+    },
+    {
+      "epoch": 0.3308704271039827,
+      "grad_norm": 0.40805482864379883,
+      "learning_rate": 8.461351726534438e-06,
+      "loss": 0.4781,
+      "step": 612
+    },
+    {
+      "epoch": 0.331411065056767,
+      "grad_norm": 0.4450664520263672,
+      "learning_rate": 8.454533381871957e-06,
+      "loss": 0.4985,
+      "step": 613
+    },
+    {
+      "epoch": 0.3319517030095513,
+      "grad_norm": 0.465251088142395,
+      "learning_rate": 8.447702723685335e-06,
+      "loss": 0.4911,
+      "step": 614
+    },
+    {
+      "epoch": 0.33249234096233554,
+      "grad_norm": 0.49976640939712524,
+      "learning_rate": 8.440859776322137e-06,
+      "loss": 0.4977,
+      "step": 615
+    },
+    {
+      "epoch": 0.33303297891511985,
+      "grad_norm": 0.5387666821479797,
+      "learning_rate": 8.43400456417373e-06,
+      "loss": 0.4938,
+      "step": 616
+    },
+    {
+      "epoch": 0.3335736168679041,
+      "grad_norm": 0.3886103630065918,
+      "learning_rate": 8.4271371116752e-06,
+      "loss": 0.4753,
+      "step": 617
+    },
+    {
+      "epoch": 0.3341142548206884,
+      "grad_norm": 0.4702761173248291,
+      "learning_rate": 8.420257443305264e-06,
+      "loss": 0.4991,
+      "step": 618
+    },
+    {
+      "epoch": 0.3346548927734727,
+      "grad_norm": 0.4505619406700134,
+      "learning_rate": 8.41336558358618e-06,
+      "loss": 0.4913,
+      "step": 619
+    },
+    {
+      "epoch": 0.33519553072625696,
+      "grad_norm": 0.4063532054424286,
+      "learning_rate": 8.406461557083666e-06,
+      "loss": 0.4866,
+      "step": 620
+    },
+    {
+      "epoch": 0.33573616867904127,
+      "grad_norm": 0.44660359621047974,
+      "learning_rate": 8.399545388406798e-06,
+      "loss": 0.5116,
+      "step": 621
+    },
+    {
+      "epoch": 0.3362768066318256,
+      "grad_norm": 0.41896873712539673,
+      "learning_rate": 8.392617102207945e-06,
+      "loss": 0.4966,
+      "step": 622
+    },
+    {
+      "epoch": 0.3368174445846098,
+      "grad_norm": 0.44783973693847656,
+      "learning_rate": 8.38567672318266e-06,
+      "loss": 0.4843,
+      "step": 623
+    },
+    {
+      "epoch": 0.33735808253739413,
+      "grad_norm": 0.4662441611289978,
+      "learning_rate": 8.3787242760696e-06,
+      "loss": 0.4646,
+      "step": 624
+    },
+    {
+      "epoch": 0.3378987204901784,
+      "grad_norm": 0.44338029623031616,
+      "learning_rate": 8.371759785650444e-06,
+      "loss": 0.4847,
+      "step": 625
+    },
+    {
+      "epoch": 0.3384393584429627,
+      "grad_norm": 0.4882574677467346,
+      "learning_rate": 8.364783276749794e-06,
+      "loss": 0.489,
+      "step": 626
+    },
+    {
+      "epoch": 0.338979996395747,
+      "grad_norm": 0.4160021245479584,
+      "learning_rate": 8.357794774235094e-06,
+      "loss": 0.504,
+      "step": 627
+    },
+    {
+      "epoch": 0.33952063434853125,
+      "grad_norm": 0.40824034810066223,
+      "learning_rate": 8.350794303016533e-06,
+      "loss": 0.4683,
+      "step": 628
+    },
+    {
+      "epoch": 0.34006127230131555,
+      "grad_norm": 0.5091646909713745,
+      "learning_rate": 8.343781888046971e-06,
+      "loss": 0.5207,
+      "step": 629
+    },
+    {
+      "epoch": 0.34060191025409986,
+      "grad_norm": 0.43365058302879333,
+      "learning_rate": 8.336757554321832e-06,
+      "loss": 0.5062,
+      "step": 630
+    },
+    {
+      "epoch": 0.3411425482068841,
+      "grad_norm": 0.4349713623523712,
+      "learning_rate": 8.329721326879032e-06,
+      "loss": 0.4745,
+      "step": 631
+    },
+    {
+      "epoch": 0.3416831861596684,
+      "grad_norm": 0.4307449460029602,
+      "learning_rate": 8.322673230798877e-06,
+      "loss": 0.4845,
+      "step": 632
+    },
+    {
+      "epoch": 0.34222382411245267,
+      "grad_norm": 0.4474685788154602,
+      "learning_rate": 8.315613291203977e-06,
+      "loss": 0.4928,
+      "step": 633
+    },
+    {
+      "epoch": 0.342764462065237,
+      "grad_norm": 0.4700278043746948,
+      "learning_rate": 8.30854153325916e-06,
+      "loss": 0.5178,
+      "step": 634
+    },
+    {
+      "epoch": 0.3433051000180213,
+      "grad_norm": 0.42883479595184326,
+      "learning_rate": 8.30145798217138e-06,
+      "loss": 0.4828,
+      "step": 635
+    },
+    {
+      "epoch": 0.34384573797080553,
+      "grad_norm": 0.45476993918418884,
+      "learning_rate": 8.294362663189626e-06,
+      "loss": 0.4882,
+      "step": 636
+    },
+    {
+      "epoch": 0.34438637592358984,
+      "grad_norm": 0.40888112783432007,
+      "learning_rate": 8.287255601604834e-06,
+      "loss": 0.4728,
+      "step": 637
+    },
+    {
+      "epoch": 0.34492701387637414,
+      "grad_norm": 0.41027674078941345,
+      "learning_rate": 8.280136822749796e-06,
+      "loss": 0.4612,
+      "step": 638
+    },
+    {
+      "epoch": 0.3454676518291584,
+      "grad_norm": 0.3990921080112457,
+      "learning_rate": 8.27300635199907e-06,
+      "loss": 0.4977,
+      "step": 639
+    },
+    {
+      "epoch": 0.3460082897819427,
+      "grad_norm": 0.39647355675697327,
+      "learning_rate": 8.265864214768883e-06,
+      "loss": 0.4817,
+      "step": 640
+    },
+    {
+      "epoch": 0.34654892773472695,
+      "grad_norm": 0.40276244282722473,
+      "learning_rate": 8.25871043651706e-06,
+      "loss": 0.4652,
+      "step": 641
+    },
+    {
+      "epoch": 0.34708956568751126,
+      "grad_norm": 0.4234772026538849,
+      "learning_rate": 8.25154504274291e-06,
+      "loss": 0.4882,
+      "step": 642
+    },
+    {
+      "epoch": 0.34763020364029557,
+      "grad_norm": 0.49208852648735046,
+      "learning_rate": 8.244368058987145e-06,
+      "loss": 0.5014,
+      "step": 643
+    },
+    {
+      "epoch": 0.3481708415930798,
+      "grad_norm": 0.3916458189487457,
+      "learning_rate": 8.237179510831792e-06,
+      "loss": 0.4678,
+      "step": 644
+    },
+    {
+      "epoch": 0.3487114795458641,
+      "grad_norm": 0.43457022309303284,
+      "learning_rate": 8.229979423900095e-06,
+      "loss": 0.4866,
+      "step": 645
+    },
+    {
+      "epoch": 0.34925211749864843,
+      "grad_norm": 0.44561585783958435,
+      "learning_rate": 8.222767823856435e-06,
+      "loss": 0.491,
+      "step": 646
+    },
+    {
+      "epoch": 0.3497927554514327,
+      "grad_norm": 0.3655422031879425,
+      "learning_rate": 8.215544736406223e-06,
+      "loss": 0.4497,
+      "step": 647
+    },
+    {
+      "epoch": 0.350333393404217,
+      "grad_norm": 0.47153764963150024,
+      "learning_rate": 8.20831018729582e-06,
+      "loss": 0.5342,
+      "step": 648
+    },
+    {
+      "epoch": 0.35087403135700124,
+      "grad_norm": 0.4188770651817322,
+      "learning_rate": 8.20106420231244e-06,
+      "loss": 0.4742,
+      "step": 649
+    },
+    {
+      "epoch": 0.35141466930978554,
+      "grad_norm": 0.45185157656669617,
+      "learning_rate": 8.193806807284064e-06,
+      "loss": 0.4623,
+      "step": 650
+    },
+    {
+      "epoch": 0.35195530726256985,
+      "grad_norm": 0.49852705001831055,
+      "learning_rate": 8.186538028079338e-06,
+      "loss": 0.4728,
+      "step": 651
+    },
+    {
+      "epoch": 0.3524959452153541,
+      "grad_norm": 0.46935099363327026,
+      "learning_rate": 8.179257890607489e-06,
+      "loss": 0.4709,
+      "step": 652
+    },
+    {
+      "epoch": 0.3530365831681384,
+      "grad_norm": 0.4840806722640991,
+      "learning_rate": 8.171966420818227e-06,
+      "loss": 0.4988,
+      "step": 653
+    },
+    {
+      "epoch": 0.3535772211209227,
+      "grad_norm": 0.45912444591522217,
+      "learning_rate": 8.164663644701662e-06,
+      "loss": 0.4933,
+      "step": 654
+    },
+    {
+      "epoch": 0.35411785907370696,
+      "grad_norm": 0.4942997694015503,
+      "learning_rate": 8.157349588288202e-06,
+      "loss": 0.4731,
+      "step": 655
+    },
+    {
+      "epoch": 0.35465849702649127,
+      "grad_norm": 0.4257151782512665,
+      "learning_rate": 8.150024277648458e-06,
+      "loss": 0.487,
+      "step": 656
+    },
+    {
+      "epoch": 0.3551991349792755,
+      "grad_norm": 0.544456958770752,
+      "learning_rate": 8.142687738893161e-06,
+      "loss": 0.5117,
+      "step": 657
+    },
+    {
+      "epoch": 0.35573977293205983,
+      "grad_norm": 0.5224266648292542,
+      "learning_rate": 8.135339998173064e-06,
+      "loss": 0.4894,
+      "step": 658
+    },
+    {
+      "epoch": 0.35628041088484413,
+      "grad_norm": 0.46919405460357666,
+      "learning_rate": 8.12798108167885e-06,
+      "loss": 0.4719,
+      "step": 659
+    },
+    {
+      "epoch": 0.3568210488376284,
+      "grad_norm": 0.7049233913421631,
+      "learning_rate": 8.120611015641036e-06,
+      "loss": 0.5216,
+      "step": 660
+    },
+    {
+      "epoch": 0.3573616867904127,
+      "grad_norm": 0.4920409619808197,
+      "learning_rate": 8.113229826329876e-06,
+      "loss": 0.4922,
+      "step": 661
+    },
+    {
+      "epoch": 0.357902324743197,
+      "grad_norm": 0.5132108330726624,
+      "learning_rate": 8.105837540055284e-06,
+      "loss": 0.4894,
+      "step": 662
+    },
+    {
+      "epoch": 0.35844296269598125,
+      "grad_norm": 0.5030608177185059,
+      "learning_rate": 8.098434183166716e-06,
+      "loss": 0.4765,
+      "step": 663
+    },
+    {
+      "epoch": 0.35898360064876556,
+      "grad_norm": 0.5627034306526184,
+      "learning_rate": 8.091019782053097e-06,
+      "loss": 0.4862,
+      "step": 664
+    },
+    {
+      "epoch": 0.3595242386015498,
+      "grad_norm": 0.5231481790542603,
+      "learning_rate": 8.083594363142717e-06,
+      "loss": 0.5224,
+      "step": 665
+    },
+    {
+      "epoch": 0.3600648765543341,
+      "grad_norm": 0.5793126225471497,
+      "learning_rate": 8.076157952903134e-06,
+      "loss": 0.5063,
+      "step": 666
+    },
+    {
+      "epoch": 0.3606055145071184,
+      "grad_norm": 0.5222432017326355,
+      "learning_rate": 8.068710577841093e-06,
+      "loss": 0.4578,
+      "step": 667
+    },
+    {
+      "epoch": 0.36114615245990267,
+      "grad_norm": 0.4852503836154938,
+      "learning_rate": 8.061252264502415e-06,
+      "loss": 0.4936,
+      "step": 668
+    },
+    {
+      "epoch": 0.361686790412687,
+      "grad_norm": 0.4867151379585266,
+      "learning_rate": 8.053783039471909e-06,
+      "loss": 0.4957,
+      "step": 669
+    },
+    {
+      "epoch": 0.3622274283654713,
+      "grad_norm": 0.45587924122810364,
+      "learning_rate": 8.046302929373286e-06,
+      "loss": 0.4655,
+      "step": 670
+    },
+    {
+      "epoch": 0.36276806631825553,
+      "grad_norm": 0.47155067324638367,
+      "learning_rate": 8.038811960869051e-06,
+      "loss": 0.487,
+      "step": 671
+    },
+    {
+      "epoch": 0.36330870427103984,
+      "grad_norm": 0.4418538510799408,
+      "learning_rate": 8.031310160660411e-06,
+      "loss": 0.4724,
+      "step": 672
+    },
+    {
+      "epoch": 0.3638493422238241,
+      "grad_norm": 0.404681533575058,
+      "learning_rate": 8.023797555487188e-06,
+      "loss": 0.4892,
+      "step": 673
+    },
+    {
+      "epoch": 0.3643899801766084,
+      "grad_norm": 0.45918798446655273,
+      "learning_rate": 8.016274172127715e-06,
+      "loss": 0.5012,
+      "step": 674
+    },
+    {
+      "epoch": 0.3649306181293927,
+      "grad_norm": 0.4333856701850891,
+      "learning_rate": 8.008740037398742e-06,
+      "loss": 0.4868,
+      "step": 675
+    },
+    {
+      "epoch": 0.36547125608217695,
+      "grad_norm": 0.3893619477748871,
+      "learning_rate": 8.001195178155344e-06,
+      "loss": 0.4684,
+      "step": 676
+    },
+    {
+      "epoch": 0.36601189403496126,
+      "grad_norm": 0.434662789106369,
+      "learning_rate": 7.99363962129082e-06,
+      "loss": 0.4814,
+      "step": 677
+    },
+    {
+      "epoch": 0.36655253198774557,
+      "grad_norm": 0.4455735981464386,
+      "learning_rate": 7.986073393736607e-06,
+      "loss": 0.4856,
+      "step": 678
+    },
+    {
+      "epoch": 0.3670931699405298,
+      "grad_norm": 0.39852529764175415,
+      "learning_rate": 7.978496522462167e-06,
+      "loss": 0.4811,
+      "step": 679
+    },
+    {
+      "epoch": 0.3676338078933141,
+      "grad_norm": 0.42947253584861755,
+      "learning_rate": 7.97090903447491e-06,
+      "loss": 0.4687,
+      "step": 680
+    },
+    {
+      "epoch": 0.3681744458460984,
+      "grad_norm": 0.4797091782093048,
+      "learning_rate": 7.963310956820085e-06,
+      "loss": 0.4892,
+      "step": 681
+    },
+    {
+      "epoch": 0.3687150837988827,
+      "grad_norm": 0.449841171503067,
+      "learning_rate": 7.955702316580686e-06,
+      "loss": 0.4918,
+      "step": 682
+    },
+    {
+      "epoch": 0.369255721751667,
+      "grad_norm": 0.4748377203941345,
+      "learning_rate": 7.94808314087736e-06,
+      "loss": 0.5122,
+      "step": 683
+    },
+    {
+      "epoch": 0.36979635970445124,
+      "grad_norm": 0.45039764046669006,
+      "learning_rate": 7.940453456868304e-06,
+      "loss": 0.4818,
+      "step": 684
+    },
+    {
+      "epoch": 0.37033699765723554,
+      "grad_norm": 0.44947609305381775,
+      "learning_rate": 7.932813291749177e-06,
+      "loss": 0.4906,
+      "step": 685
+    },
+    {
+      "epoch": 0.3708776356100198,
+      "grad_norm": 0.46139925718307495,
+      "learning_rate": 7.925162672752989e-06,
+      "loss": 0.4888,
+      "step": 686
+    },
+    {
+      "epoch": 0.3714182735628041,
+      "grad_norm": 0.49717041850090027,
+      "learning_rate": 7.917501627150019e-06,
+      "loss": 0.4981,
+      "step": 687
+    },
+    {
+      "epoch": 0.3719589115155884,
+      "grad_norm": 0.5299156904220581,
+      "learning_rate": 7.90983018224771e-06,
+      "loss": 0.4948,
+      "step": 688
+    },
+    {
+      "epoch": 0.37249954946837266,
+      "grad_norm": 0.4835863411426544,
+      "learning_rate": 7.902148365390567e-06,
+      "loss": 0.4915,
+      "step": 689
+    },
+    {
+      "epoch": 0.37304018742115697,
+      "grad_norm": 0.5481423735618591,
+      "learning_rate": 7.894456203960075e-06,
+      "loss": 0.4889,
+      "step": 690
+    },
+    {
+      "epoch": 0.37358082537394127,
+      "grad_norm": 0.419506311416626,
+      "learning_rate": 7.886753725374586e-06,
+      "loss": 0.4839,
+      "step": 691
+    },
+    {
+      "epoch": 0.3741214633267255,
+      "grad_norm": 0.429317831993103,
+      "learning_rate": 7.879040957089229e-06,
+      "loss": 0.4709,
+      "step": 692
+    },
+    {
+      "epoch": 0.37466210127950983,
+      "grad_norm": 0.47440069913864136,
+      "learning_rate": 7.871317926595804e-06,
+      "loss": 0.4777,
+      "step": 693
+    },
+    {
+      "epoch": 0.3752027392322941,
+      "grad_norm": 0.5026919841766357,
+      "learning_rate": 7.8635846614227e-06,
+      "loss": 0.4749,
+      "step": 694
+    },
+    {
+      "epoch": 0.3757433771850784,
+      "grad_norm": 0.39894619584083557,
+      "learning_rate": 7.855841189134784e-06,
+      "loss": 0.4627,
+      "step": 695
+    },
+    {
+      "epoch": 0.3762840151378627,
+      "grad_norm": 0.4291825294494629,
+      "learning_rate": 7.848087537333298e-06,
+      "loss": 0.4667,
+      "step": 696
+    },
+    {
+      "epoch": 0.37682465309064694,
+      "grad_norm": 0.3873690068721771,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.4774,
+      "step": 697
+    },
+    {
+      "epoch": 0.37736529104343125,
+      "grad_norm": 0.3892592489719391,
+      "learning_rate": 7.832549805775945e-06,
+      "loss": 0.5081,
+      "step": 698
+    },
+    {
+      "epoch": 0.37790592899621556,
+      "grad_norm": 0.40017616748809814,
+      "learning_rate": 7.8247657814036e-06,
+      "loss": 0.4783,
+      "step": 699
+    },
+    {
+      "epoch": 0.3784465669489998,
+      "grad_norm": 0.41424062848091125,
+      "learning_rate": 7.81697168828454e-06,
+      "loss": 0.4904,
+      "step": 700
+    },
+    {
+      "epoch": 0.3789872049017841,
+      "grad_norm": 0.42307886481285095,
+      "learning_rate": 7.809167554200446e-06,
+      "loss": 0.4948,
+      "step": 701
+    },
+    {
+      "epoch": 0.37952784285456836,
+      "grad_norm": 0.3783946931362152,
+      "learning_rate": 7.801353406968795e-06,
+      "loss": 0.4778,
+      "step": 702
+    },
+    {
+      "epoch": 0.38006848080735267,
+      "grad_norm": 0.39104709029197693,
+      "learning_rate": 7.793529274442753e-06,
+      "loss": 0.4661,
+      "step": 703
+    },
+    {
+      "epoch": 0.380609118760137,
+      "grad_norm": 0.4255848824977875,
+      "learning_rate": 7.785695184511074e-06,
+      "loss": 0.4629,
+      "step": 704
+    },
+    {
+      "epoch": 0.38114975671292123,
+      "grad_norm": 0.37515324354171753,
+      "learning_rate": 7.777851165098012e-06,
+      "loss": 0.4744,
+      "step": 705
+    },
+    {
+      "epoch": 0.38169039466570553,
+      "grad_norm": 0.43768253922462463,
+      "learning_rate": 7.769997244163209e-06,
+      "loss": 0.4941,
+      "step": 706
+    },
+    {
+      "epoch": 0.38223103261848984,
+      "grad_norm": 0.4131484925746918,
+      "learning_rate": 7.762133449701603e-06,
+      "loss": 0.4912,
+      "step": 707
+    },
+    {
+      "epoch": 0.3827716705712741,
+      "grad_norm": 0.42033693194389343,
+      "learning_rate": 7.754259809743325e-06,
+      "loss": 0.4606,
+      "step": 708
+    },
+    {
+      "epoch": 0.3833123085240584,
+      "grad_norm": 0.3850057125091553,
+      "learning_rate": 7.746376352353599e-06,
+      "loss": 0.4715,
+      "step": 709
+    },
+    {
+      "epoch": 0.38385294647684265,
+      "grad_norm": 0.4940873086452484,
+      "learning_rate": 7.738483105632644e-06,
+      "loss": 0.5257,
+      "step": 710
+    },
+    {
+      "epoch": 0.38439358442962696,
+      "grad_norm": 0.4676145315170288,
+      "learning_rate": 7.730580097715575e-06,
+      "loss": 0.4917,
+      "step": 711
+    },
+    {
+      "epoch": 0.38493422238241126,
+      "grad_norm": 0.41766825318336487,
+      "learning_rate": 7.722667356772291e-06,
+      "loss": 0.4738,
+      "step": 712
+    },
+    {
+      "epoch": 0.3854748603351955,
+      "grad_norm": 0.4553625285625458,
+      "learning_rate": 7.714744911007395e-06,
+      "loss": 0.4839,
+      "step": 713
+    },
+    {
+      "epoch": 0.3860154982879798,
+      "grad_norm": 0.3935423195362091,
+      "learning_rate": 7.706812788660075e-06,
+      "loss": 0.4889,
+      "step": 714
+    },
+    {
+      "epoch": 0.3865561362407641,
+      "grad_norm": 0.42542070150375366,
+      "learning_rate": 7.698871018004016e-06,
+      "loss": 0.5031,
+      "step": 715
+    },
+    {
+      "epoch": 0.3870967741935484,
+      "grad_norm": 0.48952916264533997,
+      "learning_rate": 7.690919627347292e-06,
+      "loss": 0.4594,
+      "step": 716
+    },
+    {
+      "epoch": 0.3876374121463327,
+      "grad_norm": 0.4403664469718933,
+      "learning_rate": 7.682958645032265e-06,
+      "loss": 0.458,
+      "step": 717
+    },
+    {
+      "epoch": 0.38817805009911693,
+      "grad_norm": 0.3886347711086273,
+      "learning_rate": 7.674988099435487e-06,
+      "loss": 0.471,
+      "step": 718
+    },
+    {
+      "epoch": 0.38871868805190124,
+      "grad_norm": 0.4349733889102936,
+      "learning_rate": 7.667008018967598e-06,
+      "loss": 0.4992,
+      "step": 719
+    },
+    {
+      "epoch": 0.38925932600468555,
+      "grad_norm": 0.48294758796691895,
+      "learning_rate": 7.65901843207323e-06,
+      "loss": 0.487,
+      "step": 720
+    },
+    {
+      "epoch": 0.3897999639574698,
+      "grad_norm": 0.4309777319431305,
+      "learning_rate": 7.651019367230886e-06,
+      "loss": 0.4614,
+      "step": 721
+    },
+    {
+      "epoch": 0.3903406019102541,
+      "grad_norm": 0.4028390049934387,
+      "learning_rate": 7.643010852952871e-06,
+      "loss": 0.4942,
+      "step": 722
+    },
+    {
+      "epoch": 0.3908812398630384,
+      "grad_norm": 0.4195624887943268,
+      "learning_rate": 7.634992917785156e-06,
+      "loss": 0.4883,
+      "step": 723
+    },
+    {
+      "epoch": 0.39142187781582266,
+      "grad_norm": 0.38650834560394287,
+      "learning_rate": 7.626965590307305e-06,
+      "loss": 0.5055,
+      "step": 724
+    },
+    {
+      "epoch": 0.39196251576860697,
+      "grad_norm": 0.4400707185268402,
+      "learning_rate": 7.6189288991323505e-06,
+      "loss": 0.4653,
+      "step": 725
+    },
+    {
+      "epoch": 0.3925031537213912,
+      "grad_norm": 0.3416772484779358,
+      "learning_rate": 7.610882872906709e-06,
+      "loss": 0.4676,
+      "step": 726
+    },
+    {
+      "epoch": 0.3930437916741755,
+      "grad_norm": 0.44608405232429504,
+      "learning_rate": 7.602827540310065e-06,
+      "loss": 0.4804,
+      "step": 727
+    },
+    {
+      "epoch": 0.39358442962695983,
+      "grad_norm": 0.39614206552505493,
+      "learning_rate": 7.594762930055281e-06,
+      "loss": 0.4783,
+      "step": 728
+    },
+    {
+      "epoch": 0.3941250675797441,
+      "grad_norm": 0.4028596580028534,
+      "learning_rate": 7.586689070888284e-06,
+      "loss": 0.4848,
+      "step": 729
+    },
+    {
+      "epoch": 0.3946657055325284,
+      "grad_norm": 0.37662696838378906,
+      "learning_rate": 7.578605991587974e-06,
+      "loss": 0.4751,
+      "step": 730
+    },
+    {
+      "epoch": 0.3952063434853127,
+      "grad_norm": 0.480343222618103,
+      "learning_rate": 7.570513720966108e-06,
+      "loss": 0.4842,
+      "step": 731
+    },
+    {
+      "epoch": 0.39574698143809695,
+      "grad_norm": 0.43461012840270996,
+      "learning_rate": 7.562412287867214e-06,
+      "loss": 0.445,
+      "step": 732
+    },
+    {
+      "epoch": 0.39628761939088125,
+      "grad_norm": 0.37615707516670227,
+      "learning_rate": 7.5543017211684745e-06,
+      "loss": 0.5055,
+      "step": 733
+    },
+    {
+      "epoch": 0.3968282573436655,
+      "grad_norm": 0.3956639766693115,
+      "learning_rate": 7.5461820497796255e-06,
+      "loss": 0.4696,
+      "step": 734
+    },
+    {
+      "epoch": 0.3973688952964498,
+      "grad_norm": 0.5006547570228577,
+      "learning_rate": 7.5380533026428625e-06,
+      "loss": 0.5176,
+      "step": 735
+    },
+    {
+      "epoch": 0.3979095332492341,
+      "grad_norm": 0.37995660305023193,
+      "learning_rate": 7.529915508732725e-06,
+      "loss": 0.4962,
+      "step": 736
+    },
+    {
+      "epoch": 0.39845017120201837,
+      "grad_norm": 0.41486889123916626,
+      "learning_rate": 7.521768697056004e-06,
+      "loss": 0.489,
+      "step": 737
+    },
+    {
+      "epoch": 0.3989908091548027,
+      "grad_norm": 0.49411487579345703,
+      "learning_rate": 7.513612896651632e-06,
+      "loss": 0.4543,
+      "step": 738
+    },
+    {
+      "epoch": 0.399531447107587,
+      "grad_norm": 0.4323795437812805,
+      "learning_rate": 7.505448136590583e-06,
+      "loss": 0.4707,
+      "step": 739
+    },
+    {
+      "epoch": 0.40007208506037123,
+      "grad_norm": 0.3963027596473694,
+      "learning_rate": 7.497274445975762e-06,
+      "loss": 0.4838,
+      "step": 740
+    },
+    {
+      "epoch": 0.40061272301315554,
+      "grad_norm": 0.3969276547431946,
+      "learning_rate": 7.489091853941914e-06,
+      "loss": 0.484,
+      "step": 741
+    },
+    {
+      "epoch": 0.4011533609659398,
+      "grad_norm": 0.43876469135284424,
+      "learning_rate": 7.480900389655508e-06,
+      "loss": 0.464,
+      "step": 742
+    },
+    {
+      "epoch": 0.4016939989187241,
+      "grad_norm": 0.39766424894332886,
+      "learning_rate": 7.4727000823146386e-06,
+      "loss": 0.4979,
+      "step": 743
+    },
+    {
+      "epoch": 0.4022346368715084,
+      "grad_norm": 0.3988225758075714,
+      "learning_rate": 7.464490961148921e-06,
+      "loss": 0.4645,
+      "step": 744
+    },
+    {
+      "epoch": 0.40277527482429265,
+      "grad_norm": 0.5512990951538086,
+      "learning_rate": 7.4562730554193875e-06,
+      "loss": 0.4922,
+      "step": 745
+    },
+    {
+      "epoch": 0.40331591277707696,
+      "grad_norm": 0.4058268368244171,
+      "learning_rate": 7.448046394418383e-06,
+      "loss": 0.4796,
+      "step": 746
+    },
+    {
+      "epoch": 0.40385655072986126,
+      "grad_norm": 0.3984447717666626,
+      "learning_rate": 7.439811007469457e-06,
+      "loss": 0.4694,
+      "step": 747
+    },
+    {
+      "epoch": 0.4043971886826455,
+      "grad_norm": 0.41148069500923157,
+      "learning_rate": 7.431566923927267e-06,
+      "loss": 0.4757,
+      "step": 748
+    },
+    {
+      "epoch": 0.4049378266354298,
+      "grad_norm": 0.429737389087677,
+      "learning_rate": 7.423314173177467e-06,
+      "loss": 0.4949,
+      "step": 749
+    },
+    {
+      "epoch": 0.40547846458821407,
+      "grad_norm": 0.35777562856674194,
+      "learning_rate": 7.415052784636603e-06,
+      "loss": 0.4704,
+      "step": 750
+    },
+    {
+      "epoch": 0.4060191025409984,
+      "grad_norm": 0.40855300426483154,
+      "learning_rate": 7.406782787752011e-06,
+      "loss": 0.4648,
+      "step": 751
+    },
+    {
+      "epoch": 0.4065597404937827,
+      "grad_norm": 0.4542819857597351,
+      "learning_rate": 7.398504212001714e-06,
+      "loss": 0.4896,
+      "step": 752
+    },
+    {
+      "epoch": 0.40710037844656694,
+      "grad_norm": 0.3928791880607605,
+      "learning_rate": 7.390217086894309e-06,
+      "loss": 0.5082,
+      "step": 753
+    },
+    {
+      "epoch": 0.40764101639935124,
+      "grad_norm": 0.39957326650619507,
+      "learning_rate": 7.3819214419688725e-06,
+      "loss": 0.4752,
+      "step": 754
+    },
+    {
+      "epoch": 0.4081816543521355,
+      "grad_norm": 0.4252714216709137,
+      "learning_rate": 7.373617306794844e-06,
+      "loss": 0.4828,
+      "step": 755
+    },
+    {
+      "epoch": 0.4087222923049198,
+      "grad_norm": 0.45451775193214417,
+      "learning_rate": 7.365304710971928e-06,
+      "loss": 0.4776,
+      "step": 756
+    },
+    {
+      "epoch": 0.4092629302577041,
+      "grad_norm": 0.42666348814964294,
+      "learning_rate": 7.3569836841299905e-06,
+      "loss": 0.4805,
+      "step": 757
+    },
+    {
+      "epoch": 0.40980356821048836,
+      "grad_norm": 0.44532275199890137,
+      "learning_rate": 7.348654255928941e-06,
+      "loss": 0.4683,
+      "step": 758
+    },
+    {
+      "epoch": 0.41034420616327266,
+      "grad_norm": 0.540521502494812,
+      "learning_rate": 7.340316456058644e-06,
+      "loss": 0.4717,
+      "step": 759
+    },
+    {
+      "epoch": 0.41088484411605697,
+      "grad_norm": 0.4525167942047119,
+      "learning_rate": 7.331970314238799e-06,
+      "loss": 0.4782,
+      "step": 760
+    },
+    {
+      "epoch": 0.4114254820688412,
+      "grad_norm": 0.44031041860580444,
+      "learning_rate": 7.323615860218844e-06,
+      "loss": 0.4947,
+      "step": 761
+    },
+    {
+      "epoch": 0.4119661200216255,
+      "grad_norm": 0.5076286792755127,
+      "learning_rate": 7.31525312377784e-06,
+      "loss": 0.4638,
+      "step": 762
+    },
+    {
+      "epoch": 0.4125067579744098,
+      "grad_norm": 0.4349268674850464,
+      "learning_rate": 7.306882134724376e-06,
+      "loss": 0.4782,
+      "step": 763
+    },
+    {
+      "epoch": 0.4130473959271941,
+      "grad_norm": 0.44942405819892883,
+      "learning_rate": 7.298502922896453e-06,
+      "loss": 0.5115,
+      "step": 764
+    },
+    {
+      "epoch": 0.4135880338799784,
+      "grad_norm": 0.4894666075706482,
+      "learning_rate": 7.290115518161385e-06,
+      "loss": 0.5035,
+      "step": 765
+    },
+    {
+      "epoch": 0.41412867183276264,
+      "grad_norm": 0.3696083128452301,
+      "learning_rate": 7.281719950415686e-06,
+      "loss": 0.4537,
+      "step": 766
+    },
+    {
+      "epoch": 0.41466930978554695,
+      "grad_norm": 0.40760135650634766,
+      "learning_rate": 7.273316249584969e-06,
+      "loss": 0.4418,
+      "step": 767
+    },
+    {
+      "epoch": 0.41520994773833125,
+      "grad_norm": 0.47130587697029114,
+      "learning_rate": 7.2649044456238334e-06,
+      "loss": 0.4972,
+      "step": 768
+    },
+    {
+      "epoch": 0.4157505856911155,
+      "grad_norm": 0.4279153048992157,
+      "learning_rate": 7.256484568515769e-06,
+      "loss": 0.4963,
+      "step": 769
+    },
+    {
+      "epoch": 0.4162912236438998,
+      "grad_norm": 0.4578527808189392,
+      "learning_rate": 7.248056648273034e-06,
+      "loss": 0.4888,
+      "step": 770
+    },
+    {
+      "epoch": 0.41683186159668406,
+      "grad_norm": 0.42948973178863525,
+      "learning_rate": 7.239620714936561e-06,
+      "loss": 0.4877,
+      "step": 771
+    },
+    {
+      "epoch": 0.41737249954946837,
+      "grad_norm": 0.44241803884506226,
+      "learning_rate": 7.231176798575843e-06,
+      "loss": 0.4791,
+      "step": 772
+    },
+    {
+      "epoch": 0.4179131375022527,
+      "grad_norm": 0.48078736662864685,
+      "learning_rate": 7.22272492928883e-06,
+      "loss": 0.4744,
+      "step": 773
+    },
+    {
+      "epoch": 0.4184537754550369,
+      "grad_norm": 0.4113394618034363,
+      "learning_rate": 7.214265137201817e-06,
+      "loss": 0.4807,
+      "step": 774
+    },
+    {
+      "epoch": 0.41899441340782123,
+      "grad_norm": 0.48756808042526245,
+      "learning_rate": 7.205797452469341e-06,
+      "loss": 0.5125,
+      "step": 775
+    },
+    {
+      "epoch": 0.41953505136060554,
+      "grad_norm": 0.47686880826950073,
+      "learning_rate": 7.197321905274071e-06,
+      "loss": 0.4727,
+      "step": 776
+    },
+    {
+      "epoch": 0.4200756893133898,
+      "grad_norm": 0.42904406785964966,
+      "learning_rate": 7.188838525826702e-06,
+      "loss": 0.4668,
+      "step": 777
+    },
+    {
+      "epoch": 0.4206163272661741,
+      "grad_norm": 0.4416086971759796,
+      "learning_rate": 7.18034734436585e-06,
+      "loss": 0.4768,
+      "step": 778
+    },
+    {
+      "epoch": 0.42115696521895835,
+      "grad_norm": 0.4766569137573242,
+      "learning_rate": 7.171848391157935e-06,
+      "loss": 0.4908,
+      "step": 779
+    },
+    {
+      "epoch": 0.42169760317174265,
+      "grad_norm": 0.39499804377555847,
+      "learning_rate": 7.163341696497084e-06,
+      "loss": 0.4706,
+      "step": 780
+    },
+    {
+      "epoch": 0.42223824112452696,
+      "grad_norm": 0.4001219868659973,
+      "learning_rate": 7.154827290705012e-06,
+      "loss": 0.4891,
+      "step": 781
+    },
+    {
+      "epoch": 0.4227788790773112,
+      "grad_norm": 0.4044910669326782,
+      "learning_rate": 7.146305204130928e-06,
+      "loss": 0.4542,
+      "step": 782
+    },
+    {
+      "epoch": 0.4233195170300955,
+      "grad_norm": 0.3734391927719116,
+      "learning_rate": 7.137775467151411e-06,
+      "loss": 0.4741,
+      "step": 783
+    },
+    {
+      "epoch": 0.4238601549828798,
+      "grad_norm": 0.4248086214065552,
+      "learning_rate": 7.129238110170315e-06,
+      "loss": 0.4902,
+      "step": 784
+    },
+    {
+      "epoch": 0.4244007929356641,
+      "grad_norm": 0.3903302550315857,
+      "learning_rate": 7.120693163618656e-06,
+      "loss": 0.4493,
+      "step": 785
+    },
+    {
+      "epoch": 0.4249414308884484,
+      "grad_norm": 0.412576824426651,
+      "learning_rate": 7.112140657954495e-06,
+      "loss": 0.4939,
+      "step": 786
+    },
+    {
+      "epoch": 0.42548206884123263,
+      "grad_norm": 0.39094048738479614,
+      "learning_rate": 7.103580623662845e-06,
+      "loss": 0.4403,
+      "step": 787
+    },
+    {
+      "epoch": 0.42602270679401694,
+      "grad_norm": 0.4399993419647217,
+      "learning_rate": 7.0950130912555515e-06,
+      "loss": 0.4582,
+      "step": 788
+    },
+    {
+      "epoch": 0.42656334474680124,
+      "grad_norm": 0.41360506415367126,
+      "learning_rate": 7.086438091271186e-06,
+      "loss": 0.4824,
+      "step": 789
+    },
+    {
+      "epoch": 0.4271039826995855,
+      "grad_norm": 0.44524022936820984,
+      "learning_rate": 7.077855654274939e-06,
+      "loss": 0.4711,
+      "step": 790
+    },
+    {
+      "epoch": 0.4276446206523698,
+      "grad_norm": 0.4516412913799286,
+      "learning_rate": 7.069265810858509e-06,
+      "loss": 0.4893,
+      "step": 791
+    },
+    {
+      "epoch": 0.4281852586051541,
+      "grad_norm": 0.4142738878726959,
+      "learning_rate": 7.0606685916399945e-06,
+      "loss": 0.4555,
+      "step": 792
+    },
+    {
+      "epoch": 0.42872589655793836,
+      "grad_norm": 0.4141763746738434,
+      "learning_rate": 7.052064027263785e-06,
+      "loss": 0.4387,
+      "step": 793
+    },
+    {
+      "epoch": 0.42926653451072266,
+      "grad_norm": 0.47192102670669556,
+      "learning_rate": 7.043452148400452e-06,
+      "loss": 0.4812,
+      "step": 794
+    },
+    {
+      "epoch": 0.4298071724635069,
+      "grad_norm": 0.43239060044288635,
+      "learning_rate": 7.034832985746638e-06,
+      "loss": 0.503,
+      "step": 795
+    },
+    {
+      "epoch": 0.4303478104162912,
+      "grad_norm": 0.46152111887931824,
+      "learning_rate": 7.026206570024949e-06,
+      "loss": 0.4824,
+      "step": 796
+    },
+    {
+      "epoch": 0.4308884483690755,
+      "grad_norm": 0.4815097451210022,
+      "learning_rate": 7.017572931983846e-06,
+      "loss": 0.4558,
+      "step": 797
+    },
+    {
+      "epoch": 0.4314290863218598,
+      "grad_norm": 0.4330691397190094,
+      "learning_rate": 7.00893210239753e-06,
+      "loss": 0.4493,
+      "step": 798
+    },
+    {
+      "epoch": 0.4319697242746441,
+      "grad_norm": 0.44782209396362305,
+      "learning_rate": 7.000284112065836e-06,
+      "loss": 0.4737,
+      "step": 799
+    },
+    {
+      "epoch": 0.4325103622274284,
+      "grad_norm": 0.43315109610557556,
+      "learning_rate": 6.9916289918141265e-06,
+      "loss": 0.4533,
+      "step": 800
+    },
+    {
+      "epoch": 0.43305100018021264,
+      "grad_norm": 0.42613768577575684,
+      "learning_rate": 6.982966772493176e-06,
+      "loss": 0.4517,
+      "step": 801
+    },
+    {
+      "epoch": 0.43359163813299695,
+      "grad_norm": 0.39703595638275146,
+      "learning_rate": 6.974297484979066e-06,
+      "loss": 0.4651,
+      "step": 802
+    },
+    {
+      "epoch": 0.4341322760857812,
+      "grad_norm": 0.4882420599460602,
+      "learning_rate": 6.965621160173066e-06,
+      "loss": 0.4816,
+      "step": 803
+    },
+    {
+      "epoch": 0.4346729140385655,
+      "grad_norm": 0.4013866186141968,
+      "learning_rate": 6.9569378290015375e-06,
+      "loss": 0.4545,
+      "step": 804
+    },
+    {
+      "epoch": 0.4352135519913498,
+      "grad_norm": 0.41527172923088074,
+      "learning_rate": 6.948247522415811e-06,
+      "loss": 0.4567,
+      "step": 805
+    },
+    {
+      "epoch": 0.43575418994413406,
+      "grad_norm": 0.4296116530895233,
+      "learning_rate": 6.939550271392079e-06,
+      "loss": 0.4499,
+      "step": 806
+    },
+    {
+      "epoch": 0.43629482789691837,
+      "grad_norm": 0.4318053722381592,
+      "learning_rate": 6.930846106931292e-06,
+      "loss": 0.4873,
+      "step": 807
+    },
+    {
+      "epoch": 0.4368354658497027,
+      "grad_norm": 0.39466047286987305,
+      "learning_rate": 6.922135060059043e-06,
+      "loss": 0.4705,
+      "step": 808
+    },
+    {
+      "epoch": 0.4373761038024869,
+      "grad_norm": 0.4394930601119995,
+      "learning_rate": 6.913417161825449e-06,
+      "loss": 0.4816,
+      "step": 809
+    },
+    {
+      "epoch": 0.43791674175527123,
+      "grad_norm": 0.4361076354980469,
+      "learning_rate": 6.904692443305059e-06,
+      "loss": 0.496,
+      "step": 810
+    },
+    {
+      "epoch": 0.4384573797080555,
+      "grad_norm": 0.4172879755496979,
+      "learning_rate": 6.895960935596728e-06,
+      "loss": 0.49,
+      "step": 811
+    },
+    {
+      "epoch": 0.4389980176608398,
+      "grad_norm": 0.4652804732322693,
+      "learning_rate": 6.8872226698235065e-06,
+      "loss": 0.468,
+      "step": 812
+    },
+    {
+      "epoch": 0.4395386556136241,
+      "grad_norm": 0.39598575234413147,
+      "learning_rate": 6.8784776771325426e-06,
+      "loss": 0.4487,
+      "step": 813
+    },
+    {
+      "epoch": 0.44007929356640835,
+      "grad_norm": 0.5070300102233887,
+      "learning_rate": 6.869725988694955e-06,
+      "loss": 0.453,
+      "step": 814
+    },
+    {
+      "epoch": 0.44061993151919265,
+      "grad_norm": 0.4058806598186493,
+      "learning_rate": 6.860967635705732e-06,
+      "loss": 0.4652,
+      "step": 815
+    },
+    {
+      "epoch": 0.44116056947197696,
+      "grad_norm": 0.4511767029762268,
+      "learning_rate": 6.8522026493836144e-06,
+      "loss": 0.4801,
+      "step": 816
+    },
+    {
+      "epoch": 0.4417012074247612,
+      "grad_norm": 0.435837984085083,
+      "learning_rate": 6.843431060970995e-06,
+      "loss": 0.4772,
+      "step": 817
+    },
+    {
+      "epoch": 0.4422418453775455,
+      "grad_norm": 0.4215855002403259,
+      "learning_rate": 6.834652901733789e-06,
+      "loss": 0.4633,
+      "step": 818
+    },
+    {
+      "epoch": 0.44278248333032977,
+      "grad_norm": 0.43643590807914734,
+      "learning_rate": 6.825868202961343e-06,
+      "loss": 0.4891,
+      "step": 819
+    },
+    {
+      "epoch": 0.4433231212831141,
+      "grad_norm": 0.42145484685897827,
+      "learning_rate": 6.8170769959663045e-06,
+      "loss": 0.4537,
+      "step": 820
+    },
+    {
+      "epoch": 0.4438637592358984,
+      "grad_norm": 0.44420287013053894,
+      "learning_rate": 6.808279312084525e-06,
+      "loss": 0.4894,
+      "step": 821
+    },
+    {
+      "epoch": 0.44440439718868263,
+      "grad_norm": 0.4486287534236908,
+      "learning_rate": 6.799475182674942e-06,
+      "loss": 0.4629,
+      "step": 822
+    },
+    {
+      "epoch": 0.44494503514146694,
+      "grad_norm": 0.4760923385620117,
+      "learning_rate": 6.790664639119464e-06,
+      "loss": 0.4876,
+      "step": 823
+    },
+    {
+      "epoch": 0.4454856730942512,
+      "grad_norm": 0.46989408135414124,
+      "learning_rate": 6.781847712822869e-06,
+      "loss": 0.4719,
+      "step": 824
+    },
+    {
+      "epoch": 0.4460263110470355,
+      "grad_norm": 0.46125340461730957,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.4733,
+      "step": 825
+    },
+    {
+      "epoch": 0.4465669489998198,
+      "grad_norm": 0.45252951979637146,
+      "learning_rate": 6.76419483773906e-06,
+      "loss": 0.4937,
+      "step": 826
+    },
+    {
+      "epoch": 0.44710758695260405,
+      "grad_norm": 0.40651631355285645,
+      "learning_rate": 6.755358951874701e-06,
+      "loss": 0.4674,
+      "step": 827
+    },
+    {
+      "epoch": 0.44764822490538836,
+      "grad_norm": 0.5222921967506409,
+      "learning_rate": 6.7465168091147094e-06,
+      "loss": 0.4767,
+      "step": 828
+    },
+    {
+      "epoch": 0.44818886285817267,
+      "grad_norm": 0.34470462799072266,
+      "learning_rate": 6.737668440976494e-06,
+      "loss": 0.4687,
+      "step": 829
+    },
+    {
+      "epoch": 0.4487295008109569,
+      "grad_norm": 0.5007984042167664,
+      "learning_rate": 6.728813878999652e-06,
+      "loss": 0.5117,
+      "step": 830
+    },
+    {
+      "epoch": 0.4492701387637412,
+      "grad_norm": 0.43586406111717224,
+      "learning_rate": 6.719953154745857e-06,
+      "loss": 0.4643,
+      "step": 831
+    },
+    {
+      "epoch": 0.4498107767165255,
+      "grad_norm": 0.3484852910041809,
+      "learning_rate": 6.7110862997987525e-06,
+      "loss": 0.4711,
+      "step": 832
+    },
+    {
+      "epoch": 0.4503514146693098,
+      "grad_norm": 0.5028504133224487,
+      "learning_rate": 6.70221334576383e-06,
+      "loss": 0.4679,
+      "step": 833
+    },
+    {
+      "epoch": 0.4508920526220941,
+      "grad_norm": 0.4104493260383606,
+      "learning_rate": 6.693334324268328e-06,
+      "loss": 0.4639,
+      "step": 834
+    },
+    {
+      "epoch": 0.45143269057487834,
+      "grad_norm": 0.40598517656326294,
+      "learning_rate": 6.684449266961101e-06,
+      "loss": 0.4794,
+      "step": 835
+    },
+    {
+      "epoch": 0.45197332852766264,
+      "grad_norm": 0.3934710621833801,
+      "learning_rate": 6.675558205512527e-06,
+      "loss": 0.4734,
+      "step": 836
+    },
+    {
+      "epoch": 0.45251396648044695,
+      "grad_norm": 0.39573705196380615,
+      "learning_rate": 6.666661171614382e-06,
+      "loss": 0.4828,
+      "step": 837
+    },
+    {
+      "epoch": 0.4530546044332312,
+      "grad_norm": 0.39194008708000183,
+      "learning_rate": 6.657758196979732e-06,
+      "loss": 0.4396,
+      "step": 838
+    },
+    {
+      "epoch": 0.4535952423860155,
+      "grad_norm": 0.38524988293647766,
+      "learning_rate": 6.648849313342816e-06,
+      "loss": 0.5242,
+      "step": 839
+    },
+    {
+      "epoch": 0.45413588033879976,
+      "grad_norm": 0.4037633538246155,
+      "learning_rate": 6.6399345524589366e-06,
+      "loss": 0.4575,
+      "step": 840
+    },
+    {
+      "epoch": 0.45467651829158406,
+      "grad_norm": 0.3950909376144409,
+      "learning_rate": 6.631013946104348e-06,
+      "loss": 0.4784,
+      "step": 841
+    },
+    {
+      "epoch": 0.45521715624436837,
+      "grad_norm": 0.3955569565296173,
+      "learning_rate": 6.622087526076135e-06,
+      "loss": 0.4869,
+      "step": 842
+    },
+    {
+      "epoch": 0.4557577941971526,
+      "grad_norm": 0.425361692905426,
+      "learning_rate": 6.613155324192111e-06,
+      "loss": 0.4649,
+      "step": 843
+    },
+    {
+      "epoch": 0.45629843214993693,
+      "grad_norm": 0.41891539096832275,
+      "learning_rate": 6.604217372290693e-06,
+      "loss": 0.5171,
+      "step": 844
+    },
+    {
+      "epoch": 0.45683907010272123,
+      "grad_norm": 0.41283082962036133,
+      "learning_rate": 6.5952737022308e-06,
+      "loss": 0.4919,
+      "step": 845
+    },
+    {
+      "epoch": 0.4573797080555055,
+      "grad_norm": 0.49672916531562805,
+      "learning_rate": 6.586324345891727e-06,
+      "loss": 0.472,
+      "step": 846
+    },
+    {
+      "epoch": 0.4579203460082898,
+      "grad_norm": 0.4132554829120636,
+      "learning_rate": 6.57736933517304e-06,
+      "loss": 0.4787,
+      "step": 847
+    },
+    {
+      "epoch": 0.45846098396107404,
+      "grad_norm": 0.4273938834667206,
+      "learning_rate": 6.568408701994459e-06,
+      "loss": 0.4754,
+      "step": 848
+    },
+    {
+      "epoch": 0.45900162191385835,
+      "grad_norm": 0.47689104080200195,
+      "learning_rate": 6.559442478295745e-06,
+      "loss": 0.4945,
+      "step": 849
+    },
+    {
+      "epoch": 0.45954225986664266,
+      "grad_norm": 0.4251175820827484,
+      "learning_rate": 6.550470696036591e-06,
+      "loss": 0.4706,
+      "step": 850
+    },
+    {
+      "epoch": 0.4600828978194269,
+      "grad_norm": 0.5029627680778503,
+      "learning_rate": 6.541493387196496e-06,
+      "loss": 0.464,
+      "step": 851
+    },
+    {
+      "epoch": 0.4606235357722112,
+      "grad_norm": 0.43976134061813354,
+      "learning_rate": 6.5325105837746604e-06,
+      "loss": 0.4781,
+      "step": 852
+    },
+    {
+      "epoch": 0.4611641737249955,
+      "grad_norm": 0.45708853006362915,
+      "learning_rate": 6.523522317789874e-06,
+      "loss": 0.4608,
+      "step": 853
+    },
+    {
+      "epoch": 0.46170481167777977,
+      "grad_norm": 0.47303131222724915,
+      "learning_rate": 6.514528621280391e-06,
+      "loss": 0.5097,
+      "step": 854
+    },
+    {
+      "epoch": 0.4622454496305641,
+      "grad_norm": 0.4358423948287964,
+      "learning_rate": 6.5055295263038286e-06,
+      "loss": 0.4851,
+      "step": 855
+    },
+    {
+      "epoch": 0.4627860875833483,
+      "grad_norm": 0.41981184482574463,
+      "learning_rate": 6.496525064937042e-06,
+      "loss": 0.5024,
+      "step": 856
+    },
+    {
+      "epoch": 0.46332672553613263,
+      "grad_norm": 0.4504159688949585,
+      "learning_rate": 6.487515269276015e-06,
+      "loss": 0.468,
+      "step": 857
+    },
+    {
+      "epoch": 0.46386736348891694,
+      "grad_norm": 0.5347965955734253,
+      "learning_rate": 6.478500171435751e-06,
+      "loss": 0.4995,
+      "step": 858
+    },
+    {
+      "epoch": 0.4644080014417012,
+      "grad_norm": 0.39357173442840576,
+      "learning_rate": 6.469479803550144e-06,
+      "loss": 0.4734,
+      "step": 859
+    },
+    {
+      "epoch": 0.4649486393944855,
+      "grad_norm": 0.38938194513320923,
+      "learning_rate": 6.460454197771881e-06,
+      "loss": 0.5136,
+      "step": 860
+    },
+    {
+      "epoch": 0.4654892773472698,
+      "grad_norm": 0.47099724411964417,
+      "learning_rate": 6.451423386272312e-06,
+      "loss": 0.4934,
+      "step": 861
+    },
+    {
+      "epoch": 0.46602991530005405,
+      "grad_norm": 0.4051066040992737,
+      "learning_rate": 6.442387401241349e-06,
+      "loss": 0.4642,
+      "step": 862
+    },
+    {
+      "epoch": 0.46657055325283836,
+      "grad_norm": 0.43694064021110535,
+      "learning_rate": 6.433346274887341e-06,
+      "loss": 0.484,
+      "step": 863
+    },
+    {
+      "epoch": 0.4671111912056226,
+      "grad_norm": 0.42491498589515686,
+      "learning_rate": 6.4243000394369626e-06,
+      "loss": 0.4813,
+      "step": 864
+    },
+    {
+      "epoch": 0.4676518291584069,
+      "grad_norm": 0.40995004773139954,
+      "learning_rate": 6.415248727135103e-06,
+      "loss": 0.446,
+      "step": 865
+    },
+    {
+      "epoch": 0.4681924671111912,
+      "grad_norm": 0.440571129322052,
+      "learning_rate": 6.406192370244742e-06,
+      "loss": 0.4958,
+      "step": 866
+    },
+    {
+      "epoch": 0.4687331050639755,
+      "grad_norm": 0.43054115772247314,
+      "learning_rate": 6.397131001046849e-06,
+      "loss": 0.4953,
+      "step": 867
+    },
+    {
+      "epoch": 0.4692737430167598,
+      "grad_norm": 0.4248604476451874,
+      "learning_rate": 6.38806465184025e-06,
+      "loss": 0.4782,
+      "step": 868
+    },
+    {
+      "epoch": 0.4698143809695441,
+      "grad_norm": 0.3913900852203369,
+      "learning_rate": 6.378993354941529e-06,
+      "loss": 0.4648,
+      "step": 869
+    },
+    {
+      "epoch": 0.47035501892232834,
+      "grad_norm": 0.41105592250823975,
+      "learning_rate": 6.3699171426849036e-06,
+      "loss": 0.4961,
+      "step": 870
+    },
+    {
+      "epoch": 0.47089565687511264,
+      "grad_norm": 0.3705733120441437,
+      "learning_rate": 6.3608360474221106e-06,
+      "loss": 0.4559,
+      "step": 871
+    },
+    {
+      "epoch": 0.4714362948278969,
+      "grad_norm": 0.3702208995819092,
+      "learning_rate": 6.3517501015222924e-06,
+      "loss": 0.4727,
+      "step": 872
+    },
+    {
+      "epoch": 0.4719769327806812,
+      "grad_norm": 0.3946548104286194,
+      "learning_rate": 6.342659337371884e-06,
+      "loss": 0.4848,
+      "step": 873
+    },
+    {
+      "epoch": 0.4725175707334655,
+      "grad_norm": 0.3584488034248352,
+      "learning_rate": 6.333563787374493e-06,
+      "loss": 0.458,
+      "step": 874
+    },
+    {
+      "epoch": 0.47305820868624976,
+      "grad_norm": 0.36788806319236755,
+      "learning_rate": 6.3244634839507834e-06,
+      "loss": 0.4674,
+      "step": 875
+    },
+    {
+      "epoch": 0.47359884663903407,
+      "grad_norm": 0.3621140420436859,
+      "learning_rate": 6.315358459538367e-06,
+      "loss": 0.4718,
+      "step": 876
+    },
+    {
+      "epoch": 0.47413948459181837,
+      "grad_norm": 0.375000923871994,
+      "learning_rate": 6.3062487465916825e-06,
+      "loss": 0.4812,
+      "step": 877
+    },
+    {
+      "epoch": 0.4746801225446026,
+      "grad_norm": 0.44487372040748596,
+      "learning_rate": 6.297134377581877e-06,
+      "loss": 0.4746,
+      "step": 878
+    },
+    {
+      "epoch": 0.47522076049738693,
+      "grad_norm": 0.3941539525985718,
+      "learning_rate": 6.2880153849966966e-06,
+      "loss": 0.4762,
+      "step": 879
+    },
+    {
+      "epoch": 0.4757613984501712,
+      "grad_norm": 0.4312276244163513,
+      "learning_rate": 6.2788918013403695e-06,
+      "loss": 0.4952,
+      "step": 880
+    },
+    {
+      "epoch": 0.4763020364029555,
+      "grad_norm": 0.41254860162734985,
+      "learning_rate": 6.269763659133486e-06,
+      "loss": 0.4963,
+      "step": 881
+    },
+    {
+      "epoch": 0.4768426743557398,
+      "grad_norm": 0.40565529465675354,
+      "learning_rate": 6.2606309909128845e-06,
+      "loss": 0.4787,
+      "step": 882
+    },
+    {
+      "epoch": 0.47738331230852404,
+      "grad_norm": 0.4065866470336914,
+      "learning_rate": 6.251493829231539e-06,
+      "loss": 0.4723,
+      "step": 883
+    },
+    {
+      "epoch": 0.47792395026130835,
+      "grad_norm": 0.43057599663734436,
+      "learning_rate": 6.24235220665844e-06,
+      "loss": 0.4792,
+      "step": 884
+    },
+    {
+      "epoch": 0.47846458821409266,
+      "grad_norm": 0.3824271857738495,
+      "learning_rate": 6.233206155778476e-06,
+      "loss": 0.4817,
+      "step": 885
+    },
+    {
+      "epoch": 0.4790052261668769,
+      "grad_norm": 0.46750858426094055,
+      "learning_rate": 6.224055709192323e-06,
+      "loss": 0.4613,
+      "step": 886
+    },
+    {
+      "epoch": 0.4795458641196612,
+      "grad_norm": 0.45582181215286255,
+      "learning_rate": 6.21490089951632e-06,
+      "loss": 0.4958,
+      "step": 887
+    },
+    {
+      "epoch": 0.48008650207244546,
+      "grad_norm": 0.4065682291984558,
+      "learning_rate": 6.205741759382365e-06,
+      "loss": 0.4928,
+      "step": 888
+    },
+    {
+      "epoch": 0.48062714002522977,
+      "grad_norm": 0.4958328604698181,
+      "learning_rate": 6.1965783214377895e-06,
+      "loss": 0.4648,
+      "step": 889
+    },
+    {
+      "epoch": 0.4811677779780141,
+      "grad_norm": 0.38294878602027893,
+      "learning_rate": 6.187410618345241e-06,
+      "loss": 0.4905,
+      "step": 890
+    },
+    {
+      "epoch": 0.48170841593079833,
+      "grad_norm": 0.471400648355484,
+      "learning_rate": 6.178238682782574e-06,
+      "loss": 0.4835,
+      "step": 891
+    },
+    {
+      "epoch": 0.48224905388358263,
+      "grad_norm": 0.5207462310791016,
+      "learning_rate": 6.169062547442724e-06,
+      "loss": 0.4912,
+      "step": 892
+    },
+    {
+      "epoch": 0.4827896918363669,
+      "grad_norm": 0.4611525535583496,
+      "learning_rate": 6.159882245033606e-06,
+      "loss": 0.5036,
+      "step": 893
+    },
+    {
+      "epoch": 0.4833303297891512,
+      "grad_norm": 0.4598880708217621,
+      "learning_rate": 6.150697808277979e-06,
+      "loss": 0.4472,
+      "step": 894
+    },
+    {
+      "epoch": 0.4838709677419355,
+      "grad_norm": 0.4637363851070404,
+      "learning_rate": 6.141509269913343e-06,
+      "loss": 0.4849,
+      "step": 895
+    },
+    {
+      "epoch": 0.48441160569471975,
+      "grad_norm": 0.5085048675537109,
+      "learning_rate": 6.132316662691815e-06,
+      "loss": 0.467,
+      "step": 896
+    },
+    {
+      "epoch": 0.48495224364750406,
+      "grad_norm": 0.4294266104698181,
+      "learning_rate": 6.123120019380021e-06,
+      "loss": 0.4781,
+      "step": 897
+    },
+    {
+      "epoch": 0.48549288160028836,
+      "grad_norm": 0.47353968024253845,
+      "learning_rate": 6.1139193727589665e-06,
+      "loss": 0.4969,
+      "step": 898
+    },
+    {
+      "epoch": 0.4860335195530726,
+      "grad_norm": 0.5444623827934265,
+      "learning_rate": 6.1047147556239325e-06,
+      "loss": 0.4863,
+      "step": 899
+    },
+    {
+      "epoch": 0.4865741575058569,
+      "grad_norm": 0.39669960737228394,
+      "learning_rate": 6.095506200784349e-06,
+      "loss": 0.485,
+      "step": 900
+    },
+    {
+      "epoch": 0.48711479545864117,
+      "grad_norm": 0.5122884511947632,
+      "learning_rate": 6.086293741063685e-06,
+      "loss": 0.4742,
+      "step": 901
+    },
+    {
+      "epoch": 0.4876554334114255,
+      "grad_norm": 0.4148077070713043,
+      "learning_rate": 6.077077409299323e-06,
+      "loss": 0.4846,
+      "step": 902
+    },
+    {
+      "epoch": 0.4881960713642098,
+      "grad_norm": 0.3915378153324127,
+      "learning_rate": 6.067857238342451e-06,
+      "loss": 0.4665,
+      "step": 903
+    },
+    {
+      "epoch": 0.48873670931699403,
+      "grad_norm": 0.47468093037605286,
+      "learning_rate": 6.058633261057945e-06,
+      "loss": 0.5034,
+      "step": 904
+    },
+    {
+      "epoch": 0.48927734726977834,
+      "grad_norm": 0.3712773323059082,
+      "learning_rate": 6.049405510324237e-06,
+      "loss": 0.4721,
+      "step": 905
+    },
+    {
+      "epoch": 0.48981798522256265,
+      "grad_norm": 0.4807882010936737,
+      "learning_rate": 6.040174019033226e-06,
+      "loss": 0.4779,
+      "step": 906
+    },
+    {
+      "epoch": 0.4903586231753469,
+      "grad_norm": 0.4945826530456543,
+      "learning_rate": 6.030938820090128e-06,
+      "loss": 0.4796,
+      "step": 907
+    },
+    {
+      "epoch": 0.4908992611281312,
+      "grad_norm": 0.41042837500572205,
+      "learning_rate": 6.021699946413384e-06,
+      "loss": 0.491,
+      "step": 908
+    },
+    {
+      "epoch": 0.49143989908091545,
+      "grad_norm": 0.5063190460205078,
+      "learning_rate": 6.012457430934532e-06,
+      "loss": 0.464,
+      "step": 909
+    },
+    {
+      "epoch": 0.49198053703369976,
+      "grad_norm": 0.4003697335720062,
+      "learning_rate": 6.003211306598089e-06,
+      "loss": 0.4639,
+      "step": 910
+    },
+    {
+      "epoch": 0.49252117498648407,
+      "grad_norm": 0.4417046010494232,
+      "learning_rate": 5.993961606361436e-06,
+      "loss": 0.4691,
+      "step": 911
+    },
+    {
+      "epoch": 0.4930618129392683,
+      "grad_norm": 0.45925286412239075,
+      "learning_rate": 5.984708363194702e-06,
+      "loss": 0.4819,
+      "step": 912
+    },
+    {
+      "epoch": 0.4936024508920526,
+      "grad_norm": 0.44486504793167114,
+      "learning_rate": 5.975451610080643e-06,
+      "loss": 0.5092,
+      "step": 913
+    },
+    {
+      "epoch": 0.49414308884483693,
+      "grad_norm": 0.409686803817749,
+      "learning_rate": 5.966191380014524e-06,
+      "loss": 0.4812,
+      "step": 914
+    },
+    {
+      "epoch": 0.4946837267976212,
+      "grad_norm": 0.4472230076789856,
+      "learning_rate": 5.956927706004012e-06,
+      "loss": 0.4983,
+      "step": 915
+    },
+    {
+      "epoch": 0.4952243647504055,
+      "grad_norm": 0.5017169117927551,
+      "learning_rate": 5.947660621069038e-06,
+      "loss": 0.474,
+      "step": 916
+    },
+    {
+      "epoch": 0.49576500270318974,
+      "grad_norm": 0.4999132752418518,
+      "learning_rate": 5.938390158241701e-06,
+      "loss": 0.4736,
+      "step": 917
+    },
+    {
+      "epoch": 0.49630564065597405,
+      "grad_norm": 0.42812034487724304,
+      "learning_rate": 5.929116350566132e-06,
+      "loss": 0.4802,
+      "step": 918
+    },
+    {
+      "epoch": 0.49684627860875835,
+      "grad_norm": 0.3980429470539093,
+      "learning_rate": 5.919839231098392e-06,
+      "loss": 0.4909,
+      "step": 919
+    },
+    {
+      "epoch": 0.4973869165615426,
+      "grad_norm": 0.4948299527168274,
+      "learning_rate": 5.910558832906341e-06,
+      "loss": 0.4603,
+      "step": 920
+    },
+    {
+      "epoch": 0.4979275545143269,
+      "grad_norm": 0.3767414689064026,
+      "learning_rate": 5.90127518906953e-06,
+      "loss": 0.4828,
+      "step": 921
+    },
+    {
+      "epoch": 0.4984681924671112,
+      "grad_norm": 0.4088006019592285,
+      "learning_rate": 5.891988332679075e-06,
+      "loss": 0.4912,
+      "step": 922
+    },
+    {
+      "epoch": 0.49900883041989547,
+      "grad_norm": 0.40145042538642883,
+      "learning_rate": 5.882698296837549e-06,
+      "loss": 0.4714,
+      "step": 923
+    },
+    {
+      "epoch": 0.4995494683726798,
+      "grad_norm": 0.3793385922908783,
+      "learning_rate": 5.87340511465885e-06,
+      "loss": 0.4727,
+      "step": 924
+    },
+    {
+      "epoch": 0.500090106325464,
+      "grad_norm": 0.40813931822776794,
+      "learning_rate": 5.864108819268098e-06,
+      "loss": 0.5169,
+      "step": 925
+    },
+    {
+      "epoch": 0.5006307442782484,
+      "grad_norm": 0.40652281045913696,
+      "learning_rate": 5.8548094438015065e-06,
+      "loss": 0.4698,
+      "step": 926
+    },
+    {
+      "epoch": 0.5011713822310326,
+      "grad_norm": 0.37487828731536865,
+      "learning_rate": 5.8455070214062685e-06,
+      "loss": 0.4714,
+      "step": 927
+    },
+    {
+      "epoch": 0.5017120201838169,
+      "grad_norm": 0.36619752645492554,
+      "learning_rate": 5.8362015852404365e-06,
+      "loss": 0.4425,
+      "step": 928
+    },
+    {
+      "epoch": 0.5022526581366011,
+      "grad_norm": 0.4049976170063019,
+      "learning_rate": 5.826893168472807e-06,
+      "loss": 0.4801,
+      "step": 929
+    },
+    {
+      "epoch": 0.5027932960893855,
+      "grad_norm": 0.36154064536094666,
+      "learning_rate": 5.8175818042828e-06,
+      "loss": 0.4694,
+      "step": 930
+    },
+    {
+      "epoch": 0.5033339340421698,
+      "grad_norm": 0.3709295690059662,
+      "learning_rate": 5.808267525860343e-06,
+      "loss": 0.4757,
+      "step": 931
+    },
+    {
+      "epoch": 0.503874571994954,
+      "grad_norm": 0.4130777418613434,
+      "learning_rate": 5.798950366405748e-06,
+      "loss": 0.4681,
+      "step": 932
+    },
+    {
+      "epoch": 0.5044152099477384,
+      "grad_norm": 0.42704373598098755,
+      "learning_rate": 5.789630359129599e-06,
+      "loss": 0.5039,
+      "step": 933
+    },
+    {
+      "epoch": 0.5049558479005226,
+      "grad_norm": 0.398743212223053,
+      "learning_rate": 5.780307537252629e-06,
+      "loss": 0.4582,
+      "step": 934
+    },
+    {
+      "epoch": 0.5054964858533069,
+      "grad_norm": 0.37964576482772827,
+      "learning_rate": 5.770981934005606e-06,
+      "loss": 0.4758,
+      "step": 935
+    },
+    {
+      "epoch": 0.5060371238060912,
+      "grad_norm": 0.4298136234283447,
+      "learning_rate": 5.76165358262921e-06,
+      "loss": 0.4637,
+      "step": 936
+    },
+    {
+      "epoch": 0.5065777617588755,
+      "grad_norm": 0.3951083719730377,
+      "learning_rate": 5.752322516373916e-06,
+      "loss": 0.4877,
+      "step": 937
+    },
+    {
+      "epoch": 0.5071183997116597,
+      "grad_norm": 0.4326460063457489,
+      "learning_rate": 5.742988768499879e-06,
+      "loss": 0.4555,
+      "step": 938
+    },
+    {
+      "epoch": 0.5076590376644441,
+      "grad_norm": 0.3702709376811981,
+      "learning_rate": 5.733652372276809e-06,
+      "loss": 0.495,
+      "step": 939
+    },
+    {
+      "epoch": 0.5081996756172283,
+      "grad_norm": 0.41388028860092163,
+      "learning_rate": 5.724313360983859e-06,
+      "loss": 0.4855,
+      "step": 940
+    },
+    {
+      "epoch": 0.5087403135700126,
+      "grad_norm": 0.40203750133514404,
+      "learning_rate": 5.7149717679095026e-06,
+      "loss": 0.4885,
+      "step": 941
+    },
+    {
+      "epoch": 0.509280951522797,
+      "grad_norm": 0.39383724331855774,
+      "learning_rate": 5.705627626351415e-06,
+      "loss": 0.4634,
+      "step": 942
+    },
+    {
+      "epoch": 0.5098215894755812,
+      "grad_norm": 0.37974879145622253,
+      "learning_rate": 5.6962809696163536e-06,
+      "loss": 0.4613,
+      "step": 943
+    },
+    {
+      "epoch": 0.5103622274283655,
+      "grad_norm": 0.44113990664482117,
+      "learning_rate": 5.686931831020044e-06,
+      "loss": 0.4702,
+      "step": 944
+    },
+    {
+      "epoch": 0.5109028653811497,
+      "grad_norm": 0.3676075339317322,
+      "learning_rate": 5.6775802438870596e-06,
+      "loss": 0.4384,
+      "step": 945
+    },
+    {
+      "epoch": 0.5114435033339341,
+      "grad_norm": 0.3680996894836426,
+      "learning_rate": 5.668226241550698e-06,
+      "loss": 0.4823,
+      "step": 946
+    },
+    {
+      "epoch": 0.5119841412867183,
+      "grad_norm": 0.37592819333076477,
+      "learning_rate": 5.658869857352866e-06,
+      "loss": 0.4891,
+      "step": 947
+    },
+    {
+      "epoch": 0.5125247792395026,
+      "grad_norm": 0.38988274335861206,
+      "learning_rate": 5.649511124643962e-06,
+      "loss": 0.5056,
+      "step": 948
+    },
+    {
+      "epoch": 0.5130654171922869,
+      "grad_norm": 0.3609144687652588,
+      "learning_rate": 5.640150076782755e-06,
+      "loss": 0.4541,
+      "step": 949
+    },
+    {
+      "epoch": 0.5136060551450712,
+      "grad_norm": 0.41311243176460266,
+      "learning_rate": 5.630786747136269e-06,
+      "loss": 0.4958,
+      "step": 950
+    },
+    {
+      "epoch": 0.5141466930978554,
+      "grad_norm": 0.3682233393192291,
+      "learning_rate": 5.621421169079655e-06,
+      "loss": 0.4919,
+      "step": 951
+    },
+    {
+      "epoch": 0.5146873310506398,
+      "grad_norm": 0.36114174127578735,
+      "learning_rate": 5.612053375996082e-06,
+      "loss": 0.4825,
+      "step": 952
+    },
+    {
+      "epoch": 0.515227969003424,
+      "grad_norm": 0.40716037154197693,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.4529,
+      "step": 953
+    },
+    {
+      "epoch": 0.5157686069562083,
+      "grad_norm": 0.40229591727256775,
+      "learning_rate": 5.593311278320097e-06,
+      "loss": 0.4857,
+      "step": 954
+    },
+    {
+      "epoch": 0.5163092449089927,
+      "grad_norm": 0.3767363727092743,
+      "learning_rate": 5.583937040533023e-06,
+      "loss": 0.4939,
+      "step": 955
+    },
+    {
+      "epoch": 0.5168498828617769,
+      "grad_norm": 0.3691665232181549,
+      "learning_rate": 5.574560721329431e-06,
+      "loss": 0.4705,
+      "step": 956
+    },
+    {
+      "epoch": 0.5173905208145612,
+      "grad_norm": 0.38738739490509033,
+      "learning_rate": 5.565182354130776e-06,
+      "loss": 0.4676,
+      "step": 957
+    },
+    {
+      "epoch": 0.5179311587673455,
+      "grad_norm": 0.370933473110199,
+      "learning_rate": 5.555801972365812e-06,
+      "loss": 0.4713,
+      "step": 958
+    },
+    {
+      "epoch": 0.5184717967201298,
+      "grad_norm": 0.38803374767303467,
+      "learning_rate": 5.5464196094704745e-06,
+      "loss": 0.4836,
+      "step": 959
+    },
+    {
+      "epoch": 0.519012434672914,
+      "grad_norm": 0.396415114402771,
+      "learning_rate": 5.537035298887764e-06,
+      "loss": 0.4738,
+      "step": 960
+    },
+    {
+      "epoch": 0.5195530726256983,
+      "grad_norm": 0.4398602247238159,
+      "learning_rate": 5.527649074067618e-06,
+      "loss": 0.5144,
+      "step": 961
+    },
+    {
+      "epoch": 0.5200937105784826,
+      "grad_norm": 0.3995663821697235,
+      "learning_rate": 5.5182609684668024e-06,
+      "loss": 0.5048,
+      "step": 962
+    },
+    {
+      "epoch": 0.5206343485312669,
+      "grad_norm": 0.41245123744010925,
+      "learning_rate": 5.508871015548781e-06,
+      "loss": 0.4697,
+      "step": 963
+    },
+    {
+      "epoch": 0.5211749864840511,
+      "grad_norm": 0.3744891285896301,
+      "learning_rate": 5.49947924878361e-06,
+      "loss": 0.4669,
+      "step": 964
+    },
+    {
+      "epoch": 0.5217156244368355,
+      "grad_norm": 0.42122045159339905,
+      "learning_rate": 5.490085701647805e-06,
+      "loss": 0.4859,
+      "step": 965
+    },
+    {
+      "epoch": 0.5222562623896198,
+      "grad_norm": 0.3975967466831207,
+      "learning_rate": 5.480690407624227e-06,
+      "loss": 0.4549,
+      "step": 966
+    },
+    {
+      "epoch": 0.522796900342404,
+      "grad_norm": 0.41702109575271606,
+      "learning_rate": 5.47129340020197e-06,
+      "loss": 0.4944,
+      "step": 967
+    },
+    {
+      "epoch": 0.5233375382951884,
+      "grad_norm": 0.4080027937889099,
+      "learning_rate": 5.461894712876228e-06,
+      "loss": 0.4705,
+      "step": 968
+    },
+    {
+      "epoch": 0.5238781762479726,
+      "grad_norm": 0.42316797375679016,
+      "learning_rate": 5.45249437914819e-06,
+      "loss": 0.4567,
+      "step": 969
+    },
+    {
+      "epoch": 0.5244188142007569,
+      "grad_norm": 0.3980869948863983,
+      "learning_rate": 5.443092432524906e-06,
+      "loss": 0.4738,
+      "step": 970
+    },
+    {
+      "epoch": 0.5249594521535412,
+      "grad_norm": 0.42186301946640015,
+      "learning_rate": 5.433688906519183e-06,
+      "loss": 0.4854,
+      "step": 971
+    },
+    {
+      "epoch": 0.5255000901063255,
+      "grad_norm": 0.3881399631500244,
+      "learning_rate": 5.424283834649451e-06,
+      "loss": 0.4716,
+      "step": 972
+    },
+    {
+      "epoch": 0.5260407280591097,
+      "grad_norm": 0.4174569249153137,
+      "learning_rate": 5.414877250439654e-06,
+      "loss": 0.4897,
+      "step": 973
+    },
+    {
+      "epoch": 0.526581366011894,
+      "grad_norm": 0.44264698028564453,
+      "learning_rate": 5.405469187419126e-06,
+      "loss": 0.4815,
+      "step": 974
+    },
+    {
+      "epoch": 0.5271220039646783,
+      "grad_norm": 0.32237735390663147,
+      "learning_rate": 5.39605967912247e-06,
+      "loss": 0.4508,
+      "step": 975
+    },
+    {
+      "epoch": 0.5276626419174626,
+      "grad_norm": 0.47621771693229675,
+      "learning_rate": 5.386648759089441e-06,
+      "loss": 0.5093,
+      "step": 976
+    },
+    {
+      "epoch": 0.5282032798702468,
+      "grad_norm": 0.4255550801753998,
+      "learning_rate": 5.3772364608648304e-06,
+      "loss": 0.5067,
+      "step": 977
+    },
+    {
+      "epoch": 0.5287439178230312,
+      "grad_norm": 0.35732802748680115,
+      "learning_rate": 5.367822817998338e-06,
+      "loss": 0.4611,
+      "step": 978
+    },
+    {
+      "epoch": 0.5292845557758155,
+      "grad_norm": 0.3863273859024048,
+      "learning_rate": 5.358407864044456e-06,
+      "loss": 0.4833,
+      "step": 979
+    },
+    {
+      "epoch": 0.5298251937285997,
+      "grad_norm": 0.3596065938472748,
+      "learning_rate": 5.348991632562355e-06,
+      "loss": 0.5005,
+      "step": 980
+    },
+    {
+      "epoch": 0.5303658316813841,
+      "grad_norm": 0.39100930094718933,
+      "learning_rate": 5.339574157115752e-06,
+      "loss": 0.4815,
+      "step": 981
+    },
+    {
+      "epoch": 0.5309064696341683,
+      "grad_norm": 0.38288214802742004,
+      "learning_rate": 5.330155471272804e-06,
+      "loss": 0.4522,
+      "step": 982
+    },
+    {
+      "epoch": 0.5314471075869526,
+      "grad_norm": 0.3723912835121155,
+      "learning_rate": 5.320735608605979e-06,
+      "loss": 0.4618,
+      "step": 983
+    },
+    {
+      "epoch": 0.5319877455397369,
+      "grad_norm": 0.3716428279876709,
+      "learning_rate": 5.311314602691943e-06,
+      "loss": 0.4565,
+      "step": 984
+    },
+    {
+      "epoch": 0.5325283834925212,
+      "grad_norm": 0.4149164855480194,
+      "learning_rate": 5.301892487111431e-06,
+      "loss": 0.4597,
+      "step": 985
+    },
+    {
+      "epoch": 0.5330690214453054,
+      "grad_norm": 0.45434221625328064,
+      "learning_rate": 5.292469295449141e-06,
+      "loss": 0.472,
+      "step": 986
+    },
+    {
+      "epoch": 0.5336096593980898,
+      "grad_norm": 0.35414353013038635,
+      "learning_rate": 5.2830450612936e-06,
+      "loss": 0.4724,
+      "step": 987
+    },
+    {
+      "epoch": 0.534150297350874,
+      "grad_norm": 0.41073864698410034,
+      "learning_rate": 5.273619818237058e-06,
+      "loss": 0.4936,
+      "step": 988
+    },
+    {
+      "epoch": 0.5346909353036583,
+      "grad_norm": 0.3850623667240143,
+      "learning_rate": 5.264193599875353e-06,
+      "loss": 0.4932,
+      "step": 989
+    },
+    {
+      "epoch": 0.5352315732564425,
+      "grad_norm": 0.3790433406829834,
+      "learning_rate": 5.254766439807807e-06,
+      "loss": 0.4615,
+      "step": 990
+    },
+    {
+      "epoch": 0.5357722112092269,
+      "grad_norm": 0.3931194841861725,
+      "learning_rate": 5.245338371637091e-06,
+      "loss": 0.4915,
+      "step": 991
+    },
+    {
+      "epoch": 0.5363128491620112,
+      "grad_norm": 0.36426377296447754,
+      "learning_rate": 5.235909428969119e-06,
+      "loss": 0.4598,
+      "step": 992
+    },
+    {
+      "epoch": 0.5368534871147954,
+      "grad_norm": 0.381425142288208,
+      "learning_rate": 5.226479645412923e-06,
+      "loss": 0.4681,
+      "step": 993
+    },
+    {
+      "epoch": 0.5373941250675798,
+      "grad_norm": 0.38973844051361084,
+      "learning_rate": 5.2170490545805255e-06,
+      "loss": 0.4858,
+      "step": 994
+    },
+    {
+      "epoch": 0.537934763020364,
+      "grad_norm": 0.35494348406791687,
+      "learning_rate": 5.207617690086831e-06,
+      "loss": 0.4866,
+      "step": 995
+    },
+    {
+      "epoch": 0.5384754009731483,
+      "grad_norm": 0.38420799374580383,
+      "learning_rate": 5.1981855855495035e-06,
+      "loss": 0.4588,
+      "step": 996
+    },
+    {
+      "epoch": 0.5390160389259326,
+      "grad_norm": 0.4316636621952057,
+      "learning_rate": 5.188752774588841e-06,
+      "loss": 0.4792,
+      "step": 997
+    },
+    {
+      "epoch": 0.5395566768787169,
+      "grad_norm": 0.3582872748374939,
+      "learning_rate": 5.179319290827661e-06,
+      "loss": 0.4911,
+      "step": 998
+    },
+    {
+      "epoch": 0.5400973148315011,
+      "grad_norm": 0.38559308648109436,
+      "learning_rate": 5.16988516789118e-06,
+      "loss": 0.4771,
+      "step": 999
+    },
+    {
+      "epoch": 0.5406379527842855,
+      "grad_norm": 0.4075615406036377,
+      "learning_rate": 5.16045043940689e-06,
+      "loss": 0.494,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5411785907370698,
+      "grad_norm": 0.39124950766563416,
+      "learning_rate": 5.151015139004445e-06,
+      "loss": 0.4745,
+      "step": 1001
+    },
+    {
+      "epoch": 0.541719228689854,
+      "grad_norm": 0.42799654603004456,
+      "learning_rate": 5.141579300315536e-06,
+      "loss": 0.4366,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5422598666426384,
+      "grad_norm": 0.4033052921295166,
+      "learning_rate": 5.132142956973773e-06,
+      "loss": 0.4618,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5428005045954226,
+      "grad_norm": 0.41250860691070557,
+      "learning_rate": 5.122706142614562e-06,
+      "loss": 0.4412,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5433411425482069,
+      "grad_norm": 0.37778806686401367,
+      "learning_rate": 5.113268890874994e-06,
+      "loss": 0.4824,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5438817805009911,
+      "grad_norm": 0.39247915148735046,
+      "learning_rate": 5.103831235393714e-06,
+      "loss": 0.4853,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5444224184537755,
+      "grad_norm": 0.3986344039440155,
+      "learning_rate": 5.094393209810806e-06,
+      "loss": 0.4623,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5449630564065597,
+      "grad_norm": 0.3868162930011749,
+      "learning_rate": 5.084954847767677e-06,
+      "loss": 0.4603,
+      "step": 1008
+    },
+    {
+      "epoch": 0.545503694359344,
+      "grad_norm": 0.39990749955177307,
+      "learning_rate": 5.07551618290693e-06,
+      "loss": 0.4866,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5460443323121283,
+      "grad_norm": 0.3816811442375183,
+      "learning_rate": 5.06607724887225e-06,
+      "loss": 0.4705,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5465849702649126,
+      "grad_norm": 0.39409664273262024,
+      "learning_rate": 5.056638079308277e-06,
+      "loss": 0.4685,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5471256082176968,
+      "grad_norm": 0.37422487139701843,
+      "learning_rate": 5.047198707860496e-06,
+      "loss": 0.4707,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5476662461704812,
+      "grad_norm": 0.37773483991622925,
+      "learning_rate": 5.037759168175109e-06,
+      "loss": 0.488,
+      "step": 1013
+    },
+    {
+      "epoch": 0.5482068841232655,
+      "grad_norm": 0.4065093994140625,
+      "learning_rate": 5.028319493898916e-06,
+      "loss": 0.5103,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5487475220760497,
+      "grad_norm": 0.40094494819641113,
+      "learning_rate": 5.018879718679199e-06,
+      "loss": 0.4757,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5492881600288341,
+      "grad_norm": 0.3680800199508667,
+      "learning_rate": 5.009439876163601e-06,
+      "loss": 0.4682,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5498287979816183,
+      "grad_norm": 0.3533041775226593,
+      "learning_rate": 5e-06,
+      "loss": 0.4541,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5503694359344026,
+      "grad_norm": 0.38887202739715576,
+      "learning_rate": 4.9905601238364006e-06,
+      "loss": 0.4604,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5509100738871869,
+      "grad_norm": 0.38735654950141907,
+      "learning_rate": 4.981120281320801e-06,
+      "loss": 0.474,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5514507118399712,
+      "grad_norm": 0.3694019615650177,
+      "learning_rate": 4.971680506101086e-06,
+      "loss": 0.4734,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5519913497927554,
+      "grad_norm": 0.36134111881256104,
+      "learning_rate": 4.9622408318248925e-06,
+      "loss": 0.454,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5525319877455397,
+      "grad_norm": 0.40273579955101013,
+      "learning_rate": 4.952801292139505e-06,
+      "loss": 0.5016,
+      "step": 1022
+    },
+    {
+      "epoch": 0.553072625698324,
+      "grad_norm": 0.39331087470054626,
+      "learning_rate": 4.9433619206917234e-06,
+      "loss": 0.4658,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5536132636511083,
+      "grad_norm": 0.40813592076301575,
+      "learning_rate": 4.933922751127753e-06,
+      "loss": 0.4767,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5541539016038926,
+      "grad_norm": 0.37781664729118347,
+      "learning_rate": 4.924483817093071e-06,
+      "loss": 0.4625,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5546945395566769,
+      "grad_norm": 0.420913964509964,
+      "learning_rate": 4.915045152232324e-06,
+      "loss": 0.4834,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5552351775094612,
+      "grad_norm": 0.3994501233100891,
+      "learning_rate": 4.9056067901891945e-06,
+      "loss": 0.496,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5557758154622454,
+      "grad_norm": 0.471672922372818,
+      "learning_rate": 4.896168764606289e-06,
+      "loss": 0.4774,
+      "step": 1028
+    },
+    {
+      "epoch": 0.5563164534150298,
+      "grad_norm": 0.4291553497314453,
+      "learning_rate": 4.886731109125007e-06,
+      "loss": 0.4909,
+      "step": 1029
+    },
+    {
+      "epoch": 0.556857091367814,
+      "grad_norm": 0.3480027914047241,
+      "learning_rate": 4.87729385738544e-06,
+      "loss": 0.4796,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5573977293205983,
+      "grad_norm": 0.448752760887146,
+      "learning_rate": 4.867857043026229e-06,
+      "loss": 0.4718,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5579383672733826,
+      "grad_norm": 0.429802805185318,
+      "learning_rate": 4.858420699684464e-06,
+      "loss": 0.4695,
+      "step": 1032
+    },
+    {
+      "epoch": 0.5584790052261669,
+      "grad_norm": 0.40032947063446045,
+      "learning_rate": 4.848984860995557e-06,
+      "loss": 0.4741,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5590196431789511,
+      "grad_norm": 0.35695409774780273,
+      "learning_rate": 4.839549560593111e-06,
+      "loss": 0.4628,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5595602811317354,
+      "grad_norm": 0.35745665431022644,
+      "learning_rate": 4.830114832108822e-06,
+      "loss": 0.4973,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5601009190845198,
+      "grad_norm": 0.37396466732025146,
+      "learning_rate": 4.82068070917234e-06,
+      "loss": 0.463,
+      "step": 1036
+    },
+    {
+      "epoch": 0.560641557037304,
+      "grad_norm": 0.3474443554878235,
+      "learning_rate": 4.81124722541116e-06,
+      "loss": 0.4761,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5611821949900883,
+      "grad_norm": 0.40390846133232117,
+      "learning_rate": 4.801814414450498e-06,
+      "loss": 0.4875,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5617228329428726,
+      "grad_norm": 0.3489498794078827,
+      "learning_rate": 4.7923823099131694e-06,
+      "loss": 0.4772,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5622634708956569,
+      "grad_norm": 0.4154996871948242,
+      "learning_rate": 4.782950945419475e-06,
+      "loss": 0.4778,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5628041088484411,
+      "grad_norm": 0.36867937445640564,
+      "learning_rate": 4.7735203545870794e-06,
+      "loss": 0.474,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5633447468012255,
+      "grad_norm": 0.3696093261241913,
+      "learning_rate": 4.764090571030882e-06,
+      "loss": 0.4813,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5638853847540097,
+      "grad_norm": 0.3766731023788452,
+      "learning_rate": 4.75466162836291e-06,
+      "loss": 0.4941,
+      "step": 1043
+    },
+    {
+      "epoch": 0.564426022706794,
+      "grad_norm": 0.39763715863227844,
+      "learning_rate": 4.745233560192195e-06,
+      "loss": 0.4716,
+      "step": 1044
+    },
+    {
+      "epoch": 0.5649666606595783,
+      "grad_norm": 0.3585834801197052,
+      "learning_rate": 4.735806400124648e-06,
+      "loss": 0.474,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5655072986123626,
+      "grad_norm": 0.36759787797927856,
+      "learning_rate": 4.726380181762943e-06,
+      "loss": 0.4849,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5660479365651468,
+      "grad_norm": 0.37226635217666626,
+      "learning_rate": 4.716954938706401e-06,
+      "loss": 0.4766,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5665885745179312,
+      "grad_norm": 0.392599880695343,
+      "learning_rate": 4.707530704550861e-06,
+      "loss": 0.4906,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5671292124707155,
+      "grad_norm": 0.3526148200035095,
+      "learning_rate": 4.69810751288857e-06,
+      "loss": 0.4418,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5676698504234997,
+      "grad_norm": 0.37177005410194397,
+      "learning_rate": 4.688685397308061e-06,
+      "loss": 0.5155,
+      "step": 1050
+    },
+    {
+      "epoch": 0.568210488376284,
+      "grad_norm": 0.42093127965927124,
+      "learning_rate": 4.679264391394022e-06,
+      "loss": 0.4817,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5687511263290683,
+      "grad_norm": 0.3944513499736786,
+      "learning_rate": 4.669844528727197e-06,
+      "loss": 0.4964,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5692917642818526,
+      "grad_norm": 0.339263379573822,
+      "learning_rate": 4.660425842884249e-06,
+      "loss": 0.4467,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5698324022346368,
+      "grad_norm": 0.3937620520591736,
+      "learning_rate": 4.651008367437646e-06,
+      "loss": 0.4803,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5703730401874212,
+      "grad_norm": 0.36690381169319153,
+      "learning_rate": 4.641592135955545e-06,
+      "loss": 0.4529,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5709136781402054,
+      "grad_norm": 0.373169481754303,
+      "learning_rate": 4.6321771820016635e-06,
+      "loss": 0.4899,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5714543160929897,
+      "grad_norm": 0.36590346693992615,
+      "learning_rate": 4.62276353913517e-06,
+      "loss": 0.4827,
+      "step": 1057
+    },
+    {
+      "epoch": 0.571994954045774,
+      "grad_norm": 0.3776438534259796,
+      "learning_rate": 4.6133512409105595e-06,
+      "loss": 0.4724,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5725355919985583,
+      "grad_norm": 0.3477265536785126,
+      "learning_rate": 4.603940320877533e-06,
+      "loss": 0.4912,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5730762299513426,
+      "grad_norm": 0.340049684047699,
+      "learning_rate": 4.594530812580876e-06,
+      "loss": 0.4825,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5736168679041269,
+      "grad_norm": 0.3493627905845642,
+      "learning_rate": 4.585122749560347e-06,
+      "loss": 0.4692,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5741575058569112,
+      "grad_norm": 0.34757259488105774,
+      "learning_rate": 4.575716165350549e-06,
+      "loss": 0.4381,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5746981438096954,
+      "grad_norm": 0.3675667345523834,
+      "learning_rate": 4.566311093480818e-06,
+      "loss": 0.4431,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5752387817624798,
+      "grad_norm": 0.3604763448238373,
+      "learning_rate": 4.556907567475094e-06,
+      "loss": 0.4832,
+      "step": 1064
+    },
+    {
+      "epoch": 0.575779419715264,
+      "grad_norm": 0.3719836473464966,
+      "learning_rate": 4.547505620851812e-06,
+      "loss": 0.4634,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5763200576680483,
+      "grad_norm": 0.3803561329841614,
+      "learning_rate": 4.538105287123772e-06,
+      "loss": 0.4679,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5768606956208325,
+      "grad_norm": 0.35499629378318787,
+      "learning_rate": 4.528706599798033e-06,
+      "loss": 0.4648,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5774013335736169,
+      "grad_norm": 0.3680323362350464,
+      "learning_rate": 4.5193095923757745e-06,
+      "loss": 0.4622,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5779419715264011,
+      "grad_norm": 0.3444634974002838,
+      "learning_rate": 4.509914298352197e-06,
+      "loss": 0.4707,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5784826094791854,
+      "grad_norm": 0.3498629927635193,
+      "learning_rate": 4.5005207512163914e-06,
+      "loss": 0.4618,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5790232474319698,
+      "grad_norm": 0.3526155352592468,
+      "learning_rate": 4.491128984451219e-06,
+      "loss": 0.4661,
+      "step": 1071
+    },
+    {
+      "epoch": 0.579563885384754,
+      "grad_norm": 0.3570224344730377,
+      "learning_rate": 4.481739031533201e-06,
+      "loss": 0.4846,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5801045233375383,
+      "grad_norm": 0.3742653727531433,
+      "learning_rate": 4.472350925932384e-06,
+      "loss": 0.4589,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5806451612903226,
+      "grad_norm": 0.3513067662715912,
+      "learning_rate": 4.462964701112237e-06,
+      "loss": 0.4401,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5811857992431069,
+      "grad_norm": 0.3835412561893463,
+      "learning_rate": 4.453580390529526e-06,
+      "loss": 0.4607,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5817264371958911,
+      "grad_norm": 0.35909005999565125,
+      "learning_rate": 4.444198027634191e-06,
+      "loss": 0.4408,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5822670751486755,
+      "grad_norm": 0.3801157474517822,
+      "learning_rate": 4.434817645869226e-06,
+      "loss": 0.4678,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5828077131014597,
+      "grad_norm": 0.36517688632011414,
+      "learning_rate": 4.42543927867057e-06,
+      "loss": 0.4884,
+      "step": 1078
+    },
+    {
+      "epoch": 0.583348351054244,
+      "grad_norm": 0.41169390082359314,
+      "learning_rate": 4.416062959466978e-06,
+      "loss": 0.491,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5838889890070283,
+      "grad_norm": 0.4015640616416931,
+      "learning_rate": 4.4066887216799055e-06,
+      "loss": 0.5044,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5844296269598126,
+      "grad_norm": 0.3767825663089752,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.4729,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5849702649125968,
+      "grad_norm": 0.37044841051101685,
+      "learning_rate": 4.38794662400392e-06,
+      "loss": 0.4901,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5855109028653811,
+      "grad_norm": 0.3824319839477539,
+      "learning_rate": 4.3785788309203466e-06,
+      "loss": 0.454,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5860515408181655,
+      "grad_norm": 0.42856699228286743,
+      "learning_rate": 4.369213252863733e-06,
+      "loss": 0.4726,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5865921787709497,
+      "grad_norm": 0.3541294038295746,
+      "learning_rate": 4.359849923217246e-06,
+      "loss": 0.4546,
+      "step": 1085
+    },
+    {
+      "epoch": 0.587132816723734,
+      "grad_norm": 0.4067661166191101,
+      "learning_rate": 4.350488875356041e-06,
+      "loss": 0.4584,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5876734546765183,
+      "grad_norm": 0.39768409729003906,
+      "learning_rate": 4.341130142647136e-06,
+      "loss": 0.4699,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5882140926293026,
+      "grad_norm": 0.35365110635757446,
+      "learning_rate": 4.331773758449303e-06,
+      "loss": 0.4863,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5887547305820868,
+      "grad_norm": 0.4205717444419861,
+      "learning_rate": 4.322419756112943e-06,
+      "loss": 0.4603,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5892953685348712,
+      "grad_norm": 0.3971540033817291,
+      "learning_rate": 4.313068168979957e-06,
+      "loss": 0.4485,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5898360064876554,
+      "grad_norm": 0.3554147183895111,
+      "learning_rate": 4.303719030383648e-06,
+      "loss": 0.47,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5903766444404397,
+      "grad_norm": 0.4239473044872284,
+      "learning_rate": 4.294372373648587e-06,
+      "loss": 0.5029,
+      "step": 1092
+    },
+    {
+      "epoch": 0.590917282393224,
+      "grad_norm": 0.3685072958469391,
+      "learning_rate": 4.285028232090499e-06,
+      "loss": 0.4611,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5914579203460083,
+      "grad_norm": 0.39185526967048645,
+      "learning_rate": 4.275686639016142e-06,
+      "loss": 0.454,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5919985582987926,
+      "grad_norm": 0.3944680094718933,
+      "learning_rate": 4.266347627723192e-06,
+      "loss": 0.4956,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5925391962515769,
+      "grad_norm": 0.3366733491420746,
+      "learning_rate": 4.257011231500122e-06,
+      "loss": 0.4424,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5930798342043612,
+      "grad_norm": 0.4068387746810913,
+      "learning_rate": 4.247677483626085e-06,
+      "loss": 0.4794,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5936204721571454,
+      "grad_norm": 0.4122190773487091,
+      "learning_rate": 4.238346417370793e-06,
+      "loss": 0.4769,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5941611101099297,
+      "grad_norm": 0.38368502259254456,
+      "learning_rate": 4.229018065994396e-06,
+      "loss": 0.4576,
+      "step": 1099
+    },
+    {
+      "epoch": 0.594701748062714,
+      "grad_norm": 0.33293166756629944,
+      "learning_rate": 4.2196924627473715e-06,
+      "loss": 0.4749,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5952423860154983,
+      "grad_norm": 0.43737560510635376,
+      "learning_rate": 4.210369640870403e-06,
+      "loss": 0.4989,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5957830239682825,
+      "grad_norm": 0.4038828909397125,
+      "learning_rate": 4.201049633594254e-06,
+      "loss": 0.4808,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5963236619210669,
+      "grad_norm": 0.3759760856628418,
+      "learning_rate": 4.1917324741396595e-06,
+      "loss": 0.482,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5968642998738511,
+      "grad_norm": 0.3900810778141022,
+      "learning_rate": 4.1824181957172014e-06,
+      "loss": 0.5048,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5974049378266354,
+      "grad_norm": 0.3988158106803894,
+      "learning_rate": 4.173106831527194e-06,
+      "loss": 0.4889,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5979455757794198,
+      "grad_norm": 0.393657386302948,
+      "learning_rate": 4.163798414759566e-06,
+      "loss": 0.4676,
+      "step": 1106
+    },
+    {
+      "epoch": 0.598486213732204,
+      "grad_norm": 0.3507223427295685,
+      "learning_rate": 4.154492978593733e-06,
+      "loss": 0.4462,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5990268516849883,
+      "grad_norm": 0.3565762937068939,
+      "learning_rate": 4.145190556198494e-06,
+      "loss": 0.4725,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5995674896377726,
+      "grad_norm": 0.3660755157470703,
+      "learning_rate": 4.135891180731903e-06,
+      "loss": 0.4607,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6001081275905569,
+      "grad_norm": 0.3648075461387634,
+      "learning_rate": 4.1265948853411506e-06,
+      "loss": 0.4446,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6006487655433411,
+      "grad_norm": 0.3925120532512665,
+      "learning_rate": 4.1173017031624544e-06,
+      "loss": 0.4767,
+      "step": 1111
+    },
+    {
+      "epoch": 0.6011894034961254,
+      "grad_norm": 0.35011157393455505,
+      "learning_rate": 4.108011667320926e-06,
+      "loss": 0.4739,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6017300414489097,
+      "grad_norm": 0.3578500747680664,
+      "learning_rate": 4.098724810930472e-06,
+      "loss": 0.4899,
+      "step": 1113
+    },
+    {
+      "epoch": 0.602270679401694,
+      "grad_norm": 0.40307337045669556,
+      "learning_rate": 4.08944116709366e-06,
+      "loss": 0.4923,
+      "step": 1114
+    },
+    {
+      "epoch": 0.6028113173544782,
+      "grad_norm": 0.39985835552215576,
+      "learning_rate": 4.08016076890161e-06,
+      "loss": 0.466,
+      "step": 1115
+    },
+    {
+      "epoch": 0.6033519553072626,
+      "grad_norm": 0.40574347972869873,
+      "learning_rate": 4.0708836494338695e-06,
+      "loss": 0.478,
+      "step": 1116
+    },
+    {
+      "epoch": 0.6038925932600469,
+      "grad_norm": 0.3304382264614105,
+      "learning_rate": 4.061609841758302e-06,
+      "loss": 0.4487,
+      "step": 1117
+    },
+    {
+      "epoch": 0.6044332312128311,
+      "grad_norm": 0.3854810893535614,
+      "learning_rate": 4.0523393789309625e-06,
+      "loss": 0.4704,
+      "step": 1118
+    },
+    {
+      "epoch": 0.6049738691656155,
+      "grad_norm": 0.3859521746635437,
+      "learning_rate": 4.04307229399599e-06,
+      "loss": 0.446,
+      "step": 1119
+    },
+    {
+      "epoch": 0.6055145071183997,
+      "grad_norm": 0.3449578881263733,
+      "learning_rate": 4.0338086199854765e-06,
+      "loss": 0.4592,
+      "step": 1120
+    },
+    {
+      "epoch": 0.606055145071184,
+      "grad_norm": 0.381989985704422,
+      "learning_rate": 4.02454838991936e-06,
+      "loss": 0.4553,
+      "step": 1121
+    },
+    {
+      "epoch": 0.6065957830239683,
+      "grad_norm": 0.4080401659011841,
+      "learning_rate": 4.0152916368053e-06,
+      "loss": 0.5008,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6071364209767526,
+      "grad_norm": 0.35274338722229004,
+      "learning_rate": 4.006038393638565e-06,
+      "loss": 0.4687,
+      "step": 1123
+    },
+    {
+      "epoch": 0.6076770589295368,
+      "grad_norm": 0.3596242070198059,
+      "learning_rate": 3.996788693401914e-06,
+      "loss": 0.4852,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6082176968823212,
+      "grad_norm": 0.44486021995544434,
+      "learning_rate": 3.987542569065469e-06,
+      "loss": 0.4747,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6087583348351054,
+      "grad_norm": 0.3718114495277405,
+      "learning_rate": 3.978300053586617e-06,
+      "loss": 0.5005,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6092989727878897,
+      "grad_norm": 0.4019679129123688,
+      "learning_rate": 3.969061179909872e-06,
+      "loss": 0.4566,
+      "step": 1127
+    },
+    {
+      "epoch": 0.6098396107406739,
+      "grad_norm": 0.3548070788383484,
+      "learning_rate": 3.959825980966777e-06,
+      "loss": 0.4627,
+      "step": 1128
+    },
+    {
+      "epoch": 0.6103802486934583,
+      "grad_norm": 0.35685357451438904,
+      "learning_rate": 3.9505944896757635e-06,
+      "loss": 0.491,
+      "step": 1129
+    },
+    {
+      "epoch": 0.6109208866462426,
+      "grad_norm": 0.3496491312980652,
+      "learning_rate": 3.941366738942058e-06,
+      "loss": 0.4934,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6114615245990268,
+      "grad_norm": 0.39106976985931396,
+      "learning_rate": 3.932142761657549e-06,
+      "loss": 0.4896,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6120021625518112,
+      "grad_norm": 0.39134979248046875,
+      "learning_rate": 3.922922590700679e-06,
+      "loss": 0.5008,
+      "step": 1132
+    },
+    {
+      "epoch": 0.6125428005045954,
+      "grad_norm": 0.3936554193496704,
+      "learning_rate": 3.913706258936317e-06,
+      "loss": 0.481,
+      "step": 1133
+    },
+    {
+      "epoch": 0.6130834384573797,
+      "grad_norm": 0.3407648503780365,
+      "learning_rate": 3.904493799215652e-06,
+      "loss": 0.4668,
+      "step": 1134
+    },
+    {
+      "epoch": 0.613624076410164,
+      "grad_norm": 0.3679908514022827,
+      "learning_rate": 3.895285244376068e-06,
+      "loss": 0.4934,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6141647143629483,
+      "grad_norm": 0.4318563640117645,
+      "learning_rate": 3.886080627241034e-06,
+      "loss": 0.4423,
+      "step": 1136
+    },
+    {
+      "epoch": 0.6147053523157325,
+      "grad_norm": 0.3628011643886566,
+      "learning_rate": 3.876879980619982e-06,
+      "loss": 0.4939,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6152459902685169,
+      "grad_norm": 0.375052809715271,
+      "learning_rate": 3.8676833373081864e-06,
+      "loss": 0.4655,
+      "step": 1138
+    },
+    {
+      "epoch": 0.6157866282213011,
+      "grad_norm": 0.39029353857040405,
+      "learning_rate": 3.8584907300866595e-06,
+      "loss": 0.4855,
+      "step": 1139
+    },
+    {
+      "epoch": 0.6163272661740854,
+      "grad_norm": 0.4070568084716797,
+      "learning_rate": 3.8493021917220225e-06,
+      "loss": 0.4526,
+      "step": 1140
+    },
+    {
+      "epoch": 0.6168679041268698,
+      "grad_norm": 0.4272588789463043,
+      "learning_rate": 3.840117754966396e-06,
+      "loss": 0.4892,
+      "step": 1141
+    },
+    {
+      "epoch": 0.617408542079654,
+      "grad_norm": 0.4317014217376709,
+      "learning_rate": 3.8309374525572765e-06,
+      "loss": 0.4817,
+      "step": 1142
+    },
+    {
+      "epoch": 0.6179491800324383,
+      "grad_norm": 0.4254457950592041,
+      "learning_rate": 3.821761317217428e-06,
+      "loss": 0.4631,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6184898179852225,
+      "grad_norm": 0.36125367879867554,
+      "learning_rate": 3.81258938165476e-06,
+      "loss": 0.4893,
+      "step": 1144
+    },
+    {
+      "epoch": 0.6190304559380069,
+      "grad_norm": 0.3904247581958771,
+      "learning_rate": 3.803421678562213e-06,
+      "loss": 0.4599,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6195710938907911,
+      "grad_norm": 0.41387060284614563,
+      "learning_rate": 3.794258240617636e-06,
+      "loss": 0.4982,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6201117318435754,
+      "grad_norm": 0.349608838558197,
+      "learning_rate": 3.7850991004836813e-06,
+      "loss": 0.4859,
+      "step": 1147
+    },
+    {
+      "epoch": 0.6206523697963597,
+      "grad_norm": 0.3353005647659302,
+      "learning_rate": 3.7759442908076786e-06,
+      "loss": 0.481,
+      "step": 1148
+    },
+    {
+      "epoch": 0.621193007749144,
+      "grad_norm": 0.3263114392757416,
+      "learning_rate": 3.7667938442215247e-06,
+      "loss": 0.4775,
+      "step": 1149
+    },
+    {
+      "epoch": 0.6217336457019282,
+      "grad_norm": 0.44235607981681824,
+      "learning_rate": 3.7576477933415612e-06,
+      "loss": 0.4519,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6222742836547126,
+      "grad_norm": 0.4153715670108795,
+      "learning_rate": 3.748506170768462e-06,
+      "loss": 0.474,
+      "step": 1151
+    },
+    {
+      "epoch": 0.6228149216074969,
+      "grad_norm": 0.34854984283447266,
+      "learning_rate": 3.739369009087117e-06,
+      "loss": 0.4795,
+      "step": 1152
+    },
+    {
+      "epoch": 0.6233555595602811,
+      "grad_norm": 0.37851929664611816,
+      "learning_rate": 3.7302363408665155e-06,
+      "loss": 0.5126,
+      "step": 1153
+    },
+    {
+      "epoch": 0.6238961975130655,
+      "grad_norm": 0.3372579514980316,
+      "learning_rate": 3.721108198659633e-06,
+      "loss": 0.4704,
+      "step": 1154
+    },
+    {
+      "epoch": 0.6244368354658497,
+      "grad_norm": 0.39567941427230835,
+      "learning_rate": 3.7119846150033047e-06,
+      "loss": 0.486,
+      "step": 1155
+    },
+    {
+      "epoch": 0.624977473418634,
+      "grad_norm": 0.3714936673641205,
+      "learning_rate": 3.702865622418125e-06,
+      "loss": 0.4449,
+      "step": 1156
+    },
+    {
+      "epoch": 0.6255181113714183,
+      "grad_norm": 0.35422173142433167,
+      "learning_rate": 3.693751253408319e-06,
+      "loss": 0.4686,
+      "step": 1157
+    },
+    {
+      "epoch": 0.6260587493242026,
+      "grad_norm": 0.32026010751724243,
+      "learning_rate": 3.6846415404616344e-06,
+      "loss": 0.4818,
+      "step": 1158
+    },
+    {
+      "epoch": 0.6265993872769868,
+      "grad_norm": 0.3986067771911621,
+      "learning_rate": 3.6755365160492187e-06,
+      "loss": 0.4787,
+      "step": 1159
+    },
+    {
+      "epoch": 0.6271400252297711,
+      "grad_norm": 0.330572247505188,
+      "learning_rate": 3.6664362126255087e-06,
+      "loss": 0.4725,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6276806631825554,
+      "grad_norm": 0.3532167971134186,
+      "learning_rate": 3.657340662628116e-06,
+      "loss": 0.4659,
+      "step": 1161
+    },
+    {
+      "epoch": 0.6282213011353397,
+      "grad_norm": 0.33118587732315063,
+      "learning_rate": 3.648249898477707e-06,
+      "loss": 0.4868,
+      "step": 1162
+    },
+    {
+      "epoch": 0.628761939088124,
+      "grad_norm": 0.36354872584342957,
+      "learning_rate": 3.6391639525778915e-06,
+      "loss": 0.4936,
+      "step": 1163
+    },
+    {
+      "epoch": 0.6293025770409083,
+      "grad_norm": 0.3409719467163086,
+      "learning_rate": 3.6300828573150977e-06,
+      "loss": 0.4667,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6298432149936926,
+      "grad_norm": 0.35681483149528503,
+      "learning_rate": 3.621006645058472e-06,
+      "loss": 0.4615,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6303838529464768,
+      "grad_norm": 0.3582918643951416,
+      "learning_rate": 3.6119353481597504e-06,
+      "loss": 0.4789,
+      "step": 1166
+    },
+    {
+      "epoch": 0.6309244908992612,
+      "grad_norm": 0.39356887340545654,
+      "learning_rate": 3.6028689989531533e-06,
+      "loss": 0.4494,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6314651288520454,
+      "grad_norm": 0.36148393154144287,
+      "learning_rate": 3.593807629755258e-06,
+      "loss": 0.4842,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6320057668048297,
+      "grad_norm": 0.35367411375045776,
+      "learning_rate": 3.584751272864899e-06,
+      "loss": 0.4716,
+      "step": 1169
+    },
+    {
+      "epoch": 0.632546404757614,
+      "grad_norm": 0.3774411976337433,
+      "learning_rate": 3.575699960563038e-06,
+      "loss": 0.4452,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6330870427103983,
+      "grad_norm": 0.36172568798065186,
+      "learning_rate": 3.566653725112661e-06,
+      "loss": 0.4657,
+      "step": 1171
+    },
+    {
+      "epoch": 0.6336276806631825,
+      "grad_norm": 0.368366539478302,
+      "learning_rate": 3.557612598758652e-06,
+      "loss": 0.4793,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6341683186159668,
+      "grad_norm": 0.36106520891189575,
+      "learning_rate": 3.5485766137276894e-06,
+      "loss": 0.4885,
+      "step": 1173
+    },
+    {
+      "epoch": 0.6347089565687511,
+      "grad_norm": 0.3237360119819641,
+      "learning_rate": 3.5395458022281205e-06,
+      "loss": 0.4771,
+      "step": 1174
+    },
+    {
+      "epoch": 0.6352495945215354,
+      "grad_norm": 0.3604816794395447,
+      "learning_rate": 3.5305201964498557e-06,
+      "loss": 0.4931,
+      "step": 1175
+    },
+    {
+      "epoch": 0.6357902324743196,
+      "grad_norm": 0.3982067108154297,
+      "learning_rate": 3.5214998285642517e-06,
+      "loss": 0.4677,
+      "step": 1176
+    },
+    {
+      "epoch": 0.636330870427104,
+      "grad_norm": 0.3789684474468231,
+      "learning_rate": 3.5124847307239863e-06,
+      "loss": 0.4994,
+      "step": 1177
+    },
+    {
+      "epoch": 0.6368715083798883,
+      "grad_norm": 0.3449534773826599,
+      "learning_rate": 3.5034749350629593e-06,
+      "loss": 0.4724,
+      "step": 1178
+    },
+    {
+      "epoch": 0.6374121463326725,
+      "grad_norm": 0.361224889755249,
+      "learning_rate": 3.4944704736961722e-06,
+      "loss": 0.4406,
+      "step": 1179
+    },
+    {
+      "epoch": 0.6379527842854569,
+      "grad_norm": 0.41354840993881226,
+      "learning_rate": 3.4854713787196105e-06,
+      "loss": 0.4721,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6384934222382411,
+      "grad_norm": 0.351012647151947,
+      "learning_rate": 3.4764776822101275e-06,
+      "loss": 0.4582,
+      "step": 1181
+    },
+    {
+      "epoch": 0.6390340601910254,
+      "grad_norm": 0.39335301518440247,
+      "learning_rate": 3.4674894162253404e-06,
+      "loss": 0.4826,
+      "step": 1182
+    },
+    {
+      "epoch": 0.6395746981438097,
+      "grad_norm": 0.34970593452453613,
+      "learning_rate": 3.458506612803505e-06,
+      "loss": 0.4828,
+      "step": 1183
+    },
+    {
+      "epoch": 0.640115336096594,
+      "grad_norm": 0.38310420513153076,
+      "learning_rate": 3.4495293039634113e-06,
+      "loss": 0.4851,
+      "step": 1184
+    },
+    {
+      "epoch": 0.6406559740493782,
+      "grad_norm": 0.4005867838859558,
+      "learning_rate": 3.440557521704256e-06,
+      "loss": 0.4729,
+      "step": 1185
+    },
+    {
+      "epoch": 0.6411966120021626,
+      "grad_norm": 0.4233829975128174,
+      "learning_rate": 3.4315912980055433e-06,
+      "loss": 0.4519,
+      "step": 1186
+    },
+    {
+      "epoch": 0.6417372499549469,
+      "grad_norm": 0.3797195851802826,
+      "learning_rate": 3.4226306648269616e-06,
+      "loss": 0.4758,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6422778879077311,
+      "grad_norm": 0.35736843943595886,
+      "learning_rate": 3.413675654108275e-06,
+      "loss": 0.4627,
+      "step": 1188
+    },
+    {
+      "epoch": 0.6428185258605154,
+      "grad_norm": 0.3385578989982605,
+      "learning_rate": 3.4047262977692014e-06,
+      "loss": 0.471,
+      "step": 1189
+    },
+    {
+      "epoch": 0.6433591638132997,
+      "grad_norm": 0.3547685146331787,
+      "learning_rate": 3.3957826277093074e-06,
+      "loss": 0.4813,
+      "step": 1190
+    },
+    {
+      "epoch": 0.643899801766084,
+      "grad_norm": 0.3947050869464874,
+      "learning_rate": 3.3868446758078897e-06,
+      "loss": 0.4988,
+      "step": 1191
+    },
+    {
+      "epoch": 0.6444404397188682,
+      "grad_norm": 0.3719642162322998,
+      "learning_rate": 3.3779124739238657e-06,
+      "loss": 0.4632,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6449810776716526,
+      "grad_norm": 0.3777051866054535,
+      "learning_rate": 3.3689860538956547e-06,
+      "loss": 0.4482,
+      "step": 1193
+    },
+    {
+      "epoch": 0.6455217156244368,
+      "grad_norm": 0.3597317337989807,
+      "learning_rate": 3.3600654475410643e-06,
+      "loss": 0.4687,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6460623535772211,
+      "grad_norm": 0.4228699803352356,
+      "learning_rate": 3.351150686657185e-06,
+      "loss": 0.4585,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6466029915300054,
+      "grad_norm": 0.3539244532585144,
+      "learning_rate": 3.3422418030202696e-06,
+      "loss": 0.4474,
+      "step": 1196
+    },
+    {
+      "epoch": 0.6471436294827897,
+      "grad_norm": 0.3758021891117096,
+      "learning_rate": 3.3333388283856195e-06,
+      "loss": 0.4838,
+      "step": 1197
+    },
+    {
+      "epoch": 0.647684267435574,
+      "grad_norm": 0.3555963635444641,
+      "learning_rate": 3.324441794487475e-06,
+      "loss": 0.4614,
+      "step": 1198
+    },
+    {
+      "epoch": 0.6482249053883583,
+      "grad_norm": 0.3526434302330017,
+      "learning_rate": 3.3155507330389004e-06,
+      "loss": 0.4629,
+      "step": 1199
+    },
+    {
+      "epoch": 0.6487655433411426,
+      "grad_norm": 0.38898134231567383,
+      "learning_rate": 3.306665675731674e-06,
+      "loss": 0.4907,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6493061812939268,
+      "grad_norm": 0.32450518012046814,
+      "learning_rate": 3.297786654236169e-06,
+      "loss": 0.4781,
+      "step": 1201
+    },
+    {
+      "epoch": 0.6498468192467112,
+      "grad_norm": 0.36460715532302856,
+      "learning_rate": 3.28891370020125e-06,
+      "loss": 0.4958,
+      "step": 1202
+    },
+    {
+      "epoch": 0.6503874571994954,
+      "grad_norm": 0.34277579188346863,
+      "learning_rate": 3.280046845254145e-06,
+      "loss": 0.4979,
+      "step": 1203
+    },
+    {
+      "epoch": 0.6509280951522797,
+      "grad_norm": 0.33557090163230896,
+      "learning_rate": 3.2711861210003503e-06,
+      "loss": 0.4811,
+      "step": 1204
+    },
+    {
+      "epoch": 0.6514687331050639,
+      "grad_norm": 0.3325314521789551,
+      "learning_rate": 3.2623315590235076e-06,
+      "loss": 0.4755,
+      "step": 1205
+    },
+    {
+      "epoch": 0.6520093710578483,
+      "grad_norm": 0.3976009488105774,
+      "learning_rate": 3.2534831908852914e-06,
+      "loss": 0.4736,
+      "step": 1206
+    },
+    {
+      "epoch": 0.6525500090106325,
+      "grad_norm": 0.3497442305088043,
+      "learning_rate": 3.244641048125301e-06,
+      "loss": 0.4567,
+      "step": 1207
+    },
+    {
+      "epoch": 0.6530906469634168,
+      "grad_norm": 0.3687249422073364,
+      "learning_rate": 3.235805162260942e-06,
+      "loss": 0.4592,
+      "step": 1208
+    },
+    {
+      "epoch": 0.6536312849162011,
+      "grad_norm": 0.36228495836257935,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.4593,
+      "step": 1209
+    },
+    {
+      "epoch": 0.6541719228689854,
+      "grad_norm": 0.34000781178474426,
+      "learning_rate": 3.218152287177133e-06,
+      "loss": 0.4895,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6547125608217697,
+      "grad_norm": 0.37587717175483704,
+      "learning_rate": 3.2093353608805368e-06,
+      "loss": 0.4689,
+      "step": 1211
+    },
+    {
+      "epoch": 0.655253198774554,
+      "grad_norm": 0.35287749767303467,
+      "learning_rate": 3.2005248173250593e-06,
+      "loss": 0.4846,
+      "step": 1212
+    },
+    {
+      "epoch": 0.6557938367273383,
+      "grad_norm": 0.4108916223049164,
+      "learning_rate": 3.1917206879154762e-06,
+      "loss": 0.4592,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6563344746801225,
+      "grad_norm": 0.3818206489086151,
+      "learning_rate": 3.1829230040336967e-06,
+      "loss": 0.4836,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6568751126329069,
+      "grad_norm": 0.38580086827278137,
+      "learning_rate": 3.1741317970386597e-06,
+      "loss": 0.4723,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6574157505856911,
+      "grad_norm": 0.37795159220695496,
+      "learning_rate": 3.1653470982662114e-06,
+      "loss": 0.472,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6579563885384754,
+      "grad_norm": 0.46705180406570435,
+      "learning_rate": 3.1565689390290067e-06,
+      "loss": 0.4617,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6584970264912597,
+      "grad_norm": 0.3606800138950348,
+      "learning_rate": 3.147797350616385e-06,
+      "loss": 0.494,
+      "step": 1218
+    },
+    {
+      "epoch": 0.659037664444044,
+      "grad_norm": 0.335356742143631,
+      "learning_rate": 3.139032364294271e-06,
+      "loss": 0.4466,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6595783023968282,
+      "grad_norm": 0.3441433906555176,
+      "learning_rate": 3.130274011305047e-06,
+      "loss": 0.4745,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6601189403496125,
+      "grad_norm": 0.33748725056648254,
+      "learning_rate": 3.1215223228674587e-06,
+      "loss": 0.4989,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6606595783023969,
+      "grad_norm": 0.3482097387313843,
+      "learning_rate": 3.1127773301764935e-06,
+      "loss": 0.4774,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6612002162551811,
+      "grad_norm": 0.3453892171382904,
+      "learning_rate": 3.1040390644032746e-06,
+      "loss": 0.4759,
+      "step": 1223
+    },
+    {
+      "epoch": 0.6617408542079654,
+      "grad_norm": 0.36808252334594727,
+      "learning_rate": 3.095307556694942e-06,
+      "loss": 0.4692,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6622814921607497,
+      "grad_norm": 0.4075857102870941,
+      "learning_rate": 3.0865828381745515e-06,
+      "loss": 0.4607,
+      "step": 1225
+    },
+    {
+      "epoch": 0.662822130113534,
+      "grad_norm": 0.3461955189704895,
+      "learning_rate": 3.077864939940959e-06,
+      "loss": 0.4696,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6633627680663182,
+      "grad_norm": 0.3581470847129822,
+      "learning_rate": 3.0691538930687076e-06,
+      "loss": 0.468,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6639034060191026,
+      "grad_norm": 0.34455424547195435,
+      "learning_rate": 3.0604497286079227e-06,
+      "loss": 0.4435,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6644440439718868,
+      "grad_norm": 0.35402852296829224,
+      "learning_rate": 3.051752477584191e-06,
+      "loss": 0.4831,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6649846819246711,
+      "grad_norm": 0.35135895013809204,
+      "learning_rate": 3.043062170998464e-06,
+      "loss": 0.4827,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6655253198774554,
+      "grad_norm": 0.3202992081642151,
+      "learning_rate": 3.0343788398269342e-06,
+      "loss": 0.4442,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6660659578302397,
+      "grad_norm": 0.3709985613822937,
+      "learning_rate": 3.025702515020937e-06,
+      "loss": 0.4635,
+      "step": 1232
+    },
+    {
+      "epoch": 0.666606595783024,
+      "grad_norm": 0.358951598405838,
+      "learning_rate": 3.0170332275068247e-06,
+      "loss": 0.4464,
+      "step": 1233
+    },
+    {
+      "epoch": 0.6671472337358082,
+      "grad_norm": 0.3573668897151947,
+      "learning_rate": 3.0083710081858748e-06,
+      "loss": 0.48,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6676878716885926,
+      "grad_norm": 0.34597110748291016,
+      "learning_rate": 2.9997158879341647e-06,
+      "loss": 0.4718,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6682285096413768,
+      "grad_norm": 0.3602089583873749,
+      "learning_rate": 2.9910678976024733e-06,
+      "loss": 0.449,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6687691475941611,
+      "grad_norm": 0.34547102451324463,
+      "learning_rate": 2.982427068016155e-06,
+      "loss": 0.4584,
+      "step": 1237
+    },
+    {
+      "epoch": 0.6693097855469454,
+      "grad_norm": 0.3194965124130249,
+      "learning_rate": 2.9737934299750514e-06,
+      "loss": 0.475,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6698504234997297,
+      "grad_norm": 0.37800660729408264,
+      "learning_rate": 2.965167014253363e-06,
+      "loss": 0.4739,
+      "step": 1239
+    },
+    {
+      "epoch": 0.6703910614525139,
+      "grad_norm": 0.35336002707481384,
+      "learning_rate": 2.956547851599548e-06,
+      "loss": 0.4768,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6709316994052983,
+      "grad_norm": 0.3484102785587311,
+      "learning_rate": 2.947935972736217e-06,
+      "loss": 0.4606,
+      "step": 1241
+    },
+    {
+      "epoch": 0.6714723373580825,
+      "grad_norm": 0.33098798990249634,
+      "learning_rate": 2.9393314083600076e-06,
+      "loss": 0.4737,
+      "step": 1242
+    },
+    {
+      "epoch": 0.6720129753108668,
+      "grad_norm": 0.33991244435310364,
+      "learning_rate": 2.930734189141492e-06,
+      "loss": 0.4609,
+      "step": 1243
+    },
+    {
+      "epoch": 0.6725536132636512,
+      "grad_norm": 0.3231086730957031,
+      "learning_rate": 2.922144345725062e-06,
+      "loss": 0.4668,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6730942512164354,
+      "grad_norm": 0.3015232980251312,
+      "learning_rate": 2.9135619087288153e-06,
+      "loss": 0.4378,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6736348891692197,
+      "grad_norm": 0.3297620415687561,
+      "learning_rate": 2.9049869087444493e-06,
+      "loss": 0.4843,
+      "step": 1246
+    },
+    {
+      "epoch": 0.674175527122004,
+      "grad_norm": 0.3333333134651184,
+      "learning_rate": 2.8964193763371546e-06,
+      "loss": 0.4523,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6747161650747883,
+      "grad_norm": 0.3433511257171631,
+      "learning_rate": 2.887859342045506e-06,
+      "loss": 0.4698,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6752568030275725,
+      "grad_norm": 0.3787667751312256,
+      "learning_rate": 2.879306836381345e-06,
+      "loss": 0.4556,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6757974409803568,
+      "grad_norm": 0.3427095413208008,
+      "learning_rate": 2.8707618898296864e-06,
+      "loss": 0.4689,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6763380789331411,
+      "grad_norm": 0.3421780467033386,
+      "learning_rate": 2.862224532848591e-06,
+      "loss": 0.4805,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6768787168859254,
+      "grad_norm": 0.37503376603126526,
+      "learning_rate": 2.853694795869074e-06,
+      "loss": 0.4946,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6774193548387096,
+      "grad_norm": 0.33137035369873047,
+      "learning_rate": 2.845172709294989e-06,
+      "loss": 0.4447,
+      "step": 1253
+    },
+    {
+      "epoch": 0.677959992791494,
+      "grad_norm": 0.38808560371398926,
+      "learning_rate": 2.8366583035029194e-06,
+      "loss": 0.4918,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6785006307442782,
+      "grad_norm": 0.31489184498786926,
+      "learning_rate": 2.8281516088420665e-06,
+      "loss": 0.4527,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6790412686970625,
+      "grad_norm": 0.33485811948776245,
+      "learning_rate": 2.819652655634151e-06,
+      "loss": 0.4692,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6795819066498469,
+      "grad_norm": 0.3656138777732849,
+      "learning_rate": 2.8111614741732975e-06,
+      "loss": 0.4612,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6801225446026311,
+      "grad_norm": 0.3511470854282379,
+      "learning_rate": 2.802678094725931e-06,
+      "loss": 0.4532,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6806631825554154,
+      "grad_norm": 0.3320247530937195,
+      "learning_rate": 2.794202547530661e-06,
+      "loss": 0.4452,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6812038205081997,
+      "grad_norm": 0.3250149190425873,
+      "learning_rate": 2.785734862798184e-06,
+      "loss": 0.4722,
+      "step": 1260
+    },
+    {
+      "epoch": 0.681744458460984,
+      "grad_norm": 0.34146052598953247,
+      "learning_rate": 2.77727507071117e-06,
+      "loss": 0.4769,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6822850964137682,
+      "grad_norm": 0.3707115054130554,
+      "learning_rate": 2.768823201424158e-06,
+      "loss": 0.4716,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6828257343665526,
+      "grad_norm": 0.3220188617706299,
+      "learning_rate": 2.7603792850634402e-06,
+      "loss": 0.4858,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6833663723193368,
+      "grad_norm": 0.3371463716030121,
+      "learning_rate": 2.7519433517269665e-06,
+      "loss": 0.4778,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6839070102721211,
+      "grad_norm": 0.33046388626098633,
+      "learning_rate": 2.7435154314842337e-06,
+      "loss": 0.449,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6844476482249053,
+      "grad_norm": 0.33768966794013977,
+      "learning_rate": 2.7350955543761682e-06,
+      "loss": 0.4591,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6849882861776897,
+      "grad_norm": 0.35337385535240173,
+      "learning_rate": 2.7266837504150345e-06,
+      "loss": 0.4649,
+      "step": 1267
+    },
+    {
+      "epoch": 0.685528924130474,
+      "grad_norm": 0.4103326201438904,
+      "learning_rate": 2.7182800495843166e-06,
+      "loss": 0.4778,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6860695620832582,
+      "grad_norm": 0.3523549437522888,
+      "learning_rate": 2.7098844818386164e-06,
+      "loss": 0.4656,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6866102000360426,
+      "grad_norm": 0.3538891077041626,
+      "learning_rate": 2.7014970771035474e-06,
+      "loss": 0.4821,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6871508379888268,
+      "grad_norm": 0.3851732313632965,
+      "learning_rate": 2.6931178652756262e-06,
+      "loss": 0.4928,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6876914759416111,
+      "grad_norm": 0.3899889290332794,
+      "learning_rate": 2.6847468762221616e-06,
+      "loss": 0.4867,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6882321138943954,
+      "grad_norm": 0.4064536690711975,
+      "learning_rate": 2.6763841397811576e-06,
+      "loss": 0.4629,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6887727518471797,
+      "grad_norm": 0.34950196743011475,
+      "learning_rate": 2.668029685761201e-06,
+      "loss": 0.4284,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6893133897999639,
+      "grad_norm": 0.36954036355018616,
+      "learning_rate": 2.6596835439413584e-06,
+      "loss": 0.4834,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6898540277527483,
+      "grad_norm": 0.3687545657157898,
+      "learning_rate": 2.6513457440710612e-06,
+      "loss": 0.4632,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6903946657055325,
+      "grad_norm": 0.41740676760673523,
+      "learning_rate": 2.6430163158700116e-06,
+      "loss": 0.4802,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6909353036583168,
+      "grad_norm": 0.3302917182445526,
+      "learning_rate": 2.634695289028072e-06,
+      "loss": 0.4417,
+      "step": 1278
+    },
+    {
+      "epoch": 0.6914759416111012,
+      "grad_norm": 0.3250352144241333,
+      "learning_rate": 2.6263826932051562e-06,
+      "loss": 0.465,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6920165795638854,
+      "grad_norm": 0.35972073674201965,
+      "learning_rate": 2.6180785580311284e-06,
+      "loss": 0.467,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6925572175166697,
+      "grad_norm": 0.34688135981559753,
+      "learning_rate": 2.609782913105691e-06,
+      "loss": 0.4528,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6930978554694539,
+      "grad_norm": 0.38405537605285645,
+      "learning_rate": 2.601495787998288e-06,
+      "loss": 0.4902,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6936384934222383,
+      "grad_norm": 0.38303545117378235,
+      "learning_rate": 2.59321721224799e-06,
+      "loss": 0.4536,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6941791313750225,
+      "grad_norm": 0.34185272455215454,
+      "learning_rate": 2.5849472153634003e-06,
+      "loss": 0.4441,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6947197693278068,
+      "grad_norm": 0.41703319549560547,
+      "learning_rate": 2.576685826822535e-06,
+      "loss": 0.4781,
+      "step": 1285
+    },
+    {
+      "epoch": 0.6952604072805911,
+      "grad_norm": 0.3938962519168854,
+      "learning_rate": 2.568433076072734e-06,
+      "loss": 0.4755,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6958010452333754,
+      "grad_norm": 0.36541613936424255,
+      "learning_rate": 2.5601889925305433e-06,
+      "loss": 0.4789,
+      "step": 1287
+    },
+    {
+      "epoch": 0.6963416831861596,
+      "grad_norm": 0.3563724458217621,
+      "learning_rate": 2.5519536055816194e-06,
+      "loss": 0.4781,
+      "step": 1288
+    },
+    {
+      "epoch": 0.696882321138944,
+      "grad_norm": 0.34606099128723145,
+      "learning_rate": 2.5437269445806146e-06,
+      "loss": 0.4485,
+      "step": 1289
+    },
+    {
+      "epoch": 0.6974229590917282,
+      "grad_norm": 0.42106860876083374,
+      "learning_rate": 2.5355090388510806e-06,
+      "loss": 0.4739,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6979635970445125,
+      "grad_norm": 0.4748215973377228,
+      "learning_rate": 2.527299917685362e-06,
+      "loss": 0.4659,
+      "step": 1291
+    },
+    {
+      "epoch": 0.6985042349972969,
+      "grad_norm": 0.38426297903060913,
+      "learning_rate": 2.519099610344492e-06,
+      "loss": 0.4976,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6990448729500811,
+      "grad_norm": 0.3022260069847107,
+      "learning_rate": 2.5109081460580875e-06,
+      "loss": 0.4347,
+      "step": 1293
+    },
+    {
+      "epoch": 0.6995855109028654,
+      "grad_norm": 0.35534849762916565,
+      "learning_rate": 2.502725554024239e-06,
+      "loss": 0.4635,
+      "step": 1294
+    },
+    {
+      "epoch": 0.7001261488556497,
+      "grad_norm": 0.35701191425323486,
+      "learning_rate": 2.494551863409418e-06,
+      "loss": 0.4827,
+      "step": 1295
+    },
+    {
+      "epoch": 0.700666786808434,
+      "grad_norm": 0.37284964323043823,
+      "learning_rate": 2.4863871033483693e-06,
+      "loss": 0.4857,
+      "step": 1296
+    },
+    {
+      "epoch": 0.7012074247612182,
+      "grad_norm": 0.35588300228118896,
+      "learning_rate": 2.478231302943997e-06,
+      "loss": 0.4709,
+      "step": 1297
+    },
+    {
+      "epoch": 0.7017480627140025,
+      "grad_norm": 0.3689674735069275,
+      "learning_rate": 2.470084491267278e-06,
+      "loss": 0.4734,
+      "step": 1298
+    },
+    {
+      "epoch": 0.7022887006667868,
+      "grad_norm": 0.3724137246608734,
+      "learning_rate": 2.46194669735714e-06,
+      "loss": 0.4638,
+      "step": 1299
+    },
+    {
+      "epoch": 0.7028293386195711,
+      "grad_norm": 0.42699527740478516,
+      "learning_rate": 2.4538179502203753e-06,
+      "loss": 0.5024,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7033699765723553,
+      "grad_norm": 0.35276705026626587,
+      "learning_rate": 2.445698278831528e-06,
+      "loss": 0.4693,
+      "step": 1301
+    },
+    {
+      "epoch": 0.7039106145251397,
+      "grad_norm": 0.3660089075565338,
+      "learning_rate": 2.437587712132787e-06,
+      "loss": 0.4544,
+      "step": 1302
+    },
+    {
+      "epoch": 0.704451252477924,
+      "grad_norm": 0.3719891905784607,
+      "learning_rate": 2.429486279033892e-06,
+      "loss": 0.4498,
+      "step": 1303
+    },
+    {
+      "epoch": 0.7049918904307082,
+      "grad_norm": 0.34804993867874146,
+      "learning_rate": 2.4213940084120274e-06,
+      "loss": 0.4814,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7055325283834926,
+      "grad_norm": 0.35139375925064087,
+      "learning_rate": 2.4133109291117156e-06,
+      "loss": 0.472,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7060731663362768,
+      "grad_norm": 0.3766683042049408,
+      "learning_rate": 2.405237069944721e-06,
+      "loss": 0.4688,
+      "step": 1306
+    },
+    {
+      "epoch": 0.7066138042890611,
+      "grad_norm": 0.37273478507995605,
+      "learning_rate": 2.397172459689936e-06,
+      "loss": 0.5089,
+      "step": 1307
+    },
+    {
+      "epoch": 0.7071544422418454,
+      "grad_norm": 0.33368778228759766,
+      "learning_rate": 2.3891171270932923e-06,
+      "loss": 0.4666,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7076950801946297,
+      "grad_norm": 0.3467375338077545,
+      "learning_rate": 2.3810711008676495e-06,
+      "loss": 0.467,
+      "step": 1309
+    },
+    {
+      "epoch": 0.7082357181474139,
+      "grad_norm": 0.39090535044670105,
+      "learning_rate": 2.3730344096926974e-06,
+      "loss": 0.4934,
+      "step": 1310
+    },
+    {
+      "epoch": 0.7087763561001982,
+      "grad_norm": 0.35073044896125793,
+      "learning_rate": 2.3650070822148447e-06,
+      "loss": 0.4396,
+      "step": 1311
+    },
+    {
+      "epoch": 0.7093169940529825,
+      "grad_norm": 0.321822851896286,
+      "learning_rate": 2.3569891470471308e-06,
+      "loss": 0.4375,
+      "step": 1312
+    },
+    {
+      "epoch": 0.7098576320057668,
+      "grad_norm": 0.3142220377922058,
+      "learning_rate": 2.3489806327691156e-06,
+      "loss": 0.4711,
+      "step": 1313
+    },
+    {
+      "epoch": 0.710398269958551,
+      "grad_norm": 0.3423789143562317,
+      "learning_rate": 2.3409815679267733e-06,
+      "loss": 0.4384,
+      "step": 1314
+    },
+    {
+      "epoch": 0.7109389079113354,
+      "grad_norm": 0.3237120509147644,
+      "learning_rate": 2.3329919810324036e-06,
+      "loss": 0.4744,
+      "step": 1315
+    },
+    {
+      "epoch": 0.7114795458641197,
+      "grad_norm": 0.3413024842739105,
+      "learning_rate": 2.325011900564515e-06,
+      "loss": 0.4676,
+      "step": 1316
+    },
+    {
+      "epoch": 0.7120201838169039,
+      "grad_norm": 0.3537076711654663,
+      "learning_rate": 2.3170413549677367e-06,
+      "loss": 0.4839,
+      "step": 1317
+    },
+    {
+      "epoch": 0.7125608217696883,
+      "grad_norm": 0.3629612624645233,
+      "learning_rate": 2.3090803726527083e-06,
+      "loss": 0.5076,
+      "step": 1318
+    },
+    {
+      "epoch": 0.7131014597224725,
+      "grad_norm": 0.32516616582870483,
+      "learning_rate": 2.301128981995985e-06,
+      "loss": 0.4485,
+      "step": 1319
+    },
+    {
+      "epoch": 0.7136420976752568,
+      "grad_norm": 0.38808518648147583,
+      "learning_rate": 2.293187211339926e-06,
+      "loss": 0.4638,
+      "step": 1320
+    },
+    {
+      "epoch": 0.7141827356280411,
+      "grad_norm": 0.3397766351699829,
+      "learning_rate": 2.2852550889926067e-06,
+      "loss": 0.4937,
+      "step": 1321
+    },
+    {
+      "epoch": 0.7147233735808254,
+      "grad_norm": 0.3628048896789551,
+      "learning_rate": 2.2773326432277097e-06,
+      "loss": 0.4832,
+      "step": 1322
+    },
+    {
+      "epoch": 0.7152640115336096,
+      "grad_norm": 0.32281026244163513,
+      "learning_rate": 2.2694199022844284e-06,
+      "loss": 0.4313,
+      "step": 1323
+    },
+    {
+      "epoch": 0.715804649486394,
+      "grad_norm": 0.33235985040664673,
+      "learning_rate": 2.261516894367356e-06,
+      "loss": 0.4402,
+      "step": 1324
+    },
+    {
+      "epoch": 0.7163452874391782,
+      "grad_norm": 0.3623601198196411,
+      "learning_rate": 2.2536236476464007e-06,
+      "loss": 0.4644,
+      "step": 1325
+    },
+    {
+      "epoch": 0.7168859253919625,
+      "grad_norm": 0.32898762822151184,
+      "learning_rate": 2.2457401902566745e-06,
+      "loss": 0.4539,
+      "step": 1326
+    },
+    {
+      "epoch": 0.7174265633447467,
+      "grad_norm": 0.34005796909332275,
+      "learning_rate": 2.2378665502983976e-06,
+      "loss": 0.4869,
+      "step": 1327
+    },
+    {
+      "epoch": 0.7179672012975311,
+      "grad_norm": 0.3692139983177185,
+      "learning_rate": 2.2300027558367917e-06,
+      "loss": 0.4671,
+      "step": 1328
+    },
+    {
+      "epoch": 0.7185078392503154,
+      "grad_norm": 0.3715006113052368,
+      "learning_rate": 2.2221488349019903e-06,
+      "loss": 0.4843,
+      "step": 1329
+    },
+    {
+      "epoch": 0.7190484772030996,
+      "grad_norm": 0.30792778730392456,
+      "learning_rate": 2.2143048154889272e-06,
+      "loss": 0.4696,
+      "step": 1330
+    },
+    {
+      "epoch": 0.719589115155884,
+      "grad_norm": 0.3531022071838379,
+      "learning_rate": 2.2064707255572494e-06,
+      "loss": 0.4716,
+      "step": 1331
+    },
+    {
+      "epoch": 0.7201297531086682,
+      "grad_norm": 0.3335738182067871,
+      "learning_rate": 2.1986465930312067e-06,
+      "loss": 0.4389,
+      "step": 1332
+    },
+    {
+      "epoch": 0.7206703910614525,
+      "grad_norm": 0.3675576150417328,
+      "learning_rate": 2.1908324457995556e-06,
+      "loss": 0.4436,
+      "step": 1333
+    },
+    {
+      "epoch": 0.7212110290142368,
+      "grad_norm": 0.33168819546699524,
+      "learning_rate": 2.1830283117154616e-06,
+      "loss": 0.4308,
+      "step": 1334
+    },
+    {
+      "epoch": 0.7217516669670211,
+      "grad_norm": 0.36723217368125916,
+      "learning_rate": 2.1752342185964003e-06,
+      "loss": 0.4769,
+      "step": 1335
+    },
+    {
+      "epoch": 0.7222923049198053,
+      "grad_norm": 0.36966997385025024,
+      "learning_rate": 2.1674501942240567e-06,
+      "loss": 0.4624,
+      "step": 1336
+    },
+    {
+      "epoch": 0.7228329428725897,
+      "grad_norm": 0.3729422986507416,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.4774,
+      "step": 1337
+    },
+    {
+      "epoch": 0.723373580825374,
+      "grad_norm": 0.3431287407875061,
+      "learning_rate": 2.151912462666703e-06,
+      "loss": 0.4451,
+      "step": 1338
+    },
+    {
+      "epoch": 0.7239142187781582,
+      "grad_norm": 0.3624579906463623,
+      "learning_rate": 2.144158810865217e-06,
+      "loss": 0.4711,
+      "step": 1339
+    },
+    {
+      "epoch": 0.7244548567309426,
+      "grad_norm": 0.3302533030509949,
+      "learning_rate": 2.1364153385773007e-06,
+      "loss": 0.4802,
+      "step": 1340
+    },
+    {
+      "epoch": 0.7249954946837268,
+      "grad_norm": 0.34277042746543884,
+      "learning_rate": 2.128682073404197e-06,
+      "loss": 0.4794,
+      "step": 1341
+    },
+    {
+      "epoch": 0.7255361326365111,
+      "grad_norm": 0.3795274794101715,
+      "learning_rate": 2.1209590429107734e-06,
+      "loss": 0.4614,
+      "step": 1342
+    },
+    {
+      "epoch": 0.7260767705892953,
+      "grad_norm": 0.3579147458076477,
+      "learning_rate": 2.1132462746254147e-06,
+      "loss": 0.4946,
+      "step": 1343
+    },
+    {
+      "epoch": 0.7266174085420797,
+      "grad_norm": 0.36164700984954834,
+      "learning_rate": 2.1055437960399266e-06,
+      "loss": 0.4427,
+      "step": 1344
+    },
+    {
+      "epoch": 0.7271580464948639,
+      "grad_norm": 0.3287413418292999,
+      "learning_rate": 2.0978516346094342e-06,
+      "loss": 0.4723,
+      "step": 1345
+    },
+    {
+      "epoch": 0.7276986844476482,
+      "grad_norm": 0.33054518699645996,
+      "learning_rate": 2.0901698177522944e-06,
+      "loss": 0.4589,
+      "step": 1346
+    },
+    {
+      "epoch": 0.7282393224004325,
+      "grad_norm": 0.3471773862838745,
+      "learning_rate": 2.082498372849983e-06,
+      "loss": 0.4746,
+      "step": 1347
+    },
+    {
+      "epoch": 0.7287799603532168,
+      "grad_norm": 0.34974491596221924,
+      "learning_rate": 2.074837327247012e-06,
+      "loss": 0.487,
+      "step": 1348
+    },
+    {
+      "epoch": 0.729320598306001,
+      "grad_norm": 0.3569546937942505,
+      "learning_rate": 2.067186708250826e-06,
+      "loss": 0.4852,
+      "step": 1349
+    },
+    {
+      "epoch": 0.7298612362587854,
+      "grad_norm": 0.31537672877311707,
+      "learning_rate": 2.059546543131696e-06,
+      "loss": 0.4561,
+      "step": 1350
+    },
+    {
+      "epoch": 0.7304018742115697,
+      "grad_norm": 0.2963314950466156,
+      "learning_rate": 2.051916859122641e-06,
+      "loss": 0.4487,
+      "step": 1351
+    },
+    {
+      "epoch": 0.7309425121643539,
+      "grad_norm": 0.31170910596847534,
+      "learning_rate": 2.0442976834193146e-06,
+      "loss": 0.4585,
+      "step": 1352
+    },
+    {
+      "epoch": 0.7314831501171383,
+      "grad_norm": 0.3321332037448883,
+      "learning_rate": 2.036689043179917e-06,
+      "loss": 0.4635,
+      "step": 1353
+    },
+    {
+      "epoch": 0.7320237880699225,
+      "grad_norm": 0.34618934988975525,
+      "learning_rate": 2.0290909655250913e-06,
+      "loss": 0.4654,
+      "step": 1354
+    },
+    {
+      "epoch": 0.7325644260227068,
+      "grad_norm": 0.3425121009349823,
+      "learning_rate": 2.0215034775378336e-06,
+      "loss": 0.4692,
+      "step": 1355
+    },
+    {
+      "epoch": 0.7331050639754911,
+      "grad_norm": 0.33614981174468994,
+      "learning_rate": 2.013926606263394e-06,
+      "loss": 0.473,
+      "step": 1356
+    },
+    {
+      "epoch": 0.7336457019282754,
+      "grad_norm": 0.3264980614185333,
+      "learning_rate": 2.0063603787091788e-06,
+      "loss": 0.452,
+      "step": 1357
+    },
+    {
+      "epoch": 0.7341863398810596,
+      "grad_norm": 0.33285725116729736,
+      "learning_rate": 1.9988048218446577e-06,
+      "loss": 0.4593,
+      "step": 1358
+    },
+    {
+      "epoch": 0.7347269778338439,
+      "grad_norm": 0.367969274520874,
+      "learning_rate": 1.9912599626012593e-06,
+      "loss": 0.4649,
+      "step": 1359
+    },
+    {
+      "epoch": 0.7352676157866282,
+      "grad_norm": 0.33582252264022827,
+      "learning_rate": 1.9837258278722855e-06,
+      "loss": 0.4726,
+      "step": 1360
+    },
+    {
+      "epoch": 0.7358082537394125,
+      "grad_norm": 0.3242082893848419,
+      "learning_rate": 1.976202444512813e-06,
+      "loss": 0.4946,
+      "step": 1361
+    },
+    {
+      "epoch": 0.7363488916921967,
+      "grad_norm": 0.33312639594078064,
+      "learning_rate": 1.96868983933959e-06,
+      "loss": 0.4802,
+      "step": 1362
+    },
+    {
+      "epoch": 0.7368895296449811,
+      "grad_norm": 0.3373234272003174,
+      "learning_rate": 1.9611880391309524e-06,
+      "loss": 0.4847,
+      "step": 1363
+    },
+    {
+      "epoch": 0.7374301675977654,
+      "grad_norm": 0.3237764835357666,
+      "learning_rate": 1.9536970706267156e-06,
+      "loss": 0.4641,
+      "step": 1364
+    },
+    {
+      "epoch": 0.7379708055505496,
+      "grad_norm": 0.33678045868873596,
+      "learning_rate": 1.946216960528092e-06,
+      "loss": 0.476,
+      "step": 1365
+    },
+    {
+      "epoch": 0.738511443503334,
+      "grad_norm": 0.3523315489292145,
+      "learning_rate": 1.9387477354975885e-06,
+      "loss": 0.4559,
+      "step": 1366
+    },
+    {
+      "epoch": 0.7390520814561182,
+      "grad_norm": 0.3299519121646881,
+      "learning_rate": 1.9312894221589085e-06,
+      "loss": 0.4554,
+      "step": 1367
+    },
+    {
+      "epoch": 0.7395927194089025,
+      "grad_norm": 0.3041582703590393,
+      "learning_rate": 1.9238420470968665e-06,
+      "loss": 0.4544,
+      "step": 1368
+    },
+    {
+      "epoch": 0.7401333573616868,
+      "grad_norm": 0.30898869037628174,
+      "learning_rate": 1.9164056368572847e-06,
+      "loss": 0.4635,
+      "step": 1369
+    },
+    {
+      "epoch": 0.7406739953144711,
+      "grad_norm": 0.3286183178424835,
+      "learning_rate": 1.9089802179469036e-06,
+      "loss": 0.4673,
+      "step": 1370
+    },
+    {
+      "epoch": 0.7412146332672553,
+      "grad_norm": 0.3371999263763428,
+      "learning_rate": 1.9015658168332863e-06,
+      "loss": 0.4848,
+      "step": 1371
+    },
+    {
+      "epoch": 0.7417552712200396,
+      "grad_norm": 0.33480530977249146,
+      "learning_rate": 1.8941624599447178e-06,
+      "loss": 0.4687,
+      "step": 1372
+    },
+    {
+      "epoch": 0.742295909172824,
+      "grad_norm": 0.3701348304748535,
+      "learning_rate": 1.8867701736701238e-06,
+      "loss": 0.4683,
+      "step": 1373
+    },
+    {
+      "epoch": 0.7428365471256082,
+      "grad_norm": 0.33396846055984497,
+      "learning_rate": 1.8793889843589647e-06,
+      "loss": 0.4728,
+      "step": 1374
+    },
+    {
+      "epoch": 0.7433771850783925,
+      "grad_norm": 0.3385055363178253,
+      "learning_rate": 1.87201891832115e-06,
+      "loss": 0.4715,
+      "step": 1375
+    },
+    {
+      "epoch": 0.7439178230311768,
+      "grad_norm": 0.33360961079597473,
+      "learning_rate": 1.8646600018269356e-06,
+      "loss": 0.4777,
+      "step": 1376
+    },
+    {
+      "epoch": 0.7444584609839611,
+      "grad_norm": 0.2998606264591217,
+      "learning_rate": 1.8573122611068406e-06,
+      "loss": 0.4671,
+      "step": 1377
+    },
+    {
+      "epoch": 0.7449990989367453,
+      "grad_norm": 0.3257642984390259,
+      "learning_rate": 1.8499757223515442e-06,
+      "loss": 0.4966,
+      "step": 1378
+    },
+    {
+      "epoch": 0.7455397368895297,
+      "grad_norm": 0.3353062868118286,
+      "learning_rate": 1.8426504117118011e-06,
+      "loss": 0.4495,
+      "step": 1379
+    },
+    {
+      "epoch": 0.7460803748423139,
+      "grad_norm": 0.3439493179321289,
+      "learning_rate": 1.8353363552983382e-06,
+      "loss": 0.4777,
+      "step": 1380
+    },
+    {
+      "epoch": 0.7466210127950982,
+      "grad_norm": 0.38241538405418396,
+      "learning_rate": 1.8280335791817733e-06,
+      "loss": 0.4726,
+      "step": 1381
+    },
+    {
+      "epoch": 0.7471616507478825,
+      "grad_norm": 0.3326997458934784,
+      "learning_rate": 1.8207421093925127e-06,
+      "loss": 0.4798,
+      "step": 1382
+    },
+    {
+      "epoch": 0.7477022887006668,
+      "grad_norm": 0.33782845735549927,
+      "learning_rate": 1.8134619719206624e-06,
+      "loss": 0.4439,
+      "step": 1383
+    },
+    {
+      "epoch": 0.748242926653451,
+      "grad_norm": 0.34692955017089844,
+      "learning_rate": 1.8061931927159377e-06,
+      "loss": 0.4354,
+      "step": 1384
+    },
+    {
+      "epoch": 0.7487835646062354,
+      "grad_norm": 0.3597027659416199,
+      "learning_rate": 1.7989357976875603e-06,
+      "loss": 0.4989,
+      "step": 1385
+    },
+    {
+      "epoch": 0.7493242025590197,
+      "grad_norm": 0.3161109983921051,
+      "learning_rate": 1.7916898127041815e-06,
+      "loss": 0.4341,
+      "step": 1386
+    },
+    {
+      "epoch": 0.7498648405118039,
+      "grad_norm": 0.32751956582069397,
+      "learning_rate": 1.7844552635937784e-06,
+      "loss": 0.4657,
+      "step": 1387
+    },
+    {
+      "epoch": 0.7504054784645882,
+      "grad_norm": 0.32368284463882446,
+      "learning_rate": 1.7772321761435674e-06,
+      "loss": 0.4705,
+      "step": 1388
+    },
+    {
+      "epoch": 0.7509461164173725,
+      "grad_norm": 0.3560972511768341,
+      "learning_rate": 1.7700205760999061e-06,
+      "loss": 0.4442,
+      "step": 1389
+    },
+    {
+      "epoch": 0.7514867543701568,
+      "grad_norm": 0.35852131247520447,
+      "learning_rate": 1.76282048916821e-06,
+      "loss": 0.4745,
+      "step": 1390
+    },
+    {
+      "epoch": 0.752027392322941,
+      "grad_norm": 0.35611191391944885,
+      "learning_rate": 1.7556319410128557e-06,
+      "loss": 0.4735,
+      "step": 1391
+    },
+    {
+      "epoch": 0.7525680302757254,
+      "grad_norm": 0.33755627274513245,
+      "learning_rate": 1.7484549572570913e-06,
+      "loss": 0.4822,
+      "step": 1392
+    },
+    {
+      "epoch": 0.7531086682285096,
+      "grad_norm": 0.33882173895835876,
+      "learning_rate": 1.7412895634829391e-06,
+      "loss": 0.4906,
+      "step": 1393
+    },
+    {
+      "epoch": 0.7536493061812939,
+      "grad_norm": 0.3680201768875122,
+      "learning_rate": 1.7341357852311175e-06,
+      "loss": 0.5208,
+      "step": 1394
+    },
+    {
+      "epoch": 0.7541899441340782,
+      "grad_norm": 0.3476288914680481,
+      "learning_rate": 1.726993648000933e-06,
+      "loss": 0.4545,
+      "step": 1395
+    },
+    {
+      "epoch": 0.7547305820868625,
+      "grad_norm": 0.35119006037712097,
+      "learning_rate": 1.7198631772502057e-06,
+      "loss": 0.4769,
+      "step": 1396
+    },
+    {
+      "epoch": 0.7552712200396468,
+      "grad_norm": 0.3414047360420227,
+      "learning_rate": 1.7127443983951687e-06,
+      "loss": 0.455,
+      "step": 1397
+    },
+    {
+      "epoch": 0.7558118579924311,
+      "grad_norm": 0.3647000193595886,
+      "learning_rate": 1.7056373368103756e-06,
+      "loss": 0.5124,
+      "step": 1398
+    },
+    {
+      "epoch": 0.7563524959452154,
+      "grad_norm": 0.34969907999038696,
+      "learning_rate": 1.6985420178286216e-06,
+      "loss": 0.4956,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7568931338979996,
+      "grad_norm": 0.3402901887893677,
+      "learning_rate": 1.6914584667408408e-06,
+      "loss": 0.455,
+      "step": 1400
+    },
+    {
+      "epoch": 0.757433771850784,
+      "grad_norm": 0.36687296628952026,
+      "learning_rate": 1.6843867087960252e-06,
+      "loss": 0.4694,
+      "step": 1401
+    },
+    {
+      "epoch": 0.7579744098035682,
+      "grad_norm": 0.34027478098869324,
+      "learning_rate": 1.6773267692011242e-06,
+      "loss": 0.4763,
+      "step": 1402
+    },
+    {
+      "epoch": 0.7585150477563525,
+      "grad_norm": 0.30456939339637756,
+      "learning_rate": 1.6702786731209681e-06,
+      "loss": 0.4545,
+      "step": 1403
+    },
+    {
+      "epoch": 0.7590556857091367,
+      "grad_norm": 0.32559987902641296,
+      "learning_rate": 1.6632424456781675e-06,
+      "loss": 0.484,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7595963236619211,
+      "grad_norm": 0.3651071786880493,
+      "learning_rate": 1.6562181119530314e-06,
+      "loss": 0.4981,
+      "step": 1405
+    },
+    {
+      "epoch": 0.7601369616147053,
+      "grad_norm": 0.36218157410621643,
+      "learning_rate": 1.649205696983468e-06,
+      "loss": 0.4905,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7606775995674896,
+      "grad_norm": 0.35047683119773865,
+      "learning_rate": 1.642205225764908e-06,
+      "loss": 0.4697,
+      "step": 1407
+    },
+    {
+      "epoch": 0.761218237520274,
+      "grad_norm": 0.3518739342689514,
+      "learning_rate": 1.635216723250206e-06,
+      "loss": 0.4665,
+      "step": 1408
+    },
+    {
+      "epoch": 0.7617588754730582,
+      "grad_norm": 0.3698537349700928,
+      "learning_rate": 1.6282402143495568e-06,
+      "loss": 0.4555,
+      "step": 1409
+    },
+    {
+      "epoch": 0.7622995134258425,
+      "grad_norm": 0.3216371536254883,
+      "learning_rate": 1.6212757239304e-06,
+      "loss": 0.4591,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7628401513786268,
+      "grad_norm": 0.33843740820884705,
+      "learning_rate": 1.6143232768173428e-06,
+      "loss": 0.4754,
+      "step": 1411
+    },
+    {
+      "epoch": 0.7633807893314111,
+      "grad_norm": 0.3162449300289154,
+      "learning_rate": 1.6073828977920564e-06,
+      "loss": 0.4683,
+      "step": 1412
+    },
+    {
+      "epoch": 0.7639214272841953,
+      "grad_norm": 0.3222675025463104,
+      "learning_rate": 1.6004546115932023e-06,
+      "loss": 0.4774,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7644620652369797,
+      "grad_norm": 0.3448139727115631,
+      "learning_rate": 1.5935384429163376e-06,
+      "loss": 0.4532,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7650027031897639,
+      "grad_norm": 0.3369153141975403,
+      "learning_rate": 1.5866344164138214e-06,
+      "loss": 0.4624,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7655433411425482,
+      "grad_norm": 0.3553886115550995,
+      "learning_rate": 1.5797425566947378e-06,
+      "loss": 0.4782,
+      "step": 1416
+    },
+    {
+      "epoch": 0.7660839790953325,
+      "grad_norm": 0.31446409225463867,
+      "learning_rate": 1.572862888324801e-06,
+      "loss": 0.4476,
+      "step": 1417
+    },
+    {
+      "epoch": 0.7666246170481168,
+      "grad_norm": 0.3353008031845093,
+      "learning_rate": 1.5659954358262724e-06,
+      "loss": 0.4717,
+      "step": 1418
+    },
+    {
+      "epoch": 0.767165255000901,
+      "grad_norm": 0.3172055780887604,
+      "learning_rate": 1.5591402236778647e-06,
+      "loss": 0.4599,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7677058929536853,
+      "grad_norm": 0.3034406900405884,
+      "learning_rate": 1.5522972763146653e-06,
+      "loss": 0.439,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7682465309064697,
+      "grad_norm": 0.36666351556777954,
+      "learning_rate": 1.5454666181280437e-06,
+      "loss": 0.498,
+      "step": 1421
+    },
+    {
+      "epoch": 0.7687871688592539,
+      "grad_norm": 0.36715036630630493,
+      "learning_rate": 1.5386482734655633e-06,
+      "loss": 0.489,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7693278068120382,
+      "grad_norm": 0.3549302816390991,
+      "learning_rate": 1.5318422666308997e-06,
+      "loss": 0.4685,
+      "step": 1423
+    },
+    {
+      "epoch": 0.7698684447648225,
+      "grad_norm": 0.3297741413116455,
+      "learning_rate": 1.5250486218837458e-06,
+      "loss": 0.493,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7704090827176068,
+      "grad_norm": 0.3559105396270752,
+      "learning_rate": 1.5182673634397365e-06,
+      "loss": 0.4517,
+      "step": 1425
+    },
+    {
+      "epoch": 0.770949720670391,
+      "grad_norm": 0.3720037639141083,
+      "learning_rate": 1.5114985154703505e-06,
+      "loss": 0.4604,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7714903586231754,
+      "grad_norm": 0.384896844625473,
+      "learning_rate": 1.5047421021028353e-06,
+      "loss": 0.4563,
+      "step": 1427
+    },
+    {
+      "epoch": 0.7720309965759596,
+      "grad_norm": 0.35539141297340393,
+      "learning_rate": 1.4979981474201106e-06,
+      "loss": 0.458,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7725716345287439,
+      "grad_norm": 0.3983554244041443,
+      "learning_rate": 1.4912666754606914e-06,
+      "loss": 0.4836,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7731122724815283,
+      "grad_norm": 0.34799396991729736,
+      "learning_rate": 1.4845477102185974e-06,
+      "loss": 0.4661,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7736529104343125,
+      "grad_norm": 0.3554874360561371,
+      "learning_rate": 1.4778412756432709e-06,
+      "loss": 0.479,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 0.37535446882247925,
+      "learning_rate": 1.471147395639484e-06,
+      "loss": 0.4847,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7747341863398811,
+      "grad_norm": 0.3827463686466217,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.4806,
+      "step": 1433
+    },
+    {
+      "epoch": 0.7752748242926654,
+      "grad_norm": 0.3442119359970093,
+      "learning_rate": 1.457797394741798e-06,
+      "loss": 0.5032,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7758154622454496,
+      "grad_norm": 0.35367828607559204,
+      "learning_rate": 1.451141321433358e-06,
+      "loss": 0.4948,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7763561001982339,
+      "grad_norm": 0.3889831304550171,
+      "learning_rate": 1.4444978978672103e-06,
+      "loss": 0.4688,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7768967381510182,
+      "grad_norm": 0.3154084086418152,
+      "learning_rate": 1.4378671477235268e-06,
+      "loss": 0.4786,
+      "step": 1437
+    },
+    {
+      "epoch": 0.7774373761038025,
+      "grad_norm": 0.34436705708503723,
+      "learning_rate": 1.431249094637311e-06,
+      "loss": 0.472,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7779780140565867,
+      "grad_norm": 0.3308955132961273,
+      "learning_rate": 1.4246437621983057e-06,
+      "loss": 0.4739,
+      "step": 1439
+    },
+    {
+      "epoch": 0.7785186520093711,
+      "grad_norm": 0.319321870803833,
+      "learning_rate": 1.418051173950914e-06,
+      "loss": 0.4603,
+      "step": 1440
+    },
+    {
+      "epoch": 0.7790592899621553,
+      "grad_norm": 0.31957656145095825,
+      "learning_rate": 1.4114713533941082e-06,
+      "loss": 0.4533,
+      "step": 1441
+    },
+    {
+      "epoch": 0.7795999279149396,
+      "grad_norm": 0.3338441848754883,
+      "learning_rate": 1.4049043239813575e-06,
+      "loss": 0.4805,
+      "step": 1442
+    },
+    {
+      "epoch": 0.780140565867724,
+      "grad_norm": 0.33430007100105286,
+      "learning_rate": 1.3983501091205298e-06,
+      "loss": 0.494,
+      "step": 1443
+    },
+    {
+      "epoch": 0.7806812038205082,
+      "grad_norm": 0.3184848129749298,
+      "learning_rate": 1.3918087321738244e-06,
+      "loss": 0.48,
+      "step": 1444
+    },
+    {
+      "epoch": 0.7812218417732925,
+      "grad_norm": 0.37600019574165344,
+      "learning_rate": 1.3852802164576717e-06,
+      "loss": 0.4767,
+      "step": 1445
+    },
+    {
+      "epoch": 0.7817624797260768,
+      "grad_norm": 0.3498574197292328,
+      "learning_rate": 1.3787645852426663e-06,
+      "loss": 0.4664,
+      "step": 1446
+    },
+    {
+      "epoch": 0.7823031176788611,
+      "grad_norm": 0.3013286292552948,
+      "learning_rate": 1.3722618617534727e-06,
+      "loss": 0.4444,
+      "step": 1447
+    },
+    {
+      "epoch": 0.7828437556316453,
+      "grad_norm": 0.32576611638069153,
+      "learning_rate": 1.3657720691687481e-06,
+      "loss": 0.4616,
+      "step": 1448
+    },
+    {
+      "epoch": 0.7833843935844296,
+      "grad_norm": 0.3226719796657562,
+      "learning_rate": 1.3592952306210589e-06,
+      "loss": 0.4776,
+      "step": 1449
+    },
+    {
+      "epoch": 0.7839250315372139,
+      "grad_norm": 0.30378836393356323,
+      "learning_rate": 1.3528313691967926e-06,
+      "loss": 0.4671,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7844656694899982,
+      "grad_norm": 0.36785420775413513,
+      "learning_rate": 1.3463805079360854e-06,
+      "loss": 0.4758,
+      "step": 1451
+    },
+    {
+      "epoch": 0.7850063074427824,
+      "grad_norm": 0.332956463098526,
+      "learning_rate": 1.3399426698327329e-06,
+      "loss": 0.4767,
+      "step": 1452
+    },
+    {
+      "epoch": 0.7855469453955668,
+      "grad_norm": 0.34059152007102966,
+      "learning_rate": 1.3335178778341123e-06,
+      "loss": 0.4863,
+      "step": 1453
+    },
+    {
+      "epoch": 0.786087583348351,
+      "grad_norm": 0.3034917712211609,
+      "learning_rate": 1.3271061548410947e-06,
+      "loss": 0.4506,
+      "step": 1454
+    },
+    {
+      "epoch": 0.7866282213011353,
+      "grad_norm": 0.30233845114707947,
+      "learning_rate": 1.3207075237079702e-06,
+      "loss": 0.4683,
+      "step": 1455
+    },
+    {
+      "epoch": 0.7871688592539197,
+      "grad_norm": 0.311905175447464,
+      "learning_rate": 1.3143220072423647e-06,
+      "loss": 0.44,
+      "step": 1456
+    },
+    {
+      "epoch": 0.7877094972067039,
+      "grad_norm": 0.3118400275707245,
+      "learning_rate": 1.307949628205153e-06,
+      "loss": 0.4273,
+      "step": 1457
+    },
+    {
+      "epoch": 0.7882501351594882,
+      "grad_norm": 0.3080877363681793,
+      "learning_rate": 1.301590409310387e-06,
+      "loss": 0.4283,
+      "step": 1458
+    },
+    {
+      "epoch": 0.7887907731122725,
+      "grad_norm": 0.34099653363227844,
+      "learning_rate": 1.2952443732252058e-06,
+      "loss": 0.4688,
+      "step": 1459
+    },
+    {
+      "epoch": 0.7893314110650568,
+      "grad_norm": 0.3410574495792389,
+      "learning_rate": 1.2889115425697612e-06,
+      "loss": 0.4751,
+      "step": 1460
+    },
+    {
+      "epoch": 0.789872049017841,
+      "grad_norm": 0.3065154552459717,
+      "learning_rate": 1.282591939917136e-06,
+      "loss": 0.4832,
+      "step": 1461
+    },
+    {
+      "epoch": 0.7904126869706254,
+      "grad_norm": 0.3287332355976105,
+      "learning_rate": 1.2762855877932617e-06,
+      "loss": 0.4662,
+      "step": 1462
+    },
+    {
+      "epoch": 0.7909533249234096,
+      "grad_norm": 0.3344971239566803,
+      "learning_rate": 1.269992508676835e-06,
+      "loss": 0.4654,
+      "step": 1463
+    },
+    {
+      "epoch": 0.7914939628761939,
+      "grad_norm": 0.3270774781703949,
+      "learning_rate": 1.2637127249992465e-06,
+      "loss": 0.456,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7920346008289781,
+      "grad_norm": 0.31552377343177795,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 0.4917,
+      "step": 1465
+    },
+    {
+      "epoch": 0.7925752387817625,
+      "grad_norm": 0.3612023591995239,
+      "learning_rate": 1.2511931334491068e-06,
+      "loss": 0.5151,
+      "step": 1466
+    },
+    {
+      "epoch": 0.7931158767345468,
+      "grad_norm": 0.326369047164917,
+      "learning_rate": 1.2449533702020578e-06,
+      "loss": 0.4871,
+      "step": 1467
+    },
+    {
+      "epoch": 0.793656514687331,
+      "grad_norm": 0.32858172059059143,
+      "learning_rate": 1.238726991644696e-06,
+      "loss": 0.4776,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7941971526401154,
+      "grad_norm": 0.30760154128074646,
+      "learning_rate": 1.232514019970658e-06,
+      "loss": 0.4504,
+      "step": 1469
+    },
+    {
+      "epoch": 0.7947377905928996,
+      "grad_norm": 0.3285392224788666,
+      "learning_rate": 1.2263144773257967e-06,
+      "loss": 0.4873,
+      "step": 1470
+    },
+    {
+      "epoch": 0.7952784285456839,
+      "grad_norm": 0.3165012001991272,
+      "learning_rate": 1.2201283858080903e-06,
+      "loss": 0.4435,
+      "step": 1471
+    },
+    {
+      "epoch": 0.7958190664984682,
+      "grad_norm": 0.3612854480743408,
+      "learning_rate": 1.2139557674675773e-06,
+      "loss": 0.4748,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7963597044512525,
+      "grad_norm": 0.3201451003551483,
+      "learning_rate": 1.2077966443062706e-06,
+      "loss": 0.499,
+      "step": 1473
+    },
+    {
+      "epoch": 0.7969003424040367,
+      "grad_norm": 0.3280227780342102,
+      "learning_rate": 1.2016510382780772e-06,
+      "loss": 0.4836,
+      "step": 1474
+    },
+    {
+      "epoch": 0.7974409803568211,
+      "grad_norm": 0.28860074281692505,
+      "learning_rate": 1.1955189712887272e-06,
+      "loss": 0.4607,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7979816183096053,
+      "grad_norm": 0.3293115198612213,
+      "learning_rate": 1.189400465195687e-06,
+      "loss": 0.4633,
+      "step": 1476
+    },
+    {
+      "epoch": 0.7985222562623896,
+      "grad_norm": 0.3143051862716675,
+      "learning_rate": 1.183295541808089e-06,
+      "loss": 0.4549,
+      "step": 1477
+    },
+    {
+      "epoch": 0.799062894215174,
+      "grad_norm": 0.30853116512298584,
+      "learning_rate": 1.1772042228866493e-06,
+      "loss": 0.4824,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7996035321679582,
+      "grad_norm": 0.32995665073394775,
+      "learning_rate": 1.1711265301435937e-06,
+      "loss": 0.4844,
+      "step": 1479
+    },
+    {
+      "epoch": 0.8001441701207425,
+      "grad_norm": 0.3321036100387573,
+      "learning_rate": 1.165062485242574e-06,
+      "loss": 0.4591,
+      "step": 1480
+    },
+    {
+      "epoch": 0.8006848080735267,
+      "grad_norm": 0.29632872343063354,
+      "learning_rate": 1.159012109798598e-06,
+      "loss": 0.4732,
+      "step": 1481
+    },
+    {
+      "epoch": 0.8012254460263111,
+      "grad_norm": 0.29381921887397766,
+      "learning_rate": 1.1529754253779486e-06,
+      "loss": 0.4685,
+      "step": 1482
+    },
+    {
+      "epoch": 0.8017660839790953,
+      "grad_norm": 0.321435809135437,
+      "learning_rate": 1.1469524534981091e-06,
+      "loss": 0.4575,
+      "step": 1483
+    },
+    {
+      "epoch": 0.8023067219318796,
+      "grad_norm": 0.3351998031139374,
+      "learning_rate": 1.1409432156276805e-06,
+      "loss": 0.4951,
+      "step": 1484
+    },
+    {
+      "epoch": 0.8028473598846639,
+      "grad_norm": 0.32599401473999023,
+      "learning_rate": 1.134947733186315e-06,
+      "loss": 0.4516,
+      "step": 1485
+    },
+    {
+      "epoch": 0.8033879978374482,
+      "grad_norm": 0.31513264775276184,
+      "learning_rate": 1.1289660275446318e-06,
+      "loss": 0.4816,
+      "step": 1486
+    },
+    {
+      "epoch": 0.8039286357902324,
+      "grad_norm": 0.3273223042488098,
+      "learning_rate": 1.1229981200241424e-06,
+      "loss": 0.4561,
+      "step": 1487
+    },
+    {
+      "epoch": 0.8044692737430168,
+      "grad_norm": 0.2985652685165405,
+      "learning_rate": 1.1170440318971788e-06,
+      "loss": 0.4808,
+      "step": 1488
+    },
+    {
+      "epoch": 0.805009911695801,
+      "grad_norm": 0.318036288022995,
+      "learning_rate": 1.1111037843868095e-06,
+      "loss": 0.4812,
+      "step": 1489
+    },
+    {
+      "epoch": 0.8055505496485853,
+      "grad_norm": 0.326748788356781,
+      "learning_rate": 1.1051773986667735e-06,
+      "loss": 0.4737,
+      "step": 1490
+    },
+    {
+      "epoch": 0.8060911876013697,
+      "grad_norm": 0.3135136663913727,
+      "learning_rate": 1.0992648958613961e-06,
+      "loss": 0.4618,
+      "step": 1491
+    },
+    {
+      "epoch": 0.8066318255541539,
+      "grad_norm": 0.2945253551006317,
+      "learning_rate": 1.0933662970455217e-06,
+      "loss": 0.4769,
+      "step": 1492
+    },
+    {
+      "epoch": 0.8071724635069382,
+      "grad_norm": 0.3053777813911438,
+      "learning_rate": 1.0874816232444297e-06,
+      "loss": 0.4672,
+      "step": 1493
+    },
+    {
+      "epoch": 0.8077131014597225,
+      "grad_norm": 0.3037196695804596,
+      "learning_rate": 1.081610895433769e-06,
+      "loss": 0.4399,
+      "step": 1494
+    },
+    {
+      "epoch": 0.8082537394125068,
+      "grad_norm": 0.30045562982559204,
+      "learning_rate": 1.0757541345394768e-06,
+      "loss": 0.4467,
+      "step": 1495
+    },
+    {
+      "epoch": 0.808794377365291,
+      "grad_norm": 0.2896043062210083,
+      "learning_rate": 1.0699113614377065e-06,
+      "loss": 0.4441,
+      "step": 1496
+    },
+    {
+      "epoch": 0.8093350153180753,
+      "grad_norm": 0.2978215217590332,
+      "learning_rate": 1.0640825969547498e-06,
+      "loss": 0.4531,
+      "step": 1497
+    },
+    {
+      "epoch": 0.8098756532708596,
+      "grad_norm": 0.32296404242515564,
+      "learning_rate": 1.058267861866969e-06,
+      "loss": 0.4619,
+      "step": 1498
+    },
+    {
+      "epoch": 0.8104162912236439,
+      "grad_norm": 0.32622790336608887,
+      "learning_rate": 1.0524671769007177e-06,
+      "loss": 0.4852,
+      "step": 1499
+    },
+    {
+      "epoch": 0.8109569291764281,
+      "grad_norm": 0.31101804971694946,
+      "learning_rate": 1.0466805627322685e-06,
+      "loss": 0.4534,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8114975671292125,
+      "grad_norm": 0.29956305027008057,
+      "learning_rate": 1.0409080399877413e-06,
+      "loss": 0.4594,
+      "step": 1501
+    },
+    {
+      "epoch": 0.8120382050819968,
+      "grad_norm": 0.320745050907135,
+      "learning_rate": 1.035149629243023e-06,
+      "loss": 0.4689,
+      "step": 1502
+    },
+    {
+      "epoch": 0.812578843034781,
+      "grad_norm": 0.30797460675239563,
+      "learning_rate": 1.0294053510237028e-06,
+      "loss": 0.4573,
+      "step": 1503
+    },
+    {
+      "epoch": 0.8131194809875654,
+      "grad_norm": 0.3377775549888611,
+      "learning_rate": 1.0236752258049954e-06,
+      "loss": 0.4752,
+      "step": 1504
+    },
+    {
+      "epoch": 0.8136601189403496,
+      "grad_norm": 0.3108988106250763,
+      "learning_rate": 1.017959274011665e-06,
+      "loss": 0.4693,
+      "step": 1505
+    },
+    {
+      "epoch": 0.8142007568931339,
+      "grad_norm": 0.33605459332466125,
+      "learning_rate": 1.0122575160179582e-06,
+      "loss": 0.5001,
+      "step": 1506
+    },
+    {
+      "epoch": 0.8147413948459182,
+      "grad_norm": 0.33161184191703796,
+      "learning_rate": 1.0065699721475253e-06,
+      "loss": 0.4664,
+      "step": 1507
+    },
+    {
+      "epoch": 0.8152820327987025,
+      "grad_norm": 0.31771883368492126,
+      "learning_rate": 1.0008966626733541e-06,
+      "loss": 0.4783,
+      "step": 1508
+    },
+    {
+      "epoch": 0.8158226707514867,
+      "grad_norm": 0.3252415657043457,
+      "learning_rate": 9.95237607817694e-07,
+      "loss": 0.473,
+      "step": 1509
+    },
+    {
+      "epoch": 0.816363308704271,
+      "grad_norm": 0.33228546380996704,
+      "learning_rate": 9.895928277519822e-07,
+      "loss": 0.4814,
+      "step": 1510
+    },
+    {
+      "epoch": 0.8169039466570553,
+      "grad_norm": 0.3342330753803253,
+      "learning_rate": 9.83962342596776e-07,
+      "loss": 0.4642,
+      "step": 1511
+    },
+    {
+      "epoch": 0.8174445846098396,
+      "grad_norm": 0.3376791477203369,
+      "learning_rate": 9.783461724216793e-07,
+      "loss": 0.4795,
+      "step": 1512
+    },
+    {
+      "epoch": 0.8179852225626238,
+      "grad_norm": 0.3592437505722046,
+      "learning_rate": 9.7274433724527e-07,
+      "loss": 0.4655,
+      "step": 1513
+    },
+    {
+      "epoch": 0.8185258605154082,
+      "grad_norm": 0.3259372115135193,
+      "learning_rate": 9.671568570350321e-07,
+      "loss": 0.4775,
+      "step": 1514
+    },
+    {
+      "epoch": 0.8190664984681925,
+      "grad_norm": 0.3325469493865967,
+      "learning_rate": 9.615837517072758e-07,
+      "loss": 0.476,
+      "step": 1515
+    },
+    {
+      "epoch": 0.8196071364209767,
+      "grad_norm": 0.2931354343891144,
+      "learning_rate": 9.560250411270794e-07,
+      "loss": 0.4563,
+      "step": 1516
+    },
+    {
+      "epoch": 0.8201477743737611,
+      "grad_norm": 0.320726215839386,
+      "learning_rate": 9.504807451082088e-07,
+      "loss": 0.4857,
+      "step": 1517
+    },
+    {
+      "epoch": 0.8206884123265453,
+      "grad_norm": 0.32242387533187866,
+      "learning_rate": 9.449508834130517e-07,
+      "loss": 0.4836,
+      "step": 1518
+    },
+    {
+      "epoch": 0.8212290502793296,
+      "grad_norm": 0.3245270252227783,
+      "learning_rate": 9.394354757525404e-07,
+      "loss": 0.4671,
+      "step": 1519
+    },
+    {
+      "epoch": 0.8217696882321139,
+      "grad_norm": 0.3327704668045044,
+      "learning_rate": 9.339345417860918e-07,
+      "loss": 0.485,
+      "step": 1520
+    },
+    {
+      "epoch": 0.8223103261848982,
+      "grad_norm": 0.30964258313179016,
+      "learning_rate": 9.284481011215318e-07,
+      "loss": 0.4814,
+      "step": 1521
+    },
+    {
+      "epoch": 0.8228509641376824,
+      "grad_norm": 0.3185735046863556,
+      "learning_rate": 9.229761733150205e-07,
+      "loss": 0.4372,
+      "step": 1522
+    },
+    {
+      "epoch": 0.8233916020904668,
+      "grad_norm": 0.33612483739852905,
+      "learning_rate": 9.175187778709937e-07,
+      "loss": 0.4454,
+      "step": 1523
+    },
+    {
+      "epoch": 0.823932240043251,
+      "grad_norm": 0.31606701016426086,
+      "learning_rate": 9.120759342420821e-07,
+      "loss": 0.4578,
+      "step": 1524
+    },
+    {
+      "epoch": 0.8244728779960353,
+      "grad_norm": 0.31590044498443604,
+      "learning_rate": 9.066476618290515e-07,
+      "loss": 0.4715,
+      "step": 1525
+    },
+    {
+      "epoch": 0.8250135159488196,
+      "grad_norm": 0.2985929548740387,
+      "learning_rate": 9.012339799807263e-07,
+      "loss": 0.4654,
+      "step": 1526
+    },
+    {
+      "epoch": 0.8255541539016039,
+      "grad_norm": 0.35035449266433716,
+      "learning_rate": 8.95834907993926e-07,
+      "loss": 0.4557,
+      "step": 1527
+    },
+    {
+      "epoch": 0.8260947918543882,
+      "grad_norm": 0.3264734148979187,
+      "learning_rate": 8.904504651133905e-07,
+      "loss": 0.4634,
+      "step": 1528
+    },
+    {
+      "epoch": 0.8266354298071724,
+      "grad_norm": 0.32273826003074646,
+      "learning_rate": 8.850806705317183e-07,
+      "loss": 0.4608,
+      "step": 1529
+    },
+    {
+      "epoch": 0.8271760677599568,
+      "grad_norm": 0.29928576946258545,
+      "learning_rate": 8.797255433892926e-07,
+      "loss": 0.4553,
+      "step": 1530
+    },
+    {
+      "epoch": 0.827716705712741,
+      "grad_norm": 0.30989155173301697,
+      "learning_rate": 8.743851027742172e-07,
+      "loss": 0.4553,
+      "step": 1531
+    },
+    {
+      "epoch": 0.8282573436655253,
+      "grad_norm": 0.337877482175827,
+      "learning_rate": 8.690593677222431e-07,
+      "loss": 0.442,
+      "step": 1532
+    },
+    {
+      "epoch": 0.8287979816183096,
+      "grad_norm": 0.34234437346458435,
+      "learning_rate": 8.637483572167077e-07,
+      "loss": 0.4639,
+      "step": 1533
+    },
+    {
+      "epoch": 0.8293386195710939,
+      "grad_norm": 0.3071967363357544,
+      "learning_rate": 8.584520901884608e-07,
+      "loss": 0.4784,
+      "step": 1534
+    },
+    {
+      "epoch": 0.8298792575238781,
+      "grad_norm": 0.32178065180778503,
+      "learning_rate": 8.531705855158024e-07,
+      "loss": 0.4483,
+      "step": 1535
+    },
+    {
+      "epoch": 0.8304198954766625,
+      "grad_norm": 0.33801355957984924,
+      "learning_rate": 8.479038620244089e-07,
+      "loss": 0.4655,
+      "step": 1536
+    },
+    {
+      "epoch": 0.8309605334294468,
+      "grad_norm": 0.3588986396789551,
+      "learning_rate": 8.426519384872733e-07,
+      "loss": 0.4633,
+      "step": 1537
+    },
+    {
+      "epoch": 0.831501171382231,
+      "grad_norm": 0.3052517771720886,
+      "learning_rate": 8.374148336246352e-07,
+      "loss": 0.4667,
+      "step": 1538
+    },
+    {
+      "epoch": 0.8320418093350154,
+      "grad_norm": 0.32133039832115173,
+      "learning_rate": 8.321925661039088e-07,
+      "loss": 0.4773,
+      "step": 1539
+    },
+    {
+      "epoch": 0.8325824472877996,
+      "grad_norm": 0.3321249783039093,
+      "learning_rate": 8.269851545396279e-07,
+      "loss": 0.4829,
+      "step": 1540
+    },
+    {
+      "epoch": 0.8331230852405839,
+      "grad_norm": 0.3377304673194885,
+      "learning_rate": 8.217926174933665e-07,
+      "loss": 0.4632,
+      "step": 1541
+    },
+    {
+      "epoch": 0.8336637231933681,
+      "grad_norm": 0.3176397383213043,
+      "learning_rate": 8.166149734736845e-07,
+      "loss": 0.4814,
+      "step": 1542
+    },
+    {
+      "epoch": 0.8342043611461525,
+      "grad_norm": 0.33456897735595703,
+      "learning_rate": 8.114522409360531e-07,
+      "loss": 0.4667,
+      "step": 1543
+    },
+    {
+      "epoch": 0.8347449990989367,
+      "grad_norm": 0.29794400930404663,
+      "learning_rate": 8.063044382827945e-07,
+      "loss": 0.4635,
+      "step": 1544
+    },
+    {
+      "epoch": 0.835285637051721,
+      "grad_norm": 0.3277740478515625,
+      "learning_rate": 8.011715838630107e-07,
+      "loss": 0.4805,
+      "step": 1545
+    },
+    {
+      "epoch": 0.8358262750045053,
+      "grad_norm": 0.3014332950115204,
+      "learning_rate": 7.960536959725252e-07,
+      "loss": 0.4587,
+      "step": 1546
+    },
+    {
+      "epoch": 0.8363669129572896,
+      "grad_norm": 0.3108069598674774,
+      "learning_rate": 7.909507928538107e-07,
+      "loss": 0.452,
+      "step": 1547
+    },
+    {
+      "epoch": 0.8369075509100738,
+      "grad_norm": 0.3274773061275482,
+      "learning_rate": 7.858628926959311e-07,
+      "loss": 0.4805,
+      "step": 1548
+    },
+    {
+      "epoch": 0.8374481888628582,
+      "grad_norm": 0.31421011686325073,
+      "learning_rate": 7.807900136344676e-07,
+      "loss": 0.4582,
+      "step": 1549
+    },
+    {
+      "epoch": 0.8379888268156425,
+      "grad_norm": 0.30205944180488586,
+      "learning_rate": 7.757321737514645e-07,
+      "loss": 0.4902,
+      "step": 1550
+    },
+    {
+      "epoch": 0.8385294647684267,
+      "grad_norm": 0.33638957142829895,
+      "learning_rate": 7.706893910753571e-07,
+      "loss": 0.4946,
+      "step": 1551
+    },
+    {
+      "epoch": 0.8390701027212111,
+      "grad_norm": 0.3180740475654602,
+      "learning_rate": 7.656616835809122e-07,
+      "loss": 0.4895,
+      "step": 1552
+    },
+    {
+      "epoch": 0.8396107406739953,
+      "grad_norm": 0.3404751121997833,
+      "learning_rate": 7.606490691891577e-07,
+      "loss": 0.4654,
+      "step": 1553
+    },
+    {
+      "epoch": 0.8401513786267796,
+      "grad_norm": 0.32554861903190613,
+      "learning_rate": 7.556515657673274e-07,
+      "loss": 0.4784,
+      "step": 1554
+    },
+    {
+      "epoch": 0.8406920165795639,
+      "grad_norm": 0.30992451310157776,
+      "learning_rate": 7.506691911287883e-07,
+      "loss": 0.4746,
+      "step": 1555
+    },
+    {
+      "epoch": 0.8412326545323482,
+      "grad_norm": 0.2857951521873474,
+      "learning_rate": 7.457019630329848e-07,
+      "loss": 0.435,
+      "step": 1556
+    },
+    {
+      "epoch": 0.8417732924851324,
+      "grad_norm": 0.3287476599216461,
+      "learning_rate": 7.407498991853729e-07,
+      "loss": 0.4949,
+      "step": 1557
+    },
+    {
+      "epoch": 0.8423139304379167,
+      "grad_norm": 0.3094945251941681,
+      "learning_rate": 7.358130172373523e-07,
+      "loss": 0.4399,
+      "step": 1558
+    },
+    {
+      "epoch": 0.842854568390701,
+      "grad_norm": 0.33573803305625916,
+      "learning_rate": 7.308913347862112e-07,
+      "loss": 0.4661,
+      "step": 1559
+    },
+    {
+      "epoch": 0.8433952063434853,
+      "grad_norm": 0.32553064823150635,
+      "learning_rate": 7.259848693750582e-07,
+      "loss": 0.4646,
+      "step": 1560
+    },
+    {
+      "epoch": 0.8439358442962696,
+      "grad_norm": 0.3170788288116455,
+      "learning_rate": 7.210936384927631e-07,
+      "loss": 0.4743,
+      "step": 1561
+    },
+    {
+      "epoch": 0.8444764822490539,
+      "grad_norm": 0.2907802164554596,
+      "learning_rate": 7.162176595738895e-07,
+      "loss": 0.4486,
+      "step": 1562
+    },
+    {
+      "epoch": 0.8450171202018382,
+      "grad_norm": 0.31983017921447754,
+      "learning_rate": 7.113569499986401e-07,
+      "loss": 0.4635,
+      "step": 1563
+    },
+    {
+      "epoch": 0.8455577581546224,
+      "grad_norm": 0.2787870168685913,
+      "learning_rate": 7.065115270927875e-07,
+      "loss": 0.4483,
+      "step": 1564
+    },
+    {
+      "epoch": 0.8460983961074068,
+      "grad_norm": 0.31309157609939575,
+      "learning_rate": 7.01681408127618e-07,
+      "loss": 0.4546,
+      "step": 1565
+    },
+    {
+      "epoch": 0.846639034060191,
+      "grad_norm": 0.29454049468040466,
+      "learning_rate": 6.968666103198679e-07,
+      "loss": 0.4706,
+      "step": 1566
+    },
+    {
+      "epoch": 0.8471796720129753,
+      "grad_norm": 0.31175103783607483,
+      "learning_rate": 6.920671508316584e-07,
+      "loss": 0.4844,
+      "step": 1567
+    },
+    {
+      "epoch": 0.8477203099657596,
+      "grad_norm": 0.3265729546546936,
+      "learning_rate": 6.872830467704417e-07,
+      "loss": 0.4614,
+      "step": 1568
+    },
+    {
+      "epoch": 0.8482609479185439,
+      "grad_norm": 0.310225248336792,
+      "learning_rate": 6.825143151889358e-07,
+      "loss": 0.4786,
+      "step": 1569
+    },
+    {
+      "epoch": 0.8488015858713281,
+      "grad_norm": 0.31501367688179016,
+      "learning_rate": 6.777609730850615e-07,
+      "loss": 0.4904,
+      "step": 1570
+    },
+    {
+      "epoch": 0.8493422238241125,
+      "grad_norm": 0.31551438570022583,
+      "learning_rate": 6.730230374018886e-07,
+      "loss": 0.4804,
+      "step": 1571
+    },
+    {
+      "epoch": 0.8498828617768968,
+      "grad_norm": 0.2776893079280853,
+      "learning_rate": 6.683005250275676e-07,
+      "loss": 0.4579,
+      "step": 1572
+    },
+    {
+      "epoch": 0.850423499729681,
+      "grad_norm": 0.29054975509643555,
+      "learning_rate": 6.635934527952747e-07,
+      "loss": 0.4565,
+      "step": 1573
+    },
+    {
+      "epoch": 0.8509641376824653,
+      "grad_norm": 0.30911266803741455,
+      "learning_rate": 6.589018374831529e-07,
+      "loss": 0.4731,
+      "step": 1574
+    },
+    {
+      "epoch": 0.8515047756352496,
+      "grad_norm": 0.3118568956851959,
+      "learning_rate": 6.542256958142456e-07,
+      "loss": 0.4669,
+      "step": 1575
+    },
+    {
+      "epoch": 0.8520454135880339,
+      "grad_norm": 0.29929646849632263,
+      "learning_rate": 6.495650444564433e-07,
+      "loss": 0.4549,
+      "step": 1576
+    },
+    {
+      "epoch": 0.8525860515408181,
+      "grad_norm": 0.3033123314380646,
+      "learning_rate": 6.449199000224221e-07,
+      "loss": 0.4539,
+      "step": 1577
+    },
+    {
+      "epoch": 0.8531266894936025,
+      "grad_norm": 0.30189841985702515,
+      "learning_rate": 6.402902790695842e-07,
+      "loss": 0.4844,
+      "step": 1578
+    },
+    {
+      "epoch": 0.8536673274463867,
+      "grad_norm": 0.3111442029476166,
+      "learning_rate": 6.356761980999998e-07,
+      "loss": 0.4744,
+      "step": 1579
+    },
+    {
+      "epoch": 0.854207965399171,
+      "grad_norm": 0.2925800681114197,
+      "learning_rate": 6.310776735603452e-07,
+      "loss": 0.4781,
+      "step": 1580
+    },
+    {
+      "epoch": 0.8547486033519553,
+      "grad_norm": 0.2876686453819275,
+      "learning_rate": 6.264947218418482e-07,
+      "loss": 0.4422,
+      "step": 1581
+    },
+    {
+      "epoch": 0.8552892413047396,
+      "grad_norm": 0.29705557227134705,
+      "learning_rate": 6.219273592802278e-07,
+      "loss": 0.4675,
+      "step": 1582
+    },
+    {
+      "epoch": 0.8558298792575239,
+      "grad_norm": 0.2854503095149994,
+      "learning_rate": 6.173756021556377e-07,
+      "loss": 0.4537,
+      "step": 1583
+    },
+    {
+      "epoch": 0.8563705172103082,
+      "grad_norm": 0.31599560379981995,
+      "learning_rate": 6.128394666926035e-07,
+      "loss": 0.5045,
+      "step": 1584
+    },
+    {
+      "epoch": 0.8569111551630925,
+      "grad_norm": 0.33695438504219055,
+      "learning_rate": 6.083189690599712e-07,
+      "loss": 0.4945,
+      "step": 1585
+    },
+    {
+      "epoch": 0.8574517931158767,
+      "grad_norm": 0.3071337342262268,
+      "learning_rate": 6.038141253708429e-07,
+      "loss": 0.4765,
+      "step": 1586
+    },
+    {
+      "epoch": 0.857992431068661,
+      "grad_norm": 0.3113190829753876,
+      "learning_rate": 5.993249516825278e-07,
+      "loss": 0.4864,
+      "step": 1587
+    },
+    {
+      "epoch": 0.8585330690214453,
+      "grad_norm": 0.26970189809799194,
+      "learning_rate": 5.948514639964748e-07,
+      "loss": 0.4399,
+      "step": 1588
+    },
+    {
+      "epoch": 0.8590737069742296,
+      "grad_norm": 0.3173854947090149,
+      "learning_rate": 5.903936782582253e-07,
+      "loss": 0.4575,
+      "step": 1589
+    },
+    {
+      "epoch": 0.8596143449270138,
+      "grad_norm": 0.2729853689670563,
+      "learning_rate": 5.859516103573492e-07,
+      "loss": 0.4714,
+      "step": 1590
+    },
+    {
+      "epoch": 0.8601549828797982,
+      "grad_norm": 0.3082450330257416,
+      "learning_rate": 5.815252761273927e-07,
+      "loss": 0.4524,
+      "step": 1591
+    },
+    {
+      "epoch": 0.8606956208325824,
+      "grad_norm": 0.3227364718914032,
+      "learning_rate": 5.771146913458187e-07,
+      "loss": 0.4739,
+      "step": 1592
+    },
+    {
+      "epoch": 0.8612362587853667,
+      "grad_norm": 0.3056773841381073,
+      "learning_rate": 5.727198717339511e-07,
+      "loss": 0.4696,
+      "step": 1593
+    },
+    {
+      "epoch": 0.861776896738151,
+      "grad_norm": 0.3208467364311218,
+      "learning_rate": 5.683408329569212e-07,
+      "loss": 0.4797,
+      "step": 1594
+    },
+    {
+      "epoch": 0.8623175346909353,
+      "grad_norm": 0.312785267829895,
+      "learning_rate": 5.6397759062361e-07,
+      "loss": 0.4487,
+      "step": 1595
+    },
+    {
+      "epoch": 0.8628581726437196,
+      "grad_norm": 0.31125733256340027,
+      "learning_rate": 5.596301602865938e-07,
+      "loss": 0.4564,
+      "step": 1596
+    },
+    {
+      "epoch": 0.8633988105965039,
+      "grad_norm": 0.2956276834011078,
+      "learning_rate": 5.55298557442085e-07,
+      "loss": 0.4526,
+      "step": 1597
+    },
+    {
+      "epoch": 0.8639394485492882,
+      "grad_norm": 0.30352863669395447,
+      "learning_rate": 5.509827975298809e-07,
+      "loss": 0.4756,
+      "step": 1598
+    },
+    {
+      "epoch": 0.8644800865020724,
+      "grad_norm": 0.30768123269081116,
+      "learning_rate": 5.466828959333087e-07,
+      "loss": 0.4793,
+      "step": 1599
+    },
+    {
+      "epoch": 0.8650207244548568,
+      "grad_norm": 0.32290348410606384,
+      "learning_rate": 5.423988679791686e-07,
+      "loss": 0.4746,
+      "step": 1600
+    },
+    {
+      "epoch": 0.865561362407641,
+      "grad_norm": 0.32862550020217896,
+      "learning_rate": 5.381307289376786e-07,
+      "loss": 0.4604,
+      "step": 1601
+    },
+    {
+      "epoch": 0.8661020003604253,
+      "grad_norm": 0.32285216450691223,
+      "learning_rate": 5.338784940224239e-07,
+      "loss": 0.4757,
+      "step": 1602
+    },
+    {
+      "epoch": 0.8666426383132095,
+      "grad_norm": 0.31145691871643066,
+      "learning_rate": 5.296421783902972e-07,
+      "loss": 0.4591,
+      "step": 1603
+    },
+    {
+      "epoch": 0.8671832762659939,
+      "grad_norm": 0.28722888231277466,
+      "learning_rate": 5.254217971414499e-07,
+      "loss": 0.4738,
+      "step": 1604
+    },
+    {
+      "epoch": 0.8677239142187781,
+      "grad_norm": 0.32165801525115967,
+      "learning_rate": 5.212173653192365e-07,
+      "loss": 0.498,
+      "step": 1605
+    },
+    {
+      "epoch": 0.8682645521715624,
+      "grad_norm": 0.3288798928260803,
+      "learning_rate": 5.170288979101573e-07,
+      "loss": 0.4666,
+      "step": 1606
+    },
+    {
+      "epoch": 0.8688051901243468,
+      "grad_norm": 0.3006567656993866,
+      "learning_rate": 5.128564098438116e-07,
+      "loss": 0.4625,
+      "step": 1607
+    },
+    {
+      "epoch": 0.869345828077131,
+      "grad_norm": 0.3048226833343506,
+      "learning_rate": 5.086999159928391e-07,
+      "loss": 0.4746,
+      "step": 1608
+    },
+    {
+      "epoch": 0.8698864660299153,
+      "grad_norm": 0.2663028836250305,
+      "learning_rate": 5.045594311728708e-07,
+      "loss": 0.4526,
+      "step": 1609
+    },
+    {
+      "epoch": 0.8704271039826996,
+      "grad_norm": 0.35595017671585083,
+      "learning_rate": 5.00434970142471e-07,
+      "loss": 0.4808,
+      "step": 1610
+    },
+    {
+      "epoch": 0.8709677419354839,
+      "grad_norm": 0.33726397156715393,
+      "learning_rate": 4.963265476030916e-07,
+      "loss": 0.4678,
+      "step": 1611
+    },
+    {
+      "epoch": 0.8715083798882681,
+      "grad_norm": 0.29663407802581787,
+      "learning_rate": 4.922341781990131e-07,
+      "loss": 0.4845,
+      "step": 1612
+    },
+    {
+      "epoch": 0.8720490178410525,
+      "grad_norm": 0.31134849786758423,
+      "learning_rate": 4.881578765172979e-07,
+      "loss": 0.5056,
+      "step": 1613
+    },
+    {
+      "epoch": 0.8725896557938367,
+      "grad_norm": 0.3079214096069336,
+      "learning_rate": 4.840976570877332e-07,
+      "loss": 0.4556,
+      "step": 1614
+    },
+    {
+      "epoch": 0.873130293746621,
+      "grad_norm": 0.3087759017944336,
+      "learning_rate": 4.800535343827834e-07,
+      "loss": 0.4654,
+      "step": 1615
+    },
+    {
+      "epoch": 0.8736709316994054,
+      "grad_norm": 0.3534053862094879,
+      "learning_rate": 4.7602552281753647e-07,
+      "loss": 0.4545,
+      "step": 1616
+    },
+    {
+      "epoch": 0.8742115696521896,
+      "grad_norm": 0.30727800726890564,
+      "learning_rate": 4.720136367496536e-07,
+      "loss": 0.4578,
+      "step": 1617
+    },
+    {
+      "epoch": 0.8747522076049739,
+      "grad_norm": 0.3113386631011963,
+      "learning_rate": 4.6801789047931535e-07,
+      "loss": 0.4606,
+      "step": 1618
+    },
+    {
+      "epoch": 0.8752928455577581,
+      "grad_norm": 0.30020859837532043,
+      "learning_rate": 4.6403829824917643e-07,
+      "loss": 0.4627,
+      "step": 1619
+    },
+    {
+      "epoch": 0.8758334835105425,
+      "grad_norm": 0.3038908541202545,
+      "learning_rate": 4.6007487424430565e-07,
+      "loss": 0.4521,
+      "step": 1620
+    },
+    {
+      "epoch": 0.8763741214633267,
+      "grad_norm": 0.28542560338974,
+      "learning_rate": 4.5612763259214653e-07,
+      "loss": 0.4602,
+      "step": 1621
+    },
+    {
+      "epoch": 0.876914759416111,
+      "grad_norm": 0.29688113927841187,
+      "learning_rate": 4.52196587362459e-07,
+      "loss": 0.461,
+      "step": 1622
+    },
+    {
+      "epoch": 0.8774553973688953,
+      "grad_norm": 0.30379223823547363,
+      "learning_rate": 4.4828175256727056e-07,
+      "loss": 0.4739,
+      "step": 1623
+    },
+    {
+      "epoch": 0.8779960353216796,
+      "grad_norm": 0.27639734745025635,
+      "learning_rate": 4.4438314216082856e-07,
+      "loss": 0.4268,
+      "step": 1624
+    },
+    {
+      "epoch": 0.8785366732744638,
+      "grad_norm": 0.2714707553386688,
+      "learning_rate": 4.405007700395497e-07,
+      "loss": 0.4475,
+      "step": 1625
+    },
+    {
+      "epoch": 0.8790773112272482,
+      "grad_norm": 0.31498268246650696,
+      "learning_rate": 4.3663465004196995e-07,
+      "loss": 0.4623,
+      "step": 1626
+    },
+    {
+      "epoch": 0.8796179491800324,
+      "grad_norm": 0.29611799120903015,
+      "learning_rate": 4.3278479594869307e-07,
+      "loss": 0.4435,
+      "step": 1627
+    },
+    {
+      "epoch": 0.8801585871328167,
+      "grad_norm": 0.3016403615474701,
+      "learning_rate": 4.289512214823466e-07,
+      "loss": 0.4643,
+      "step": 1628
+    },
+    {
+      "epoch": 0.8806992250856011,
+      "grad_norm": 0.3306654393672943,
+      "learning_rate": 4.251339403075294e-07,
+      "loss": 0.4731,
+      "step": 1629
+    },
+    {
+      "epoch": 0.8812398630383853,
+      "grad_norm": 0.2894127368927002,
+      "learning_rate": 4.21332966030763e-07,
+      "loss": 0.4743,
+      "step": 1630
+    },
+    {
+      "epoch": 0.8817805009911696,
+      "grad_norm": 0.296880304813385,
+      "learning_rate": 4.175483122004448e-07,
+      "loss": 0.4616,
+      "step": 1631
+    },
+    {
+      "epoch": 0.8823211389439539,
+      "grad_norm": 0.2968302369117737,
+      "learning_rate": 4.1377999230679646e-07,
+      "loss": 0.4585,
+      "step": 1632
+    },
+    {
+      "epoch": 0.8828617768967382,
+      "grad_norm": 0.3259296119213104,
+      "learning_rate": 4.100280197818207e-07,
+      "loss": 0.4732,
+      "step": 1633
+    },
+    {
+      "epoch": 0.8834024148495224,
+      "grad_norm": 0.301943838596344,
+      "learning_rate": 4.062924079992492e-07,
+      "loss": 0.4769,
+      "step": 1634
+    },
+    {
+      "epoch": 0.8839430528023067,
+      "grad_norm": 0.31166207790374756,
+      "learning_rate": 4.025731702744978e-07,
+      "loss": 0.4528,
+      "step": 1635
+    },
+    {
+      "epoch": 0.884483690755091,
+      "grad_norm": 0.3266367018222809,
+      "learning_rate": 3.9887031986461546e-07,
+      "loss": 0.4766,
+      "step": 1636
+    },
+    {
+      "epoch": 0.8850243287078753,
+      "grad_norm": 0.29861709475517273,
+      "learning_rate": 3.9518386996824196e-07,
+      "loss": 0.4517,
+      "step": 1637
+    },
+    {
+      "epoch": 0.8855649666606595,
+      "grad_norm": 0.2750012278556824,
+      "learning_rate": 3.9151383372555696e-07,
+      "loss": 0.4419,
+      "step": 1638
+    },
+    {
+      "epoch": 0.8861056046134439,
+      "grad_norm": 0.29816916584968567,
+      "learning_rate": 3.8786022421823497e-07,
+      "loss": 0.4679,
+      "step": 1639
+    },
+    {
+      "epoch": 0.8866462425662281,
+      "grad_norm": 0.29293686151504517,
+      "learning_rate": 3.84223054469397e-07,
+      "loss": 0.4898,
+      "step": 1640
+    },
+    {
+      "epoch": 0.8871868805190124,
+      "grad_norm": 0.3161197900772095,
+      "learning_rate": 3.8060233744356634e-07,
+      "loss": 0.4795,
+      "step": 1641
+    },
+    {
+      "epoch": 0.8877275184717968,
+      "grad_norm": 0.2860853970050812,
+      "learning_rate": 3.76998086046621e-07,
+      "loss": 0.4689,
+      "step": 1642
+    },
+    {
+      "epoch": 0.888268156424581,
+      "grad_norm": 0.29909902811050415,
+      "learning_rate": 3.7341031312574827e-07,
+      "loss": 0.4548,
+      "step": 1643
+    },
+    {
+      "epoch": 0.8888087943773653,
+      "grad_norm": 0.3215901553630829,
+      "learning_rate": 3.6983903146939894e-07,
+      "loss": 0.4874,
+      "step": 1644
+    },
+    {
+      "epoch": 0.8893494323301496,
+      "grad_norm": 0.2928643524646759,
+      "learning_rate": 3.6628425380723975e-07,
+      "loss": 0.4469,
+      "step": 1645
+    },
+    {
+      "epoch": 0.8898900702829339,
+      "grad_norm": 0.3444850444793701,
+      "learning_rate": 3.627459928101118e-07,
+      "loss": 0.4671,
+      "step": 1646
+    },
+    {
+      "epoch": 0.8904307082357181,
+      "grad_norm": 0.30591514706611633,
+      "learning_rate": 3.5922426108998154e-07,
+      "loss": 0.4835,
+      "step": 1647
+    },
+    {
+      "epoch": 0.8909713461885024,
+      "grad_norm": 0.33051443099975586,
+      "learning_rate": 3.5571907119990033e-07,
+      "loss": 0.453,
+      "step": 1648
+    },
+    {
+      "epoch": 0.8915119841412867,
+      "grad_norm": 0.29916903376579285,
+      "learning_rate": 3.522304356339529e-07,
+      "loss": 0.4541,
+      "step": 1649
+    },
+    {
+      "epoch": 0.892052622094071,
+      "grad_norm": 0.31293609738349915,
+      "learning_rate": 3.4875836682722096e-07,
+      "loss": 0.4572,
+      "step": 1650
+    },
+    {
+      "epoch": 0.8925932600468552,
+      "grad_norm": 0.3154950439929962,
+      "learning_rate": 3.45302877155731e-07,
+      "loss": 0.4799,
+      "step": 1651
+    },
+    {
+      "epoch": 0.8931338979996396,
+      "grad_norm": 0.2948947846889496,
+      "learning_rate": 3.418639789364175e-07,
+      "loss": 0.4914,
+      "step": 1652
+    },
+    {
+      "epoch": 0.8936745359524239,
+      "grad_norm": 0.32319676876068115,
+      "learning_rate": 3.3844168442707213e-07,
+      "loss": 0.4846,
+      "step": 1653
+    },
+    {
+      "epoch": 0.8942151739052081,
+      "grad_norm": 0.30142778158187866,
+      "learning_rate": 3.350360058263058e-07,
+      "loss": 0.4656,
+      "step": 1654
+    },
+    {
+      "epoch": 0.8947558118579925,
+      "grad_norm": 0.31285157799720764,
+      "learning_rate": 3.3164695527350244e-07,
+      "loss": 0.4622,
+      "step": 1655
+    },
+    {
+      "epoch": 0.8952964498107767,
+      "grad_norm": 0.3004051744937897,
+      "learning_rate": 3.2827454484877564e-07,
+      "loss": 0.4715,
+      "step": 1656
+    },
+    {
+      "epoch": 0.895837087763561,
+      "grad_norm": 0.3092837631702423,
+      "learning_rate": 3.2491878657292643e-07,
+      "loss": 0.486,
+      "step": 1657
+    },
+    {
+      "epoch": 0.8963777257163453,
+      "grad_norm": 0.31189343333244324,
+      "learning_rate": 3.215796924073983e-07,
+      "loss": 0.4275,
+      "step": 1658
+    },
+    {
+      "epoch": 0.8969183636691296,
+      "grad_norm": 0.2800220549106598,
+      "learning_rate": 3.1825727425423837e-07,
+      "loss": 0.4642,
+      "step": 1659
+    },
+    {
+      "epoch": 0.8974590016219138,
+      "grad_norm": 0.31216147541999817,
+      "learning_rate": 3.149515439560524e-07,
+      "loss": 0.4516,
+      "step": 1660
+    },
+    {
+      "epoch": 0.8979996395746982,
+      "grad_norm": 0.30997899174690247,
+      "learning_rate": 3.116625132959633e-07,
+      "loss": 0.4794,
+      "step": 1661
+    },
+    {
+      "epoch": 0.8985402775274824,
+      "grad_norm": 0.31042930483818054,
+      "learning_rate": 3.083901939975675e-07,
+      "loss": 0.4699,
+      "step": 1662
+    },
+    {
+      "epoch": 0.8990809154802667,
+      "grad_norm": 0.3064749836921692,
+      "learning_rate": 3.051345977248954e-07,
+      "loss": 0.4366,
+      "step": 1663
+    },
+    {
+      "epoch": 0.899621553433051,
+      "grad_norm": 0.27594202756881714,
+      "learning_rate": 3.018957360823699e-07,
+      "loss": 0.4572,
+      "step": 1664
+    },
+    {
+      "epoch": 0.9001621913858353,
+      "grad_norm": 0.2951033115386963,
+      "learning_rate": 2.986736206147628e-07,
+      "loss": 0.4475,
+      "step": 1665
+    },
+    {
+      "epoch": 0.9007028293386196,
+      "grad_norm": 0.29796889424324036,
+      "learning_rate": 2.9546826280715536e-07,
+      "loss": 0.4708,
+      "step": 1666
+    },
+    {
+      "epoch": 0.9012434672914038,
+      "grad_norm": 0.32082822918891907,
+      "learning_rate": 2.9227967408489653e-07,
+      "loss": 0.4723,
+      "step": 1667
+    },
+    {
+      "epoch": 0.9017841052441882,
+      "grad_norm": 0.31180939078330994,
+      "learning_rate": 2.891078658135632e-07,
+      "loss": 0.4659,
+      "step": 1668
+    },
+    {
+      "epoch": 0.9023247431969724,
+      "grad_norm": 0.30132412910461426,
+      "learning_rate": 2.859528492989194e-07,
+      "loss": 0.4867,
+      "step": 1669
+    },
+    {
+      "epoch": 0.9028653811497567,
+      "grad_norm": 0.3208564221858978,
+      "learning_rate": 2.828146357868755e-07,
+      "loss": 0.4589,
+      "step": 1670
+    },
+    {
+      "epoch": 0.903406019102541,
+      "grad_norm": 0.2799170911312103,
+      "learning_rate": 2.796932364634475e-07,
+      "loss": 0.4563,
+      "step": 1671
+    },
+    {
+      "epoch": 0.9039466570553253,
+      "grad_norm": 0.2884426712989807,
+      "learning_rate": 2.765886624547182e-07,
+      "loss": 0.4782,
+      "step": 1672
+    },
+    {
+      "epoch": 0.9044872950081095,
+      "grad_norm": 0.28514596819877625,
+      "learning_rate": 2.7350092482679836e-07,
+      "loss": 0.4549,
+      "step": 1673
+    },
+    {
+      "epoch": 0.9050279329608939,
+      "grad_norm": 0.3076411485671997,
+      "learning_rate": 2.7043003458578685e-07,
+      "loss": 0.4737,
+      "step": 1674
+    },
+    {
+      "epoch": 0.9055685709136782,
+      "grad_norm": 0.34531399607658386,
+      "learning_rate": 2.673760026777272e-07,
+      "loss": 0.479,
+      "step": 1675
+    },
+    {
+      "epoch": 0.9061092088664624,
+      "grad_norm": 0.32933852076530457,
+      "learning_rate": 2.6433883998857657e-07,
+      "loss": 0.4612,
+      "step": 1676
+    },
+    {
+      "epoch": 0.9066498468192468,
+      "grad_norm": 0.2830871045589447,
+      "learning_rate": 2.61318557344159e-07,
+      "loss": 0.4248,
+      "step": 1677
+    },
+    {
+      "epoch": 0.907190484772031,
+      "grad_norm": 0.33129066228866577,
+      "learning_rate": 2.5831516551013405e-07,
+      "loss": 0.4706,
+      "step": 1678
+    },
+    {
+      "epoch": 0.9077311227248153,
+      "grad_norm": 0.3337078392505646,
+      "learning_rate": 2.553286751919509e-07,
+      "loss": 0.4877,
+      "step": 1679
+    },
+    {
+      "epoch": 0.9082717606775995,
+      "grad_norm": 0.3119748830795288,
+      "learning_rate": 2.523590970348166e-07,
+      "loss": 0.4523,
+      "step": 1680
+    },
+    {
+      "epoch": 0.9088123986303839,
+      "grad_norm": 0.29456672072410583,
+      "learning_rate": 2.4940644162365523e-07,
+      "loss": 0.4758,
+      "step": 1681
+    },
+    {
+      "epoch": 0.9093530365831681,
+      "grad_norm": 0.3045862317085266,
+      "learning_rate": 2.46470719483069e-07,
+      "loss": 0.4765,
+      "step": 1682
+    },
+    {
+      "epoch": 0.9098936745359524,
+      "grad_norm": 0.2948693037033081,
+      "learning_rate": 2.435519410773052e-07,
+      "loss": 0.4615,
+      "step": 1683
+    },
+    {
+      "epoch": 0.9104343124887367,
+      "grad_norm": 0.30374738574028015,
+      "learning_rate": 2.4065011681021266e-07,
+      "loss": 0.4846,
+      "step": 1684
+    },
+    {
+      "epoch": 0.910974950441521,
+      "grad_norm": 0.3169853985309601,
+      "learning_rate": 2.3776525702520925e-07,
+      "loss": 0.4746,
+      "step": 1685
+    },
+    {
+      "epoch": 0.9115155883943052,
+      "grad_norm": 0.297050803899765,
+      "learning_rate": 2.3489737200524498e-07,
+      "loss": 0.4785,
+      "step": 1686
+    },
+    {
+      "epoch": 0.9120562263470896,
+      "grad_norm": 0.3017227351665497,
+      "learning_rate": 2.3204647197276387e-07,
+      "loss": 0.4769,
+      "step": 1687
+    },
+    {
+      "epoch": 0.9125968642998739,
+      "grad_norm": 0.3048754334449768,
+      "learning_rate": 2.29212567089665e-07,
+      "loss": 0.4772,
+      "step": 1688
+    },
+    {
+      "epoch": 0.9131375022526581,
+      "grad_norm": 0.3351805508136749,
+      "learning_rate": 2.2639566745727203e-07,
+      "loss": 0.4579,
+      "step": 1689
+    },
+    {
+      "epoch": 0.9136781402054425,
+      "grad_norm": 0.32740268111228943,
+      "learning_rate": 2.2359578311629272e-07,
+      "loss": 0.4945,
+      "step": 1690
+    },
+    {
+      "epoch": 0.9142187781582267,
+      "grad_norm": 0.3174039125442505,
+      "learning_rate": 2.2081292404678655e-07,
+      "loss": 0.4647,
+      "step": 1691
+    },
+    {
+      "epoch": 0.914759416111011,
+      "grad_norm": 0.30359378457069397,
+      "learning_rate": 2.1804710016812337e-07,
+      "loss": 0.469,
+      "step": 1692
+    },
+    {
+      "epoch": 0.9153000540637953,
+      "grad_norm": 0.29310646653175354,
+      "learning_rate": 2.152983213389559e-07,
+      "loss": 0.4867,
+      "step": 1693
+    },
+    {
+      "epoch": 0.9158406920165796,
+      "grad_norm": 0.3022826910018921,
+      "learning_rate": 2.1256659735717777e-07,
+      "loss": 0.4552,
+      "step": 1694
+    },
+    {
+      "epoch": 0.9163813299693638,
+      "grad_norm": 0.2924525737762451,
+      "learning_rate": 2.0985193795989345e-07,
+      "loss": 0.4371,
+      "step": 1695
+    },
+    {
+      "epoch": 0.9169219679221481,
+      "grad_norm": 0.3041470944881439,
+      "learning_rate": 2.071543528233805e-07,
+      "loss": 0.5015,
+      "step": 1696
+    },
+    {
+      "epoch": 0.9174626058749324,
+      "grad_norm": 0.299441397190094,
+      "learning_rate": 2.0447385156305565e-07,
+      "loss": 0.4557,
+      "step": 1697
+    },
+    {
+      "epoch": 0.9180032438277167,
+      "grad_norm": 0.29456791281700134,
+      "learning_rate": 2.0181044373344172e-07,
+      "loss": 0.4744,
+      "step": 1698
+    },
+    {
+      "epoch": 0.918543881780501,
+      "grad_norm": 0.31137824058532715,
+      "learning_rate": 1.9916413882813235e-07,
+      "loss": 0.4564,
+      "step": 1699
+    },
+    {
+      "epoch": 0.9190845197332853,
+      "grad_norm": 0.29632532596588135,
+      "learning_rate": 1.9653494627975888e-07,
+      "loss": 0.4657,
+      "step": 1700
+    },
+    {
+      "epoch": 0.9196251576860696,
+      "grad_norm": 0.28566405177116394,
+      "learning_rate": 1.9392287545995536e-07,
+      "loss": 0.4983,
+      "step": 1701
+    },
+    {
+      "epoch": 0.9201657956388538,
+      "grad_norm": 0.31092479825019836,
+      "learning_rate": 1.913279356793285e-07,
+      "loss": 0.4584,
+      "step": 1702
+    },
+    {
+      "epoch": 0.9207064335916382,
+      "grad_norm": 0.2855190634727478,
+      "learning_rate": 1.8875013618742e-07,
+      "loss": 0.4444,
+      "step": 1703
+    },
+    {
+      "epoch": 0.9212470715444224,
+      "grad_norm": 0.29378846287727356,
+      "learning_rate": 1.8618948617267764e-07,
+      "loss": 0.4697,
+      "step": 1704
+    },
+    {
+      "epoch": 0.9217877094972067,
+      "grad_norm": 0.30169007182121277,
+      "learning_rate": 1.8364599476241862e-07,
+      "loss": 0.4616,
+      "step": 1705
+    },
+    {
+      "epoch": 0.922328347449991,
+      "grad_norm": 0.3052110970020294,
+      "learning_rate": 1.8111967102280082e-07,
+      "loss": 0.4854,
+      "step": 1706
+    },
+    {
+      "epoch": 0.9228689854027753,
+      "grad_norm": 0.3236134350299835,
+      "learning_rate": 1.7861052395878764e-07,
+      "loss": 0.4437,
+      "step": 1707
+    },
+    {
+      "epoch": 0.9234096233555595,
+      "grad_norm": 0.29827356338500977,
+      "learning_rate": 1.7611856251411818e-07,
+      "loss": 0.4458,
+      "step": 1708
+    },
+    {
+      "epoch": 0.9239502613083439,
+      "grad_norm": 0.2997409701347351,
+      "learning_rate": 1.7364379557127387e-07,
+      "loss": 0.4891,
+      "step": 1709
+    },
+    {
+      "epoch": 0.9244908992611282,
+      "grad_norm": 0.2794349491596222,
+      "learning_rate": 1.711862319514457e-07,
+      "loss": 0.4808,
+      "step": 1710
+    },
+    {
+      "epoch": 0.9250315372139124,
+      "grad_norm": 0.299027681350708,
+      "learning_rate": 1.6874588041450535e-07,
+      "loss": 0.4743,
+      "step": 1711
+    },
+    {
+      "epoch": 0.9255721751666967,
+      "grad_norm": 0.291069895029068,
+      "learning_rate": 1.6632274965897365e-07,
+      "loss": 0.4791,
+      "step": 1712
+    },
+    {
+      "epoch": 0.926112813119481,
+      "grad_norm": 0.3018452227115631,
+      "learning_rate": 1.639168483219872e-07,
+      "loss": 0.479,
+      "step": 1713
+    },
+    {
+      "epoch": 0.9266534510722653,
+      "grad_norm": 0.2994323670864105,
+      "learning_rate": 1.6152818497926993e-07,
+      "loss": 0.4573,
+      "step": 1714
+    },
+    {
+      "epoch": 0.9271940890250495,
+      "grad_norm": 0.27843761444091797,
+      "learning_rate": 1.5915676814510173e-07,
+      "loss": 0.4585,
+      "step": 1715
+    },
+    {
+      "epoch": 0.9277347269778339,
+      "grad_norm": 0.30757150053977966,
+      "learning_rate": 1.5680260627228772e-07,
+      "loss": 0.4708,
+      "step": 1716
+    },
+    {
+      "epoch": 0.9282753649306181,
+      "grad_norm": 0.31326669454574585,
+      "learning_rate": 1.5446570775212944e-07,
+      "loss": 0.4671,
+      "step": 1717
+    },
+    {
+      "epoch": 0.9288160028834024,
+      "grad_norm": 0.31248635053634644,
+      "learning_rate": 1.5214608091439265e-07,
+      "loss": 0.4639,
+      "step": 1718
+    },
+    {
+      "epoch": 0.9293566408361867,
+      "grad_norm": 0.29233023524284363,
+      "learning_rate": 1.4984373402728014e-07,
+      "loss": 0.4691,
+      "step": 1719
+    },
+    {
+      "epoch": 0.929897278788971,
+      "grad_norm": 0.32531052827835083,
+      "learning_rate": 1.4755867529740064e-07,
+      "loss": 0.5093,
+      "step": 1720
+    },
+    {
+      "epoch": 0.9304379167417552,
+      "grad_norm": 0.35300230979919434,
+      "learning_rate": 1.4529091286973994e-07,
+      "loss": 0.488,
+      "step": 1721
+    },
+    {
+      "epoch": 0.9309785546945396,
+      "grad_norm": 0.3052767813205719,
+      "learning_rate": 1.4304045482763263e-07,
+      "loss": 0.4674,
+      "step": 1722
+    },
+    {
+      "epoch": 0.9315191926473239,
+      "grad_norm": 0.2952291667461395,
+      "learning_rate": 1.408073091927309e-07,
+      "loss": 0.4572,
+      "step": 1723
+    },
+    {
+      "epoch": 0.9320598306001081,
+      "grad_norm": 0.2960074841976166,
+      "learning_rate": 1.3859148392498023e-07,
+      "loss": 0.464,
+      "step": 1724
+    },
+    {
+      "epoch": 0.9326004685528924,
+      "grad_norm": 0.28832143545150757,
+      "learning_rate": 1.3639298692258606e-07,
+      "loss": 0.4699,
+      "step": 1725
+    },
+    {
+      "epoch": 0.9331411065056767,
+      "grad_norm": 0.29948270320892334,
+      "learning_rate": 1.342118260219899e-07,
+      "loss": 0.4722,
+      "step": 1726
+    },
+    {
+      "epoch": 0.933681744458461,
+      "grad_norm": 0.28257155418395996,
+      "learning_rate": 1.320480089978382e-07,
+      "loss": 0.4768,
+      "step": 1727
+    },
+    {
+      "epoch": 0.9342223824112452,
+      "grad_norm": 0.2848031520843506,
+      "learning_rate": 1.2990154356295636e-07,
+      "loss": 0.453,
+      "step": 1728
+    },
+    {
+      "epoch": 0.9347630203640296,
+      "grad_norm": 0.29781103134155273,
+      "learning_rate": 1.2777243736832202e-07,
+      "loss": 0.4798,
+      "step": 1729
+    },
+    {
+      "epoch": 0.9353036583168138,
+      "grad_norm": 0.30655109882354736,
+      "learning_rate": 1.2566069800303393e-07,
+      "loss": 0.4732,
+      "step": 1730
+    },
+    {
+      "epoch": 0.9358442962695981,
+      "grad_norm": 0.30582815408706665,
+      "learning_rate": 1.2356633299429044e-07,
+      "loss": 0.4601,
+      "step": 1731
+    },
+    {
+      "epoch": 0.9363849342223824,
+      "grad_norm": 0.2904635965824127,
+      "learning_rate": 1.2148934980735772e-07,
+      "loss": 0.4922,
+      "step": 1732
+    },
+    {
+      "epoch": 0.9369255721751667,
+      "grad_norm": 0.30475184321403503,
+      "learning_rate": 1.1942975584554594e-07,
+      "loss": 0.4747,
+      "step": 1733
+    },
+    {
+      "epoch": 0.937466210127951,
+      "grad_norm": 0.29005348682403564,
+      "learning_rate": 1.1738755845018323e-07,
+      "loss": 0.4601,
+      "step": 1734
+    },
+    {
+      "epoch": 0.9380068480807353,
+      "grad_norm": 0.29658186435699463,
+      "learning_rate": 1.1536276490058784e-07,
+      "loss": 0.4417,
+      "step": 1735
+    },
+    {
+      "epoch": 0.9385474860335196,
+      "grad_norm": 0.2936530113220215,
+      "learning_rate": 1.1335538241404099e-07,
+      "loss": 0.4629,
+      "step": 1736
+    },
+    {
+      "epoch": 0.9390881239863038,
+      "grad_norm": 0.3370441198348999,
+      "learning_rate": 1.1136541814576574e-07,
+      "loss": 0.4861,
+      "step": 1737
+    },
+    {
+      "epoch": 0.9396287619390882,
+      "grad_norm": 0.2809031009674072,
+      "learning_rate": 1.0939287918889652e-07,
+      "loss": 0.4676,
+      "step": 1738
+    },
+    {
+      "epoch": 0.9401693998918724,
+      "grad_norm": 0.29533958435058594,
+      "learning_rate": 1.0743777257445853e-07,
+      "loss": 0.4947,
+      "step": 1739
+    },
+    {
+      "epoch": 0.9407100378446567,
+      "grad_norm": 0.2839764654636383,
+      "learning_rate": 1.055001052713378e-07,
+      "loss": 0.4924,
+      "step": 1740
+    },
+    {
+      "epoch": 0.9412506757974409,
+      "grad_norm": 0.31136807799339294,
+      "learning_rate": 1.0357988418625897e-07,
+      "loss": 0.467,
+      "step": 1741
+    },
+    {
+      "epoch": 0.9417913137502253,
+      "grad_norm": 0.3142279386520386,
+      "learning_rate": 1.0167711616376196e-07,
+      "loss": 0.4629,
+      "step": 1742
+    },
+    {
+      "epoch": 0.9423319517030095,
+      "grad_norm": 0.27500444650650024,
+      "learning_rate": 9.979180798617538e-08,
+      "loss": 0.4686,
+      "step": 1743
+    },
+    {
+      "epoch": 0.9428725896557938,
+      "grad_norm": 0.3138892650604248,
+      "learning_rate": 9.792396637359203e-08,
+      "loss": 0.4638,
+      "step": 1744
+    },
+    {
+      "epoch": 0.9434132276085782,
+      "grad_norm": 0.3079321086406708,
+      "learning_rate": 9.607359798384785e-08,
+      "loss": 0.4491,
+      "step": 1745
+    },
+    {
+      "epoch": 0.9439538655613624,
+      "grad_norm": 0.29620155692100525,
+      "learning_rate": 9.424070941249419e-08,
+      "loss": 0.4593,
+      "step": 1746
+    },
+    {
+      "epoch": 0.9444945035141467,
+      "grad_norm": 0.32409822940826416,
+      "learning_rate": 9.242530719277776e-08,
+      "loss": 0.4248,
+      "step": 1747
+    },
+    {
+      "epoch": 0.945035141466931,
+      "grad_norm": 0.2947947680950165,
+      "learning_rate": 9.062739779561624e-08,
+      "loss": 0.4769,
+      "step": 1748
+    },
+    {
+      "epoch": 0.9455757794197153,
+      "grad_norm": 0.28672558069229126,
+      "learning_rate": 8.884698762957334e-08,
+      "loss": 0.4895,
+      "step": 1749
+    },
+    {
+      "epoch": 0.9461164173724995,
+      "grad_norm": 0.2779400050640106,
+      "learning_rate": 8.708408304083927e-08,
+      "loss": 0.4694,
+      "step": 1750
+    },
+    {
+      "epoch": 0.9466570553252839,
+      "grad_norm": 0.3198831379413605,
+      "learning_rate": 8.53386903132053e-08,
+      "loss": 0.4591,
+      "step": 1751
+    },
+    {
+      "epoch": 0.9471976932780681,
+      "grad_norm": 0.3286699652671814,
+      "learning_rate": 8.361081566804318e-08,
+      "loss": 0.4965,
+      "step": 1752
+    },
+    {
+      "epoch": 0.9477383312308524,
+      "grad_norm": 0.2861204743385315,
+      "learning_rate": 8.190046526428241e-08,
+      "loss": 0.4464,
+      "step": 1753
+    },
+    {
+      "epoch": 0.9482789691836367,
+      "grad_norm": 0.28904879093170166,
+      "learning_rate": 8.020764519838686e-08,
+      "loss": 0.4665,
+      "step": 1754
+    },
+    {
+      "epoch": 0.948819607136421,
+      "grad_norm": 0.308715283870697,
+      "learning_rate": 7.853236150433541e-08,
+      "loss": 0.4604,
+      "step": 1755
+    },
+    {
+      "epoch": 0.9493602450892052,
+      "grad_norm": 0.2850310206413269,
+      "learning_rate": 7.687462015360026e-08,
+      "loss": 0.4536,
+      "step": 1756
+    },
+    {
+      "epoch": 0.9499008830419895,
+      "grad_norm": 0.32719627022743225,
+      "learning_rate": 7.523442705512196e-08,
+      "loss": 0.4873,
+      "step": 1757
+    },
+    {
+      "epoch": 0.9504415209947739,
+      "grad_norm": 0.29945439100265503,
+      "learning_rate": 7.36117880552939e-08,
+      "loss": 0.4827,
+      "step": 1758
+    },
+    {
+      "epoch": 0.9509821589475581,
+      "grad_norm": 0.31181010603904724,
+      "learning_rate": 7.200670893793727e-08,
+      "loss": 0.4435,
+      "step": 1759
+    },
+    {
+      "epoch": 0.9515227969003424,
+      "grad_norm": 0.29942837357521057,
+      "learning_rate": 7.041919542428221e-08,
+      "loss": 0.4747,
+      "step": 1760
+    },
+    {
+      "epoch": 0.9520634348531267,
+      "grad_norm": 0.3144790232181549,
+      "learning_rate": 6.884925317294678e-08,
+      "loss": 0.481,
+      "step": 1761
+    },
+    {
+      "epoch": 0.952604072805911,
+      "grad_norm": 0.321038156747818,
+      "learning_rate": 6.72968877799185e-08,
+      "loss": 0.4891,
+      "step": 1762
+    },
+    {
+      "epoch": 0.9531447107586952,
+      "grad_norm": 0.29290351271629333,
+      "learning_rate": 6.576210477853007e-08,
+      "loss": 0.4755,
+      "step": 1763
+    },
+    {
+      "epoch": 0.9536853487114796,
+      "grad_norm": 0.31532248854637146,
+      "learning_rate": 6.424490963944597e-08,
+      "loss": 0.4916,
+      "step": 1764
+    },
+    {
+      "epoch": 0.9542259866642638,
+      "grad_norm": 0.29246556758880615,
+      "learning_rate": 6.274530777063747e-08,
+      "loss": 0.4765,
+      "step": 1765
+    },
+    {
+      "epoch": 0.9547666246170481,
+      "grad_norm": 0.28569135069847107,
+      "learning_rate": 6.126330451736495e-08,
+      "loss": 0.4446,
+      "step": 1766
+    },
+    {
+      "epoch": 0.9553072625698324,
+      "grad_norm": 0.2923862338066101,
+      "learning_rate": 5.97989051621617e-08,
+      "loss": 0.4793,
+      "step": 1767
+    },
+    {
+      "epoch": 0.9558479005226167,
+      "grad_norm": 0.27991148829460144,
+      "learning_rate": 5.835211492481063e-08,
+      "loss": 0.4756,
+      "step": 1768
+    },
+    {
+      "epoch": 0.956388538475401,
+      "grad_norm": 0.30270445346832275,
+      "learning_rate": 5.6922938962329364e-08,
+      "loss": 0.4498,
+      "step": 1769
+    },
+    {
+      "epoch": 0.9569291764281853,
+      "grad_norm": 0.29934385418891907,
+      "learning_rate": 5.551138236894793e-08,
+      "loss": 0.4677,
+      "step": 1770
+    },
+    {
+      "epoch": 0.9574698143809696,
+      "grad_norm": 0.2954845428466797,
+      "learning_rate": 5.411745017609493e-08,
+      "loss": 0.4762,
+      "step": 1771
+    },
+    {
+      "epoch": 0.9580104523337538,
+      "grad_norm": 0.29608166217803955,
+      "learning_rate": 5.274114735237812e-08,
+      "loss": 0.4584,
+      "step": 1772
+    },
+    {
+      "epoch": 0.9585510902865381,
+      "grad_norm": 0.3020345866680145,
+      "learning_rate": 5.138247880356384e-08,
+      "loss": 0.4686,
+      "step": 1773
+    },
+    {
+      "epoch": 0.9590917282393224,
+      "grad_norm": 0.3264354169368744,
+      "learning_rate": 5.004144937256372e-08,
+      "loss": 0.4808,
+      "step": 1774
+    },
+    {
+      "epoch": 0.9596323661921067,
+      "grad_norm": 0.29560211300849915,
+      "learning_rate": 4.8718063839414683e-08,
+      "loss": 0.4731,
+      "step": 1775
+    },
+    {
+      "epoch": 0.9601730041448909,
+      "grad_norm": 0.28680717945098877,
+      "learning_rate": 4.741232692126396e-08,
+      "loss": 0.4502,
+      "step": 1776
+    },
+    {
+      "epoch": 0.9607136420976753,
+      "grad_norm": 0.30334657430648804,
+      "learning_rate": 4.612424327234966e-08,
+      "loss": 0.4602,
+      "step": 1777
+    },
+    {
+      "epoch": 0.9612542800504595,
+      "grad_norm": 0.28252384066581726,
+      "learning_rate": 4.485381748398576e-08,
+      "loss": 0.4527,
+      "step": 1778
+    },
+    {
+      "epoch": 0.9617949180032438,
+      "grad_norm": 0.3101551830768585,
+      "learning_rate": 4.360105408454718e-08,
+      "loss": 0.4967,
+      "step": 1779
+    },
+    {
+      "epoch": 0.9623355559560282,
+      "grad_norm": 0.2830714285373688,
+      "learning_rate": 4.236595753944972e-08,
+      "loss": 0.4613,
+      "step": 1780
+    },
+    {
+      "epoch": 0.9628761939088124,
+      "grad_norm": 0.298048198223114,
+      "learning_rate": 4.114853225113902e-08,
+      "loss": 0.459,
+      "step": 1781
+    },
+    {
+      "epoch": 0.9634168318615967,
+      "grad_norm": 0.30758899450302124,
+      "learning_rate": 3.994878255907053e-08,
+      "loss": 0.4355,
+      "step": 1782
+    },
+    {
+      "epoch": 0.963957469814381,
+      "grad_norm": 0.31392624974250793,
+      "learning_rate": 3.8766712739696786e-08,
+      "loss": 0.434,
+      "step": 1783
+    },
+    {
+      "epoch": 0.9644981077671653,
+      "grad_norm": 0.3108672797679901,
+      "learning_rate": 3.7602327006450166e-08,
+      "loss": 0.4323,
+      "step": 1784
+    },
+    {
+      "epoch": 0.9650387457199495,
+      "grad_norm": 0.3078506290912628,
+      "learning_rate": 3.645562950973014e-08,
+      "loss": 0.4671,
+      "step": 1785
+    },
+    {
+      "epoch": 0.9655793836727338,
+      "grad_norm": 0.29575785994529724,
+      "learning_rate": 3.5326624336886604e-08,
+      "loss": 0.4592,
+      "step": 1786
+    },
+    {
+      "epoch": 0.9661200216255181,
+      "grad_norm": 0.2994788885116577,
+      "learning_rate": 3.4215315512206584e-08,
+      "loss": 0.4864,
+      "step": 1787
+    },
+    {
+      "epoch": 0.9666606595783024,
+      "grad_norm": 0.2929299771785736,
+      "learning_rate": 3.312170699689865e-08,
+      "loss": 0.4609,
+      "step": 1788
+    },
+    {
+      "epoch": 0.9672012975310866,
+      "grad_norm": 0.31516411900520325,
+      "learning_rate": 3.204580268907909e-08,
+      "loss": 0.4905,
+      "step": 1789
+    },
+    {
+      "epoch": 0.967741935483871,
+      "grad_norm": 0.31366169452667236,
+      "learning_rate": 3.0987606423759644e-08,
+      "loss": 0.4793,
+      "step": 1790
+    },
+    {
+      "epoch": 0.9682825734366552,
+      "grad_norm": 0.30144020915031433,
+      "learning_rate": 2.9947121972832e-08,
+      "loss": 0.4619,
+      "step": 1791
+    },
+    {
+      "epoch": 0.9688232113894395,
+      "grad_norm": 0.2976379692554474,
+      "learning_rate": 2.8924353045054475e-08,
+      "loss": 0.4555,
+      "step": 1792
+    },
+    {
+      "epoch": 0.9693638493422239,
+      "grad_norm": 0.3140285909175873,
+      "learning_rate": 2.7919303286039202e-08,
+      "loss": 0.4592,
+      "step": 1793
+    },
+    {
+      "epoch": 0.9699044872950081,
+      "grad_norm": 0.28508275747299194,
+      "learning_rate": 2.693197627823996e-08,
+      "loss": 0.4784,
+      "step": 1794
+    },
+    {
+      "epoch": 0.9704451252477924,
+      "grad_norm": 0.29108139872550964,
+      "learning_rate": 2.5962375540937724e-08,
+      "loss": 0.4857,
+      "step": 1795
+    },
+    {
+      "epoch": 0.9709857632005767,
+      "grad_norm": 0.3323630094528198,
+      "learning_rate": 2.5010504530229574e-08,
+      "loss": 0.4616,
+      "step": 1796
+    },
+    {
+      "epoch": 0.971526401153361,
+      "grad_norm": 0.31090638041496277,
+      "learning_rate": 2.4076366639015914e-08,
+      "loss": 0.4782,
+      "step": 1797
+    },
+    {
+      "epoch": 0.9720670391061452,
+      "grad_norm": 0.31540656089782715,
+      "learning_rate": 2.3159965196987156e-08,
+      "loss": 0.4851,
+      "step": 1798
+    },
+    {
+      "epoch": 0.9726076770589296,
+      "grad_norm": 0.283552885055542,
+      "learning_rate": 2.2261303470614282e-08,
+      "loss": 0.4473,
+      "step": 1799
+    },
+    {
+      "epoch": 0.9731483150117138,
+      "grad_norm": 0.2891945540904999,
+      "learning_rate": 2.1380384663135523e-08,
+      "loss": 0.4574,
+      "step": 1800
+    },
+    {
+      "epoch": 0.9736889529644981,
+      "grad_norm": 0.2970227897167206,
+      "learning_rate": 2.0517211914545254e-08,
+      "loss": 0.4785,
+      "step": 1801
+    },
+    {
+      "epoch": 0.9742295909172823,
+      "grad_norm": 0.30604010820388794,
+      "learning_rate": 1.967178830158234e-08,
+      "loss": 0.4464,
+      "step": 1802
+    },
+    {
+      "epoch": 0.9747702288700667,
+      "grad_norm": 0.29695698618888855,
+      "learning_rate": 1.8844116837719582e-08,
+      "loss": 0.498,
+      "step": 1803
+    },
+    {
+      "epoch": 0.975310866822851,
+      "grad_norm": 0.29107484221458435,
+      "learning_rate": 1.803420047315485e-08,
+      "loss": 0.4752,
+      "step": 1804
+    },
+    {
+      "epoch": 0.9758515047756352,
+      "grad_norm": 0.2813703119754791,
+      "learning_rate": 1.724204209479663e-08,
+      "loss": 0.4509,
+      "step": 1805
+    },
+    {
+      "epoch": 0.9763921427284196,
+      "grad_norm": 0.2851557731628418,
+      "learning_rate": 1.646764452625682e-08,
+      "loss": 0.4704,
+      "step": 1806
+    },
+    {
+      "epoch": 0.9769327806812038,
+      "grad_norm": 0.2960273325443268,
+      "learning_rate": 1.5711010527839633e-08,
+      "loss": 0.4505,
+      "step": 1807
+    },
+    {
+      "epoch": 0.9774734186339881,
+      "grad_norm": 0.3089202642440796,
+      "learning_rate": 1.4972142796532696e-08,
+      "loss": 0.4747,
+      "step": 1808
+    },
+    {
+      "epoch": 0.9780140565867724,
+      "grad_norm": 0.30507782101631165,
+      "learning_rate": 1.4251043965994304e-08,
+      "loss": 0.4496,
+      "step": 1809
+    },
+    {
+      "epoch": 0.9785546945395567,
+      "grad_norm": 0.29044288396835327,
+      "learning_rate": 1.3547716606548967e-08,
+      "loss": 0.4645,
+      "step": 1810
+    },
+    {
+      "epoch": 0.9790953324923409,
+      "grad_norm": 0.2851908802986145,
+      "learning_rate": 1.2862163225174084e-08,
+      "loss": 0.4587,
+      "step": 1811
+    },
+    {
+      "epoch": 0.9796359704451253,
+      "grad_norm": 0.2960631251335144,
+      "learning_rate": 1.2194386265492742e-08,
+      "loss": 0.4429,
+      "step": 1812
+    },
+    {
+      "epoch": 0.9801766083979095,
+      "grad_norm": 0.3134407103061676,
+      "learning_rate": 1.1544388107765924e-08,
+      "loss": 0.4452,
+      "step": 1813
+    },
+    {
+      "epoch": 0.9807172463506938,
+      "grad_norm": 0.29251259565353394,
+      "learning_rate": 1.0912171068880318e-08,
+      "loss": 0.4647,
+      "step": 1814
+    },
+    {
+      "epoch": 0.9812578843034782,
+      "grad_norm": 0.3051975667476654,
+      "learning_rate": 1.029773740234552e-08,
+      "loss": 0.4604,
+      "step": 1815
+    },
+    {
+      "epoch": 0.9817985222562624,
+      "grad_norm": 0.31170353293418884,
+      "learning_rate": 9.701089298281285e-09,
+      "loss": 0.453,
+      "step": 1816
+    },
+    {
+      "epoch": 0.9823391602090467,
+      "grad_norm": 0.2902314066886902,
+      "learning_rate": 9.12222888341252e-09,
+      "loss": 0.4533,
+      "step": 1817
+    },
+    {
+      "epoch": 0.9828797981618309,
+      "grad_norm": 0.3045901656150818,
+      "learning_rate": 8.561158221060406e-09,
+      "loss": 0.4514,
+      "step": 1818
+    },
+    {
+      "epoch": 0.9834204361146153,
+      "grad_norm": 0.29368510842323303,
+      "learning_rate": 8.017879311134624e-09,
+      "loss": 0.4808,
+      "step": 1819
+    },
+    {
+      "epoch": 0.9839610740673995,
+      "grad_norm": 0.27795591950416565,
+      "learning_rate": 7.492394090128364e-09,
+      "loss": 0.4974,
+      "step": 1820
+    },
+    {
+      "epoch": 0.9845017120201838,
+      "grad_norm": 0.29451608657836914,
+      "learning_rate": 6.98470443110888e-09,
+      "loss": 0.4874,
+      "step": 1821
+    },
+    {
+      "epoch": 0.9850423499729681,
+      "grad_norm": 0.29039186239242554,
+      "learning_rate": 6.4948121437125035e-09,
+      "loss": 0.4587,
+      "step": 1822
+    },
+    {
+      "epoch": 0.9855829879257524,
+      "grad_norm": 0.29998692870140076,
+      "learning_rate": 6.022718974137976e-09,
+      "loss": 0.4895,
+      "step": 1823
+    },
+    {
+      "epoch": 0.9861236258785366,
+      "grad_norm": 0.27220648527145386,
+      "learning_rate": 5.568426605139232e-09,
+      "loss": 0.475,
+      "step": 1824
+    },
+    {
+      "epoch": 0.986664263831321,
+      "grad_norm": 0.3113320767879486,
+      "learning_rate": 5.131936656020409e-09,
+      "loss": 0.4675,
+      "step": 1825
+    },
+    {
+      "epoch": 0.9872049017841052,
+      "grad_norm": 0.2876143157482147,
+      "learning_rate": 4.713250682629733e-09,
+      "loss": 0.445,
+      "step": 1826
+    },
+    {
+      "epoch": 0.9877455397368895,
+      "grad_norm": 0.2922397553920746,
+      "learning_rate": 4.312370177353975e-09,
+      "loss": 0.4666,
+      "step": 1827
+    },
+    {
+      "epoch": 0.9882861776896739,
+      "grad_norm": 0.3193817436695099,
+      "learning_rate": 3.929296569112895e-09,
+      "loss": 0.4823,
+      "step": 1828
+    },
+    {
+      "epoch": 0.9888268156424581,
+      "grad_norm": 0.30332955718040466,
+      "learning_rate": 3.5640312233548024e-09,
+      "loss": 0.4637,
+      "step": 1829
+    },
+    {
+      "epoch": 0.9893674535952424,
+      "grad_norm": 0.28622302412986755,
+      "learning_rate": 3.2165754420510063e-09,
+      "loss": 0.4491,
+      "step": 1830
+    },
+    {
+      "epoch": 0.9899080915480267,
+      "grad_norm": 0.2920747399330139,
+      "learning_rate": 2.886930463691928e-09,
+      "loss": 0.4712,
+      "step": 1831
+    },
+    {
+      "epoch": 0.990448729500811,
+      "grad_norm": 0.2950655221939087,
+      "learning_rate": 2.5750974632809955e-09,
+      "loss": 0.4571,
+      "step": 1832
+    },
+    {
+      "epoch": 0.9909893674535952,
+      "grad_norm": 0.30665168166160583,
+      "learning_rate": 2.2810775523329775e-09,
+      "loss": 0.4675,
+      "step": 1833
+    },
+    {
+      "epoch": 0.9915300054063795,
+      "grad_norm": 0.30462467670440674,
+      "learning_rate": 2.0048717788684335e-09,
+      "loss": 0.4618,
+      "step": 1834
+    },
+    {
+      "epoch": 0.9920706433591638,
+      "grad_norm": 0.28570374846458435,
+      "learning_rate": 1.746481127409827e-09,
+      "loss": 0.4415,
+      "step": 1835
+    },
+    {
+      "epoch": 0.9926112813119481,
+      "grad_norm": 0.33271685242652893,
+      "learning_rate": 1.5059065189787502e-09,
+      "loss": 0.4603,
+      "step": 1836
+    },
+    {
+      "epoch": 0.9931519192647323,
+      "grad_norm": 0.3160949647426605,
+      "learning_rate": 1.2831488110920386e-09,
+      "loss": 0.4551,
+      "step": 1837
+    },
+    {
+      "epoch": 0.9936925572175167,
+      "grad_norm": 0.2587985694408417,
+      "learning_rate": 1.07820879775955e-09,
+      "loss": 0.4744,
+      "step": 1838
+    },
+    {
+      "epoch": 0.994233195170301,
+      "grad_norm": 0.32415860891342163,
+      "learning_rate": 8.910872094802792e-10,
+      "loss": 0.4525,
+      "step": 1839
+    },
+    {
+      "epoch": 0.9947738331230852,
+      "grad_norm": 0.27735623717308044,
+      "learning_rate": 7.217847132401367e-10,
+      "loss": 0.4508,
+      "step": 1840
+    },
+    {
+      "epoch": 0.9953144710758696,
+      "grad_norm": 0.2976240813732147,
+      "learning_rate": 5.703019125102849e-10,
+      "loss": 0.5082,
+      "step": 1841
+    },
+    {
+      "epoch": 0.9958551090286538,
+      "grad_norm": 0.28493812680244446,
+      "learning_rate": 4.3663934724436086e-10,
+      "loss": 0.4465,
+      "step": 1842
+    },
+    {
+      "epoch": 0.9963957469814381,
+      "grad_norm": 0.3043738603591919,
+      "learning_rate": 3.20797493876257e-10,
+      "loss": 0.4566,
+      "step": 1843
+    },
+    {
+      "epoch": 0.9969363849342224,
+      "grad_norm": 0.3110690116882324,
+      "learning_rate": 2.227767653190105e-10,
+      "loss": 0.5034,
+      "step": 1844
+    },
+    {
+      "epoch": 0.9974770228870067,
+      "grad_norm": 0.29455140233039856,
+      "learning_rate": 1.4257751096202755e-10,
+      "loss": 0.4548,
+      "step": 1845
+    },
+    {
+      "epoch": 0.9980176608397909,
+      "grad_norm": 0.2990035116672516,
+      "learning_rate": 8.020001667330412e-11,
+      "loss": 0.4927,
+      "step": 1846
+    },
+    {
+      "epoch": 0.9985582987925753,
+      "grad_norm": 0.29310280084609985,
+      "learning_rate": 3.564450479387471e-11,
+      "loss": 0.4643,
+      "step": 1847
+    },
+    {
+      "epoch": 0.9990989367453595,
+      "grad_norm": 0.2967851161956787,
+      "learning_rate": 8.911134139477639e-12,
+      "loss": 0.4731,
+      "step": 1848
+    },
+    {
+      "epoch": 0.9996395746981438,
+      "grad_norm": 0.3127877712249756,
+      "learning_rate": 0.0,
+      "loss": 0.4589,
+      "step": 1849
+    },
+    {
+      "epoch": 0.9996395746981438,
+      "step": 1849,
+      "total_flos": 2827754950098944.0,
+      "train_loss": 0.4939856140145487,
+      "train_runtime": 65758.0763,
+      "train_samples_per_second": 2.7,
+      "train_steps_per_second": 0.028
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1849,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2827754950098944.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}