diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,25893 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9983761840324763,
+  "eval_steps": 500,
+  "global_step": 3693,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008119079837618403,
+      "grad_norm": 2.4709081649780273,
+      "learning_rate": 2.702702702702703e-08,
+      "loss": 0.7272,
+      "step": 1
+    },
+    {
+      "epoch": 0.0016238159675236806,
+      "grad_norm": 2.6747725009918213,
+      "learning_rate": 5.405405405405406e-08,
+      "loss": 0.7719,
+      "step": 2
+    },
+    {
+      "epoch": 0.002435723951285521,
+      "grad_norm": 2.624976634979248,
+      "learning_rate": 8.108108108108109e-08,
+      "loss": 0.7657,
+      "step": 3
+    },
+    {
+      "epoch": 0.003247631935047361,
+      "grad_norm": 2.5566775798797607,
+      "learning_rate": 1.0810810810810812e-07,
+      "loss": 0.7069,
+      "step": 4
+    },
+    {
+      "epoch": 0.0040595399188092015,
+      "grad_norm": 2.694103717803955,
+      "learning_rate": 1.3513513513513515e-07,
+      "loss": 0.7369,
+      "step": 5
+    },
+    {
+      "epoch": 0.004871447902571042,
+      "grad_norm": 2.5976076126098633,
+      "learning_rate": 1.6216216216216218e-07,
+      "loss": 0.7199,
+      "step": 6
+    },
+    {
+      "epoch": 0.005683355886332882,
+      "grad_norm": 2.674605131149292,
+      "learning_rate": 1.8918918918918921e-07,
+      "loss": 0.7703,
+      "step": 7
+    },
+    {
+      "epoch": 0.006495263870094722,
+      "grad_norm": 2.6879048347473145,
+      "learning_rate": 2.1621621621621625e-07,
+      "loss": 0.737,
+      "step": 8
+    },
+    {
+      "epoch": 0.007307171853856563,
+      "grad_norm": 2.684795618057251,
+      "learning_rate": 2.4324324324324326e-07,
+      "loss": 0.7437,
+      "step": 9
+    },
+    {
+      "epoch": 0.008119079837618403,
+      "grad_norm": 2.7520954608917236,
+      "learning_rate": 2.702702702702703e-07,
+      "loss": 0.7897,
+      "step": 10
+    },
+    {
+      "epoch": 0.008930987821380243,
+      "grad_norm": 2.616285562515259,
+      "learning_rate": 2.972972972972973e-07,
+      "loss": 0.7403,
+      "step": 11
+    },
+    {
+      "epoch": 0.009742895805142085,
+      "grad_norm": 2.5858795642852783,
+      "learning_rate": 3.2432432432432436e-07,
+      "loss": 0.7354,
+      "step": 12
+    },
+    {
+      "epoch": 0.010554803788903925,
+      "grad_norm": 2.6487784385681152,
+      "learning_rate": 3.513513513513514e-07,
+      "loss": 0.7522,
+      "step": 13
+    },
+    {
+      "epoch": 0.011366711772665765,
+      "grad_norm": 2.5158751010894775,
+      "learning_rate": 3.7837837837837843e-07,
+      "loss": 0.7716,
+      "step": 14
+    },
+    {
+      "epoch": 0.012178619756427604,
+      "grad_norm": 2.548631429672241,
+      "learning_rate": 4.0540540540540546e-07,
+      "loss": 0.7454,
+      "step": 15
+    },
+    {
+      "epoch": 0.012990527740189444,
+      "grad_norm": 2.4671082496643066,
+      "learning_rate": 4.324324324324325e-07,
+      "loss": 0.7273,
+      "step": 16
+    },
+    {
+      "epoch": 0.013802435723951286,
+      "grad_norm": 2.4273195266723633,
+      "learning_rate": 4.5945945945945953e-07,
+      "loss": 0.7347,
+      "step": 17
+    },
+    {
+      "epoch": 0.014614343707713126,
+      "grad_norm": 2.57314133644104,
+      "learning_rate": 4.864864864864865e-07,
+      "loss": 0.7603,
+      "step": 18
+    },
+    {
+      "epoch": 0.015426251691474966,
+      "grad_norm": 2.0447757244110107,
+      "learning_rate": 5.135135135135135e-07,
+      "loss": 0.722,
+      "step": 19
+    },
+    {
+      "epoch": 0.016238159675236806,
+      "grad_norm": 2.146315336227417,
+      "learning_rate": 5.405405405405406e-07,
+      "loss": 0.7293,
+      "step": 20
+    },
+    {
+      "epoch": 0.017050067658998646,
+      "grad_norm": 2.086336135864258,
+      "learning_rate": 5.675675675675676e-07,
+      "loss": 0.7125,
+      "step": 21
+    },
+    {
+      "epoch": 0.017861975642760486,
+      "grad_norm": 1.9912855625152588,
+      "learning_rate": 5.945945945945947e-07,
+      "loss": 0.7243,
+      "step": 22
+    },
+    {
+      "epoch": 0.018673883626522326,
+      "grad_norm": 2.073620080947876,
+      "learning_rate": 6.216216216216217e-07,
+      "loss": 0.719,
+      "step": 23
+    },
+    {
+      "epoch": 0.01948579161028417,
+      "grad_norm": 1.9811540842056274,
+      "learning_rate": 6.486486486486487e-07,
+      "loss": 0.7175,
+      "step": 24
+    },
+    {
+      "epoch": 0.02029769959404601,
+      "grad_norm": 1.9595799446105957,
+      "learning_rate": 6.756756756756758e-07,
+      "loss": 0.7404,
+      "step": 25
+    },
+    {
+      "epoch": 0.02110960757780785,
+      "grad_norm": 1.374733567237854,
+      "learning_rate": 7.027027027027028e-07,
+      "loss": 0.668,
+      "step": 26
+    },
+    {
+      "epoch": 0.02192151556156969,
+      "grad_norm": 1.344380497932434,
+      "learning_rate": 7.297297297297298e-07,
+      "loss": 0.6995,
+      "step": 27
+    },
+    {
+      "epoch": 0.02273342354533153,
+      "grad_norm": 1.335472822189331,
+      "learning_rate": 7.567567567567569e-07,
+      "loss": 0.6711,
+      "step": 28
+    },
+    {
+      "epoch": 0.02354533152909337,
+      "grad_norm": 1.3130022287368774,
+      "learning_rate": 7.837837837837839e-07,
+      "loss": 0.6925,
+      "step": 29
+    },
+    {
+      "epoch": 0.02435723951285521,
+      "grad_norm": 1.3389763832092285,
+      "learning_rate": 8.108108108108109e-07,
+      "loss": 0.686,
+      "step": 30
+    },
+    {
+      "epoch": 0.02516914749661705,
+      "grad_norm": 1.3016369342803955,
+      "learning_rate": 8.37837837837838e-07,
+      "loss": 0.6846,
+      "step": 31
+    },
+    {
+      "epoch": 0.02598105548037889,
+      "grad_norm": 1.2279702425003052,
+      "learning_rate": 8.64864864864865e-07,
+      "loss": 0.6525,
+      "step": 32
+    },
+    {
+      "epoch": 0.026792963464140732,
+      "grad_norm": 1.2366715669631958,
+      "learning_rate": 8.91891891891892e-07,
+      "loss": 0.6973,
+      "step": 33
+    },
+    {
+      "epoch": 0.027604871447902572,
+      "grad_norm": 1.1468015909194946,
+      "learning_rate": 9.189189189189191e-07,
+      "loss": 0.6703,
+      "step": 34
+    },
+    {
+      "epoch": 0.028416779431664412,
+      "grad_norm": 0.9961360096931458,
+      "learning_rate": 9.459459459459461e-07,
+      "loss": 0.663,
+      "step": 35
+    },
+    {
+      "epoch": 0.029228687415426252,
+      "grad_norm": 1.0331542491912842,
+      "learning_rate": 9.72972972972973e-07,
+      "loss": 0.6425,
+      "step": 36
+    },
+    {
+      "epoch": 0.030040595399188092,
+      "grad_norm": 1.0154340267181396,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.6436,
+      "step": 37
+    },
+    {
+      "epoch": 0.030852503382949932,
+      "grad_norm": 1.0031628608703613,
+      "learning_rate": 1.027027027027027e-06,
+      "loss": 0.6707,
+      "step": 38
+    },
+    {
+      "epoch": 0.031664411366711775,
+      "grad_norm": 0.9457878470420837,
+      "learning_rate": 1.0540540540540542e-06,
+      "loss": 0.6615,
+      "step": 39
+    },
+    {
+      "epoch": 0.03247631935047361,
+      "grad_norm": 0.9153221249580383,
+      "learning_rate": 1.0810810810810812e-06,
+      "loss": 0.6566,
+      "step": 40
+    },
+    {
+      "epoch": 0.033288227334235455,
+      "grad_norm": 0.8965979814529419,
+      "learning_rate": 1.1081081081081083e-06,
+      "loss": 0.6463,
+      "step": 41
+    },
+    {
+      "epoch": 0.03410013531799729,
+      "grad_norm": 0.82919842004776,
+      "learning_rate": 1.1351351351351352e-06,
+      "loss": 0.6205,
+      "step": 42
+    },
+    {
+      "epoch": 0.034912043301759135,
+      "grad_norm": 0.7856370806694031,
+      "learning_rate": 1.1621621621621624e-06,
+      "loss": 0.6049,
+      "step": 43
+    },
+    {
+      "epoch": 0.03572395128552097,
+      "grad_norm": 0.745934784412384,
+      "learning_rate": 1.1891891891891893e-06,
+      "loss": 0.6004,
+      "step": 44
+    },
+    {
+      "epoch": 0.036535859269282815,
+      "grad_norm": 0.7870489954948425,
+      "learning_rate": 1.2162162162162164e-06,
+      "loss": 0.6525,
+      "step": 45
+    },
+    {
+      "epoch": 0.03734776725304465,
+      "grad_norm": 0.6815828084945679,
+      "learning_rate": 1.2432432432432434e-06,
+      "loss": 0.6056,
+      "step": 46
+    },
+    {
+      "epoch": 0.038159675236806495,
+      "grad_norm": 0.6834678649902344,
+      "learning_rate": 1.2702702702702705e-06,
+      "loss": 0.6222,
+      "step": 47
+    },
+    {
+      "epoch": 0.03897158322056834,
+      "grad_norm": 0.6657995581626892,
+      "learning_rate": 1.2972972972972974e-06,
+      "loss": 0.6066,
+      "step": 48
+    },
+    {
+      "epoch": 0.039783491204330175,
+      "grad_norm": 0.605292558670044,
+      "learning_rate": 1.3243243243243246e-06,
+      "loss": 0.5961,
+      "step": 49
+    },
+    {
+      "epoch": 0.04059539918809202,
+      "grad_norm": 0.6393334865570068,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 0.5683,
+      "step": 50
+    },
+    {
+      "epoch": 0.041407307171853855,
+      "grad_norm": 0.53614741563797,
+      "learning_rate": 1.3783783783783786e-06,
+      "loss": 0.5808,
+      "step": 51
+    },
+    {
+      "epoch": 0.0422192151556157,
+      "grad_norm": 0.5120183825492859,
+      "learning_rate": 1.4054054054054056e-06,
+      "loss": 0.5729,
+      "step": 52
+    },
+    {
+      "epoch": 0.043031123139377535,
+      "grad_norm": 0.5057471394538879,
+      "learning_rate": 1.4324324324324327e-06,
+      "loss": 0.5728,
+      "step": 53
+    },
+    {
+      "epoch": 0.04384303112313938,
+      "grad_norm": 0.527693510055542,
+      "learning_rate": 1.4594594594594596e-06,
+      "loss": 0.5596,
+      "step": 54
+    },
+    {
+      "epoch": 0.044654939106901215,
+      "grad_norm": 0.5691236853599548,
+      "learning_rate": 1.4864864864864868e-06,
+      "loss": 0.5943,
+      "step": 55
+    },
+    {
+      "epoch": 0.04546684709066306,
+      "grad_norm": 0.4679834842681885,
+      "learning_rate": 1.5135135135135137e-06,
+      "loss": 0.5699,
+      "step": 56
+    },
+    {
+      "epoch": 0.0462787550744249,
+      "grad_norm": 0.49469879269599915,
+      "learning_rate": 1.5405405405405409e-06,
+      "loss": 0.5702,
+      "step": 57
+    },
+    {
+      "epoch": 0.04709066305818674,
+      "grad_norm": 0.5563521981239319,
+      "learning_rate": 1.5675675675675678e-06,
+      "loss": 0.5903,
+      "step": 58
+    },
+    {
+      "epoch": 0.04790257104194858,
+      "grad_norm": 0.5197617411613464,
+      "learning_rate": 1.5945945945945947e-06,
+      "loss": 0.5823,
+      "step": 59
+    },
+    {
+      "epoch": 0.04871447902571042,
+      "grad_norm": 0.42747747898101807,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 0.5656,
+      "step": 60
+    },
+    {
+      "epoch": 0.04952638700947226,
+      "grad_norm": 0.4303089380264282,
+      "learning_rate": 1.6486486486486488e-06,
+      "loss": 0.5714,
+      "step": 61
+    },
+    {
+      "epoch": 0.0503382949932341,
+      "grad_norm": 0.43793538212776184,
+      "learning_rate": 1.675675675675676e-06,
+      "loss": 0.5538,
+      "step": 62
+    },
+    {
+      "epoch": 0.05115020297699594,
+      "grad_norm": 0.3736399710178375,
+      "learning_rate": 1.7027027027027028e-06,
+      "loss": 0.5425,
+      "step": 63
+    },
+    {
+      "epoch": 0.05196211096075778,
+      "grad_norm": 0.35722148418426514,
+      "learning_rate": 1.72972972972973e-06,
+      "loss": 0.5391,
+      "step": 64
+    },
+    {
+      "epoch": 0.05277401894451962,
+      "grad_norm": 0.36116600036621094,
+      "learning_rate": 1.756756756756757e-06,
+      "loss": 0.5645,
+      "step": 65
+    },
+    {
+      "epoch": 0.053585926928281465,
+      "grad_norm": 0.37125980854034424,
+      "learning_rate": 1.783783783783784e-06,
+      "loss": 0.5438,
+      "step": 66
+    },
+    {
+      "epoch": 0.0543978349120433,
+      "grad_norm": 0.36771395802497864,
+      "learning_rate": 1.810810810810811e-06,
+      "loss": 0.5683,
+      "step": 67
+    },
+    {
+      "epoch": 0.055209742895805144,
+      "grad_norm": 0.35073021054267883,
+      "learning_rate": 1.8378378378378381e-06,
+      "loss": 0.5348,
+      "step": 68
+    },
+    {
+      "epoch": 0.05602165087956698,
+      "grad_norm": 0.39187294244766235,
+      "learning_rate": 1.864864864864865e-06,
+      "loss": 0.5354,
+      "step": 69
+    },
+    {
+      "epoch": 0.056833558863328824,
+      "grad_norm": 0.34322622418403625,
+      "learning_rate": 1.8918918918918922e-06,
+      "loss": 0.5348,
+      "step": 70
+    },
+    {
+      "epoch": 0.05764546684709066,
+      "grad_norm": 0.36688125133514404,
+      "learning_rate": 1.918918918918919e-06,
+      "loss": 0.5422,
+      "step": 71
+    },
+    {
+      "epoch": 0.058457374830852504,
+      "grad_norm": 0.3508293330669403,
+      "learning_rate": 1.945945945945946e-06,
+      "loss": 0.5523,
+      "step": 72
+    },
+    {
+      "epoch": 0.05926928281461434,
+      "grad_norm": 0.3525550961494446,
+      "learning_rate": 1.9729729729729734e-06,
+      "loss": 0.5399,
+      "step": 73
+    },
+    {
+      "epoch": 0.060081190798376184,
+      "grad_norm": 0.3013868033885956,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.5,
+      "step": 74
+    },
+    {
+      "epoch": 0.06089309878213803,
+      "grad_norm": 0.3025849163532257,
+      "learning_rate": 2.0270270270270273e-06,
+      "loss": 0.4978,
+      "step": 75
+    },
+    {
+      "epoch": 0.061705006765899864,
+      "grad_norm": 0.2955538034439087,
+      "learning_rate": 2.054054054054054e-06,
+      "loss": 0.5129,
+      "step": 76
+    },
+    {
+      "epoch": 0.0625169147496617,
+      "grad_norm": 0.29706647992134094,
+      "learning_rate": 2.0810810810810815e-06,
+      "loss": 0.5017,
+      "step": 77
+    },
+    {
+      "epoch": 0.06332882273342355,
+      "grad_norm": 0.2925974726676941,
+      "learning_rate": 2.1081081081081085e-06,
+      "loss": 0.5295,
+      "step": 78
+    },
+    {
+      "epoch": 0.06414073071718539,
+      "grad_norm": 0.3114009499549866,
+      "learning_rate": 2.1351351351351354e-06,
+      "loss": 0.5201,
+      "step": 79
+    },
+    {
+      "epoch": 0.06495263870094722,
+      "grad_norm": 0.3007862865924835,
+      "learning_rate": 2.1621621621621623e-06,
+      "loss": 0.5167,
+      "step": 80
+    },
+    {
+      "epoch": 0.06576454668470906,
+      "grad_norm": 0.2881142199039459,
+      "learning_rate": 2.1891891891891897e-06,
+      "loss": 0.5121,
+      "step": 81
+    },
+    {
+      "epoch": 0.06657645466847091,
+      "grad_norm": 0.2987614572048187,
+      "learning_rate": 2.2162162162162166e-06,
+      "loss": 0.4822,
+      "step": 82
+    },
+    {
+      "epoch": 0.06738836265223275,
+      "grad_norm": 0.29465195536613464,
+      "learning_rate": 2.2432432432432435e-06,
+      "loss": 0.5285,
+      "step": 83
+    },
+    {
+      "epoch": 0.06820027063599458,
+      "grad_norm": 0.3099493980407715,
+      "learning_rate": 2.2702702702702705e-06,
+      "loss": 0.5266,
+      "step": 84
+    },
+    {
+      "epoch": 0.06901217861975643,
+      "grad_norm": 0.28734275698661804,
+      "learning_rate": 2.297297297297298e-06,
+      "loss": 0.5116,
+      "step": 85
+    },
+    {
+      "epoch": 0.06982408660351827,
+      "grad_norm": 0.2799994945526123,
+      "learning_rate": 2.3243243243243247e-06,
+      "loss": 0.5164,
+      "step": 86
+    },
+    {
+      "epoch": 0.07063599458728011,
+      "grad_norm": 0.33441606163978577,
+      "learning_rate": 2.3513513513513517e-06,
+      "loss": 0.5055,
+      "step": 87
+    },
+    {
+      "epoch": 0.07144790257104194,
+      "grad_norm": 0.29269319772720337,
+      "learning_rate": 2.3783783783783786e-06,
+      "loss": 0.5167,
+      "step": 88
+    },
+    {
+      "epoch": 0.0722598105548038,
+      "grad_norm": 0.2932380735874176,
+      "learning_rate": 2.4054054054054055e-06,
+      "loss": 0.5076,
+      "step": 89
+    },
+    {
+      "epoch": 0.07307171853856563,
+      "grad_norm": 0.26718634366989136,
+      "learning_rate": 2.432432432432433e-06,
+      "loss": 0.5195,
+      "step": 90
+    },
+    {
+      "epoch": 0.07388362652232747,
+      "grad_norm": 0.27515503764152527,
+      "learning_rate": 2.45945945945946e-06,
+      "loss": 0.4772,
+      "step": 91
+    },
+    {
+      "epoch": 0.0746955345060893,
+      "grad_norm": 0.2766149938106537,
+      "learning_rate": 2.4864864864864867e-06,
+      "loss": 0.4975,
+      "step": 92
+    },
+    {
+      "epoch": 0.07550744248985115,
+      "grad_norm": 0.3020631670951843,
+      "learning_rate": 2.5135135135135137e-06,
+      "loss": 0.5061,
+      "step": 93
+    },
+    {
+      "epoch": 0.07631935047361299,
+      "grad_norm": 0.547136664390564,
+      "learning_rate": 2.540540540540541e-06,
+      "loss": 0.501,
+      "step": 94
+    },
+    {
+      "epoch": 0.07713125845737483,
+      "grad_norm": 0.27995291352272034,
+      "learning_rate": 2.5675675675675675e-06,
+      "loss": 0.4783,
+      "step": 95
+    },
+    {
+      "epoch": 0.07794316644113668,
+      "grad_norm": 0.24574805796146393,
+      "learning_rate": 2.594594594594595e-06,
+      "loss": 0.4759,
+      "step": 96
+    },
+    {
+      "epoch": 0.07875507442489851,
+      "grad_norm": 0.27462834119796753,
+      "learning_rate": 2.621621621621622e-06,
+      "loss": 0.5056,
+      "step": 97
+    },
+    {
+      "epoch": 0.07956698240866035,
+      "grad_norm": 0.2782885730266571,
+      "learning_rate": 2.648648648648649e-06,
+      "loss": 0.4941,
+      "step": 98
+    },
+    {
+      "epoch": 0.08037889039242219,
+      "grad_norm": 0.880927562713623,
+      "learning_rate": 2.6756756756756757e-06,
+      "loss": 0.4863,
+      "step": 99
+    },
+    {
+      "epoch": 0.08119079837618404,
+      "grad_norm": 0.2892383635044098,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 0.4763,
+      "step": 100
+    },
+    {
+      "epoch": 0.08200270635994587,
+      "grad_norm": 0.32562583684921265,
+      "learning_rate": 2.72972972972973e-06,
+      "loss": 0.4926,
+      "step": 101
+    },
+    {
+      "epoch": 0.08281461434370771,
+      "grad_norm": 0.2722681760787964,
+      "learning_rate": 2.7567567567567573e-06,
+      "loss": 0.5049,
+      "step": 102
+    },
+    {
+      "epoch": 0.08362652232746956,
+      "grad_norm": 0.2793631851673126,
+      "learning_rate": 2.783783783783784e-06,
+      "loss": 0.49,
+      "step": 103
+    },
+    {
+      "epoch": 0.0844384303112314,
+      "grad_norm": 0.25650158524513245,
+      "learning_rate": 2.810810810810811e-06,
+      "loss": 0.4747,
+      "step": 104
+    },
+    {
+      "epoch": 0.08525033829499323,
+      "grad_norm": 0.25994375348091125,
+      "learning_rate": 2.837837837837838e-06,
+      "loss": 0.4921,
+      "step": 105
+    },
+    {
+      "epoch": 0.08606224627875507,
+      "grad_norm": 0.26915740966796875,
+      "learning_rate": 2.8648648648648654e-06,
+      "loss": 0.4937,
+      "step": 106
+    },
+    {
+      "epoch": 0.08687415426251692,
+      "grad_norm": 0.24270494282245636,
+      "learning_rate": 2.891891891891892e-06,
+      "loss": 0.4618,
+      "step": 107
+    },
+    {
+      "epoch": 0.08768606224627876,
+      "grad_norm": 0.2629684805870056,
+      "learning_rate": 2.9189189189189193e-06,
+      "loss": 0.5003,
+      "step": 108
+    },
+    {
+      "epoch": 0.08849797023004059,
+      "grad_norm": 0.26809972524642944,
+      "learning_rate": 2.9459459459459462e-06,
+      "loss": 0.4787,
+      "step": 109
+    },
+    {
+      "epoch": 0.08930987821380243,
+      "grad_norm": 0.2704669237136841,
+      "learning_rate": 2.9729729729729736e-06,
+      "loss": 0.4736,
+      "step": 110
+    },
+    {
+      "epoch": 0.09012178619756428,
+      "grad_norm": 0.2432822734117508,
+      "learning_rate": 3e-06,
+      "loss": 0.4877,
+      "step": 111
+    },
+    {
+      "epoch": 0.09093369418132612,
+      "grad_norm": 0.28277257084846497,
+      "learning_rate": 3.0270270270270274e-06,
+      "loss": 0.4788,
+      "step": 112
+    },
+    {
+      "epoch": 0.09174560216508795,
+      "grad_norm": 0.251797616481781,
+      "learning_rate": 3.0540540540540544e-06,
+      "loss": 0.4953,
+      "step": 113
+    },
+    {
+      "epoch": 0.0925575101488498,
+      "grad_norm": 0.2718489468097687,
+      "learning_rate": 3.0810810810810817e-06,
+      "loss": 0.4844,
+      "step": 114
+    },
+    {
+      "epoch": 0.09336941813261164,
+      "grad_norm": 0.2669789791107178,
+      "learning_rate": 3.1081081081081082e-06,
+      "loss": 0.4572,
+      "step": 115
+    },
+    {
+      "epoch": 0.09418132611637348,
+      "grad_norm": 0.27514809370040894,
+      "learning_rate": 3.1351351351351356e-06,
+      "loss": 0.4861,
+      "step": 116
+    },
+    {
+      "epoch": 0.09499323410013531,
+      "grad_norm": 0.2771575152873993,
+      "learning_rate": 3.1621621621621625e-06,
+      "loss": 0.4706,
+      "step": 117
+    },
+    {
+      "epoch": 0.09580514208389716,
+      "grad_norm": 0.2914608120918274,
+      "learning_rate": 3.1891891891891894e-06,
+      "loss": 0.4718,
+      "step": 118
+    },
+    {
+      "epoch": 0.096617050067659,
+      "grad_norm": 0.2665104568004608,
+      "learning_rate": 3.2162162162162164e-06,
+      "loss": 0.455,
+      "step": 119
+    },
+    {
+      "epoch": 0.09742895805142084,
+      "grad_norm": 0.25711777806282043,
+      "learning_rate": 3.2432432432432437e-06,
+      "loss": 0.4688,
+      "step": 120
+    },
+    {
+      "epoch": 0.09824086603518269,
+      "grad_norm": 0.25264760851860046,
+      "learning_rate": 3.2702702702702706e-06,
+      "loss": 0.4838,
+      "step": 121
+    },
+    {
+      "epoch": 0.09905277401894452,
+      "grad_norm": 0.2712831199169159,
+      "learning_rate": 3.2972972972972976e-06,
+      "loss": 0.4708,
+      "step": 122
+    },
+    {
+      "epoch": 0.09986468200270636,
+      "grad_norm": 0.2638511061668396,
+      "learning_rate": 3.3243243243243245e-06,
+      "loss": 0.4687,
+      "step": 123
+    },
+    {
+      "epoch": 0.1006765899864682,
+      "grad_norm": 0.2573316991329193,
+      "learning_rate": 3.351351351351352e-06,
+      "loss": 0.4621,
+      "step": 124
+    },
+    {
+      "epoch": 0.10148849797023005,
+      "grad_norm": 0.27992698550224304,
+      "learning_rate": 3.3783783783783788e-06,
+      "loss": 0.4854,
+      "step": 125
+    },
+    {
+      "epoch": 0.10230040595399188,
+      "grad_norm": 0.2589249014854431,
+      "learning_rate": 3.4054054054054057e-06,
+      "loss": 0.4778,
+      "step": 126
+    },
+    {
+      "epoch": 0.10311231393775372,
+      "grad_norm": 0.26866415143013,
+      "learning_rate": 3.4324324324324326e-06,
+      "loss": 0.4824,
+      "step": 127
+    },
+    {
+      "epoch": 0.10392422192151556,
+      "grad_norm": 0.3012371063232422,
+      "learning_rate": 3.45945945945946e-06,
+      "loss": 0.5003,
+      "step": 128
+    },
+    {
+      "epoch": 0.1047361299052774,
+      "grad_norm": 0.2884964048862457,
+      "learning_rate": 3.4864864864864865e-06,
+      "loss": 0.4729,
+      "step": 129
+    },
+    {
+      "epoch": 0.10554803788903924,
+      "grad_norm": 0.2777898609638214,
+      "learning_rate": 3.513513513513514e-06,
+      "loss": 0.4672,
+      "step": 130
+    },
+    {
+      "epoch": 0.10635994587280108,
+      "grad_norm": 0.2576296627521515,
+      "learning_rate": 3.5405405405405408e-06,
+      "loss": 0.4859,
+      "step": 131
+    },
+    {
+      "epoch": 0.10717185385656293,
+      "grad_norm": 0.26763930916786194,
+      "learning_rate": 3.567567567567568e-06,
+      "loss": 0.4694,
+      "step": 132
+    },
+    {
+      "epoch": 0.10798376184032477,
+      "grad_norm": 0.2965649366378784,
+      "learning_rate": 3.5945945945945946e-06,
+      "loss": 0.4612,
+      "step": 133
+    },
+    {
+      "epoch": 0.1087956698240866,
+      "grad_norm": 0.2903744578361511,
+      "learning_rate": 3.621621621621622e-06,
+      "loss": 0.4518,
+      "step": 134
+    },
+    {
+      "epoch": 0.10960757780784844,
+      "grad_norm": 0.2623900771141052,
+      "learning_rate": 3.648648648648649e-06,
+      "loss": 0.454,
+      "step": 135
+    },
+    {
+      "epoch": 0.11041948579161029,
+      "grad_norm": 0.28886106610298157,
+      "learning_rate": 3.6756756756756763e-06,
+      "loss": 0.4341,
+      "step": 136
+    },
+    {
+      "epoch": 0.11123139377537213,
+      "grad_norm": 0.284225195646286,
+      "learning_rate": 3.7027027027027028e-06,
+      "loss": 0.471,
+      "step": 137
+    },
+    {
+      "epoch": 0.11204330175913396,
+      "grad_norm": 0.2710971534252167,
+      "learning_rate": 3.72972972972973e-06,
+      "loss": 0.4948,
+      "step": 138
+    },
+    {
+      "epoch": 0.1128552097428958,
+      "grad_norm": 0.2795678675174713,
+      "learning_rate": 3.756756756756757e-06,
+      "loss": 0.4944,
+      "step": 139
+    },
+    {
+      "epoch": 0.11366711772665765,
+      "grad_norm": 0.25956296920776367,
+      "learning_rate": 3.7837837837837844e-06,
+      "loss": 0.442,
+      "step": 140
+    },
+    {
+      "epoch": 0.11447902571041949,
+      "grad_norm": 0.2820890247821808,
+      "learning_rate": 3.810810810810811e-06,
+      "loss": 0.4767,
+      "step": 141
+    },
+    {
+      "epoch": 0.11529093369418132,
+      "grad_norm": 0.3063553273677826,
+      "learning_rate": 3.837837837837838e-06,
+      "loss": 0.4704,
+      "step": 142
+    },
+    {
+      "epoch": 0.11610284167794317,
+      "grad_norm": 0.3929675221443176,
+      "learning_rate": 3.864864864864865e-06,
+      "loss": 0.4519,
+      "step": 143
+    },
+    {
+      "epoch": 0.11691474966170501,
+      "grad_norm": 0.2783176302909851,
+      "learning_rate": 3.891891891891892e-06,
+      "loss": 0.4543,
+      "step": 144
+    },
+    {
+      "epoch": 0.11772665764546685,
+      "grad_norm": 0.29280924797058105,
+      "learning_rate": 3.918918918918919e-06,
+      "loss": 0.4721,
+      "step": 145
+    },
+    {
+      "epoch": 0.11853856562922868,
+      "grad_norm": 0.2664378881454468,
+      "learning_rate": 3.945945945945947e-06,
+      "loss": 0.463,
+      "step": 146
+    },
+    {
+      "epoch": 0.11935047361299053,
+      "grad_norm": 0.25492623448371887,
+      "learning_rate": 3.972972972972973e-06,
+      "loss": 0.4736,
+      "step": 147
+    },
+    {
+      "epoch": 0.12016238159675237,
+      "grad_norm": 0.28595370054244995,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.4638,
+      "step": 148
+    },
+    {
+      "epoch": 0.1209742895805142,
+      "grad_norm": 0.2609878480434418,
+      "learning_rate": 4.027027027027028e-06,
+      "loss": 0.4736,
+      "step": 149
+    },
+    {
+      "epoch": 0.12178619756427606,
+      "grad_norm": 0.2587873339653015,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 0.4597,
+      "step": 150
+    },
+    {
+      "epoch": 0.12259810554803789,
+      "grad_norm": 0.26544898748397827,
+      "learning_rate": 4.0810810810810815e-06,
+      "loss": 0.4566,
+      "step": 151
+    },
+    {
+      "epoch": 0.12341001353179973,
+      "grad_norm": 0.25257131457328796,
+      "learning_rate": 4.108108108108108e-06,
+      "loss": 0.459,
+      "step": 152
+    },
+    {
+      "epoch": 0.12422192151556156,
+      "grad_norm": 0.26189160346984863,
+      "learning_rate": 4.135135135135135e-06,
+      "loss": 0.4625,
+      "step": 153
+    },
+    {
+      "epoch": 0.1250338294993234,
+      "grad_norm": 0.27410003542900085,
+      "learning_rate": 4.162162162162163e-06,
+      "loss": 0.475,
+      "step": 154
+    },
+    {
+      "epoch": 0.12584573748308525,
+      "grad_norm": 0.29529687762260437,
+      "learning_rate": 4.189189189189189e-06,
+      "loss": 0.4465,
+      "step": 155
+    },
+    {
+      "epoch": 0.1266576454668471,
+      "grad_norm": 0.279422789812088,
+      "learning_rate": 4.216216216216217e-06,
+      "loss": 0.5021,
+      "step": 156
+    },
+    {
+      "epoch": 0.12746955345060892,
+      "grad_norm": 0.2716977596282959,
+      "learning_rate": 4.243243243243244e-06,
+      "loss": 0.4435,
+      "step": 157
+    },
+    {
+      "epoch": 0.12828146143437077,
+      "grad_norm": 0.28507885336875916,
+      "learning_rate": 4.270270270270271e-06,
+      "loss": 0.4582,
+      "step": 158
+    },
+    {
+      "epoch": 0.12909336941813263,
+      "grad_norm": 0.27989527583122253,
+      "learning_rate": 4.297297297297298e-06,
+      "loss": 0.4592,
+      "step": 159
+    },
+    {
+      "epoch": 0.12990527740189445,
+      "grad_norm": 0.28802230954170227,
+      "learning_rate": 4.324324324324325e-06,
+      "loss": 0.4814,
+      "step": 160
+    },
+    {
+      "epoch": 0.1307171853856563,
+      "grad_norm": 0.2725646495819092,
+      "learning_rate": 4.351351351351352e-06,
+      "loss": 0.4714,
+      "step": 161
+    },
+    {
+      "epoch": 0.13152909336941812,
+      "grad_norm": 0.30618712306022644,
+      "learning_rate": 4.378378378378379e-06,
+      "loss": 0.4784,
+      "step": 162
+    },
+    {
+      "epoch": 0.13234100135317997,
+      "grad_norm": 0.30851316452026367,
+      "learning_rate": 4.4054054054054054e-06,
+      "loss": 0.4773,
+      "step": 163
+    },
+    {
+      "epoch": 0.13315290933694182,
+      "grad_norm": 0.26055771112442017,
+      "learning_rate": 4.432432432432433e-06,
+      "loss": 0.4325,
+      "step": 164
+    },
+    {
+      "epoch": 0.13396481732070364,
+      "grad_norm": 0.2664947509765625,
+      "learning_rate": 4.45945945945946e-06,
+      "loss": 0.4238,
+      "step": 165
+    },
+    {
+      "epoch": 0.1347767253044655,
+      "grad_norm": 0.2826455235481262,
+      "learning_rate": 4.486486486486487e-06,
+      "loss": 0.458,
+      "step": 166
+    },
+    {
+      "epoch": 0.13558863328822734,
+      "grad_norm": 0.2854776680469513,
+      "learning_rate": 4.513513513513514e-06,
+      "loss": 0.4488,
+      "step": 167
+    },
+    {
+      "epoch": 0.13640054127198917,
+      "grad_norm": 0.29865017533302307,
+      "learning_rate": 4.540540540540541e-06,
+      "loss": 0.4408,
+      "step": 168
+    },
+    {
+      "epoch": 0.13721244925575102,
+      "grad_norm": 0.269780695438385,
+      "learning_rate": 4.567567567567568e-06,
+      "loss": 0.4653,
+      "step": 169
+    },
+    {
+      "epoch": 0.13802435723951287,
+      "grad_norm": 0.27224722504615784,
+      "learning_rate": 4.594594594594596e-06,
+      "loss": 0.4524,
+      "step": 170
+    },
+    {
+      "epoch": 0.1388362652232747,
+      "grad_norm": 0.2690134346485138,
+      "learning_rate": 4.621621621621622e-06,
+      "loss": 0.4479,
+      "step": 171
+    },
+    {
+      "epoch": 0.13964817320703654,
+      "grad_norm": 0.29436424374580383,
+      "learning_rate": 4.6486486486486495e-06,
+      "loss": 0.4639,
+      "step": 172
+    },
+    {
+      "epoch": 0.14046008119079836,
+      "grad_norm": 0.2664104998111725,
+      "learning_rate": 4.675675675675676e-06,
+      "loss": 0.4376,
+      "step": 173
+    },
+    {
+      "epoch": 0.14127198917456021,
+      "grad_norm": 0.2912067770957947,
+      "learning_rate": 4.702702702702703e-06,
+      "loss": 0.4413,
+      "step": 174
+    },
+    {
+      "epoch": 0.14208389715832206,
+      "grad_norm": 0.2660939395427704,
+      "learning_rate": 4.72972972972973e-06,
+      "loss": 0.4576,
+      "step": 175
+    },
+    {
+      "epoch": 0.1428958051420839,
+      "grad_norm": 0.25915104150772095,
+      "learning_rate": 4.756756756756757e-06,
+      "loss": 0.4663,
+      "step": 176
+    },
+    {
+      "epoch": 0.14370771312584574,
+      "grad_norm": 0.28762301802635193,
+      "learning_rate": 4.783783783783784e-06,
+      "loss": 0.406,
+      "step": 177
+    },
+    {
+      "epoch": 0.1445196211096076,
+      "grad_norm": 0.2768748700618744,
+      "learning_rate": 4.810810810810811e-06,
+      "loss": 0.4615,
+      "step": 178
+    },
+    {
+      "epoch": 0.1453315290933694,
+      "grad_norm": 0.2841472923755646,
+      "learning_rate": 4.837837837837838e-06,
+      "loss": 0.4362,
+      "step": 179
+    },
+    {
+      "epoch": 0.14614343707713126,
+      "grad_norm": 0.2866773307323456,
+      "learning_rate": 4.864864864864866e-06,
+      "loss": 0.4432,
+      "step": 180
+    },
+    {
+      "epoch": 0.1469553450608931,
+      "grad_norm": 0.25606611371040344,
+      "learning_rate": 4.891891891891893e-06,
+      "loss": 0.4381,
+      "step": 181
+    },
+    {
+      "epoch": 0.14776725304465493,
+      "grad_norm": 0.24891367554664612,
+      "learning_rate": 4.91891891891892e-06,
+      "loss": 0.4298,
+      "step": 182
+    },
+    {
+      "epoch": 0.14857916102841678,
+      "grad_norm": 0.2688259184360504,
+      "learning_rate": 4.9459459459459466e-06,
+      "loss": 0.4306,
+      "step": 183
+    },
+    {
+      "epoch": 0.1493910690121786,
+      "grad_norm": 0.26982131600379944,
+      "learning_rate": 4.9729729729729735e-06,
+      "loss": 0.4405,
+      "step": 184
+    },
+    {
+      "epoch": 0.15020297699594046,
+      "grad_norm": 0.9926776885986328,
+      "learning_rate": 5e-06,
+      "loss": 0.4489,
+      "step": 185
+    },
+    {
+      "epoch": 0.1510148849797023,
+      "grad_norm": 0.2671455144882202,
+      "learning_rate": 5.027027027027027e-06,
+      "loss": 0.4548,
+      "step": 186
+    },
+    {
+      "epoch": 0.15182679296346413,
+      "grad_norm": 0.27227431535720825,
+      "learning_rate": 5.054054054054054e-06,
+      "loss": 0.4234,
+      "step": 187
+    },
+    {
+      "epoch": 0.15263870094722598,
+      "grad_norm": 0.2794792056083679,
+      "learning_rate": 5.081081081081082e-06,
+      "loss": 0.4636,
+      "step": 188
+    },
+    {
+      "epoch": 0.15345060893098783,
+      "grad_norm": 0.3312227129936218,
+      "learning_rate": 5.108108108108108e-06,
+      "loss": 0.4661,
+      "step": 189
+    },
+    {
+      "epoch": 0.15426251691474965,
+      "grad_norm": 0.26594364643096924,
+      "learning_rate": 5.135135135135135e-06,
+      "loss": 0.4441,
+      "step": 190
+    },
+    {
+      "epoch": 0.1550744248985115,
+      "grad_norm": 0.2596721649169922,
+      "learning_rate": 5.162162162162162e-06,
+      "loss": 0.4567,
+      "step": 191
+    },
+    {
+      "epoch": 0.15588633288227335,
+      "grad_norm": 0.26411551237106323,
+      "learning_rate": 5.18918918918919e-06,
+      "loss": 0.4465,
+      "step": 192
+    },
+    {
+      "epoch": 0.15669824086603518,
+      "grad_norm": 0.26920297741889954,
+      "learning_rate": 5.216216216216217e-06,
+      "loss": 0.4383,
+      "step": 193
+    },
+    {
+      "epoch": 0.15751014884979703,
+      "grad_norm": 0.2484126091003418,
+      "learning_rate": 5.243243243243244e-06,
+      "loss": 0.4419,
+      "step": 194
+    },
+    {
+      "epoch": 0.15832205683355885,
+      "grad_norm": 0.282625675201416,
+      "learning_rate": 5.2702702702702705e-06,
+      "loss": 0.4419,
+      "step": 195
+    },
+    {
+      "epoch": 0.1591339648173207,
+      "grad_norm": 0.27681341767311096,
+      "learning_rate": 5.297297297297298e-06,
+      "loss": 0.4013,
+      "step": 196
+    },
+    {
+      "epoch": 0.15994587280108255,
+      "grad_norm": 0.25536930561065674,
+      "learning_rate": 5.324324324324324e-06,
+      "loss": 0.4414,
+      "step": 197
+    },
+    {
+      "epoch": 0.16075778078484437,
+      "grad_norm": 0.2457224577665329,
+      "learning_rate": 5.351351351351351e-06,
+      "loss": 0.4331,
+      "step": 198
+    },
+    {
+      "epoch": 0.16156968876860622,
+      "grad_norm": 0.2755725681781769,
+      "learning_rate": 5.378378378378378e-06,
+      "loss": 0.4417,
+      "step": 199
+    },
+    {
+      "epoch": 0.16238159675236807,
+      "grad_norm": 0.3056943714618683,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 0.431,
+      "step": 200
+    },
+    {
+      "epoch": 0.1631935047361299,
+      "grad_norm": 0.24364936351776123,
+      "learning_rate": 5.432432432432433e-06,
+      "loss": 0.4287,
+      "step": 201
+    },
+    {
+      "epoch": 0.16400541271989175,
+      "grad_norm": 0.35343047976493835,
+      "learning_rate": 5.45945945945946e-06,
+      "loss": 0.4527,
+      "step": 202
+    },
+    {
+      "epoch": 0.1648173207036536,
+      "grad_norm": 0.28926339745521545,
+      "learning_rate": 5.486486486486487e-06,
+      "loss": 0.4431,
+      "step": 203
+    },
+    {
+      "epoch": 0.16562922868741542,
+      "grad_norm": 0.3241758644580841,
+      "learning_rate": 5.513513513513515e-06,
+      "loss": 0.4295,
+      "step": 204
+    },
+    {
+      "epoch": 0.16644113667117727,
+      "grad_norm": 0.25906625390052795,
+      "learning_rate": 5.540540540540541e-06,
+      "loss": 0.4255,
+      "step": 205
+    },
+    {
+      "epoch": 0.16725304465493912,
+      "grad_norm": 0.24108658730983734,
+      "learning_rate": 5.567567567567568e-06,
+      "loss": 0.4253,
+      "step": 206
+    },
+    {
+      "epoch": 0.16806495263870094,
+      "grad_norm": 0.27738070487976074,
+      "learning_rate": 5.5945945945945945e-06,
+      "loss": 0.4416,
+      "step": 207
+    },
+    {
+      "epoch": 0.1688768606224628,
+      "grad_norm": 0.2622399628162384,
+      "learning_rate": 5.621621621621622e-06,
+      "loss": 0.4498,
+      "step": 208
+    },
+    {
+      "epoch": 0.16968876860622462,
+      "grad_norm": 0.26075997948646545,
+      "learning_rate": 5.648648648648649e-06,
+      "loss": 0.419,
+      "step": 209
+    },
+    {
+      "epoch": 0.17050067658998647,
+      "grad_norm": 0.2707868218421936,
+      "learning_rate": 5.675675675675676e-06,
+      "loss": 0.4536,
+      "step": 210
+    },
+    {
+      "epoch": 0.17131258457374832,
+      "grad_norm": 0.2925918698310852,
+      "learning_rate": 5.702702702702702e-06,
+      "loss": 0.4463,
+      "step": 211
+    },
+    {
+      "epoch": 0.17212449255751014,
+      "grad_norm": 0.2600102424621582,
+      "learning_rate": 5.729729729729731e-06,
+      "loss": 0.4194,
+      "step": 212
+    },
+    {
+      "epoch": 0.172936400541272,
+      "grad_norm": 0.2773691713809967,
+      "learning_rate": 5.756756756756757e-06,
+      "loss": 0.4543,
+      "step": 213
+    },
+    {
+      "epoch": 0.17374830852503384,
+      "grad_norm": 0.27895769476890564,
+      "learning_rate": 5.783783783783784e-06,
+      "loss": 0.4404,
+      "step": 214
+    },
+    {
+      "epoch": 0.17456021650879566,
+      "grad_norm": 0.24456042051315308,
+      "learning_rate": 5.810810810810811e-06,
+      "loss": 0.4429,
+      "step": 215
+    },
+    {
+      "epoch": 0.1753721244925575,
+      "grad_norm": 0.2892736494541168,
+      "learning_rate": 5.837837837837839e-06,
+      "loss": 0.4653,
+      "step": 216
+    },
+    {
+      "epoch": 0.17618403247631936,
+      "grad_norm": 0.2391236424446106,
+      "learning_rate": 5.8648648648648655e-06,
+      "loss": 0.4239,
+      "step": 217
+    },
+    {
+      "epoch": 0.17699594046008119,
+      "grad_norm": 0.2870897948741913,
+      "learning_rate": 5.8918918918918924e-06,
+      "loss": 0.44,
+      "step": 218
+    },
+    {
+      "epoch": 0.17780784844384304,
+      "grad_norm": 0.24295304715633392,
+      "learning_rate": 5.9189189189189185e-06,
+      "loss": 0.4335,
+      "step": 219
+    },
+    {
+      "epoch": 0.17861975642760486,
+      "grad_norm": 0.24816599488258362,
+      "learning_rate": 5.945945945945947e-06,
+      "loss": 0.4391,
+      "step": 220
+    },
+    {
+      "epoch": 0.1794316644113667,
+      "grad_norm": 0.2316385805606842,
+      "learning_rate": 5.972972972972973e-06,
+      "loss": 0.4297,
+      "step": 221
+    },
+    {
+      "epoch": 0.18024357239512856,
+      "grad_norm": 0.25425034761428833,
+      "learning_rate": 6e-06,
+      "loss": 0.4527,
+      "step": 222
+    },
+    {
+      "epoch": 0.18105548037889038,
+      "grad_norm": 0.2506188452243805,
+      "learning_rate": 6.027027027027027e-06,
+      "loss": 0.4125,
+      "step": 223
+    },
+    {
+      "epoch": 0.18186738836265223,
+      "grad_norm": 0.22972145676612854,
+      "learning_rate": 6.054054054054055e-06,
+      "loss": 0.435,
+      "step": 224
+    },
+    {
+      "epoch": 0.18267929634641408,
+      "grad_norm": 0.25031954050064087,
+      "learning_rate": 6.081081081081082e-06,
+      "loss": 0.399,
+      "step": 225
+    },
+    {
+      "epoch": 0.1834912043301759,
+      "grad_norm": 0.26745155453681946,
+      "learning_rate": 6.108108108108109e-06,
+      "loss": 0.4337,
+      "step": 226
+    },
+    {
+      "epoch": 0.18430311231393776,
+      "grad_norm": 0.2435240000486374,
+      "learning_rate": 6.135135135135135e-06,
+      "loss": 0.4397,
+      "step": 227
+    },
+    {
+      "epoch": 0.1851150202976996,
+      "grad_norm": 0.25708040595054626,
+      "learning_rate": 6.162162162162163e-06,
+      "loss": 0.414,
+      "step": 228
+    },
+    {
+      "epoch": 0.18592692828146143,
+      "grad_norm": 0.24750782549381256,
+      "learning_rate": 6.1891891891891895e-06,
+      "loss": 0.4223,
+      "step": 229
+    },
+    {
+      "epoch": 0.18673883626522328,
+      "grad_norm": 0.2839811444282532,
+      "learning_rate": 6.2162162162162164e-06,
+      "loss": 0.4395,
+      "step": 230
+    },
+    {
+      "epoch": 0.1875507442489851,
+      "grad_norm": 0.26136255264282227,
+      "learning_rate": 6.243243243243243e-06,
+      "loss": 0.4461,
+      "step": 231
+    },
+    {
+      "epoch": 0.18836265223274695,
+      "grad_norm": 0.272703617811203,
+      "learning_rate": 6.270270270270271e-06,
+      "loss": 0.4528,
+      "step": 232
+    },
+    {
+      "epoch": 0.1891745602165088,
+      "grad_norm": 0.29748833179473877,
+      "learning_rate": 6.297297297297298e-06,
+      "loss": 0.4206,
+      "step": 233
+    },
+    {
+      "epoch": 0.18998646820027063,
+      "grad_norm": 0.2871575653553009,
+      "learning_rate": 6.324324324324325e-06,
+      "loss": 0.425,
+      "step": 234
+    },
+    {
+      "epoch": 0.19079837618403248,
+      "grad_norm": 0.2743903398513794,
+      "learning_rate": 6.351351351351351e-06,
+      "loss": 0.4649,
+      "step": 235
+    },
+    {
+      "epoch": 0.19161028416779433,
+      "grad_norm": 0.251460999250412,
+      "learning_rate": 6.378378378378379e-06,
+      "loss": 0.4535,
+      "step": 236
+    },
+    {
+      "epoch": 0.19242219215155615,
+      "grad_norm": 0.2984464764595032,
+      "learning_rate": 6.405405405405406e-06,
+      "loss": 0.4092,
+      "step": 237
+    },
+    {
+      "epoch": 0.193234100135318,
+      "grad_norm": 0.261911541223526,
+      "learning_rate": 6.432432432432433e-06,
+      "loss": 0.4302,
+      "step": 238
+    },
+    {
+      "epoch": 0.19404600811907985,
+      "grad_norm": 0.2641757130622864,
+      "learning_rate": 6.45945945945946e-06,
+      "loss": 0.4093,
+      "step": 239
+    },
+    {
+      "epoch": 0.19485791610284167,
+      "grad_norm": 0.2685806453227997,
+      "learning_rate": 6.486486486486487e-06,
+      "loss": 0.409,
+      "step": 240
+    },
+    {
+      "epoch": 0.19566982408660352,
+      "grad_norm": 0.2919831871986389,
+      "learning_rate": 6.513513513513514e-06,
+      "loss": 0.4415,
+      "step": 241
+    },
+    {
+      "epoch": 0.19648173207036537,
+      "grad_norm": 0.24788640439510345,
+      "learning_rate": 6.540540540540541e-06,
+      "loss": 0.4428,
+      "step": 242
+    },
+    {
+      "epoch": 0.1972936400541272,
+      "grad_norm": 0.26327651739120483,
+      "learning_rate": 6.567567567567567e-06,
+      "loss": 0.4224,
+      "step": 243
+    },
+    {
+      "epoch": 0.19810554803788905,
+      "grad_norm": 0.2687091529369354,
+      "learning_rate": 6.594594594594595e-06,
+      "loss": 0.4266,
+      "step": 244
+    },
+    {
+      "epoch": 0.19891745602165087,
+      "grad_norm": 0.2728613018989563,
+      "learning_rate": 6.621621621621622e-06,
+      "loss": 0.4161,
+      "step": 245
+    },
+    {
+      "epoch": 0.19972936400541272,
+      "grad_norm": 0.29980242252349854,
+      "learning_rate": 6.648648648648649e-06,
+      "loss": 0.4423,
+      "step": 246
+    },
+    {
+      "epoch": 0.20054127198917457,
+      "grad_norm": 0.27139437198638916,
+      "learning_rate": 6.675675675675676e-06,
+      "loss": 0.4588,
+      "step": 247
+    },
+    {
+      "epoch": 0.2013531799729364,
+      "grad_norm": 0.2737210690975189,
+      "learning_rate": 6.702702702702704e-06,
+      "loss": 0.4291,
+      "step": 248
+    },
+    {
+      "epoch": 0.20216508795669824,
+      "grad_norm": 0.3038538992404938,
+      "learning_rate": 6.729729729729731e-06,
+      "loss": 0.4494,
+      "step": 249
+    },
+    {
+      "epoch": 0.2029769959404601,
+      "grad_norm": 0.2711346745491028,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 0.4347,
+      "step": 250
+    },
+    {
+      "epoch": 0.20378890392422191,
+      "grad_norm": 0.2780447006225586,
+      "learning_rate": 6.783783783783784e-06,
+      "loss": 0.4175,
+      "step": 251
+    },
+    {
+      "epoch": 0.20460081190798377,
+      "grad_norm": 0.28645268082618713,
+      "learning_rate": 6.810810810810811e-06,
+      "loss": 0.4319,
+      "step": 252
+    },
+    {
+      "epoch": 0.20541271989174562,
+      "grad_norm": 0.2975715398788452,
+      "learning_rate": 6.837837837837838e-06,
+      "loss": 0.4143,
+      "step": 253
+    },
+    {
+      "epoch": 0.20622462787550744,
+      "grad_norm": 0.2550729811191559,
+      "learning_rate": 6.864864864864865e-06,
+      "loss": 0.4375,
+      "step": 254
+    },
+    {
+      "epoch": 0.2070365358592693,
+      "grad_norm": 0.29103562235832214,
+      "learning_rate": 6.891891891891892e-06,
+      "loss": 0.4239,
+      "step": 255
+    },
+    {
+      "epoch": 0.2078484438430311,
+      "grad_norm": 0.2507869005203247,
+      "learning_rate": 6.91891891891892e-06,
+      "loss": 0.4216,
+      "step": 256
+    },
+    {
+      "epoch": 0.20866035182679296,
+      "grad_norm": 0.2504599690437317,
+      "learning_rate": 6.945945945945947e-06,
+      "loss": 0.4245,
+      "step": 257
+    },
+    {
+      "epoch": 0.2094722598105548,
+      "grad_norm": 0.3094748258590698,
+      "learning_rate": 6.972972972972973e-06,
+      "loss": 0.4225,
+      "step": 258
+    },
+    {
+      "epoch": 0.21028416779431663,
+      "grad_norm": 0.257976770401001,
+      "learning_rate": 7e-06,
+      "loss": 0.4415,
+      "step": 259
+    },
+    {
+      "epoch": 0.21109607577807848,
+      "grad_norm": 0.25533437728881836,
+      "learning_rate": 7.027027027027028e-06,
+      "loss": 0.4166,
+      "step": 260
+    },
+    {
+      "epoch": 0.21190798376184034,
+      "grad_norm": 0.2664169371128082,
+      "learning_rate": 7.054054054054055e-06,
+      "loss": 0.4348,
+      "step": 261
+    },
+    {
+      "epoch": 0.21271989174560216,
+      "grad_norm": 0.2527692914009094,
+      "learning_rate": 7.0810810810810815e-06,
+      "loss": 0.4488,
+      "step": 262
+    },
+    {
+      "epoch": 0.213531799729364,
+      "grad_norm": 0.2904249131679535,
+      "learning_rate": 7.1081081081081085e-06,
+      "loss": 0.4368,
+      "step": 263
+    },
+    {
+      "epoch": 0.21434370771312586,
+      "grad_norm": 0.2809026539325714,
+      "learning_rate": 7.135135135135136e-06,
+      "loss": 0.4321,
+      "step": 264
+    },
+    {
+      "epoch": 0.21515561569688768,
+      "grad_norm": 0.292337566614151,
+      "learning_rate": 7.162162162162163e-06,
+      "loss": 0.435,
+      "step": 265
+    },
+    {
+      "epoch": 0.21596752368064953,
+      "grad_norm": 0.2708173990249634,
+      "learning_rate": 7.189189189189189e-06,
+      "loss": 0.4041,
+      "step": 266
+    },
+    {
+      "epoch": 0.21677943166441135,
+      "grad_norm": 0.27476823329925537,
+      "learning_rate": 7.216216216216216e-06,
+      "loss": 0.4214,
+      "step": 267
+    },
+    {
+      "epoch": 0.2175913396481732,
+      "grad_norm": 0.29490163922309875,
+      "learning_rate": 7.243243243243244e-06,
+      "loss": 0.4418,
+      "step": 268
+    },
+    {
+      "epoch": 0.21840324763193505,
+      "grad_norm": 0.27122682332992554,
+      "learning_rate": 7.270270270270271e-06,
+      "loss": 0.4268,
+      "step": 269
+    },
+    {
+      "epoch": 0.21921515561569688,
+      "grad_norm": 0.3094821572303772,
+      "learning_rate": 7.297297297297298e-06,
+      "loss": 0.4269,
+      "step": 270
+    },
+    {
+      "epoch": 0.22002706359945873,
+      "grad_norm": 0.2790597677230835,
+      "learning_rate": 7.324324324324325e-06,
+      "loss": 0.4389,
+      "step": 271
+    },
+    {
+      "epoch": 0.22083897158322058,
+      "grad_norm": 0.26837489008903503,
+      "learning_rate": 7.3513513513513525e-06,
+      "loss": 0.4367,
+      "step": 272
+    },
+    {
+      "epoch": 0.2216508795669824,
+      "grad_norm": 0.2677345871925354,
+      "learning_rate": 7.3783783783783794e-06,
+      "loss": 0.4127,
+      "step": 273
+    },
+    {
+      "epoch": 0.22246278755074425,
+      "grad_norm": 0.27498430013656616,
+      "learning_rate": 7.4054054054054055e-06,
+      "loss": 0.4237,
+      "step": 274
+    },
+    {
+      "epoch": 0.2232746955345061,
+      "grad_norm": 0.28092265129089355,
+      "learning_rate": 7.4324324324324324e-06,
+      "loss": 0.4193,
+      "step": 275
+    },
+    {
+      "epoch": 0.22408660351826792,
+      "grad_norm": 0.348589152097702,
+      "learning_rate": 7.45945945945946e-06,
+      "loss": 0.4387,
+      "step": 276
+    },
+    {
+      "epoch": 0.22489851150202977,
+      "grad_norm": 0.2704857885837555,
+      "learning_rate": 7.486486486486487e-06,
+      "loss": 0.4064,
+      "step": 277
+    },
+    {
+      "epoch": 0.2257104194857916,
+      "grad_norm": 0.30964452028274536,
+      "learning_rate": 7.513513513513514e-06,
+      "loss": 0.4307,
+      "step": 278
+    },
+    {
+      "epoch": 0.22652232746955345,
+      "grad_norm": 0.30235254764556885,
+      "learning_rate": 7.540540540540541e-06,
+      "loss": 0.4365,
+      "step": 279
+    },
+    {
+      "epoch": 0.2273342354533153,
+      "grad_norm": 0.2706587612628937,
+      "learning_rate": 7.567567567567569e-06,
+      "loss": 0.4154,
+      "step": 280
+    },
+    {
+      "epoch": 0.22814614343707712,
+      "grad_norm": 0.2671191394329071,
+      "learning_rate": 7.594594594594596e-06,
+      "loss": 0.4122,
+      "step": 281
+    },
+    {
+      "epoch": 0.22895805142083897,
+      "grad_norm": 0.27113959193229675,
+      "learning_rate": 7.621621621621622e-06,
+      "loss": 0.4437,
+      "step": 282
+    },
+    {
+      "epoch": 0.22976995940460082,
+      "grad_norm": 0.27566424012184143,
+      "learning_rate": 7.648648648648649e-06,
+      "loss": 0.4001,
+      "step": 283
+    },
+    {
+      "epoch": 0.23058186738836264,
+      "grad_norm": 0.28094545006752014,
+      "learning_rate": 7.675675675675676e-06,
+      "loss": 0.4229,
+      "step": 284
+    },
+    {
+      "epoch": 0.2313937753721245,
+      "grad_norm": 0.28368356823921204,
+      "learning_rate": 7.702702702702704e-06,
+      "loss": 0.4112,
+      "step": 285
+    },
+    {
+      "epoch": 0.23220568335588634,
+      "grad_norm": 0.30077770352363586,
+      "learning_rate": 7.72972972972973e-06,
+      "loss": 0.4393,
+      "step": 286
+    },
+    {
+      "epoch": 0.23301759133964817,
+      "grad_norm": 0.3043121099472046,
+      "learning_rate": 7.756756756756756e-06,
+      "loss": 0.409,
+      "step": 287
+    },
+    {
+      "epoch": 0.23382949932341002,
+      "grad_norm": 0.2876274585723877,
+      "learning_rate": 7.783783783783784e-06,
+      "loss": 0.4304,
+      "step": 288
+    },
+    {
+      "epoch": 0.23464140730717187,
+      "grad_norm": 0.2997243106365204,
+      "learning_rate": 7.810810810810812e-06,
+      "loss": 0.4186,
+      "step": 289
+    },
+    {
+      "epoch": 0.2354533152909337,
+      "grad_norm": 0.27215775847435,
+      "learning_rate": 7.837837837837838e-06,
+      "loss": 0.4215,
+      "step": 290
+    },
+    {
+      "epoch": 0.23626522327469554,
+      "grad_norm": 0.3135278522968292,
+      "learning_rate": 7.864864864864866e-06,
+      "loss": 0.4322,
+      "step": 291
+    },
+    {
+      "epoch": 0.23707713125845736,
+      "grad_norm": 0.2993195652961731,
+      "learning_rate": 7.891891891891894e-06,
+      "loss": 0.4189,
+      "step": 292
+    },
+    {
+      "epoch": 0.2378890392422192,
+      "grad_norm": 0.26529794931411743,
+      "learning_rate": 7.91891891891892e-06,
+      "loss": 0.4279,
+      "step": 293
+    },
+    {
+      "epoch": 0.23870094722598106,
+      "grad_norm": 0.2905847728252411,
+      "learning_rate": 7.945945945945946e-06,
+      "loss": 0.4334,
+      "step": 294
+    },
+    {
+      "epoch": 0.2395128552097429,
+      "grad_norm": 0.3000369966030121,
+      "learning_rate": 7.972972972972974e-06,
+      "loss": 0.4084,
+      "step": 295
+    },
+    {
+      "epoch": 0.24032476319350474,
+      "grad_norm": 0.2811955511569977,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.429,
+      "step": 296
+    },
+    {
+      "epoch": 0.2411366711772666,
+      "grad_norm": 0.30162736773490906,
+      "learning_rate": 8.027027027027027e-06,
+      "loss": 0.4281,
+      "step": 297
+    },
+    {
+      "epoch": 0.2419485791610284,
+      "grad_norm": 0.27000707387924194,
+      "learning_rate": 8.054054054054055e-06,
+      "loss": 0.4237,
+      "step": 298
+    },
+    {
+      "epoch": 0.24276048714479026,
+      "grad_norm": 0.28867992758750916,
+      "learning_rate": 8.081081081081081e-06,
+      "loss": 0.4257,
+      "step": 299
+    },
+    {
+      "epoch": 0.2435723951285521,
+      "grad_norm": 0.31571611762046814,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 0.4442,
+      "step": 300
+    },
+    {
+      "epoch": 0.24438430311231393,
+      "grad_norm": 0.2796787917613983,
+      "learning_rate": 8.135135135135137e-06,
+      "loss": 0.407,
+      "step": 301
+    },
+    {
+      "epoch": 0.24519621109607578,
+      "grad_norm": 0.31239089369773865,
+      "learning_rate": 8.162162162162163e-06,
+      "loss": 0.4333,
+      "step": 302
+    },
+    {
+      "epoch": 0.2460081190798376,
+      "grad_norm": 0.29210394620895386,
+      "learning_rate": 8.189189189189189e-06,
+      "loss": 0.4258,
+      "step": 303
+    },
+    {
+      "epoch": 0.24682002706359946,
+      "grad_norm": 0.312454491853714,
+      "learning_rate": 8.216216216216217e-06,
+      "loss": 0.408,
+      "step": 304
+    },
+    {
+      "epoch": 0.2476319350473613,
+      "grad_norm": 0.28502053022384644,
+      "learning_rate": 8.243243243243245e-06,
+      "loss": 0.4326,
+      "step": 305
+    },
+    {
+      "epoch": 0.24844384303112313,
+      "grad_norm": 0.3296396732330322,
+      "learning_rate": 8.27027027027027e-06,
+      "loss": 0.4331,
+      "step": 306
+    },
+    {
+      "epoch": 0.24925575101488498,
+      "grad_norm": 0.2515292465686798,
+      "learning_rate": 8.297297297297298e-06,
+      "loss": 0.4159,
+      "step": 307
+    },
+    {
+      "epoch": 0.2500676589986468,
+      "grad_norm": 0.29174137115478516,
+      "learning_rate": 8.324324324324326e-06,
+      "loss": 0.4438,
+      "step": 308
+    },
+    {
+      "epoch": 0.25087956698240865,
+      "grad_norm": 0.2825658917427063,
+      "learning_rate": 8.351351351351352e-06,
+      "loss": 0.4233,
+      "step": 309
+    },
+    {
+      "epoch": 0.2516914749661705,
+      "grad_norm": 0.29513809084892273,
+      "learning_rate": 8.378378378378378e-06,
+      "loss": 0.4142,
+      "step": 310
+    },
+    {
+      "epoch": 0.25250338294993235,
+      "grad_norm": 0.27558788657188416,
+      "learning_rate": 8.405405405405406e-06,
+      "loss": 0.4191,
+      "step": 311
+    },
+    {
+      "epoch": 0.2533152909336942,
+      "grad_norm": 0.27788442373275757,
+      "learning_rate": 8.432432432432434e-06,
+      "loss": 0.4142,
+      "step": 312
+    },
+    {
+      "epoch": 0.254127198917456,
+      "grad_norm": 0.3359508514404297,
+      "learning_rate": 8.45945945945946e-06,
+      "loss": 0.4396,
+      "step": 313
+    },
+    {
+      "epoch": 0.25493910690121785,
+      "grad_norm": 0.2504373788833618,
+      "learning_rate": 8.486486486486488e-06,
+      "loss": 0.3999,
+      "step": 314
+    },
+    {
+      "epoch": 0.2557510148849797,
+      "grad_norm": 0.2846215069293976,
+      "learning_rate": 8.513513513513514e-06,
+      "loss": 0.407,
+      "step": 315
+    },
+    {
+      "epoch": 0.25656292286874155,
+      "grad_norm": 0.3023104667663574,
+      "learning_rate": 8.540540540540542e-06,
+      "loss": 0.4136,
+      "step": 316
+    },
+    {
+      "epoch": 0.2573748308525034,
+      "grad_norm": 0.2777184247970581,
+      "learning_rate": 8.567567567567568e-06,
+      "loss": 0.4069,
+      "step": 317
+    },
+    {
+      "epoch": 0.25818673883626525,
+      "grad_norm": 0.27417460083961487,
+      "learning_rate": 8.594594594594595e-06,
+      "loss": 0.4205,
+      "step": 318
+    },
+    {
+      "epoch": 0.25899864682002705,
+      "grad_norm": 0.27624189853668213,
+      "learning_rate": 8.621621621621622e-06,
+      "loss": 0.3892,
+      "step": 319
+    },
+    {
+      "epoch": 0.2598105548037889,
+      "grad_norm": 0.2745577394962311,
+      "learning_rate": 8.64864864864865e-06,
+      "loss": 0.3993,
+      "step": 320
+    },
+    {
+      "epoch": 0.26062246278755075,
+      "grad_norm": 0.30521059036254883,
+      "learning_rate": 8.675675675675677e-06,
+      "loss": 0.4124,
+      "step": 321
+    },
+    {
+      "epoch": 0.2614343707713126,
+      "grad_norm": 0.2839369773864746,
+      "learning_rate": 8.702702702702703e-06,
+      "loss": 0.4415,
+      "step": 322
+    },
+    {
+      "epoch": 0.26224627875507445,
+      "grad_norm": 0.25499430298805237,
+      "learning_rate": 8.72972972972973e-06,
+      "loss": 0.4161,
+      "step": 323
+    },
+    {
+      "epoch": 0.26305818673883624,
+      "grad_norm": 0.24711433053016663,
+      "learning_rate": 8.756756756756759e-06,
+      "loss": 0.4043,
+      "step": 324
+    },
+    {
+      "epoch": 0.2638700947225981,
+      "grad_norm": 0.25903162360191345,
+      "learning_rate": 8.783783783783785e-06,
+      "loss": 0.4249,
+      "step": 325
+    },
+    {
+      "epoch": 0.26468200270635994,
+      "grad_norm": 0.2625133693218231,
+      "learning_rate": 8.810810810810811e-06,
+      "loss": 0.4129,
+      "step": 326
+    },
+    {
+      "epoch": 0.2654939106901218,
+      "grad_norm": 0.24447953701019287,
+      "learning_rate": 8.837837837837839e-06,
+      "loss": 0.4369,
+      "step": 327
+    },
+    {
+      "epoch": 0.26630581867388364,
+      "grad_norm": 0.3021467924118042,
+      "learning_rate": 8.864864864864866e-06,
+      "loss": 0.4212,
+      "step": 328
+    },
+    {
+      "epoch": 0.2671177266576455,
+      "grad_norm": 0.2817029356956482,
+      "learning_rate": 8.891891891891893e-06,
+      "loss": 0.3964,
+      "step": 329
+    },
+    {
+      "epoch": 0.2679296346414073,
+      "grad_norm": 0.27256298065185547,
+      "learning_rate": 8.91891891891892e-06,
+      "loss": 0.4203,
+      "step": 330
+    },
+    {
+      "epoch": 0.26874154262516914,
+      "grad_norm": 0.31614595651626587,
+      "learning_rate": 8.945945945945946e-06,
+      "loss": 0.4105,
+      "step": 331
+    },
+    {
+      "epoch": 0.269553450608931,
+      "grad_norm": 0.29067346453666687,
+      "learning_rate": 8.972972972972974e-06,
+      "loss": 0.4138,
+      "step": 332
+    },
+    {
+      "epoch": 0.27036535859269284,
+      "grad_norm": 0.28187695145606995,
+      "learning_rate": 9e-06,
+      "loss": 0.427,
+      "step": 333
+    },
+    {
+      "epoch": 0.2711772665764547,
+      "grad_norm": 0.3028711676597595,
+      "learning_rate": 9.027027027027028e-06,
+      "loss": 0.4361,
+      "step": 334
+    },
+    {
+      "epoch": 0.2719891745602165,
+      "grad_norm": 0.3457663059234619,
+      "learning_rate": 9.054054054054054e-06,
+      "loss": 0.4315,
+      "step": 335
+    },
+    {
+      "epoch": 0.27280108254397833,
+      "grad_norm": 0.27318572998046875,
+      "learning_rate": 9.081081081081082e-06,
+      "loss": 0.4243,
+      "step": 336
+    },
+    {
+      "epoch": 0.2736129905277402,
+      "grad_norm": 0.34319356083869934,
+      "learning_rate": 9.10810810810811e-06,
+      "loss": 0.4119,
+      "step": 337
+    },
+    {
+      "epoch": 0.27442489851150204,
+      "grad_norm": 0.293912410736084,
+      "learning_rate": 9.135135135135136e-06,
+      "loss": 0.386,
+      "step": 338
+    },
+    {
+      "epoch": 0.2752368064952639,
+      "grad_norm": 0.2892298102378845,
+      "learning_rate": 9.162162162162162e-06,
+      "loss": 0.4064,
+      "step": 339
+    },
+    {
+      "epoch": 0.27604871447902574,
+      "grad_norm": 0.3144510090351105,
+      "learning_rate": 9.189189189189191e-06,
+      "loss": 0.4182,
+      "step": 340
+    },
+    {
+      "epoch": 0.27686062246278753,
+      "grad_norm": 0.29586517810821533,
+      "learning_rate": 9.216216216216217e-06,
+      "loss": 0.4008,
+      "step": 341
+    },
+    {
+      "epoch": 0.2776725304465494,
+      "grad_norm": 0.31982049345970154,
+      "learning_rate": 9.243243243243243e-06,
+      "loss": 0.4168,
+      "step": 342
+    },
+    {
+      "epoch": 0.27848443843031123,
+      "grad_norm": 0.31422844529151917,
+      "learning_rate": 9.270270270270271e-06,
+      "loss": 0.4245,
+      "step": 343
+    },
+    {
+      "epoch": 0.2792963464140731,
+      "grad_norm": 0.3336861729621887,
+      "learning_rate": 9.297297297297299e-06,
+      "loss": 0.4241,
+      "step": 344
+    },
+    {
+      "epoch": 0.28010825439783493,
+      "grad_norm": 0.28966692090034485,
+      "learning_rate": 9.324324324324325e-06,
+      "loss": 0.4287,
+      "step": 345
+    },
+    {
+      "epoch": 0.2809201623815967,
+      "grad_norm": 0.30232182145118713,
+      "learning_rate": 9.351351351351353e-06,
+      "loss": 0.4234,
+      "step": 346
+    },
+    {
+      "epoch": 0.2817320703653586,
+      "grad_norm": 0.2580353617668152,
+      "learning_rate": 9.378378378378379e-06,
+      "loss": 0.398,
+      "step": 347
+    },
+    {
+      "epoch": 0.28254397834912043,
+      "grad_norm": 0.3279270529747009,
+      "learning_rate": 9.405405405405407e-06,
+      "loss": 0.413,
+      "step": 348
+    },
+    {
+      "epoch": 0.2833558863328823,
+      "grad_norm": 0.2678486704826355,
+      "learning_rate": 9.432432432432433e-06,
+      "loss": 0.401,
+      "step": 349
+    },
+    {
+      "epoch": 0.28416779431664413,
+      "grad_norm": 0.339856892824173,
+      "learning_rate": 9.45945945945946e-06,
+      "loss": 0.4162,
+      "step": 350
+    },
+    {
+      "epoch": 0.284979702300406,
+      "grad_norm": 0.2643829584121704,
+      "learning_rate": 9.486486486486487e-06,
+      "loss": 0.419,
+      "step": 351
+    },
+    {
+      "epoch": 0.2857916102841678,
+      "grad_norm": 0.3013765513896942,
+      "learning_rate": 9.513513513513514e-06,
+      "loss": 0.4102,
+      "step": 352
+    },
+    {
+      "epoch": 0.2866035182679296,
+      "grad_norm": 0.2833913266658783,
+      "learning_rate": 9.540540540540542e-06,
+      "loss": 0.387,
+      "step": 353
+    },
+    {
+      "epoch": 0.2874154262516915,
+      "grad_norm": 0.2898872494697571,
+      "learning_rate": 9.567567567567568e-06,
+      "loss": 0.4347,
+      "step": 354
+    },
+    {
+      "epoch": 0.2882273342354533,
+      "grad_norm": 0.2897942364215851,
+      "learning_rate": 9.594594594594594e-06,
+      "loss": 0.4188,
+      "step": 355
+    },
+    {
+      "epoch": 0.2890392422192152,
+      "grad_norm": 0.2966800630092621,
+      "learning_rate": 9.621621621621622e-06,
+      "loss": 0.416,
+      "step": 356
+    },
+    {
+      "epoch": 0.28985115020297697,
+      "grad_norm": 0.281484991312027,
+      "learning_rate": 9.64864864864865e-06,
+      "loss": 0.3991,
+      "step": 357
+    },
+    {
+      "epoch": 0.2906630581867388,
+      "grad_norm": 0.28956663608551025,
+      "learning_rate": 9.675675675675676e-06,
+      "loss": 0.4281,
+      "step": 358
+    },
+    {
+      "epoch": 0.29147496617050067,
+      "grad_norm": 0.2592754662036896,
+      "learning_rate": 9.702702702702704e-06,
+      "loss": 0.4344,
+      "step": 359
+    },
+    {
+      "epoch": 0.2922868741542625,
+      "grad_norm": 0.2672978341579437,
+      "learning_rate": 9.729729729729732e-06,
+      "loss": 0.4409,
+      "step": 360
+    },
+    {
+      "epoch": 0.29309878213802437,
+      "grad_norm": 0.30128002166748047,
+      "learning_rate": 9.756756756756758e-06,
+      "loss": 0.4177,
+      "step": 361
+    },
+    {
+      "epoch": 0.2939106901217862,
+      "grad_norm": 0.28915122151374817,
+      "learning_rate": 9.783783783783785e-06,
+      "loss": 0.421,
+      "step": 362
+    },
+    {
+      "epoch": 0.294722598105548,
+      "grad_norm": 0.28276965022087097,
+      "learning_rate": 9.810810810810811e-06,
+      "loss": 0.4217,
+      "step": 363
+    },
+    {
+      "epoch": 0.29553450608930987,
+      "grad_norm": 0.2785509526729584,
+      "learning_rate": 9.83783783783784e-06,
+      "loss": 0.4139,
+      "step": 364
+    },
+    {
+      "epoch": 0.2963464140730717,
+      "grad_norm": 0.2847628593444824,
+      "learning_rate": 9.864864864864865e-06,
+      "loss": 0.4095,
+      "step": 365
+    },
+    {
+      "epoch": 0.29715832205683357,
+      "grad_norm": 0.2873031497001648,
+      "learning_rate": 9.891891891891893e-06,
+      "loss": 0.4072,
+      "step": 366
+    },
+    {
+      "epoch": 0.2979702300405954,
+      "grad_norm": 0.2595895826816559,
+      "learning_rate": 9.91891891891892e-06,
+      "loss": 0.4355,
+      "step": 367
+    },
+    {
+      "epoch": 0.2987821380243572,
+      "grad_norm": 0.32075682282447815,
+      "learning_rate": 9.945945945945947e-06,
+      "loss": 0.4235,
+      "step": 368
+    },
+    {
+      "epoch": 0.29959404600811906,
+      "grad_norm": 0.2869185507297516,
+      "learning_rate": 9.972972972972975e-06,
+      "loss": 0.4125,
+      "step": 369
+    },
+    {
+      "epoch": 0.3004059539918809,
+      "grad_norm": 0.2791314125061035,
+      "learning_rate": 1e-05,
+      "loss": 0.4062,
+      "step": 370
+    },
+    {
+      "epoch": 0.30121786197564276,
+      "grad_norm": 0.2743418216705322,
+      "learning_rate": 9.999997765506791e-06,
+      "loss": 0.3914,
+      "step": 371
+    },
+    {
+      "epoch": 0.3020297699594046,
+      "grad_norm": 0.2648778259754181,
+      "learning_rate": 9.99999106202916e-06,
+      "loss": 0.419,
+      "step": 372
+    },
+    {
+      "epoch": 0.30284167794316647,
+      "grad_norm": 0.2914637625217438,
+      "learning_rate": 9.9999798895731e-06,
+      "loss": 0.4247,
+      "step": 373
+    },
+    {
+      "epoch": 0.30365358592692826,
+      "grad_norm": 0.266376793384552,
+      "learning_rate": 9.999964248148597e-06,
+      "loss": 0.401,
+      "step": 374
+    },
+    {
+      "epoch": 0.3044654939106901,
+      "grad_norm": 0.2855246961116791,
+      "learning_rate": 9.999944137769629e-06,
+      "loss": 0.4108,
+      "step": 375
+    },
+    {
+      "epoch": 0.30527740189445196,
+      "grad_norm": 0.2781916558742523,
+      "learning_rate": 9.999919558454171e-06,
+      "loss": 0.3844,
+      "step": 376
+    },
+    {
+      "epoch": 0.3060893098782138,
+      "grad_norm": 0.2806081771850586,
+      "learning_rate": 9.999890510224195e-06,
+      "loss": 0.4173,
+      "step": 377
+    },
+    {
+      "epoch": 0.30690121786197566,
+      "grad_norm": 0.2794731557369232,
+      "learning_rate": 9.999856993105661e-06,
+      "loss": 0.4136,
+      "step": 378
+    },
+    {
+      "epoch": 0.30771312584573746,
+      "grad_norm": 0.277649849653244,
+      "learning_rate": 9.999819007128527e-06,
+      "loss": 0.3943,
+      "step": 379
+    },
+    {
+      "epoch": 0.3085250338294993,
+      "grad_norm": 0.2836255431175232,
+      "learning_rate": 9.999776552326747e-06,
+      "loss": 0.4332,
+      "step": 380
+    },
+    {
+      "epoch": 0.30933694181326116,
+      "grad_norm": 0.3226708471775055,
+      "learning_rate": 9.999729628738263e-06,
+      "loss": 0.4219,
+      "step": 381
+    },
+    {
+      "epoch": 0.310148849797023,
+      "grad_norm": 0.2904399633407593,
+      "learning_rate": 9.99967823640502e-06,
+      "loss": 0.4164,
+      "step": 382
+    },
+    {
+      "epoch": 0.31096075778078486,
+      "grad_norm": 0.31306928396224976,
+      "learning_rate": 9.99962237537295e-06,
+      "loss": 0.3992,
+      "step": 383
+    },
+    {
+      "epoch": 0.3117726657645467,
+      "grad_norm": 0.30289721488952637,
+      "learning_rate": 9.999562045691979e-06,
+      "loss": 0.4095,
+      "step": 384
+    },
+    {
+      "epoch": 0.3125845737483085,
+      "grad_norm": 0.27280378341674805,
+      "learning_rate": 9.999497247416035e-06,
+      "loss": 0.4062,
+      "step": 385
+    },
+    {
+      "epoch": 0.31339648173207035,
+      "grad_norm": 0.3146529495716095,
+      "learning_rate": 9.99942798060303e-06,
+      "loss": 0.4208,
+      "step": 386
+    },
+    {
+      "epoch": 0.3142083897158322,
+      "grad_norm": 0.2850913405418396,
+      "learning_rate": 9.999354245314874e-06,
+      "loss": 0.4491,
+      "step": 387
+    },
+    {
+      "epoch": 0.31502029769959405,
+      "grad_norm": 0.3229399621486664,
+      "learning_rate": 9.999276041617477e-06,
+      "loss": 0.4127,
+      "step": 388
+    },
+    {
+      "epoch": 0.3158322056833559,
+      "grad_norm": 0.3087291717529297,
+      "learning_rate": 9.999193369580729e-06,
+      "loss": 0.4082,
+      "step": 389
+    },
+    {
+      "epoch": 0.3166441136671177,
+      "grad_norm": 0.3293628990650177,
+      "learning_rate": 9.999106229278531e-06,
+      "loss": 0.4326,
+      "step": 390
+    },
+    {
+      "epoch": 0.31745602165087955,
+      "grad_norm": 0.2849918603897095,
+      "learning_rate": 9.99901462078876e-06,
+      "loss": 0.4215,
+      "step": 391
+    },
+    {
+      "epoch": 0.3182679296346414,
+      "grad_norm": 0.3028421998023987,
+      "learning_rate": 9.998918544193303e-06,
+      "loss": 0.4233,
+      "step": 392
+    },
+    {
+      "epoch": 0.31907983761840325,
+      "grad_norm": 0.2924360930919647,
+      "learning_rate": 9.998817999578028e-06,
+      "loss": 0.4413,
+      "step": 393
+    },
+    {
+      "epoch": 0.3198917456021651,
+      "grad_norm": 0.3041113615036011,
+      "learning_rate": 9.998712987032803e-06,
+      "loss": 0.4384,
+      "step": 394
+    },
+    {
+      "epoch": 0.32070365358592695,
+      "grad_norm": 0.28165844082832336,
+      "learning_rate": 9.99860350665149e-06,
+      "loss": 0.4041,
+      "step": 395
+    },
+    {
+      "epoch": 0.32151556156968875,
+      "grad_norm": 0.3048534393310547,
+      "learning_rate": 9.99848955853194e-06,
+      "loss": 0.4128,
+      "step": 396
+    },
+    {
+      "epoch": 0.3223274695534506,
+      "grad_norm": 0.29435670375823975,
+      "learning_rate": 9.998371142776e-06,
+      "loss": 0.4374,
+      "step": 397
+    },
+    {
+      "epoch": 0.32313937753721245,
+      "grad_norm": 0.3139035999774933,
+      "learning_rate": 9.998248259489509e-06,
+      "loss": 0.427,
+      "step": 398
+    },
+    {
+      "epoch": 0.3239512855209743,
+      "grad_norm": 0.2909335792064667,
+      "learning_rate": 9.998120908782301e-06,
+      "loss": 0.4149,
+      "step": 399
+    },
+    {
+      "epoch": 0.32476319350473615,
+      "grad_norm": 0.34057843685150146,
+      "learning_rate": 9.9979890907682e-06,
+      "loss": 0.4306,
+      "step": 400
+    },
+    {
+      "epoch": 0.325575101488498,
+      "grad_norm": 0.2985652983188629,
+      "learning_rate": 9.997852805565026e-06,
+      "loss": 0.4087,
+      "step": 401
+    },
+    {
+      "epoch": 0.3263870094722598,
+      "grad_norm": 0.3349083662033081,
+      "learning_rate": 9.99771205329459e-06,
+      "loss": 0.3833,
+      "step": 402
+    },
+    {
+      "epoch": 0.32719891745602164,
+      "grad_norm": 0.3146043121814728,
+      "learning_rate": 9.997566834082697e-06,
+      "loss": 0.4362,
+      "step": 403
+    },
+    {
+      "epoch": 0.3280108254397835,
+      "grad_norm": 0.2737433910369873,
+      "learning_rate": 9.997417148059142e-06,
+      "loss": 0.4064,
+      "step": 404
+    },
+    {
+      "epoch": 0.32882273342354534,
+      "grad_norm": 0.28521665930747986,
+      "learning_rate": 9.997262995357714e-06,
+      "loss": 0.4098,
+      "step": 405
+    },
+    {
+      "epoch": 0.3296346414073072,
+      "grad_norm": 0.33004093170166016,
+      "learning_rate": 9.997104376116195e-06,
+      "loss": 0.4305,
+      "step": 406
+    },
+    {
+      "epoch": 0.330446549391069,
+      "grad_norm": 0.301505982875824,
+      "learning_rate": 9.996941290476359e-06,
+      "loss": 0.4381,
+      "step": 407
+    },
+    {
+      "epoch": 0.33125845737483084,
+      "grad_norm": 0.2945079505443573,
+      "learning_rate": 9.99677373858397e-06,
+      "loss": 0.42,
+      "step": 408
+    },
+    {
+      "epoch": 0.3320703653585927,
+      "grad_norm": 0.32049235701560974,
+      "learning_rate": 9.996601720588787e-06,
+      "loss": 0.424,
+      "step": 409
+    },
+    {
+      "epoch": 0.33288227334235454,
+      "grad_norm": 0.35567617416381836,
+      "learning_rate": 9.996425236644558e-06,
+      "loss": 0.4072,
+      "step": 410
+    },
+    {
+      "epoch": 0.3336941813261164,
+      "grad_norm": 0.26636794209480286,
+      "learning_rate": 9.996244286909022e-06,
+      "loss": 0.4291,
+      "step": 411
+    },
+    {
+      "epoch": 0.33450608930987824,
+      "grad_norm": 0.3060134947299957,
+      "learning_rate": 9.996058871543917e-06,
+      "loss": 0.4036,
+      "step": 412
+    },
+    {
+      "epoch": 0.33531799729364004,
+      "grad_norm": 0.33968493342399597,
+      "learning_rate": 9.995868990714963e-06,
+      "loss": 0.4056,
+      "step": 413
+    },
+    {
+      "epoch": 0.3361299052774019,
+      "grad_norm": 0.31462815403938293,
+      "learning_rate": 9.995674644591873e-06,
+      "loss": 0.4261,
+      "step": 414
+    },
+    {
+      "epoch": 0.33694181326116374,
+      "grad_norm": 0.25864580273628235,
+      "learning_rate": 9.995475833348359e-06,
+      "loss": 0.4065,
+      "step": 415
+    },
+    {
+      "epoch": 0.3377537212449256,
+      "grad_norm": 0.2932393252849579,
+      "learning_rate": 9.99527255716211e-06,
+      "loss": 0.4003,
+      "step": 416
+    },
+    {
+      "epoch": 0.33856562922868744,
+      "grad_norm": 0.33115771412849426,
+      "learning_rate": 9.99506481621482e-06,
+      "loss": 0.3978,
+      "step": 417
+    },
+    {
+      "epoch": 0.33937753721244923,
+      "grad_norm": 0.2847278416156769,
+      "learning_rate": 9.994852610692166e-06,
+      "loss": 0.4026,
+      "step": 418
+    },
+    {
+      "epoch": 0.3401894451962111,
+      "grad_norm": 0.291806161403656,
+      "learning_rate": 9.994635940783816e-06,
+      "loss": 0.415,
+      "step": 419
+    },
+    {
+      "epoch": 0.34100135317997293,
+      "grad_norm": 0.27313894033432007,
+      "learning_rate": 9.994414806683429e-06,
+      "loss": 0.3826,
+      "step": 420
+    },
+    {
+      "epoch": 0.3418132611637348,
+      "grad_norm": 0.28268569707870483,
+      "learning_rate": 9.994189208588655e-06,
+      "loss": 0.3975,
+      "step": 421
+    },
+    {
+      "epoch": 0.34262516914749663,
+      "grad_norm": 0.28169697523117065,
+      "learning_rate": 9.993959146701128e-06,
+      "loss": 0.4022,
+      "step": 422
+    },
+    {
+      "epoch": 0.3434370771312585,
+      "grad_norm": 0.27267321944236755,
+      "learning_rate": 9.993724621226484e-06,
+      "loss": 0.3813,
+      "step": 423
+    },
+    {
+      "epoch": 0.3442489851150203,
+      "grad_norm": 0.25187477469444275,
+      "learning_rate": 9.993485632374338e-06,
+      "loss": 0.3885,
+      "step": 424
+    },
+    {
+      "epoch": 0.34506089309878213,
+      "grad_norm": 0.3331622779369354,
+      "learning_rate": 9.993242180358298e-06,
+      "loss": 0.4187,
+      "step": 425
+    },
+    {
+      "epoch": 0.345872801082544,
+      "grad_norm": 0.2578586935997009,
+      "learning_rate": 9.992994265395959e-06,
+      "loss": 0.3847,
+      "step": 426
+    },
+    {
+      "epoch": 0.34668470906630583,
+      "grad_norm": 0.30197644233703613,
+      "learning_rate": 9.992741887708908e-06,
+      "loss": 0.3955,
+      "step": 427
+    },
+    {
+      "epoch": 0.3474966170500677,
+      "grad_norm": 0.5474060773849487,
+      "learning_rate": 9.99248504752272e-06,
+      "loss": 0.4258,
+      "step": 428
+    },
+    {
+      "epoch": 0.3483085250338295,
+      "grad_norm": 0.3022882640361786,
+      "learning_rate": 9.992223745066959e-06,
+      "loss": 0.4074,
+      "step": 429
+    },
+    {
+      "epoch": 0.3491204330175913,
+      "grad_norm": 0.2710326313972473,
+      "learning_rate": 9.991957980575172e-06,
+      "loss": 0.4147,
+      "step": 430
+    },
+    {
+      "epoch": 0.3499323410013532,
+      "grad_norm": 0.30201712250709534,
+      "learning_rate": 9.991687754284904e-06,
+      "loss": 0.4003,
+      "step": 431
+    },
+    {
+      "epoch": 0.350744248985115,
+      "grad_norm": 0.28015461564064026,
+      "learning_rate": 9.99141306643768e-06,
+      "loss": 0.4207,
+      "step": 432
+    },
+    {
+      "epoch": 0.3515561569688769,
+      "grad_norm": 0.32233279943466187,
+      "learning_rate": 9.991133917279013e-06,
+      "loss": 0.4051,
+      "step": 433
+    },
+    {
+      "epoch": 0.3523680649526387,
+      "grad_norm": 0.28556424379348755,
+      "learning_rate": 9.990850307058411e-06,
+      "loss": 0.4087,
+      "step": 434
+    },
+    {
+      "epoch": 0.3531799729364005,
+      "grad_norm": 0.30146023631095886,
+      "learning_rate": 9.99056223602936e-06,
+      "loss": 0.4091,
+      "step": 435
+    },
+    {
+      "epoch": 0.35399188092016237,
+      "grad_norm": 0.2997048497200012,
+      "learning_rate": 9.990269704449338e-06,
+      "loss": 0.3953,
+      "step": 436
+    },
+    {
+      "epoch": 0.3548037889039242,
+      "grad_norm": 0.29788514971733093,
+      "learning_rate": 9.989972712579809e-06,
+      "loss": 0.4055,
+      "step": 437
+    },
+    {
+      "epoch": 0.3556156968876861,
+      "grad_norm": 0.29539647698402405,
+      "learning_rate": 9.989671260686225e-06,
+      "loss": 0.4023,
+      "step": 438
+    },
+    {
+      "epoch": 0.3564276048714479,
+      "grad_norm": 0.2750176787376404,
+      "learning_rate": 9.989365349038019e-06,
+      "loss": 0.4119,
+      "step": 439
+    },
+    {
+      "epoch": 0.3572395128552097,
+      "grad_norm": 0.26545006036758423,
+      "learning_rate": 9.98905497790862e-06,
+      "loss": 0.4071,
+      "step": 440
+    },
+    {
+      "epoch": 0.35805142083897157,
+      "grad_norm": 0.2633037567138672,
+      "learning_rate": 9.98874014757543e-06,
+      "loss": 0.3963,
+      "step": 441
+    },
+    {
+      "epoch": 0.3588633288227334,
+      "grad_norm": 0.3042871356010437,
+      "learning_rate": 9.98842085831985e-06,
+      "loss": 0.4163,
+      "step": 442
+    },
+    {
+      "epoch": 0.35967523680649527,
+      "grad_norm": 0.297606498003006,
+      "learning_rate": 9.988097110427255e-06,
+      "loss": 0.4488,
+      "step": 443
+    },
+    {
+      "epoch": 0.3604871447902571,
+      "grad_norm": 0.3045158386230469,
+      "learning_rate": 9.987768904187014e-06,
+      "loss": 0.4169,
+      "step": 444
+    },
+    {
+      "epoch": 0.36129905277401897,
+      "grad_norm": 0.26727333664894104,
+      "learning_rate": 9.987436239892472e-06,
+      "loss": 0.4068,
+      "step": 445
+    },
+    {
+      "epoch": 0.36211096075778076,
+      "grad_norm": 0.2862260639667511,
+      "learning_rate": 9.987099117840969e-06,
+      "loss": 0.3953,
+      "step": 446
+    },
+    {
+      "epoch": 0.3629228687415426,
+      "grad_norm": 0.27677246928215027,
+      "learning_rate": 9.986757538333817e-06,
+      "loss": 0.412,
+      "step": 447
+    },
+    {
+      "epoch": 0.36373477672530447,
+      "grad_norm": 0.26836535334587097,
+      "learning_rate": 9.986411501676327e-06,
+      "loss": 0.3925,
+      "step": 448
+    },
+    {
+      "epoch": 0.3645466847090663,
+      "grad_norm": 0.2904950678348541,
+      "learning_rate": 9.986061008177779e-06,
+      "loss": 0.397,
+      "step": 449
+    },
+    {
+      "epoch": 0.36535859269282817,
+      "grad_norm": 0.30477994680404663,
+      "learning_rate": 9.985706058151446e-06,
+      "loss": 0.4431,
+      "step": 450
+    },
+    {
+      "epoch": 0.36617050067658996,
+      "grad_norm": 0.29113322496414185,
+      "learning_rate": 9.985346651914581e-06,
+      "loss": 0.4008,
+      "step": 451
+    },
+    {
+      "epoch": 0.3669824086603518,
+      "grad_norm": 0.2730692923069,
+      "learning_rate": 9.98498278978842e-06,
+      "loss": 0.3723,
+      "step": 452
+    },
+    {
+      "epoch": 0.36779431664411366,
+      "grad_norm": 0.2783091068267822,
+      "learning_rate": 9.984614472098181e-06,
+      "loss": 0.4072,
+      "step": 453
+    },
+    {
+      "epoch": 0.3686062246278755,
+      "grad_norm": 0.25962334871292114,
+      "learning_rate": 9.984241699173069e-06,
+      "loss": 0.4182,
+      "step": 454
+    },
+    {
+      "epoch": 0.36941813261163736,
+      "grad_norm": 0.28452181816101074,
+      "learning_rate": 9.983864471346263e-06,
+      "loss": 0.412,
+      "step": 455
+    },
+    {
+      "epoch": 0.3702300405953992,
+      "grad_norm": 0.31326985359191895,
+      "learning_rate": 9.983482788954931e-06,
+      "loss": 0.421,
+      "step": 456
+    },
+    {
+      "epoch": 0.371041948579161,
+      "grad_norm": 0.27150681614875793,
+      "learning_rate": 9.983096652340219e-06,
+      "loss": 0.3982,
+      "step": 457
+    },
+    {
+      "epoch": 0.37185385656292286,
+      "grad_norm": 0.2711666226387024,
+      "learning_rate": 9.982706061847254e-06,
+      "loss": 0.3858,
+      "step": 458
+    },
+    {
+      "epoch": 0.3726657645466847,
+      "grad_norm": 0.2907959222793579,
+      "learning_rate": 9.982311017825145e-06,
+      "loss": 0.4398,
+      "step": 459
+    },
+    {
+      "epoch": 0.37347767253044656,
+      "grad_norm": 0.3147335946559906,
+      "learning_rate": 9.981911520626983e-06,
+      "loss": 0.4147,
+      "step": 460
+    },
+    {
+      "epoch": 0.3742895805142084,
+      "grad_norm": 0.28171485662460327,
+      "learning_rate": 9.981507570609837e-06,
+      "loss": 0.4116,
+      "step": 461
+    },
+    {
+      "epoch": 0.3751014884979702,
+      "grad_norm": 0.28558969497680664,
+      "learning_rate": 9.981099168134753e-06,
+      "loss": 0.4077,
+      "step": 462
+    },
+    {
+      "epoch": 0.37591339648173205,
+      "grad_norm": 0.2836032807826996,
+      "learning_rate": 9.980686313566765e-06,
+      "loss": 0.4076,
+      "step": 463
+    },
+    {
+      "epoch": 0.3767253044654939,
+      "grad_norm": 0.26544904708862305,
+      "learning_rate": 9.980269007274878e-06,
+      "loss": 0.3974,
+      "step": 464
+    },
+    {
+      "epoch": 0.37753721244925575,
+      "grad_norm": 0.26577988266944885,
+      "learning_rate": 9.97984724963208e-06,
+      "loss": 0.3955,
+      "step": 465
+    },
+    {
+      "epoch": 0.3783491204330176,
+      "grad_norm": 0.3191268742084503,
+      "learning_rate": 9.979421041015336e-06,
+      "loss": 0.4159,
+      "step": 466
+    },
+    {
+      "epoch": 0.37916102841677946,
+      "grad_norm": 0.31662705540657043,
+      "learning_rate": 9.978990381805593e-06,
+      "loss": 0.4128,
+      "step": 467
+    },
+    {
+      "epoch": 0.37997293640054125,
+      "grad_norm": 0.29183229804039,
+      "learning_rate": 9.978555272387771e-06,
+      "loss": 0.4083,
+      "step": 468
+    },
+    {
+      "epoch": 0.3807848443843031,
+      "grad_norm": 0.27413347363471985,
+      "learning_rate": 9.978115713150768e-06,
+      "loss": 0.4257,
+      "step": 469
+    },
+    {
+      "epoch": 0.38159675236806495,
+      "grad_norm": 0.3053770959377289,
+      "learning_rate": 9.977671704487465e-06,
+      "loss": 0.3918,
+      "step": 470
+    },
+    {
+      "epoch": 0.3824086603518268,
+      "grad_norm": 0.2907295227050781,
+      "learning_rate": 9.977223246794712e-06,
+      "loss": 0.415,
+      "step": 471
+    },
+    {
+      "epoch": 0.38322056833558865,
+      "grad_norm": 0.28701066970825195,
+      "learning_rate": 9.97677034047334e-06,
+      "loss": 0.4151,
+      "step": 472
+    },
+    {
+      "epoch": 0.38403247631935045,
+      "grad_norm": 0.3182099461555481,
+      "learning_rate": 9.976312985928159e-06,
+      "loss": 0.418,
+      "step": 473
+    },
+    {
+      "epoch": 0.3848443843031123,
+      "grad_norm": 0.27537721395492554,
+      "learning_rate": 9.975851183567945e-06,
+      "loss": 0.3834,
+      "step": 474
+    },
+    {
+      "epoch": 0.38565629228687415,
+      "grad_norm": 0.2982698082923889,
+      "learning_rate": 9.975384933805461e-06,
+      "loss": 0.4248,
+      "step": 475
+    },
+    {
+      "epoch": 0.386468200270636,
+      "grad_norm": 0.32861489057540894,
+      "learning_rate": 9.974914237057435e-06,
+      "loss": 0.4273,
+      "step": 476
+    },
+    {
+      "epoch": 0.38728010825439785,
+      "grad_norm": 0.3048606514930725,
+      "learning_rate": 9.974439093744581e-06,
+      "loss": 0.4338,
+      "step": 477
+    },
+    {
+      "epoch": 0.3880920162381597,
+      "grad_norm": 0.30172622203826904,
+      "learning_rate": 9.973959504291574e-06,
+      "loss": 0.4018,
+      "step": 478
+    },
+    {
+      "epoch": 0.3889039242219215,
+      "grad_norm": 0.2939899265766144,
+      "learning_rate": 9.973475469127074e-06,
+      "loss": 0.3999,
+      "step": 479
+    },
+    {
+      "epoch": 0.38971583220568334,
+      "grad_norm": 0.3174692690372467,
+      "learning_rate": 9.972986988683711e-06,
+      "loss": 0.4237,
+      "step": 480
+    },
+    {
+      "epoch": 0.3905277401894452,
+      "grad_norm": 0.27286937832832336,
+      "learning_rate": 9.972494063398083e-06,
+      "loss": 0.3932,
+      "step": 481
+    },
+    {
+      "epoch": 0.39133964817320704,
+      "grad_norm": 0.32066410779953003,
+      "learning_rate": 9.971996693710768e-06,
+      "loss": 0.4129,
+      "step": 482
+    },
+    {
+      "epoch": 0.3921515561569689,
+      "grad_norm": 0.29693475365638733,
+      "learning_rate": 9.971494880066315e-06,
+      "loss": 0.4027,
+      "step": 483
+    },
+    {
+      "epoch": 0.39296346414073074,
+      "grad_norm": 0.29818621277809143,
+      "learning_rate": 9.97098862291324e-06,
+      "loss": 0.3846,
+      "step": 484
+    },
+    {
+      "epoch": 0.39377537212449254,
+      "grad_norm": 0.28604191541671753,
+      "learning_rate": 9.97047792270404e-06,
+      "loss": 0.4052,
+      "step": 485
+    },
+    {
+      "epoch": 0.3945872801082544,
+      "grad_norm": 0.2798047363758087,
+      "learning_rate": 9.969962779895172e-06,
+      "loss": 0.4192,
+      "step": 486
+    },
+    {
+      "epoch": 0.39539918809201624,
+      "grad_norm": 0.317839652299881,
+      "learning_rate": 9.96944319494707e-06,
+      "loss": 0.4025,
+      "step": 487
+    },
+    {
+      "epoch": 0.3962110960757781,
+      "grad_norm": 0.29114970564842224,
+      "learning_rate": 9.96891916832414e-06,
+      "loss": 0.4222,
+      "step": 488
+    },
+    {
+      "epoch": 0.39702300405953994,
+      "grad_norm": 0.3236350119113922,
+      "learning_rate": 9.968390700494755e-06,
+      "loss": 0.4165,
+      "step": 489
+    },
+    {
+      "epoch": 0.39783491204330174,
+      "grad_norm": 0.27579963207244873,
+      "learning_rate": 9.967857791931257e-06,
+      "loss": 0.408,
+      "step": 490
+    },
+    {
+      "epoch": 0.3986468200270636,
+      "grad_norm": 0.33093470335006714,
+      "learning_rate": 9.967320443109958e-06,
+      "loss": 0.4257,
+      "step": 491
+    },
+    {
+      "epoch": 0.39945872801082544,
+      "grad_norm": 0.2989307641983032,
+      "learning_rate": 9.966778654511143e-06,
+      "loss": 0.425,
+      "step": 492
+    },
+    {
+      "epoch": 0.4002706359945873,
+      "grad_norm": 0.2659866213798523,
+      "learning_rate": 9.966232426619054e-06,
+      "loss": 0.3853,
+      "step": 493
+    },
+    {
+      "epoch": 0.40108254397834914,
+      "grad_norm": 0.28991463780403137,
+      "learning_rate": 9.965681759921912e-06,
+      "loss": 0.3762,
+      "step": 494
+    },
+    {
+      "epoch": 0.401894451962111,
+      "grad_norm": 0.2772791087627411,
+      "learning_rate": 9.965126654911904e-06,
+      "loss": 0.3861,
+      "step": 495
+    },
+    {
+      "epoch": 0.4027063599458728,
+      "grad_norm": 0.2828364968299866,
+      "learning_rate": 9.964567112085175e-06,
+      "loss": 0.4157,
+      "step": 496
+    },
+    {
+      "epoch": 0.40351826792963463,
+      "grad_norm": 0.3041248917579651,
+      "learning_rate": 9.96400313194185e-06,
+      "loss": 0.3781,
+      "step": 497
+    },
+    {
+      "epoch": 0.4043301759133965,
+      "grad_norm": 0.2666463255882263,
+      "learning_rate": 9.963434714986006e-06,
+      "loss": 0.3887,
+      "step": 498
+    },
+    {
+      "epoch": 0.40514208389715833,
+      "grad_norm": 0.26858797669410706,
+      "learning_rate": 9.962861861725698e-06,
+      "loss": 0.416,
+      "step": 499
+    },
+    {
+      "epoch": 0.4059539918809202,
+      "grad_norm": 0.28757184743881226,
+      "learning_rate": 9.96228457267294e-06,
+      "loss": 0.3918,
+      "step": 500
+    },
+    {
+      "epoch": 0.406765899864682,
+      "grad_norm": 0.28327131271362305,
+      "learning_rate": 9.961702848343709e-06,
+      "loss": 0.4266,
+      "step": 501
+    },
+    {
+      "epoch": 0.40757780784844383,
+      "grad_norm": 0.26191478967666626,
+      "learning_rate": 9.961116689257949e-06,
+      "loss": 0.3994,
+      "step": 502
+    },
+    {
+      "epoch": 0.4083897158322057,
+      "grad_norm": 0.31872013211250305,
+      "learning_rate": 9.960526095939566e-06,
+      "loss": 0.4239,
+      "step": 503
+    },
+    {
+      "epoch": 0.40920162381596753,
+      "grad_norm": 0.27068060636520386,
+      "learning_rate": 9.959931068916436e-06,
+      "loss": 0.4114,
+      "step": 504
+    },
+    {
+      "epoch": 0.4100135317997294,
+      "grad_norm": 0.2793078124523163,
+      "learning_rate": 9.959331608720386e-06,
+      "loss": 0.3939,
+      "step": 505
+    },
+    {
+      "epoch": 0.41082543978349123,
+      "grad_norm": 0.2680758833885193,
+      "learning_rate": 9.958727715887218e-06,
+      "loss": 0.3996,
+      "step": 506
+    },
+    {
+      "epoch": 0.411637347767253,
+      "grad_norm": 0.28231486678123474,
+      "learning_rate": 9.958119390956685e-06,
+      "loss": 0.4031,
+      "step": 507
+    },
+    {
+      "epoch": 0.4124492557510149,
+      "grad_norm": 0.25295841693878174,
+      "learning_rate": 9.95750663447251e-06,
+      "loss": 0.409,
+      "step": 508
+    },
+    {
+      "epoch": 0.4132611637347767,
+      "grad_norm": 0.2809412479400635,
+      "learning_rate": 9.956889446982369e-06,
+      "loss": 0.4095,
+      "step": 509
+    },
+    {
+      "epoch": 0.4140730717185386,
+      "grad_norm": 0.27103403210639954,
+      "learning_rate": 9.956267829037906e-06,
+      "loss": 0.4235,
+      "step": 510
+    },
+    {
+      "epoch": 0.4148849797023004,
+      "grad_norm": 0.24615657329559326,
+      "learning_rate": 9.95564178119472e-06,
+      "loss": 0.3921,
+      "step": 511
+    },
+    {
+      "epoch": 0.4156968876860622,
+      "grad_norm": 0.3086927533149719,
+      "learning_rate": 9.95501130401237e-06,
+      "loss": 0.4174,
+      "step": 512
+    },
+    {
+      "epoch": 0.4165087956698241,
+      "grad_norm": 0.27201151847839355,
+      "learning_rate": 9.954376398054377e-06,
+      "loss": 0.4154,
+      "step": 513
+    },
+    {
+      "epoch": 0.4173207036535859,
+      "grad_norm": 0.29445183277130127,
+      "learning_rate": 9.953737063888216e-06,
+      "loss": 0.3909,
+      "step": 514
+    },
+    {
+      "epoch": 0.4181326116373478,
+      "grad_norm": 0.27061527967453003,
+      "learning_rate": 9.953093302085324e-06,
+      "loss": 0.4206,
+      "step": 515
+    },
+    {
+      "epoch": 0.4189445196211096,
+      "grad_norm": 0.28709256649017334,
+      "learning_rate": 9.952445113221093e-06,
+      "loss": 0.4211,
+      "step": 516
+    },
+    {
+      "epoch": 0.4197564276048715,
+      "grad_norm": 0.2507060468196869,
+      "learning_rate": 9.95179249787487e-06,
+      "loss": 0.3832,
+      "step": 517
+    },
+    {
+      "epoch": 0.42056833558863327,
+      "grad_norm": 0.262490838766098,
+      "learning_rate": 9.951135456629966e-06,
+      "loss": 0.3931,
+      "step": 518
+    },
+    {
+      "epoch": 0.4213802435723951,
+      "grad_norm": 0.25955870747566223,
+      "learning_rate": 9.950473990073637e-06,
+      "loss": 0.3788,
+      "step": 519
+    },
+    {
+      "epoch": 0.42219215155615697,
+      "grad_norm": 0.29438304901123047,
+      "learning_rate": 9.949808098797104e-06,
+      "loss": 0.382,
+      "step": 520
+    },
+    {
+      "epoch": 0.4230040595399188,
+      "grad_norm": 0.26844847202301025,
+      "learning_rate": 9.949137783395537e-06,
+      "loss": 0.4133,
+      "step": 521
+    },
+    {
+      "epoch": 0.42381596752368067,
+      "grad_norm": 0.2878880202770233,
+      "learning_rate": 9.948463044468063e-06,
+      "loss": 0.4059,
+      "step": 522
+    },
+    {
+      "epoch": 0.42462787550744246,
+      "grad_norm": 0.24729745090007782,
+      "learning_rate": 9.94778388261776e-06,
+      "loss": 0.4328,
+      "step": 523
+    },
+    {
+      "epoch": 0.4254397834912043,
+      "grad_norm": 0.2775348424911499,
+      "learning_rate": 9.947100298451663e-06,
+      "loss": 0.3883,
+      "step": 524
+    },
+    {
+      "epoch": 0.42625169147496617,
+      "grad_norm": 0.32374122738838196,
+      "learning_rate": 9.946412292580757e-06,
+      "loss": 0.4223,
+      "step": 525
+    },
+    {
+      "epoch": 0.427063599458728,
+      "grad_norm": 0.26501405239105225,
+      "learning_rate": 9.94571986561998e-06,
+      "loss": 0.4195,
+      "step": 526
+    },
+    {
+      "epoch": 0.42787550744248987,
+      "grad_norm": 0.2845929265022278,
+      "learning_rate": 9.945023018188222e-06,
+      "loss": 0.4041,
+      "step": 527
+    },
+    {
+      "epoch": 0.4286874154262517,
+      "grad_norm": 0.28472813963890076,
+      "learning_rate": 9.944321750908321e-06,
+      "loss": 0.4088,
+      "step": 528
+    },
+    {
+      "epoch": 0.4294993234100135,
+      "grad_norm": 0.2770460844039917,
+      "learning_rate": 9.94361606440707e-06,
+      "loss": 0.392,
+      "step": 529
+    },
+    {
+      "epoch": 0.43031123139377536,
+      "grad_norm": 0.30595073103904724,
+      "learning_rate": 9.94290595931521e-06,
+      "loss": 0.4001,
+      "step": 530
+    },
+    {
+      "epoch": 0.4311231393775372,
+      "grad_norm": 0.25412067770957947,
+      "learning_rate": 9.942191436267428e-06,
+      "loss": 0.4095,
+      "step": 531
+    },
+    {
+      "epoch": 0.43193504736129906,
+      "grad_norm": 0.28710073232650757,
+      "learning_rate": 9.941472495902366e-06,
+      "loss": 0.409,
+      "step": 532
+    },
+    {
+      "epoch": 0.4327469553450609,
+      "grad_norm": 0.28940242528915405,
+      "learning_rate": 9.940749138862609e-06,
+      "loss": 0.3769,
+      "step": 533
+    },
+    {
+      "epoch": 0.4335588633288227,
+      "grad_norm": 0.2601233124732971,
+      "learning_rate": 9.940021365794692e-06,
+      "loss": 0.3814,
+      "step": 534
+    },
+    {
+      "epoch": 0.43437077131258456,
+      "grad_norm": 0.26906105875968933,
+      "learning_rate": 9.939289177349097e-06,
+      "loss": 0.3864,
+      "step": 535
+    },
+    {
+      "epoch": 0.4351826792963464,
+      "grad_norm": 0.26609525084495544,
+      "learning_rate": 9.938552574180252e-06,
+      "loss": 0.3982,
+      "step": 536
+    },
+    {
+      "epoch": 0.43599458728010826,
+      "grad_norm": 0.30542659759521484,
+      "learning_rate": 9.937811556946531e-06,
+      "loss": 0.4167,
+      "step": 537
+    },
+    {
+      "epoch": 0.4368064952638701,
+      "grad_norm": 0.3012215793132782,
+      "learning_rate": 9.937066126310253e-06,
+      "loss": 0.4161,
+      "step": 538
+    },
+    {
+      "epoch": 0.43761840324763196,
+      "grad_norm": 0.25374212861061096,
+      "learning_rate": 9.936316282937682e-06,
+      "loss": 0.4109,
+      "step": 539
+    },
+    {
+      "epoch": 0.43843031123139375,
+      "grad_norm": 0.30464813113212585,
+      "learning_rate": 9.935562027499026e-06,
+      "loss": 0.39,
+      "step": 540
+    },
+    {
+      "epoch": 0.4392422192151556,
+      "grad_norm": 0.27063754200935364,
+      "learning_rate": 9.934803360668435e-06,
+      "loss": 0.3923,
+      "step": 541
+    },
+    {
+      "epoch": 0.44005412719891746,
+      "grad_norm": 0.30924689769744873,
+      "learning_rate": 9.934040283124006e-06,
+      "loss": 0.4031,
+      "step": 542
+    },
+    {
+      "epoch": 0.4408660351826793,
+      "grad_norm": 0.23840641975402832,
+      "learning_rate": 9.933272795547773e-06,
+      "loss": 0.3999,
+      "step": 543
+    },
+    {
+      "epoch": 0.44167794316644116,
+      "grad_norm": 0.2702539265155792,
+      "learning_rate": 9.932500898625716e-06,
+      "loss": 0.3869,
+      "step": 544
+    },
+    {
+      "epoch": 0.44248985115020295,
+      "grad_norm": 0.27259889245033264,
+      "learning_rate": 9.931724593047754e-06,
+      "loss": 0.4093,
+      "step": 545
+    },
+    {
+      "epoch": 0.4433017591339648,
+      "grad_norm": 0.28210800886154175,
+      "learning_rate": 9.930943879507748e-06,
+      "loss": 0.3946,
+      "step": 546
+    },
+    {
+      "epoch": 0.44411366711772665,
+      "grad_norm": 0.27400898933410645,
+      "learning_rate": 9.930158758703495e-06,
+      "loss": 0.3932,
+      "step": 547
+    },
+    {
+      "epoch": 0.4449255751014885,
+      "grad_norm": 0.27335163950920105,
+      "learning_rate": 9.929369231336735e-06,
+      "loss": 0.3941,
+      "step": 548
+    },
+    {
+      "epoch": 0.44573748308525035,
+      "grad_norm": 0.3006235659122467,
+      "learning_rate": 9.928575298113146e-06,
+      "loss": 0.4124,
+      "step": 549
+    },
+    {
+      "epoch": 0.4465493910690122,
+      "grad_norm": 0.2245517075061798,
+      "learning_rate": 9.927776959742344e-06,
+      "loss": 0.3862,
+      "step": 550
+    },
+    {
+      "epoch": 0.447361299052774,
+      "grad_norm": 0.34785765409469604,
+      "learning_rate": 9.92697421693788e-06,
+      "loss": 0.4099,
+      "step": 551
+    },
+    {
+      "epoch": 0.44817320703653585,
+      "grad_norm": 0.27095261216163635,
+      "learning_rate": 9.926167070417243e-06,
+      "loss": 0.4141,
+      "step": 552
+    },
+    {
+      "epoch": 0.4489851150202977,
+      "grad_norm": 0.2670780122280121,
+      "learning_rate": 9.925355520901861e-06,
+      "loss": 0.382,
+      "step": 553
+    },
+    {
+      "epoch": 0.44979702300405955,
+      "grad_norm": 0.2916838526725769,
+      "learning_rate": 9.924539569117092e-06,
+      "loss": 0.4226,
+      "step": 554
+    },
+    {
+      "epoch": 0.4506089309878214,
+      "grad_norm": 0.2687590420246124,
+      "learning_rate": 9.923719215792233e-06,
+      "loss": 0.4098,
+      "step": 555
+    },
+    {
+      "epoch": 0.4514208389715832,
+      "grad_norm": 0.2712917625904083,
+      "learning_rate": 9.922894461660513e-06,
+      "loss": 0.4157,
+      "step": 556
+    },
+    {
+      "epoch": 0.45223274695534504,
+      "grad_norm": 0.2955559492111206,
+      "learning_rate": 9.922065307459096e-06,
+      "loss": 0.403,
+      "step": 557
+    },
+    {
+      "epoch": 0.4530446549391069,
+      "grad_norm": 0.2563600242137909,
+      "learning_rate": 9.921231753929077e-06,
+      "loss": 0.3903,
+      "step": 558
+    },
+    {
+      "epoch": 0.45385656292286874,
+      "grad_norm": 0.2697726786136627,
+      "learning_rate": 9.920393801815482e-06,
+      "loss": 0.4217,
+      "step": 559
+    },
+    {
+      "epoch": 0.4546684709066306,
+      "grad_norm": 0.2866057753562927,
+      "learning_rate": 9.919551451867275e-06,
+      "loss": 0.4237,
+      "step": 560
+    },
+    {
+      "epoch": 0.45548037889039245,
+      "grad_norm": 0.2483537644147873,
+      "learning_rate": 9.91870470483734e-06,
+      "loss": 0.3944,
+      "step": 561
+    },
+    {
+      "epoch": 0.45629228687415424,
+      "grad_norm": 0.32340553402900696,
+      "learning_rate": 9.917853561482503e-06,
+      "loss": 0.4142,
+      "step": 562
+    },
+    {
+      "epoch": 0.4571041948579161,
+      "grad_norm": 0.3224326968193054,
+      "learning_rate": 9.91699802256351e-06,
+      "loss": 0.3965,
+      "step": 563
+    },
+    {
+      "epoch": 0.45791610284167794,
+      "grad_norm": 0.291079580783844,
+      "learning_rate": 9.916138088845042e-06,
+      "loss": 0.401,
+      "step": 564
+    },
+    {
+      "epoch": 0.4587280108254398,
+      "grad_norm": 0.25762370228767395,
+      "learning_rate": 9.915273761095701e-06,
+      "loss": 0.3966,
+      "step": 565
+    },
+    {
+      "epoch": 0.45953991880920164,
+      "grad_norm": 0.26920315623283386,
+      "learning_rate": 9.914405040088026e-06,
+      "loss": 0.37,
+      "step": 566
+    },
+    {
+      "epoch": 0.46035182679296344,
+      "grad_norm": 0.27145928144454956,
+      "learning_rate": 9.913531926598475e-06,
+      "loss": 0.3925,
+      "step": 567
+    },
+    {
+      "epoch": 0.4611637347767253,
+      "grad_norm": 0.25682514905929565,
+      "learning_rate": 9.912654421407434e-06,
+      "loss": 0.4172,
+      "step": 568
+    },
+    {
+      "epoch": 0.46197564276048714,
+      "grad_norm": 0.2607958912849426,
+      "learning_rate": 9.911772525299214e-06,
+      "loss": 0.3913,
+      "step": 569
+    },
+    {
+      "epoch": 0.462787550744249,
+      "grad_norm": 0.2675032913684845,
+      "learning_rate": 9.910886239062054e-06,
+      "loss": 0.4153,
+      "step": 570
+    },
+    {
+      "epoch": 0.46359945872801084,
+      "grad_norm": 0.2699697017669678,
+      "learning_rate": 9.909995563488116e-06,
+      "loss": 0.3892,
+      "step": 571
+    },
+    {
+      "epoch": 0.4644113667117727,
+      "grad_norm": 0.2583911716938019,
+      "learning_rate": 9.909100499373476e-06,
+      "loss": 0.402,
+      "step": 572
+    },
+    {
+      "epoch": 0.4652232746955345,
+      "grad_norm": 0.2585200369358063,
+      "learning_rate": 9.908201047518146e-06,
+      "loss": 0.4088,
+      "step": 573
+    },
+    {
+      "epoch": 0.46603518267929633,
+      "grad_norm": 0.2860269248485565,
+      "learning_rate": 9.907297208726051e-06,
+      "loss": 0.3828,
+      "step": 574
+    },
+    {
+      "epoch": 0.4668470906630582,
+      "grad_norm": 0.27281343936920166,
+      "learning_rate": 9.90638898380504e-06,
+      "loss": 0.4207,
+      "step": 575
+    },
+    {
+      "epoch": 0.46765899864682003,
+      "grad_norm": 0.28957298398017883,
+      "learning_rate": 9.905476373566884e-06,
+      "loss": 0.4366,
+      "step": 576
+    },
+    {
+      "epoch": 0.4684709066305819,
+      "grad_norm": 0.26523536443710327,
+      "learning_rate": 9.90455937882727e-06,
+      "loss": 0.3879,
+      "step": 577
+    },
+    {
+      "epoch": 0.46928281461434374,
+      "grad_norm": 0.3027326166629791,
+      "learning_rate": 9.903638000405805e-06,
+      "loss": 0.3939,
+      "step": 578
+    },
+    {
+      "epoch": 0.47009472259810553,
+      "grad_norm": 0.24961911141872406,
+      "learning_rate": 9.902712239126016e-06,
+      "loss": 0.4094,
+      "step": 579
+    },
+    {
+      "epoch": 0.4709066305818674,
+      "grad_norm": 0.27050891518592834,
+      "learning_rate": 9.901782095815343e-06,
+      "loss": 0.4112,
+      "step": 580
+    },
+    {
+      "epoch": 0.47171853856562923,
+      "grad_norm": 0.2851492464542389,
+      "learning_rate": 9.900847571305148e-06,
+      "loss": 0.3887,
+      "step": 581
+    },
+    {
+      "epoch": 0.4725304465493911,
+      "grad_norm": 0.26577696204185486,
+      "learning_rate": 9.899908666430707e-06,
+      "loss": 0.3589,
+      "step": 582
+    },
+    {
+      "epoch": 0.47334235453315293,
+      "grad_norm": 0.26485592126846313,
+      "learning_rate": 9.898965382031208e-06,
+      "loss": 0.3911,
+      "step": 583
+    },
+    {
+      "epoch": 0.4741542625169147,
+      "grad_norm": 0.2588464617729187,
+      "learning_rate": 9.898017718949759e-06,
+      "loss": 0.4188,
+      "step": 584
+    },
+    {
+      "epoch": 0.4749661705006766,
+      "grad_norm": 0.3090347945690155,
+      "learning_rate": 9.897065678033376e-06,
+      "loss": 0.4027,
+      "step": 585
+    },
+    {
+      "epoch": 0.4757780784844384,
+      "grad_norm": 0.27240392565727234,
+      "learning_rate": 9.896109260132993e-06,
+      "loss": 0.3725,
+      "step": 586
+    },
+    {
+      "epoch": 0.4765899864682003,
+      "grad_norm": 0.2436683475971222,
+      "learning_rate": 9.895148466103451e-06,
+      "loss": 0.3583,
+      "step": 587
+    },
+    {
+      "epoch": 0.4774018944519621,
+      "grad_norm": 0.26293447613716125,
+      "learning_rate": 9.894183296803509e-06,
+      "loss": 0.3727,
+      "step": 588
+    },
+    {
+      "epoch": 0.478213802435724,
+      "grad_norm": 0.2522885501384735,
+      "learning_rate": 9.893213753095829e-06,
+      "loss": 0.3862,
+      "step": 589
+    },
+    {
+      "epoch": 0.4790257104194858,
+      "grad_norm": 0.2749432325363159,
+      "learning_rate": 9.892239835846988e-06,
+      "loss": 0.3836,
+      "step": 590
+    },
+    {
+      "epoch": 0.4798376184032476,
+      "grad_norm": 0.24647930264472961,
+      "learning_rate": 9.891261545927469e-06,
+      "loss": 0.4028,
+      "step": 591
+    },
+    {
+      "epoch": 0.4806495263870095,
+      "grad_norm": 0.25447267293930054,
+      "learning_rate": 9.890278884211668e-06,
+      "loss": 0.4033,
+      "step": 592
+    },
+    {
+      "epoch": 0.4814614343707713,
+      "grad_norm": 0.27552664279937744,
+      "learning_rate": 9.889291851577885e-06,
+      "loss": 0.3996,
+      "step": 593
+    },
+    {
+      "epoch": 0.4822733423545332,
+      "grad_norm": 0.25885629653930664,
+      "learning_rate": 9.888300448908324e-06,
+      "loss": 0.4039,
+      "step": 594
+    },
+    {
+      "epoch": 0.48308525033829497,
+      "grad_norm": 0.2651670575141907,
+      "learning_rate": 9.8873046770891e-06,
+      "loss": 0.3696,
+      "step": 595
+    },
+    {
+      "epoch": 0.4838971583220568,
+      "grad_norm": 0.26470398902893066,
+      "learning_rate": 9.88630453701023e-06,
+      "loss": 0.4075,
+      "step": 596
+    },
+    {
+      "epoch": 0.48470906630581867,
+      "grad_norm": 0.2844875454902649,
+      "learning_rate": 9.88530002956564e-06,
+      "loss": 0.4067,
+      "step": 597
+    },
+    {
+      "epoch": 0.4855209742895805,
+      "grad_norm": 0.28087031841278076,
+      "learning_rate": 9.884291155653151e-06,
+      "loss": 0.4109,
+      "step": 598
+    },
+    {
+      "epoch": 0.48633288227334237,
+      "grad_norm": 0.2842080593109131,
+      "learning_rate": 9.883277916174496e-06,
+      "loss": 0.3833,
+      "step": 599
+    },
+    {
+      "epoch": 0.4871447902571042,
+      "grad_norm": 0.2502003312110901,
+      "learning_rate": 9.882260312035304e-06,
+      "loss": 0.3997,
+      "step": 600
+    },
+    {
+      "epoch": 0.487956698240866,
+      "grad_norm": 0.2580205202102661,
+      "learning_rate": 9.881238344145106e-06,
+      "loss": 0.4258,
+      "step": 601
+    },
+    {
+      "epoch": 0.48876860622462787,
+      "grad_norm": 0.24465477466583252,
+      "learning_rate": 9.880212013417334e-06,
+      "loss": 0.3997,
+      "step": 602
+    },
+    {
+      "epoch": 0.4895805142083897,
+      "grad_norm": 0.28527599573135376,
+      "learning_rate": 9.879181320769322e-06,
+      "loss": 0.3931,
+      "step": 603
+    },
+    {
+      "epoch": 0.49039242219215157,
+      "grad_norm": 0.28538769483566284,
+      "learning_rate": 9.878146267122298e-06,
+      "loss": 0.4103,
+      "step": 604
+    },
+    {
+      "epoch": 0.4912043301759134,
+      "grad_norm": 0.2732044458389282,
+      "learning_rate": 9.877106853401392e-06,
+      "loss": 0.3919,
+      "step": 605
+    },
+    {
+      "epoch": 0.4920162381596752,
+      "grad_norm": 0.2676577568054199,
+      "learning_rate": 9.876063080535627e-06,
+      "loss": 0.4025,
+      "step": 606
+    },
+    {
+      "epoch": 0.49282814614343706,
+      "grad_norm": 0.2781953811645508,
+      "learning_rate": 9.875014949457926e-06,
+      "loss": 0.3817,
+      "step": 607
+    },
+    {
+      "epoch": 0.4936400541271989,
+      "grad_norm": 0.2806389331817627,
+      "learning_rate": 9.873962461105104e-06,
+      "loss": 0.3988,
+      "step": 608
+    },
+    {
+      "epoch": 0.49445196211096076,
+      "grad_norm": 0.24419797956943512,
+      "learning_rate": 9.872905616417875e-06,
+      "loss": 0.4165,
+      "step": 609
+    },
+    {
+      "epoch": 0.4952638700947226,
+      "grad_norm": 0.2636622488498688,
+      "learning_rate": 9.87184441634084e-06,
+      "loss": 0.3967,
+      "step": 610
+    },
+    {
+      "epoch": 0.49607577807848446,
+      "grad_norm": 0.2683873474597931,
+      "learning_rate": 9.870778861822502e-06,
+      "loss": 0.3977,
+      "step": 611
+    },
+    {
+      "epoch": 0.49688768606224626,
+      "grad_norm": 0.27110394835472107,
+      "learning_rate": 9.869708953815247e-06,
+      "loss": 0.4009,
+      "step": 612
+    },
+    {
+      "epoch": 0.4976995940460081,
+      "grad_norm": 0.2694375514984131,
+      "learning_rate": 9.868634693275356e-06,
+      "loss": 0.3946,
+      "step": 613
+    },
+    {
+      "epoch": 0.49851150202976996,
+      "grad_norm": 0.2592358887195587,
+      "learning_rate": 9.867556081163002e-06,
+      "loss": 0.366,
+      "step": 614
+    },
+    {
+      "epoch": 0.4993234100135318,
+      "grad_norm": 0.26441365480422974,
+      "learning_rate": 9.86647311844224e-06,
+      "loss": 0.4183,
+      "step": 615
+    },
+    {
+      "epoch": 0.5001353179972936,
+      "grad_norm": 0.268669068813324,
+      "learning_rate": 9.865385806081028e-06,
+      "loss": 0.3871,
+      "step": 616
+    },
+    {
+      "epoch": 0.5009472259810555,
+      "grad_norm": 0.2881656587123871,
+      "learning_rate": 9.864294145051197e-06,
+      "loss": 0.4053,
+      "step": 617
+    },
+    {
+      "epoch": 0.5017591339648173,
+      "grad_norm": 0.2792668640613556,
+      "learning_rate": 9.863198136328474e-06,
+      "loss": 0.4104,
+      "step": 618
+    },
+    {
+      "epoch": 0.5025710419485792,
+      "grad_norm": 0.25814157724380493,
+      "learning_rate": 9.862097780892463e-06,
+      "loss": 0.4084,
+      "step": 619
+    },
+    {
+      "epoch": 0.503382949932341,
+      "grad_norm": 0.2680467665195465,
+      "learning_rate": 9.860993079726665e-06,
+      "loss": 0.4078,
+      "step": 620
+    },
+    {
+      "epoch": 0.5041948579161029,
+      "grad_norm": 0.2779659330844879,
+      "learning_rate": 9.859884033818454e-06,
+      "loss": 0.4079,
+      "step": 621
+    },
+    {
+      "epoch": 0.5050067658998647,
+      "grad_norm": 0.276141881942749,
+      "learning_rate": 9.858770644159098e-06,
+      "loss": 0.3943,
+      "step": 622
+    },
+    {
+      "epoch": 0.5058186738836266,
+      "grad_norm": 0.2546170949935913,
+      "learning_rate": 9.857652911743734e-06,
+      "loss": 0.3873,
+      "step": 623
+    },
+    {
+      "epoch": 0.5066305818673884,
+      "grad_norm": 0.2978194057941437,
+      "learning_rate": 9.856530837571394e-06,
+      "loss": 0.3751,
+      "step": 624
+    },
+    {
+      "epoch": 0.5074424898511503,
+      "grad_norm": 0.283349871635437,
+      "learning_rate": 9.855404422644983e-06,
+      "loss": 0.4054,
+      "step": 625
+    },
+    {
+      "epoch": 0.508254397834912,
+      "grad_norm": 0.24305357038974762,
+      "learning_rate": 9.85427366797129e-06,
+      "loss": 0.3748,
+      "step": 626
+    },
+    {
+      "epoch": 0.5090663058186738,
+      "grad_norm": 0.26549088954925537,
+      "learning_rate": 9.853138574560976e-06,
+      "loss": 0.4122,
+      "step": 627
+    },
+    {
+      "epoch": 0.5098782138024357,
+      "grad_norm": 0.26904577016830444,
+      "learning_rate": 9.851999143428587e-06,
+      "loss": 0.3838,
+      "step": 628
+    },
+    {
+      "epoch": 0.5106901217861975,
+      "grad_norm": 0.26965466141700745,
+      "learning_rate": 9.850855375592543e-06,
+      "loss": 0.4271,
+      "step": 629
+    },
+    {
+      "epoch": 0.5115020297699594,
+      "grad_norm": 0.2440609335899353,
+      "learning_rate": 9.84970727207514e-06,
+      "loss": 0.4017,
+      "step": 630
+    },
+    {
+      "epoch": 0.5123139377537212,
+      "grad_norm": 0.24865713715553284,
+      "learning_rate": 9.848554833902551e-06,
+      "loss": 0.4031,
+      "step": 631
+    },
+    {
+      "epoch": 0.5131258457374831,
+      "grad_norm": 0.3070017695426941,
+      "learning_rate": 9.847398062104823e-06,
+      "loss": 0.3922,
+      "step": 632
+    },
+    {
+      "epoch": 0.513937753721245,
+      "grad_norm": 0.2637052536010742,
+      "learning_rate": 9.846236957715872e-06,
+      "loss": 0.4043,
+      "step": 633
+    },
+    {
+      "epoch": 0.5147496617050068,
+      "grad_norm": 0.27455443143844604,
+      "learning_rate": 9.845071521773492e-06,
+      "loss": 0.3766,
+      "step": 634
+    },
+    {
+      "epoch": 0.5155615696887687,
+      "grad_norm": 0.31241312623023987,
+      "learning_rate": 9.843901755319347e-06,
+      "loss": 0.4103,
+      "step": 635
+    },
+    {
+      "epoch": 0.5163734776725305,
+      "grad_norm": 0.32152727246284485,
+      "learning_rate": 9.842727659398971e-06,
+      "loss": 0.4227,
+      "step": 636
+    },
+    {
+      "epoch": 0.5171853856562922,
+      "grad_norm": 0.27412205934524536,
+      "learning_rate": 9.841549235061766e-06,
+      "loss": 0.3907,
+      "step": 637
+    },
+    {
+      "epoch": 0.5179972936400541,
+      "grad_norm": 0.32343533635139465,
+      "learning_rate": 9.840366483361008e-06,
+      "loss": 0.3894,
+      "step": 638
+    },
+    {
+      "epoch": 0.5188092016238159,
+      "grad_norm": 0.2638799548149109,
+      "learning_rate": 9.839179405353834e-06,
+      "loss": 0.4026,
+      "step": 639
+    },
+    {
+      "epoch": 0.5196211096075778,
+      "grad_norm": 0.29649946093559265,
+      "learning_rate": 9.837988002101253e-06,
+      "loss": 0.4055,
+      "step": 640
+    },
+    {
+      "epoch": 0.5204330175913396,
+      "grad_norm": 0.23794426023960114,
+      "learning_rate": 9.836792274668135e-06,
+      "loss": 0.4077,
+      "step": 641
+    },
+    {
+      "epoch": 0.5212449255751015,
+      "grad_norm": 0.2977282404899597,
+      "learning_rate": 9.835592224123221e-06,
+      "loss": 0.3992,
+      "step": 642
+    },
+    {
+      "epoch": 0.5220568335588633,
+      "grad_norm": 0.2726401388645172,
+      "learning_rate": 9.834387851539113e-06,
+      "loss": 0.3981,
+      "step": 643
+    },
+    {
+      "epoch": 0.5228687415426252,
+      "grad_norm": 0.29016032814979553,
+      "learning_rate": 9.833179157992274e-06,
+      "loss": 0.3991,
+      "step": 644
+    },
+    {
+      "epoch": 0.523680649526387,
+      "grad_norm": 0.288035124540329,
+      "learning_rate": 9.831966144563032e-06,
+      "loss": 0.381,
+      "step": 645
+    },
+    {
+      "epoch": 0.5244925575101489,
+      "grad_norm": 0.2654731273651123,
+      "learning_rate": 9.830748812335576e-06,
+      "loss": 0.3887,
+      "step": 646
+    },
+    {
+      "epoch": 0.5253044654939107,
+      "grad_norm": 0.2857693135738373,
+      "learning_rate": 9.829527162397951e-06,
+      "loss": 0.4192,
+      "step": 647
+    },
+    {
+      "epoch": 0.5261163734776725,
+      "grad_norm": 0.2955917716026306,
+      "learning_rate": 9.828301195842069e-06,
+      "loss": 0.4462,
+      "step": 648
+    },
+    {
+      "epoch": 0.5269282814614343,
+      "grad_norm": 0.2845597565174103,
+      "learning_rate": 9.827070913763693e-06,
+      "loss": 0.4008,
+      "step": 649
+    },
+    {
+      "epoch": 0.5277401894451962,
+      "grad_norm": 0.2988428473472595,
+      "learning_rate": 9.825836317262445e-06,
+      "loss": 0.3904,
+      "step": 650
+    },
+    {
+      "epoch": 0.528552097428958,
+      "grad_norm": 0.28510186076164246,
+      "learning_rate": 9.824597407441805e-06,
+      "loss": 0.3773,
+      "step": 651
+    },
+    {
+      "epoch": 0.5293640054127199,
+      "grad_norm": 0.2716323137283325,
+      "learning_rate": 9.823354185409108e-06,
+      "loss": 0.4011,
+      "step": 652
+    },
+    {
+      "epoch": 0.5301759133964817,
+      "grad_norm": 0.3209831118583679,
+      "learning_rate": 9.822106652275541e-06,
+      "loss": 0.4048,
+      "step": 653
+    },
+    {
+      "epoch": 0.5309878213802436,
+      "grad_norm": 0.24772131443023682,
+      "learning_rate": 9.820854809156146e-06,
+      "loss": 0.3887,
+      "step": 654
+    },
+    {
+      "epoch": 0.5317997293640054,
+      "grad_norm": 0.2859378457069397,
+      "learning_rate": 9.819598657169819e-06,
+      "loss": 0.3964,
+      "step": 655
+    },
+    {
+      "epoch": 0.5326116373477673,
+      "grad_norm": 0.31785473227500916,
+      "learning_rate": 9.818338197439304e-06,
+      "loss": 0.4002,
+      "step": 656
+    },
+    {
+      "epoch": 0.5334235453315291,
+      "grad_norm": 0.269234836101532,
+      "learning_rate": 9.817073431091194e-06,
+      "loss": 0.3797,
+      "step": 657
+    },
+    {
+      "epoch": 0.534235453315291,
+      "grad_norm": 0.3626851737499237,
+      "learning_rate": 9.815804359255937e-06,
+      "loss": 0.3721,
+      "step": 658
+    },
+    {
+      "epoch": 0.5350473612990527,
+      "grad_norm": 0.3065357506275177,
+      "learning_rate": 9.814530983067825e-06,
+      "loss": 0.4134,
+      "step": 659
+    },
+    {
+      "epoch": 0.5358592692828146,
+      "grad_norm": 0.29417091608047485,
+      "learning_rate": 9.813253303664997e-06,
+      "loss": 0.4118,
+      "step": 660
+    },
+    {
+      "epoch": 0.5366711772665764,
+      "grad_norm": 0.3026554584503174,
+      "learning_rate": 9.811971322189442e-06,
+      "loss": 0.4165,
+      "step": 661
+    },
+    {
+      "epoch": 0.5374830852503383,
+      "grad_norm": 0.262397825717926,
+      "learning_rate": 9.810685039786989e-06,
+      "loss": 0.4027,
+      "step": 662
+    },
+    {
+      "epoch": 0.5382949932341001,
+      "grad_norm": 0.3032433092594147,
+      "learning_rate": 9.809394457607315e-06,
+      "loss": 0.3833,
+      "step": 663
+    },
+    {
+      "epoch": 0.539106901217862,
+      "grad_norm": 0.2987179160118103,
+      "learning_rate": 9.808099576803937e-06,
+      "loss": 0.4077,
+      "step": 664
+    },
+    {
+      "epoch": 0.5399188092016238,
+      "grad_norm": 0.2752896547317505,
+      "learning_rate": 9.80680039853422e-06,
+      "loss": 0.3979,
+      "step": 665
+    },
+    {
+      "epoch": 0.5407307171853857,
+      "grad_norm": 0.22436368465423584,
+      "learning_rate": 9.805496923959363e-06,
+      "loss": 0.4019,
+      "step": 666
+    },
+    {
+      "epoch": 0.5415426251691475,
+      "grad_norm": 0.28384217619895935,
+      "learning_rate": 9.804189154244408e-06,
+      "loss": 0.3839,
+      "step": 667
+    },
+    {
+      "epoch": 0.5423545331529094,
+      "grad_norm": 0.2612059712409973,
+      "learning_rate": 9.802877090558237e-06,
+      "loss": 0.4092,
+      "step": 668
+    },
+    {
+      "epoch": 0.5431664411366712,
+      "grad_norm": 0.26217183470726013,
+      "learning_rate": 9.801560734073567e-06,
+      "loss": 0.3755,
+      "step": 669
+    },
+    {
+      "epoch": 0.543978349120433,
+      "grad_norm": 0.2505117356777191,
+      "learning_rate": 9.800240085966957e-06,
+      "loss": 0.3951,
+      "step": 670
+    },
+    {
+      "epoch": 0.5447902571041948,
+      "grad_norm": 0.24618709087371826,
+      "learning_rate": 9.798915147418795e-06,
+      "loss": 0.392,
+      "step": 671
+    },
+    {
+      "epoch": 0.5456021650879567,
+      "grad_norm": 0.3006080090999603,
+      "learning_rate": 9.797585919613311e-06,
+      "loss": 0.3849,
+      "step": 672
+    },
+    {
+      "epoch": 0.5464140730717185,
+      "grad_norm": 0.261544793844223,
+      "learning_rate": 9.796252403738564e-06,
+      "loss": 0.4105,
+      "step": 673
+    },
+    {
+      "epoch": 0.5472259810554804,
+      "grad_norm": 0.2564762830734253,
+      "learning_rate": 9.794914600986446e-06,
+      "loss": 0.3872,
+      "step": 674
+    },
+    {
+      "epoch": 0.5480378890392422,
+      "grad_norm": 0.260647177696228,
+      "learning_rate": 9.793572512552683e-06,
+      "loss": 0.4029,
+      "step": 675
+    },
+    {
+      "epoch": 0.5488497970230041,
+      "grad_norm": 0.28803110122680664,
+      "learning_rate": 9.792226139636827e-06,
+      "loss": 0.4124,
+      "step": 676
+    },
+    {
+      "epoch": 0.5496617050067659,
+      "grad_norm": 0.23791493475437164,
+      "learning_rate": 9.790875483442265e-06,
+      "loss": 0.3986,
+      "step": 677
+    },
+    {
+      "epoch": 0.5504736129905278,
+      "grad_norm": 0.31256410479545593,
+      "learning_rate": 9.78952054517621e-06,
+      "loss": 0.4082,
+      "step": 678
+    },
+    {
+      "epoch": 0.5512855209742896,
+      "grad_norm": 0.28937309980392456,
+      "learning_rate": 9.788161326049702e-06,
+      "loss": 0.4045,
+      "step": 679
+    },
+    {
+      "epoch": 0.5520974289580515,
+      "grad_norm": 0.25063779950141907,
+      "learning_rate": 9.786797827277607e-06,
+      "loss": 0.4103,
+      "step": 680
+    },
+    {
+      "epoch": 0.5529093369418132,
+      "grad_norm": 0.2724301517009735,
+      "learning_rate": 9.785430050078616e-06,
+      "loss": 0.4028,
+      "step": 681
+    },
+    {
+      "epoch": 0.5537212449255751,
+      "grad_norm": 0.2946489155292511,
+      "learning_rate": 9.784057995675245e-06,
+      "loss": 0.4215,
+      "step": 682
+    },
+    {
+      "epoch": 0.5545331529093369,
+      "grad_norm": 0.27117615938186646,
+      "learning_rate": 9.782681665293831e-06,
+      "loss": 0.413,
+      "step": 683
+    },
+    {
+      "epoch": 0.5553450608930988,
+      "grad_norm": 0.27362489700317383,
+      "learning_rate": 9.781301060164537e-06,
+      "loss": 0.4061,
+      "step": 684
+    },
+    {
+      "epoch": 0.5561569688768606,
+      "grad_norm": 0.28726473450660706,
+      "learning_rate": 9.779916181521342e-06,
+      "loss": 0.4014,
+      "step": 685
+    },
+    {
+      "epoch": 0.5569688768606225,
+      "grad_norm": 0.2812362313270569,
+      "learning_rate": 9.778527030602049e-06,
+      "loss": 0.3923,
+      "step": 686
+    },
+    {
+      "epoch": 0.5577807848443843,
+      "grad_norm": 0.26063573360443115,
+      "learning_rate": 9.777133608648276e-06,
+      "loss": 0.4002,
+      "step": 687
+    },
+    {
+      "epoch": 0.5585926928281462,
+      "grad_norm": 0.30372869968414307,
+      "learning_rate": 9.775735916905458e-06,
+      "loss": 0.426,
+      "step": 688
+    },
+    {
+      "epoch": 0.559404600811908,
+      "grad_norm": 0.2947506308555603,
+      "learning_rate": 9.77433395662285e-06,
+      "loss": 0.4089,
+      "step": 689
+    },
+    {
+      "epoch": 0.5602165087956699,
+      "grad_norm": 0.25095275044441223,
+      "learning_rate": 9.77292772905352e-06,
+      "loss": 0.4015,
+      "step": 690
+    },
+    {
+      "epoch": 0.5610284167794317,
+      "grad_norm": 0.284509539604187,
+      "learning_rate": 9.77151723545435e-06,
+      "loss": 0.403,
+      "step": 691
+    },
+    {
+      "epoch": 0.5618403247631935,
+      "grad_norm": 0.28761130571365356,
+      "learning_rate": 9.77010247708604e-06,
+      "loss": 0.3922,
+      "step": 692
+    },
+    {
+      "epoch": 0.5626522327469553,
+      "grad_norm": 0.3075431287288666,
+      "learning_rate": 9.768683455213089e-06,
+      "loss": 0.3844,
+      "step": 693
+    },
+    {
+      "epoch": 0.5634641407307172,
+      "grad_norm": 0.2755295932292938,
+      "learning_rate": 9.76726017110382e-06,
+      "loss": 0.3914,
+      "step": 694
+    },
+    {
+      "epoch": 0.564276048714479,
+      "grad_norm": 0.25007662177085876,
+      "learning_rate": 9.765832626030359e-06,
+      "loss": 0.4167,
+      "step": 695
+    },
+    {
+      "epoch": 0.5650879566982409,
+      "grad_norm": 0.2931385636329651,
+      "learning_rate": 9.764400821268642e-06,
+      "loss": 0.3902,
+      "step": 696
+    },
+    {
+      "epoch": 0.5658998646820027,
+      "grad_norm": 0.27402058243751526,
+      "learning_rate": 9.762964758098412e-06,
+      "loss": 0.3981,
+      "step": 697
+    },
+    {
+      "epoch": 0.5667117726657646,
+      "grad_norm": 0.25080975890159607,
+      "learning_rate": 9.761524437803221e-06,
+      "loss": 0.3995,
+      "step": 698
+    },
+    {
+      "epoch": 0.5675236806495264,
+      "grad_norm": 0.2998807728290558,
+      "learning_rate": 9.76007986167042e-06,
+      "loss": 0.388,
+      "step": 699
+    },
+    {
+      "epoch": 0.5683355886332883,
+      "grad_norm": 0.2640250325202942,
+      "learning_rate": 9.758631030991168e-06,
+      "loss": 0.4053,
+      "step": 700
+    },
+    {
+      "epoch": 0.5691474966170501,
+      "grad_norm": 0.2860358655452728,
+      "learning_rate": 9.757177947060426e-06,
+      "loss": 0.3977,
+      "step": 701
+    },
+    {
+      "epoch": 0.569959404600812,
+      "grad_norm": 0.3119959831237793,
+      "learning_rate": 9.755720611176957e-06,
+      "loss": 0.398,
+      "step": 702
+    },
+    {
+      "epoch": 0.5707713125845737,
+      "grad_norm": 0.3096456229686737,
+      "learning_rate": 9.754259024643324e-06,
+      "loss": 0.3953,
+      "step": 703
+    },
+    {
+      "epoch": 0.5715832205683355,
+      "grad_norm": 0.26966220140457153,
+      "learning_rate": 9.75279318876589e-06,
+      "loss": 0.3817,
+      "step": 704
+    },
+    {
+      "epoch": 0.5723951285520974,
+      "grad_norm": 0.29072538018226624,
+      "learning_rate": 9.751323104854812e-06,
+      "loss": 0.4065,
+      "step": 705
+    },
+    {
+      "epoch": 0.5732070365358592,
+      "grad_norm": 0.28483906388282776,
+      "learning_rate": 9.74984877422405e-06,
+      "loss": 0.374,
+      "step": 706
+    },
+    {
+      "epoch": 0.5740189445196211,
+      "grad_norm": 0.2920861840248108,
+      "learning_rate": 9.748370198191355e-06,
+      "loss": 0.3832,
+      "step": 707
+    },
+    {
+      "epoch": 0.574830852503383,
+      "grad_norm": 0.27481260895729065,
+      "learning_rate": 9.746887378078276e-06,
+      "loss": 0.4024,
+      "step": 708
+    },
+    {
+      "epoch": 0.5756427604871448,
+      "grad_norm": 0.2871198058128357,
+      "learning_rate": 9.74540031521015e-06,
+      "loss": 0.4025,
+      "step": 709
+    },
+    {
+      "epoch": 0.5764546684709067,
+      "grad_norm": 0.2981361448764801,
+      "learning_rate": 9.743909010916115e-06,
+      "loss": 0.4152,
+      "step": 710
+    },
+    {
+      "epoch": 0.5772665764546685,
+      "grad_norm": 0.24472342431545258,
+      "learning_rate": 9.74241346652909e-06,
+      "loss": 0.384,
+      "step": 711
+    },
+    {
+      "epoch": 0.5780784844384304,
+      "grad_norm": 0.2617102563381195,
+      "learning_rate": 9.740913683385791e-06,
+      "loss": 0.3975,
+      "step": 712
+    },
+    {
+      "epoch": 0.5788903924221922,
+      "grad_norm": 0.2603019177913666,
+      "learning_rate": 9.73940966282672e-06,
+      "loss": 0.3794,
+      "step": 713
+    },
+    {
+      "epoch": 0.5797023004059539,
+      "grad_norm": 0.24799233675003052,
+      "learning_rate": 9.737901406196164e-06,
+      "loss": 0.3913,
+      "step": 714
+    },
+    {
+      "epoch": 0.5805142083897158,
+      "grad_norm": 0.3101574182510376,
+      "learning_rate": 9.736388914842203e-06,
+      "loss": 0.3839,
+      "step": 715
+    },
+    {
+      "epoch": 0.5813261163734776,
+      "grad_norm": 0.2570565640926361,
+      "learning_rate": 9.734872190116695e-06,
+      "loss": 0.4164,
+      "step": 716
+    },
+    {
+      "epoch": 0.5821380243572395,
+      "grad_norm": 0.26475048065185547,
+      "learning_rate": 9.733351233375283e-06,
+      "loss": 0.4133,
+      "step": 717
+    },
+    {
+      "epoch": 0.5829499323410013,
+      "grad_norm": 0.272061288356781,
+      "learning_rate": 9.731826045977397e-06,
+      "loss": 0.4112,
+      "step": 718
+    },
+    {
+      "epoch": 0.5837618403247632,
+      "grad_norm": 0.27314600348472595,
+      "learning_rate": 9.730296629286245e-06,
+      "loss": 0.4234,
+      "step": 719
+    },
+    {
+      "epoch": 0.584573748308525,
+      "grad_norm": 0.2585921287536621,
+      "learning_rate": 9.728762984668813e-06,
+      "loss": 0.3969,
+      "step": 720
+    },
+    {
+      "epoch": 0.5853856562922869,
+      "grad_norm": 0.24929925799369812,
+      "learning_rate": 9.727225113495871e-06,
+      "loss": 0.3993,
+      "step": 721
+    },
+    {
+      "epoch": 0.5861975642760487,
+      "grad_norm": 0.25315889716148376,
+      "learning_rate": 9.725683017141964e-06,
+      "loss": 0.3724,
+      "step": 722
+    },
+    {
+      "epoch": 0.5870094722598106,
+      "grad_norm": 0.24312494695186615,
+      "learning_rate": 9.724136696985412e-06,
+      "loss": 0.3634,
+      "step": 723
+    },
+    {
+      "epoch": 0.5878213802435724,
+      "grad_norm": 0.2752605974674225,
+      "learning_rate": 9.722586154408312e-06,
+      "loss": 0.3912,
+      "step": 724
+    },
+    {
+      "epoch": 0.5886332882273342,
+      "grad_norm": 0.29730653762817383,
+      "learning_rate": 9.721031390796535e-06,
+      "loss": 0.4017,
+      "step": 725
+    },
+    {
+      "epoch": 0.589445196211096,
+      "grad_norm": 0.23849959671497345,
+      "learning_rate": 9.719472407539725e-06,
+      "loss": 0.4023,
+      "step": 726
+    },
+    {
+      "epoch": 0.5902571041948579,
+      "grad_norm": 0.26989737153053284,
+      "learning_rate": 9.717909206031295e-06,
+      "loss": 0.3861,
+      "step": 727
+    },
+    {
+      "epoch": 0.5910690121786197,
+      "grad_norm": 0.2595803737640381,
+      "learning_rate": 9.716341787668434e-06,
+      "loss": 0.3754,
+      "step": 728
+    },
+    {
+      "epoch": 0.5918809201623816,
+      "grad_norm": 0.2775750160217285,
+      "learning_rate": 9.714770153852093e-06,
+      "loss": 0.3921,
+      "step": 729
+    },
+    {
+      "epoch": 0.5926928281461434,
+      "grad_norm": 0.25772228837013245,
+      "learning_rate": 9.713194305986995e-06,
+      "loss": 0.4116,
+      "step": 730
+    },
+    {
+      "epoch": 0.5935047361299053,
+      "grad_norm": 0.25167158246040344,
+      "learning_rate": 9.71161424548163e-06,
+      "loss": 0.3805,
+      "step": 731
+    },
+    {
+      "epoch": 0.5943166441136671,
+      "grad_norm": 0.26059457659721375,
+      "learning_rate": 9.710029973748249e-06,
+      "loss": 0.3804,
+      "step": 732
+    },
+    {
+      "epoch": 0.595128552097429,
+      "grad_norm": 0.2562505006790161,
+      "learning_rate": 9.708441492202874e-06,
+      "loss": 0.3974,
+      "step": 733
+    },
+    {
+      "epoch": 0.5959404600811908,
+      "grad_norm": 0.2823938727378845,
+      "learning_rate": 9.70684880226528e-06,
+      "loss": 0.3783,
+      "step": 734
+    },
+    {
+      "epoch": 0.5967523680649527,
+      "grad_norm": 0.26797738671302795,
+      "learning_rate": 9.705251905359014e-06,
+      "loss": 0.3945,
+      "step": 735
+    },
+    {
+      "epoch": 0.5975642760487144,
+      "grad_norm": 0.27570703625679016,
+      "learning_rate": 9.703650802911374e-06,
+      "loss": 0.3929,
+      "step": 736
+    },
+    {
+      "epoch": 0.5983761840324763,
+      "grad_norm": 0.2930142879486084,
+      "learning_rate": 9.702045496353422e-06,
+      "loss": 0.3949,
+      "step": 737
+    },
+    {
+      "epoch": 0.5991880920162381,
+      "grad_norm": 0.26298847794532776,
+      "learning_rate": 9.700435987119981e-06,
+      "loss": 0.4132,
+      "step": 738
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2939494848251343,
+      "learning_rate": 9.69882227664962e-06,
+      "loss": 0.3981,
+      "step": 739
+    },
+    {
+      "epoch": 0.6008119079837618,
+      "grad_norm": 0.28670206665992737,
+      "learning_rate": 9.69720436638467e-06,
+      "loss": 0.3925,
+      "step": 740
+    },
+    {
+      "epoch": 0.6016238159675237,
+      "grad_norm": 0.2595313787460327,
+      "learning_rate": 9.69558225777122e-06,
+      "loss": 0.3862,
+      "step": 741
+    },
+    {
+      "epoch": 0.6024357239512855,
+      "grad_norm": 0.2743270993232727,
+      "learning_rate": 9.693955952259099e-06,
+      "loss": 0.4242,
+      "step": 742
+    },
+    {
+      "epoch": 0.6032476319350474,
+      "grad_norm": 0.2691590189933777,
+      "learning_rate": 9.6923254513019e-06,
+      "loss": 0.4112,
+      "step": 743
+    },
+    {
+      "epoch": 0.6040595399188092,
+      "grad_norm": 0.27362626791000366,
+      "learning_rate": 9.69069075635696e-06,
+      "loss": 0.4015,
+      "step": 744
+    },
+    {
+      "epoch": 0.6048714479025711,
+      "grad_norm": 0.2664232850074768,
+      "learning_rate": 9.689051868885362e-06,
+      "loss": 0.3952,
+      "step": 745
+    },
+    {
+      "epoch": 0.6056833558863329,
+      "grad_norm": 0.28029754757881165,
+      "learning_rate": 9.68740879035194e-06,
+      "loss": 0.3966,
+      "step": 746
+    },
+    {
+      "epoch": 0.6064952638700947,
+      "grad_norm": 0.3053135573863983,
+      "learning_rate": 9.685761522225271e-06,
+      "loss": 0.4405,
+      "step": 747
+    },
+    {
+      "epoch": 0.6073071718538565,
+      "grad_norm": 0.2768215835094452,
+      "learning_rate": 9.684110065977685e-06,
+      "loss": 0.3947,
+      "step": 748
+    },
+    {
+      "epoch": 0.6081190798376184,
+      "grad_norm": 0.26351386308670044,
+      "learning_rate": 9.682454423085244e-06,
+      "loss": 0.3727,
+      "step": 749
+    },
+    {
+      "epoch": 0.6089309878213802,
+      "grad_norm": 0.2557833194732666,
+      "learning_rate": 9.680794595027761e-06,
+      "loss": 0.4149,
+      "step": 750
+    },
+    {
+      "epoch": 0.6097428958051421,
+      "grad_norm": 0.2391970306634903,
+      "learning_rate": 9.67913058328878e-06,
+      "loss": 0.3945,
+      "step": 751
+    },
+    {
+      "epoch": 0.6105548037889039,
+      "grad_norm": 0.3080144226551056,
+      "learning_rate": 9.677462389355594e-06,
+      "loss": 0.4136,
+      "step": 752
+    },
+    {
+      "epoch": 0.6113667117726658,
+      "grad_norm": 0.2842309772968292,
+      "learning_rate": 9.675790014719231e-06,
+      "loss": 0.4003,
+      "step": 753
+    },
+    {
+      "epoch": 0.6121786197564276,
+      "grad_norm": 0.2598082721233368,
+      "learning_rate": 9.674113460874453e-06,
+      "loss": 0.4006,
+      "step": 754
+    },
+    {
+      "epoch": 0.6129905277401895,
+      "grad_norm": 0.27199339866638184,
+      "learning_rate": 9.67243272931976e-06,
+      "loss": 0.3784,
+      "step": 755
+    },
+    {
+      "epoch": 0.6138024357239513,
+      "grad_norm": 0.25725775957107544,
+      "learning_rate": 9.670747821557387e-06,
+      "loss": 0.3668,
+      "step": 756
+    },
+    {
+      "epoch": 0.6146143437077132,
+      "grad_norm": 0.2551143765449524,
+      "learning_rate": 9.669058739093296e-06,
+      "loss": 0.3931,
+      "step": 757
+    },
+    {
+      "epoch": 0.6154262516914749,
+      "grad_norm": 0.2750457525253296,
+      "learning_rate": 9.667365483437188e-06,
+      "loss": 0.4123,
+      "step": 758
+    },
+    {
+      "epoch": 0.6162381596752368,
+      "grad_norm": 0.2570054531097412,
+      "learning_rate": 9.665668056102487e-06,
+      "loss": 0.3851,
+      "step": 759
+    },
+    {
+      "epoch": 0.6170500676589986,
+      "grad_norm": 0.28647738695144653,
+      "learning_rate": 9.663966458606353e-06,
+      "loss": 0.3978,
+      "step": 760
+    },
+    {
+      "epoch": 0.6178619756427605,
+      "grad_norm": 0.28455913066864014,
+      "learning_rate": 9.662260692469664e-06,
+      "loss": 0.4031,
+      "step": 761
+    },
+    {
+      "epoch": 0.6186738836265223,
+      "grad_norm": 0.3203304708003998,
+      "learning_rate": 9.660550759217035e-06,
+      "loss": 0.4133,
+      "step": 762
+    },
+    {
+      "epoch": 0.6194857916102842,
+      "grad_norm": 0.24532243609428406,
+      "learning_rate": 9.658836660376796e-06,
+      "loss": 0.3939,
+      "step": 763
+    },
+    {
+      "epoch": 0.620297699594046,
+      "grad_norm": 0.29281824827194214,
+      "learning_rate": 9.657118397481004e-06,
+      "loss": 0.3799,
+      "step": 764
+    },
+    {
+      "epoch": 0.6211096075778079,
+      "grad_norm": 0.31822606921195984,
+      "learning_rate": 9.65539597206544e-06,
+      "loss": 0.4067,
+      "step": 765
+    },
+    {
+      "epoch": 0.6219215155615697,
+      "grad_norm": 0.2593955099582672,
+      "learning_rate": 9.6536693856696e-06,
+      "loss": 0.4293,
+      "step": 766
+    },
+    {
+      "epoch": 0.6227334235453316,
+      "grad_norm": 0.26911666989326477,
+      "learning_rate": 9.651938639836705e-06,
+      "loss": 0.3746,
+      "step": 767
+    },
+    {
+      "epoch": 0.6235453315290934,
+      "grad_norm": 0.2574180066585541,
+      "learning_rate": 9.650203736113689e-06,
+      "loss": 0.3924,
+      "step": 768
+    },
+    {
+      "epoch": 0.6243572395128552,
+      "grad_norm": 0.2955131530761719,
+      "learning_rate": 9.648464676051206e-06,
+      "loss": 0.3826,
+      "step": 769
+    },
+    {
+      "epoch": 0.625169147496617,
+      "grad_norm": 0.24141432344913483,
+      "learning_rate": 9.64672146120362e-06,
+      "loss": 0.38,
+      "step": 770
+    },
+    {
+      "epoch": 0.6259810554803789,
+      "grad_norm": 0.2561118006706238,
+      "learning_rate": 9.644974093129017e-06,
+      "loss": 0.4294,
+      "step": 771
+    },
+    {
+      "epoch": 0.6267929634641407,
+      "grad_norm": 0.2567097842693329,
+      "learning_rate": 9.643222573389184e-06,
+      "loss": 0.421,
+      "step": 772
+    },
+    {
+      "epoch": 0.6276048714479026,
+      "grad_norm": 0.24068014323711395,
+      "learning_rate": 9.641466903549628e-06,
+      "loss": 0.3948,
+      "step": 773
+    },
+    {
+      "epoch": 0.6284167794316644,
+      "grad_norm": 0.27005648612976074,
+      "learning_rate": 9.639707085179562e-06,
+      "loss": 0.3881,
+      "step": 774
+    },
+    {
+      "epoch": 0.6292286874154263,
+      "grad_norm": 0.23928485810756683,
+      "learning_rate": 9.637943119851905e-06,
+      "loss": 0.38,
+      "step": 775
+    },
+    {
+      "epoch": 0.6300405953991881,
+      "grad_norm": 0.2637043595314026,
+      "learning_rate": 9.636175009143286e-06,
+      "loss": 0.387,
+      "step": 776
+    },
+    {
+      "epoch": 0.63085250338295,
+      "grad_norm": 0.27847790718078613,
+      "learning_rate": 9.634402754634037e-06,
+      "loss": 0.4042,
+      "step": 777
+    },
+    {
+      "epoch": 0.6316644113667118,
+      "grad_norm": 0.2506670653820038,
+      "learning_rate": 9.632626357908194e-06,
+      "loss": 0.3814,
+      "step": 778
+    },
+    {
+      "epoch": 0.6324763193504737,
+      "grad_norm": 0.27196523547172546,
+      "learning_rate": 9.630845820553495e-06,
+      "loss": 0.4122,
+      "step": 779
+    },
+    {
+      "epoch": 0.6332882273342354,
+      "grad_norm": 0.2716112732887268,
+      "learning_rate": 9.629061144161383e-06,
+      "loss": 0.4026,
+      "step": 780
+    },
+    {
+      "epoch": 0.6341001353179972,
+      "grad_norm": 0.26494401693344116,
+      "learning_rate": 9.627272330326993e-06,
+      "loss": 0.4164,
+      "step": 781
+    },
+    {
+      "epoch": 0.6349120433017591,
+      "grad_norm": 0.2657059133052826,
+      "learning_rate": 9.625479380649163e-06,
+      "loss": 0.3766,
+      "step": 782
+    },
+    {
+      "epoch": 0.635723951285521,
+      "grad_norm": 0.2604874074459076,
+      "learning_rate": 9.623682296730427e-06,
+      "loss": 0.3763,
+      "step": 783
+    },
+    {
+      "epoch": 0.6365358592692828,
+      "grad_norm": 0.2501976788043976,
+      "learning_rate": 9.621881080177013e-06,
+      "loss": 0.3917,
+      "step": 784
+    },
+    {
+      "epoch": 0.6373477672530447,
+      "grad_norm": 0.23539717495441437,
+      "learning_rate": 9.620075732598843e-06,
+      "loss": 0.391,
+      "step": 785
+    },
+    {
+      "epoch": 0.6381596752368065,
+      "grad_norm": 0.25820279121398926,
+      "learning_rate": 9.618266255609533e-06,
+      "loss": 0.3821,
+      "step": 786
+    },
+    {
+      "epoch": 0.6389715832205684,
+      "grad_norm": 0.25813332200050354,
+      "learning_rate": 9.616452650826392e-06,
+      "loss": 0.3785,
+      "step": 787
+    },
+    {
+      "epoch": 0.6397834912043302,
+      "grad_norm": 0.25521135330200195,
+      "learning_rate": 9.614634919870407e-06,
+      "loss": 0.4104,
+      "step": 788
+    },
+    {
+      "epoch": 0.640595399188092,
+      "grad_norm": 0.2235318124294281,
+      "learning_rate": 9.612813064366267e-06,
+      "loss": 0.4042,
+      "step": 789
+    },
+    {
+      "epoch": 0.6414073071718539,
+      "grad_norm": 0.2524764835834503,
+      "learning_rate": 9.61098708594234e-06,
+      "loss": 0.3857,
+      "step": 790
+    },
+    {
+      "epoch": 0.6422192151556156,
+      "grad_norm": 0.231026291847229,
+      "learning_rate": 9.60915698623068e-06,
+      "loss": 0.3804,
+      "step": 791
+    },
+    {
+      "epoch": 0.6430311231393775,
+      "grad_norm": 0.2385670244693756,
+      "learning_rate": 9.607322766867026e-06,
+      "loss": 0.3922,
+      "step": 792
+    },
+    {
+      "epoch": 0.6438430311231393,
+      "grad_norm": 0.24389398097991943,
+      "learning_rate": 9.605484429490796e-06,
+      "loss": 0.3886,
+      "step": 793
+    },
+    {
+      "epoch": 0.6446549391069012,
+      "grad_norm": 0.24471168220043182,
+      "learning_rate": 9.603641975745095e-06,
+      "loss": 0.4295,
+      "step": 794
+    },
+    {
+      "epoch": 0.645466847090663,
+      "grad_norm": 0.25666460394859314,
+      "learning_rate": 9.601795407276699e-06,
+      "loss": 0.4118,
+      "step": 795
+    },
+    {
+      "epoch": 0.6462787550744249,
+      "grad_norm": 0.2565913498401642,
+      "learning_rate": 9.59994472573607e-06,
+      "loss": 0.3909,
+      "step": 796
+    },
+    {
+      "epoch": 0.6470906630581867,
+      "grad_norm": 0.28901827335357666,
+      "learning_rate": 9.598089932777339e-06,
+      "loss": 0.3837,
+      "step": 797
+    },
+    {
+      "epoch": 0.6479025710419486,
+      "grad_norm": 0.23624546825885773,
+      "learning_rate": 9.596231030058315e-06,
+      "loss": 0.3947,
+      "step": 798
+    },
+    {
+      "epoch": 0.6487144790257104,
+      "grad_norm": 0.2706625461578369,
+      "learning_rate": 9.594368019240484e-06,
+      "loss": 0.3793,
+      "step": 799
+    },
+    {
+      "epoch": 0.6495263870094723,
+      "grad_norm": 0.2454807609319687,
+      "learning_rate": 9.592500901988994e-06,
+      "loss": 0.4012,
+      "step": 800
+    },
+    {
+      "epoch": 0.6503382949932341,
+      "grad_norm": 0.23770014941692352,
+      "learning_rate": 9.590629679972673e-06,
+      "loss": 0.4048,
+      "step": 801
+    },
+    {
+      "epoch": 0.651150202976996,
+      "grad_norm": 0.28094348311424255,
+      "learning_rate": 9.588754354864014e-06,
+      "loss": 0.3874,
+      "step": 802
+    },
+    {
+      "epoch": 0.6519621109607577,
+      "grad_norm": 0.24728763103485107,
+      "learning_rate": 9.586874928339176e-06,
+      "loss": 0.3944,
+      "step": 803
+    },
+    {
+      "epoch": 0.6527740189445196,
+      "grad_norm": 0.2462068796157837,
+      "learning_rate": 9.584991402077988e-06,
+      "loss": 0.3806,
+      "step": 804
+    },
+    {
+      "epoch": 0.6535859269282814,
+      "grad_norm": 0.26098212599754333,
+      "learning_rate": 9.583103777763938e-06,
+      "loss": 0.4009,
+      "step": 805
+    },
+    {
+      "epoch": 0.6543978349120433,
+      "grad_norm": 0.24661263823509216,
+      "learning_rate": 9.58121205708418e-06,
+      "loss": 0.3979,
+      "step": 806
+    },
+    {
+      "epoch": 0.6552097428958051,
+      "grad_norm": 0.25520509481430054,
+      "learning_rate": 9.57931624172953e-06,
+      "loss": 0.379,
+      "step": 807
+    },
+    {
+      "epoch": 0.656021650879567,
+      "grad_norm": 0.2596292793750763,
+      "learning_rate": 9.577416333394463e-06,
+      "loss": 0.418,
+      "step": 808
+    },
+    {
+      "epoch": 0.6568335588633288,
+      "grad_norm": 0.28905919194221497,
+      "learning_rate": 9.575512333777108e-06,
+      "loss": 0.4031,
+      "step": 809
+    },
+    {
+      "epoch": 0.6576454668470907,
+      "grad_norm": 0.26445189118385315,
+      "learning_rate": 9.57360424457926e-06,
+      "loss": 0.376,
+      "step": 810
+    },
+    {
+      "epoch": 0.6584573748308525,
+      "grad_norm": 0.2513701617717743,
+      "learning_rate": 9.571692067506363e-06,
+      "loss": 0.3925,
+      "step": 811
+    },
+    {
+      "epoch": 0.6592692828146144,
+      "grad_norm": 0.27707213163375854,
+      "learning_rate": 9.56977580426751e-06,
+      "loss": 0.3912,
+      "step": 812
+    },
+    {
+      "epoch": 0.6600811907983762,
+      "grad_norm": 0.2527226507663727,
+      "learning_rate": 9.567855456575459e-06,
+      "loss": 0.3848,
+      "step": 813
+    },
+    {
+      "epoch": 0.660893098782138,
+      "grad_norm": 0.24332165718078613,
+      "learning_rate": 9.565931026146607e-06,
+      "loss": 0.3745,
+      "step": 814
+    },
+    {
+      "epoch": 0.6617050067658998,
+      "grad_norm": 0.276545912027359,
+      "learning_rate": 9.564002514701006e-06,
+      "loss": 0.423,
+      "step": 815
+    },
+    {
+      "epoch": 0.6625169147496617,
+      "grad_norm": 0.26829278469085693,
+      "learning_rate": 9.562069923962355e-06,
+      "loss": 0.3862,
+      "step": 816
+    },
+    {
+      "epoch": 0.6633288227334235,
+      "grad_norm": 0.2558344304561615,
+      "learning_rate": 9.560133255657997e-06,
+      "loss": 0.3903,
+      "step": 817
+    },
+    {
+      "epoch": 0.6641407307171854,
+      "grad_norm": 0.2592118978500366,
+      "learning_rate": 9.558192511518925e-06,
+      "loss": 0.4057,
+      "step": 818
+    },
+    {
+      "epoch": 0.6649526387009472,
+      "grad_norm": 0.2938236594200134,
+      "learning_rate": 9.556247693279764e-06,
+      "loss": 0.4216,
+      "step": 819
+    },
+    {
+      "epoch": 0.6657645466847091,
+      "grad_norm": 0.25571298599243164,
+      "learning_rate": 9.554298802678793e-06,
+      "loss": 0.4013,
+      "step": 820
+    },
+    {
+      "epoch": 0.6665764546684709,
+      "grad_norm": 0.27169671654701233,
+      "learning_rate": 9.552345841457922e-06,
+      "loss": 0.4099,
+      "step": 821
+    },
+    {
+      "epoch": 0.6673883626522328,
+      "grad_norm": 0.2796476483345032,
+      "learning_rate": 9.550388811362704e-06,
+      "loss": 0.3828,
+      "step": 822
+    },
+    {
+      "epoch": 0.6682002706359946,
+      "grad_norm": 0.2858635485172272,
+      "learning_rate": 9.548427714142326e-06,
+      "loss": 0.4127,
+      "step": 823
+    },
+    {
+      "epoch": 0.6690121786197565,
+      "grad_norm": 0.2862201929092407,
+      "learning_rate": 9.546462551549612e-06,
+      "loss": 0.3863,
+      "step": 824
+    },
+    {
+      "epoch": 0.6698240866035182,
+      "grad_norm": 0.30897125601768494,
+      "learning_rate": 9.54449332534102e-06,
+      "loss": 0.3862,
+      "step": 825
+    },
+    {
+      "epoch": 0.6706359945872801,
+      "grad_norm": 0.2528267502784729,
+      "learning_rate": 9.542520037276636e-06,
+      "loss": 0.3819,
+      "step": 826
+    },
+    {
+      "epoch": 0.6714479025710419,
+      "grad_norm": 0.28384026885032654,
+      "learning_rate": 9.540542689120184e-06,
+      "loss": 0.4089,
+      "step": 827
+    },
+    {
+      "epoch": 0.6722598105548038,
+      "grad_norm": 0.269479900598526,
+      "learning_rate": 9.538561282639008e-06,
+      "loss": 0.409,
+      "step": 828
+    },
+    {
+      "epoch": 0.6730717185385656,
+      "grad_norm": 0.2516125738620758,
+      "learning_rate": 9.536575819604087e-06,
+      "loss": 0.3783,
+      "step": 829
+    },
+    {
+      "epoch": 0.6738836265223275,
+      "grad_norm": 0.2534337639808655,
+      "learning_rate": 9.534586301790021e-06,
+      "loss": 0.3891,
+      "step": 830
+    },
+    {
+      "epoch": 0.6746955345060893,
+      "grad_norm": 0.28992563486099243,
+      "learning_rate": 9.532592730975035e-06,
+      "loss": 0.3905,
+      "step": 831
+    },
+    {
+      "epoch": 0.6755074424898512,
+      "grad_norm": 0.25423532724380493,
+      "learning_rate": 9.530595108940978e-06,
+      "loss": 0.388,
+      "step": 832
+    },
+    {
+      "epoch": 0.676319350473613,
+      "grad_norm": 0.24912771582603455,
+      "learning_rate": 9.52859343747332e-06,
+      "loss": 0.3958,
+      "step": 833
+    },
+    {
+      "epoch": 0.6771312584573749,
+      "grad_norm": 0.2401212602853775,
+      "learning_rate": 9.526587718361147e-06,
+      "loss": 0.3924,
+      "step": 834
+    },
+    {
+      "epoch": 0.6779431664411367,
+      "grad_norm": 0.2993282973766327,
+      "learning_rate": 9.524577953397167e-06,
+      "loss": 0.3815,
+      "step": 835
+    },
+    {
+      "epoch": 0.6787550744248985,
+      "grad_norm": 0.25689002871513367,
+      "learning_rate": 9.522564144377703e-06,
+      "loss": 0.4038,
+      "step": 836
+    },
+    {
+      "epoch": 0.6795669824086603,
+      "grad_norm": 0.23488833010196686,
+      "learning_rate": 9.52054629310269e-06,
+      "loss": 0.3936,
+      "step": 837
+    },
+    {
+      "epoch": 0.6803788903924222,
+      "grad_norm": 0.30573493242263794,
+      "learning_rate": 9.518524401375682e-06,
+      "loss": 0.4281,
+      "step": 838
+    },
+    {
+      "epoch": 0.681190798376184,
+      "grad_norm": 0.27808699011802673,
+      "learning_rate": 9.516498471003837e-06,
+      "loss": 0.3661,
+      "step": 839
+    },
+    {
+      "epoch": 0.6820027063599459,
+      "grad_norm": 0.24241849780082703,
+      "learning_rate": 9.514468503797927e-06,
+      "loss": 0.4021,
+      "step": 840
+    },
+    {
+      "epoch": 0.6828146143437077,
+      "grad_norm": 0.26007118821144104,
+      "learning_rate": 9.51243450157233e-06,
+      "loss": 0.3974,
+      "step": 841
+    },
+    {
+      "epoch": 0.6836265223274696,
+      "grad_norm": 0.27422502636909485,
+      "learning_rate": 9.510396466145032e-06,
+      "loss": 0.4034,
+      "step": 842
+    },
+    {
+      "epoch": 0.6844384303112314,
+      "grad_norm": 0.2590222656726837,
+      "learning_rate": 9.508354399337625e-06,
+      "loss": 0.4064,
+      "step": 843
+    },
+    {
+      "epoch": 0.6852503382949933,
+      "grad_norm": 0.24739545583724976,
+      "learning_rate": 9.5063083029753e-06,
+      "loss": 0.3927,
+      "step": 844
+    },
+    {
+      "epoch": 0.6860622462787551,
+      "grad_norm": 0.27892038226127625,
+      "learning_rate": 9.504258178886858e-06,
+      "loss": 0.4084,
+      "step": 845
+    },
+    {
+      "epoch": 0.686874154262517,
+      "grad_norm": 0.27396920323371887,
+      "learning_rate": 9.502204028904687e-06,
+      "loss": 0.3876,
+      "step": 846
+    },
+    {
+      "epoch": 0.6876860622462787,
+      "grad_norm": 0.27307480573654175,
+      "learning_rate": 9.500145854864785e-06,
+      "loss": 0.4075,
+      "step": 847
+    },
+    {
+      "epoch": 0.6884979702300406,
+      "grad_norm": 0.2743874788284302,
+      "learning_rate": 9.498083658606744e-06,
+      "loss": 0.4052,
+      "step": 848
+    },
+    {
+      "epoch": 0.6893098782138024,
+      "grad_norm": 0.2821621596813202,
+      "learning_rate": 9.496017441973747e-06,
+      "loss": 0.4086,
+      "step": 849
+    },
+    {
+      "epoch": 0.6901217861975643,
+      "grad_norm": 0.26029956340789795,
+      "learning_rate": 9.49394720681257e-06,
+      "loss": 0.4126,
+      "step": 850
+    },
+    {
+      "epoch": 0.6909336941813261,
+      "grad_norm": 0.25640571117401123,
+      "learning_rate": 9.49187295497359e-06,
+      "loss": 0.402,
+      "step": 851
+    },
+    {
+      "epoch": 0.691745602165088,
+      "grad_norm": 0.24865415692329407,
+      "learning_rate": 9.48979468831076e-06,
+      "loss": 0.3883,
+      "step": 852
+    },
+    {
+      "epoch": 0.6925575101488498,
+      "grad_norm": 0.2797751724720001,
+      "learning_rate": 9.487712408681635e-06,
+      "loss": 0.3935,
+      "step": 853
+    },
+    {
+      "epoch": 0.6933694181326117,
+      "grad_norm": 0.24718858301639557,
+      "learning_rate": 9.485626117947351e-06,
+      "loss": 0.375,
+      "step": 854
+    },
+    {
+      "epoch": 0.6941813261163735,
+      "grad_norm": 0.23954224586486816,
+      "learning_rate": 9.483535817972625e-06,
+      "loss": 0.3893,
+      "step": 855
+    },
+    {
+      "epoch": 0.6949932341001354,
+      "grad_norm": 0.266996294260025,
+      "learning_rate": 9.481441510625765e-06,
+      "loss": 0.373,
+      "step": 856
+    },
+    {
+      "epoch": 0.6958051420838972,
+      "grad_norm": 0.2541523575782776,
+      "learning_rate": 9.479343197778653e-06,
+      "loss": 0.3906,
+      "step": 857
+    },
+    {
+      "epoch": 0.696617050067659,
+      "grad_norm": 0.2585209906101227,
+      "learning_rate": 9.477240881306759e-06,
+      "loss": 0.3833,
+      "step": 858
+    },
+    {
+      "epoch": 0.6974289580514208,
+      "grad_norm": 0.2782779633998871,
+      "learning_rate": 9.475134563089128e-06,
+      "loss": 0.4008,
+      "step": 859
+    },
+    {
+      "epoch": 0.6982408660351827,
+      "grad_norm": 0.24906453490257263,
+      "learning_rate": 9.47302424500838e-06,
+      "loss": 0.3976,
+      "step": 860
+    },
+    {
+      "epoch": 0.6990527740189445,
+      "grad_norm": 0.2578105330467224,
+      "learning_rate": 9.47090992895071e-06,
+      "loss": 0.4047,
+      "step": 861
+    },
+    {
+      "epoch": 0.6998646820027064,
+      "grad_norm": 0.2682112157344818,
+      "learning_rate": 9.468791616805893e-06,
+      "loss": 0.4008,
+      "step": 862
+    },
+    {
+      "epoch": 0.7006765899864682,
+      "grad_norm": 0.27572986483573914,
+      "learning_rate": 9.466669310467266e-06,
+      "loss": 0.3842,
+      "step": 863
+    },
+    {
+      "epoch": 0.70148849797023,
+      "grad_norm": 0.2380996197462082,
+      "learning_rate": 9.464543011831742e-06,
+      "loss": 0.3744,
+      "step": 864
+    },
+    {
+      "epoch": 0.7023004059539919,
+      "grad_norm": 0.2699928879737854,
+      "learning_rate": 9.462412722799801e-06,
+      "loss": 0.3859,
+      "step": 865
+    },
+    {
+      "epoch": 0.7031123139377538,
+      "grad_norm": 0.24437807500362396,
+      "learning_rate": 9.46027844527549e-06,
+      "loss": 0.4031,
+      "step": 866
+    },
+    {
+      "epoch": 0.7039242219215156,
+      "grad_norm": 0.2767066955566406,
+      "learning_rate": 9.45814018116642e-06,
+      "loss": 0.3746,
+      "step": 867
+    },
+    {
+      "epoch": 0.7047361299052775,
+      "grad_norm": 0.26939821243286133,
+      "learning_rate": 9.455997932383766e-06,
+      "loss": 0.3953,
+      "step": 868
+    },
+    {
+      "epoch": 0.7055480378890392,
+      "grad_norm": 0.3108167052268982,
+      "learning_rate": 9.453851700842262e-06,
+      "loss": 0.4091,
+      "step": 869
+    },
+    {
+      "epoch": 0.706359945872801,
+      "grad_norm": 0.2633616626262665,
+      "learning_rate": 9.451701488460207e-06,
+      "loss": 0.3967,
+      "step": 870
+    },
+    {
+      "epoch": 0.7071718538565629,
+      "grad_norm": 0.29193559288978577,
+      "learning_rate": 9.449547297159453e-06,
+      "loss": 0.3893,
+      "step": 871
+    },
+    {
+      "epoch": 0.7079837618403247,
+      "grad_norm": 0.2598331570625305,
+      "learning_rate": 9.447389128865413e-06,
+      "loss": 0.3727,
+      "step": 872
+    },
+    {
+      "epoch": 0.7087956698240866,
+      "grad_norm": 0.27109217643737793,
+      "learning_rate": 9.445226985507047e-06,
+      "loss": 0.3836,
+      "step": 873
+    },
+    {
+      "epoch": 0.7096075778078484,
+      "grad_norm": 0.25920167565345764,
+      "learning_rate": 9.44306086901688e-06,
+      "loss": 0.3791,
+      "step": 874
+    },
+    {
+      "epoch": 0.7104194857916103,
+      "grad_norm": 0.2699492871761322,
+      "learning_rate": 9.440890781330974e-06,
+      "loss": 0.3723,
+      "step": 875
+    },
+    {
+      "epoch": 0.7112313937753721,
+      "grad_norm": 0.28936323523521423,
+      "learning_rate": 9.43871672438895e-06,
+      "loss": 0.4093,
+      "step": 876
+    },
+    {
+      "epoch": 0.712043301759134,
+      "grad_norm": 0.25994059443473816,
+      "learning_rate": 9.436538700133977e-06,
+      "loss": 0.4045,
+      "step": 877
+    },
+    {
+      "epoch": 0.7128552097428958,
+      "grad_norm": 0.2501068115234375,
+      "learning_rate": 9.434356710512763e-06,
+      "loss": 0.3928,
+      "step": 878
+    },
+    {
+      "epoch": 0.7136671177266577,
+      "grad_norm": 0.30248770117759705,
+      "learning_rate": 9.432170757475566e-06,
+      "loss": 0.4079,
+      "step": 879
+    },
+    {
+      "epoch": 0.7144790257104194,
+      "grad_norm": 0.27035707235336304,
+      "learning_rate": 9.429980842976186e-06,
+      "loss": 0.4056,
+      "step": 880
+    },
+    {
+      "epoch": 0.7152909336941813,
+      "grad_norm": 0.28653231263160706,
+      "learning_rate": 9.427786968971961e-06,
+      "loss": 0.4277,
+      "step": 881
+    },
+    {
+      "epoch": 0.7161028416779431,
+      "grad_norm": 0.24470244348049164,
+      "learning_rate": 9.425589137423772e-06,
+      "loss": 0.3792,
+      "step": 882
+    },
+    {
+      "epoch": 0.716914749661705,
+      "grad_norm": 0.25580763816833496,
+      "learning_rate": 9.423387350296032e-06,
+      "loss": 0.3755,
+      "step": 883
+    },
+    {
+      "epoch": 0.7177266576454668,
+      "grad_norm": 0.26169440150260925,
+      "learning_rate": 9.421181609556693e-06,
+      "loss": 0.4001,
+      "step": 884
+    },
+    {
+      "epoch": 0.7185385656292287,
+      "grad_norm": 0.23789216578006744,
+      "learning_rate": 9.418971917177241e-06,
+      "loss": 0.4097,
+      "step": 885
+    },
+    {
+      "epoch": 0.7193504736129905,
+      "grad_norm": 0.25004270672798157,
+      "learning_rate": 9.416758275132693e-06,
+      "loss": 0.4098,
+      "step": 886
+    },
+    {
+      "epoch": 0.7201623815967524,
+      "grad_norm": 0.26923874020576477,
+      "learning_rate": 9.414540685401596e-06,
+      "loss": 0.3888,
+      "step": 887
+    },
+    {
+      "epoch": 0.7209742895805142,
+      "grad_norm": 0.25549858808517456,
+      "learning_rate": 9.412319149966025e-06,
+      "loss": 0.3725,
+      "step": 888
+    },
+    {
+      "epoch": 0.7217861975642761,
+      "grad_norm": 0.2700618803501129,
+      "learning_rate": 9.410093670811582e-06,
+      "loss": 0.3891,
+      "step": 889
+    },
+    {
+      "epoch": 0.7225981055480379,
+      "grad_norm": 0.2732279896736145,
+      "learning_rate": 9.407864249927396e-06,
+      "loss": 0.3805,
+      "step": 890
+    },
+    {
+      "epoch": 0.7234100135317997,
+      "grad_norm": 0.259591281414032,
+      "learning_rate": 9.405630889306116e-06,
+      "loss": 0.3891,
+      "step": 891
+    },
+    {
+      "epoch": 0.7242219215155615,
+      "grad_norm": 0.2855673134326935,
+      "learning_rate": 9.403393590943916e-06,
+      "loss": 0.3973,
+      "step": 892
+    },
+    {
+      "epoch": 0.7250338294993234,
+      "grad_norm": 0.2759544551372528,
+      "learning_rate": 9.401152356840484e-06,
+      "loss": 0.3913,
+      "step": 893
+    },
+    {
+      "epoch": 0.7258457374830852,
+      "grad_norm": 0.27817124128341675,
+      "learning_rate": 9.39890718899903e-06,
+      "loss": 0.3913,
+      "step": 894
+    },
+    {
+      "epoch": 0.7266576454668471,
+      "grad_norm": 0.3125092685222626,
+      "learning_rate": 9.39665808942628e-06,
+      "loss": 0.3763,
+      "step": 895
+    },
+    {
+      "epoch": 0.7274695534506089,
+      "grad_norm": 0.30287715792655945,
+      "learning_rate": 9.394405060132473e-06,
+      "loss": 0.4136,
+      "step": 896
+    },
+    {
+      "epoch": 0.7282814614343708,
+      "grad_norm": 0.2583794891834259,
+      "learning_rate": 9.392148103131358e-06,
+      "loss": 0.3937,
+      "step": 897
+    },
+    {
+      "epoch": 0.7290933694181326,
+      "grad_norm": 0.3228049874305725,
+      "learning_rate": 9.3898872204402e-06,
+      "loss": 0.4124,
+      "step": 898
+    },
+    {
+      "epoch": 0.7299052774018945,
+      "grad_norm": 0.32627710700035095,
+      "learning_rate": 9.38762241407977e-06,
+      "loss": 0.4181,
+      "step": 899
+    },
+    {
+      "epoch": 0.7307171853856563,
+      "grad_norm": 0.2713007628917694,
+      "learning_rate": 9.38535368607434e-06,
+      "loss": 0.4015,
+      "step": 900
+    },
+    {
+      "epoch": 0.7315290933694182,
+      "grad_norm": 0.26987311244010925,
+      "learning_rate": 9.3830810384517e-06,
+      "loss": 0.3893,
+      "step": 901
+    },
+    {
+      "epoch": 0.7323410013531799,
+      "grad_norm": 0.31110334396362305,
+      "learning_rate": 9.380804473243133e-06,
+      "loss": 0.3989,
+      "step": 902
+    },
+    {
+      "epoch": 0.7331529093369418,
+      "grad_norm": 0.24174535274505615,
+      "learning_rate": 9.378523992483429e-06,
+      "loss": 0.3751,
+      "step": 903
+    },
+    {
+      "epoch": 0.7339648173207036,
+      "grad_norm": 0.29198309779167175,
+      "learning_rate": 9.376239598210873e-06,
+      "loss": 0.3909,
+      "step": 904
+    },
+    {
+      "epoch": 0.7347767253044655,
+      "grad_norm": 0.35126492381095886,
+      "learning_rate": 9.37395129246725e-06,
+      "loss": 0.4092,
+      "step": 905
+    },
+    {
+      "epoch": 0.7355886332882273,
+      "grad_norm": 0.3240865468978882,
+      "learning_rate": 9.371659077297843e-06,
+      "loss": 0.3912,
+      "step": 906
+    },
+    {
+      "epoch": 0.7364005412719892,
+      "grad_norm": 0.27511653304100037,
+      "learning_rate": 9.369362954751428e-06,
+      "loss": 0.4173,
+      "step": 907
+    },
+    {
+      "epoch": 0.737212449255751,
+      "grad_norm": 0.31036898493766785,
+      "learning_rate": 9.367062926880273e-06,
+      "loss": 0.4143,
+      "step": 908
+    },
+    {
+      "epoch": 0.7380243572395129,
+      "grad_norm": 0.2945440411567688,
+      "learning_rate": 9.364758995740136e-06,
+      "loss": 0.3708,
+      "step": 909
+    },
+    {
+      "epoch": 0.7388362652232747,
+      "grad_norm": 0.24441523849964142,
+      "learning_rate": 9.362451163390265e-06,
+      "loss": 0.3849,
+      "step": 910
+    },
+    {
+      "epoch": 0.7396481732070366,
+      "grad_norm": 0.2710442841053009,
+      "learning_rate": 9.360139431893394e-06,
+      "loss": 0.4063,
+      "step": 911
+    },
+    {
+      "epoch": 0.7404600811907984,
+      "grad_norm": 0.22298097610473633,
+      "learning_rate": 9.35782380331574e-06,
+      "loss": 0.376,
+      "step": 912
+    },
+    {
+      "epoch": 0.7412719891745602,
+      "grad_norm": 0.30962908267974854,
+      "learning_rate": 9.35550427972701e-06,
+      "loss": 0.3993,
+      "step": 913
+    },
+    {
+      "epoch": 0.742083897158322,
+      "grad_norm": 0.25685498118400574,
+      "learning_rate": 9.353180863200385e-06,
+      "loss": 0.3831,
+      "step": 914
+    },
+    {
+      "epoch": 0.7428958051420839,
+      "grad_norm": 0.27784019708633423,
+      "learning_rate": 9.350853555812529e-06,
+      "loss": 0.3904,
+      "step": 915
+    },
+    {
+      "epoch": 0.7437077131258457,
+      "grad_norm": 0.2735278308391571,
+      "learning_rate": 9.348522359643583e-06,
+      "loss": 0.3984,
+      "step": 916
+    },
+    {
+      "epoch": 0.7445196211096076,
+      "grad_norm": 0.23908422887325287,
+      "learning_rate": 9.346187276777163e-06,
+      "loss": 0.3715,
+      "step": 917
+    },
+    {
+      "epoch": 0.7453315290933694,
+      "grad_norm": 0.2641873061656952,
+      "learning_rate": 9.34384830930036e-06,
+      "loss": 0.3932,
+      "step": 918
+    },
+    {
+      "epoch": 0.7461434370771313,
+      "grad_norm": 0.2791314423084259,
+      "learning_rate": 9.34150545930374e-06,
+      "loss": 0.3916,
+      "step": 919
+    },
+    {
+      "epoch": 0.7469553450608931,
+      "grad_norm": 0.23333701491355896,
+      "learning_rate": 9.339158728881332e-06,
+      "loss": 0.3867,
+      "step": 920
+    },
+    {
+      "epoch": 0.747767253044655,
+      "grad_norm": 0.27976715564727783,
+      "learning_rate": 9.336808120130638e-06,
+      "loss": 0.3981,
+      "step": 921
+    },
+    {
+      "epoch": 0.7485791610284168,
+      "grad_norm": 0.2522328495979309,
+      "learning_rate": 9.334453635152628e-06,
+      "loss": 0.3828,
+      "step": 922
+    },
+    {
+      "epoch": 0.7493910690121787,
+      "grad_norm": 0.2677200138568878,
+      "learning_rate": 9.332095276051729e-06,
+      "loss": 0.3889,
+      "step": 923
+    },
+    {
+      "epoch": 0.7502029769959404,
+      "grad_norm": 0.27865439653396606,
+      "learning_rate": 9.329733044935843e-06,
+      "loss": 0.3898,
+      "step": 924
+    },
+    {
+      "epoch": 0.7510148849797023,
+      "grad_norm": 0.24510467052459717,
+      "learning_rate": 9.32736694391632e-06,
+      "loss": 0.3711,
+      "step": 925
+    },
+    {
+      "epoch": 0.7518267929634641,
+      "grad_norm": 0.257647305727005,
+      "learning_rate": 9.324996975107978e-06,
+      "loss": 0.3958,
+      "step": 926
+    },
+    {
+      "epoch": 0.752638700947226,
+      "grad_norm": 0.2517211139202118,
+      "learning_rate": 9.322623140629088e-06,
+      "loss": 0.3781,
+      "step": 927
+    },
+    {
+      "epoch": 0.7534506089309878,
+      "grad_norm": 0.24808689951896667,
+      "learning_rate": 9.320245442601377e-06,
+      "loss": 0.3817,
+      "step": 928
+    },
+    {
+      "epoch": 0.7542625169147497,
+      "grad_norm": 0.2618692219257355,
+      "learning_rate": 9.317863883150022e-06,
+      "loss": 0.4106,
+      "step": 929
+    },
+    {
+      "epoch": 0.7550744248985115,
+      "grad_norm": 0.23472526669502258,
+      "learning_rate": 9.31547846440366e-06,
+      "loss": 0.3985,
+      "step": 930
+    },
+    {
+      "epoch": 0.7558863328822734,
+      "grad_norm": 0.2363179475069046,
+      "learning_rate": 9.313089188494366e-06,
+      "loss": 0.378,
+      "step": 931
+    },
+    {
+      "epoch": 0.7566982408660352,
+      "grad_norm": 0.2915582060813904,
+      "learning_rate": 9.31069605755767e-06,
+      "loss": 0.4052,
+      "step": 932
+    },
+    {
+      "epoch": 0.7575101488497971,
+      "grad_norm": 0.2445308268070221,
+      "learning_rate": 9.30829907373255e-06,
+      "loss": 0.3824,
+      "step": 933
+    },
+    {
+      "epoch": 0.7583220568335589,
+      "grad_norm": 0.2758992314338684,
+      "learning_rate": 9.30589823916142e-06,
+      "loss": 0.3906,
+      "step": 934
+    },
+    {
+      "epoch": 0.7591339648173207,
+      "grad_norm": 0.28089702129364014,
+      "learning_rate": 9.303493555990139e-06,
+      "loss": 0.3889,
+      "step": 935
+    },
+    {
+      "epoch": 0.7599458728010825,
+      "grad_norm": 0.3191051185131073,
+      "learning_rate": 9.301085026368006e-06,
+      "loss": 0.3843,
+      "step": 936
+    },
+    {
+      "epoch": 0.7607577807848444,
+      "grad_norm": 0.2730553150177002,
+      "learning_rate": 9.298672652447763e-06,
+      "loss": 0.3816,
+      "step": 937
+    },
+    {
+      "epoch": 0.7615696887686062,
+      "grad_norm": 0.26440948247909546,
+      "learning_rate": 9.296256436385576e-06,
+      "loss": 0.3858,
+      "step": 938
+    },
+    {
+      "epoch": 0.762381596752368,
+      "grad_norm": 0.30479004979133606,
+      "learning_rate": 9.293836380341059e-06,
+      "loss": 0.4073,
+      "step": 939
+    },
+    {
+      "epoch": 0.7631935047361299,
+      "grad_norm": 0.2642437517642975,
+      "learning_rate": 9.291412486477247e-06,
+      "loss": 0.3691,
+      "step": 940
+    },
+    {
+      "epoch": 0.7640054127198918,
+      "grad_norm": 0.3002397119998932,
+      "learning_rate": 9.28898475696061e-06,
+      "loss": 0.3811,
+      "step": 941
+    },
+    {
+      "epoch": 0.7648173207036536,
+      "grad_norm": 0.3140433728694916,
+      "learning_rate": 9.28655319396105e-06,
+      "loss": 0.396,
+      "step": 942
+    },
+    {
+      "epoch": 0.7656292286874155,
+      "grad_norm": 0.28609123826026917,
+      "learning_rate": 9.284117799651887e-06,
+      "loss": 0.4043,
+      "step": 943
+    },
+    {
+      "epoch": 0.7664411366711773,
+      "grad_norm": 0.2929842174053192,
+      "learning_rate": 9.281678576209873e-06,
+      "loss": 0.3975,
+      "step": 944
+    },
+    {
+      "epoch": 0.7672530446549392,
+      "grad_norm": 0.25411462783813477,
+      "learning_rate": 9.279235525815177e-06,
+      "loss": 0.387,
+      "step": 945
+    },
+    {
+      "epoch": 0.7680649526387009,
+      "grad_norm": 0.28058159351348877,
+      "learning_rate": 9.276788650651392e-06,
+      "loss": 0.3673,
+      "step": 946
+    },
+    {
+      "epoch": 0.7688768606224627,
+      "grad_norm": 0.2674712538719177,
+      "learning_rate": 9.274337952905528e-06,
+      "loss": 0.4096,
+      "step": 947
+    },
+    {
+      "epoch": 0.7696887686062246,
+      "grad_norm": 0.24106945097446442,
+      "learning_rate": 9.271883434768012e-06,
+      "loss": 0.3977,
+      "step": 948
+    },
+    {
+      "epoch": 0.7705006765899864,
+      "grad_norm": 0.28799790143966675,
+      "learning_rate": 9.269425098432686e-06,
+      "loss": 0.3859,
+      "step": 949
+    },
+    {
+      "epoch": 0.7713125845737483,
+      "grad_norm": 0.31067049503326416,
+      "learning_rate": 9.266962946096802e-06,
+      "loss": 0.4158,
+      "step": 950
+    },
+    {
+      "epoch": 0.7721244925575101,
+      "grad_norm": 0.27122172713279724,
+      "learning_rate": 9.264496979961031e-06,
+      "loss": 0.3893,
+      "step": 951
+    },
+    {
+      "epoch": 0.772936400541272,
+      "grad_norm": 0.2872089743614197,
+      "learning_rate": 9.26202720222944e-06,
+      "loss": 0.4018,
+      "step": 952
+    },
+    {
+      "epoch": 0.7737483085250338,
+      "grad_norm": 0.27677980065345764,
+      "learning_rate": 9.259553615109514e-06,
+      "loss": 0.3969,
+      "step": 953
+    },
+    {
+      "epoch": 0.7745602165087957,
+      "grad_norm": 0.25516343116760254,
+      "learning_rate": 9.257076220812136e-06,
+      "loss": 0.3994,
+      "step": 954
+    },
+    {
+      "epoch": 0.7753721244925575,
+      "grad_norm": 0.2539294362068176,
+      "learning_rate": 9.254595021551595e-06,
+      "loss": 0.4029,
+      "step": 955
+    },
+    {
+      "epoch": 0.7761840324763194,
+      "grad_norm": 0.27131420373916626,
+      "learning_rate": 9.252110019545581e-06,
+      "loss": 0.3915,
+      "step": 956
+    },
+    {
+      "epoch": 0.7769959404600811,
+      "grad_norm": 0.27614831924438477,
+      "learning_rate": 9.249621217015182e-06,
+      "loss": 0.4287,
+      "step": 957
+    },
+    {
+      "epoch": 0.777807848443843,
+      "grad_norm": 0.24743741750717163,
+      "learning_rate": 9.247128616184881e-06,
+      "loss": 0.3892,
+      "step": 958
+    },
+    {
+      "epoch": 0.7786197564276048,
+      "grad_norm": 0.2628540098667145,
+      "learning_rate": 9.244632219282561e-06,
+      "loss": 0.3826,
+      "step": 959
+    },
+    {
+      "epoch": 0.7794316644113667,
+      "grad_norm": 0.29091012477874756,
+      "learning_rate": 9.242132028539493e-06,
+      "loss": 0.4098,
+      "step": 960
+    },
+    {
+      "epoch": 0.7802435723951285,
+      "grad_norm": 0.26664280891418457,
+      "learning_rate": 9.239628046190342e-06,
+      "loss": 0.3912,
+      "step": 961
+    },
+    {
+      "epoch": 0.7810554803788904,
+      "grad_norm": 0.2686353325843811,
+      "learning_rate": 9.237120274473157e-06,
+      "loss": 0.4093,
+      "step": 962
+    },
+    {
+      "epoch": 0.7818673883626522,
+      "grad_norm": 0.2596835494041443,
+      "learning_rate": 9.23460871562938e-06,
+      "loss": 0.3731,
+      "step": 963
+    },
+    {
+      "epoch": 0.7826792963464141,
+      "grad_norm": 0.26707789301872253,
+      "learning_rate": 9.232093371903836e-06,
+      "loss": 0.38,
+      "step": 964
+    },
+    {
+      "epoch": 0.7834912043301759,
+      "grad_norm": 0.26132428646087646,
+      "learning_rate": 9.229574245544732e-06,
+      "loss": 0.3851,
+      "step": 965
+    },
+    {
+      "epoch": 0.7843031123139378,
+      "grad_norm": 0.32610023021698,
+      "learning_rate": 9.227051338803656e-06,
+      "loss": 0.4148,
+      "step": 966
+    },
+    {
+      "epoch": 0.7851150202976996,
+      "grad_norm": 0.30999842286109924,
+      "learning_rate": 9.224524653935577e-06,
+      "loss": 0.3956,
+      "step": 967
+    },
+    {
+      "epoch": 0.7859269282814615,
+      "grad_norm": 0.27831462025642395,
+      "learning_rate": 9.221994193198835e-06,
+      "loss": 0.3949,
+      "step": 968
+    },
+    {
+      "epoch": 0.7867388362652232,
+      "grad_norm": 0.32212093472480774,
+      "learning_rate": 9.219459958855152e-06,
+      "loss": 0.404,
+      "step": 969
+    },
+    {
+      "epoch": 0.7875507442489851,
+      "grad_norm": 0.2826819121837616,
+      "learning_rate": 9.216921953169618e-06,
+      "loss": 0.3789,
+      "step": 970
+    },
+    {
+      "epoch": 0.7883626522327469,
+      "grad_norm": 0.24644902348518372,
+      "learning_rate": 9.214380178410698e-06,
+      "loss": 0.3911,
+      "step": 971
+    },
+    {
+      "epoch": 0.7891745602165088,
+      "grad_norm": 0.2618508040904999,
+      "learning_rate": 9.211834636850222e-06,
+      "loss": 0.3943,
+      "step": 972
+    },
+    {
+      "epoch": 0.7899864682002706,
+      "grad_norm": 0.2676730453968048,
+      "learning_rate": 9.209285330763388e-06,
+      "loss": 0.4205,
+      "step": 973
+    },
+    {
+      "epoch": 0.7907983761840325,
+      "grad_norm": 0.2913203239440918,
+      "learning_rate": 9.20673226242876e-06,
+      "loss": 0.3763,
+      "step": 974
+    },
+    {
+      "epoch": 0.7916102841677943,
+      "grad_norm": 0.23103663325309753,
+      "learning_rate": 9.204175434128259e-06,
+      "loss": 0.3662,
+      "step": 975
+    },
+    {
+      "epoch": 0.7924221921515562,
+      "grad_norm": 0.26732227206230164,
+      "learning_rate": 9.201614848147179e-06,
+      "loss": 0.3959,
+      "step": 976
+    },
+    {
+      "epoch": 0.793234100135318,
+      "grad_norm": 0.3175308108329773,
+      "learning_rate": 9.199050506774159e-06,
+      "loss": 0.387,
+      "step": 977
+    },
+    {
+      "epoch": 0.7940460081190799,
+      "grad_norm": 0.23762473464012146,
+      "learning_rate": 9.196482412301202e-06,
+      "loss": 0.3787,
+      "step": 978
+    },
+    {
+      "epoch": 0.7948579161028417,
+      "grad_norm": 0.26041853427886963,
+      "learning_rate": 9.193910567023663e-06,
+      "loss": 0.3791,
+      "step": 979
+    },
+    {
+      "epoch": 0.7956698240866035,
+      "grad_norm": 0.29760414361953735,
+      "learning_rate": 9.19133497324025e-06,
+      "loss": 0.3948,
+      "step": 980
+    },
+    {
+      "epoch": 0.7964817320703653,
+      "grad_norm": 0.2656368017196655,
+      "learning_rate": 9.188755633253025e-06,
+      "loss": 0.3925,
+      "step": 981
+    },
+    {
+      "epoch": 0.7972936400541272,
+      "grad_norm": 0.266494482755661,
+      "learning_rate": 9.186172549367394e-06,
+      "loss": 0.3696,
+      "step": 982
+    },
+    {
+      "epoch": 0.798105548037889,
+      "grad_norm": 0.2550007402896881,
+      "learning_rate": 9.183585723892107e-06,
+      "loss": 0.3749,
+      "step": 983
+    },
+    {
+      "epoch": 0.7989174560216509,
+      "grad_norm": 0.2610156536102295,
+      "learning_rate": 9.180995159139265e-06,
+      "loss": 0.4047,
+      "step": 984
+    },
+    {
+      "epoch": 0.7997293640054127,
+      "grad_norm": 0.24212345480918884,
+      "learning_rate": 9.178400857424305e-06,
+      "loss": 0.3782,
+      "step": 985
+    },
+    {
+      "epoch": 0.8005412719891746,
+      "grad_norm": 0.25455161929130554,
+      "learning_rate": 9.175802821066009e-06,
+      "loss": 0.3915,
+      "step": 986
+    },
+    {
+      "epoch": 0.8013531799729364,
+      "grad_norm": 0.2709045112133026,
+      "learning_rate": 9.173201052386496e-06,
+      "loss": 0.3926,
+      "step": 987
+    },
+    {
+      "epoch": 0.8021650879566983,
+      "grad_norm": 0.2679954171180725,
+      "learning_rate": 9.170595553711216e-06,
+      "loss": 0.3886,
+      "step": 988
+    },
+    {
+      "epoch": 0.8029769959404601,
+      "grad_norm": 0.24335704743862152,
+      "learning_rate": 9.167986327368958e-06,
+      "loss": 0.3912,
+      "step": 989
+    },
+    {
+      "epoch": 0.803788903924222,
+      "grad_norm": 0.2922064960002899,
+      "learning_rate": 9.165373375691845e-06,
+      "loss": 0.3798,
+      "step": 990
+    },
+    {
+      "epoch": 0.8046008119079837,
+      "grad_norm": 0.27894696593284607,
+      "learning_rate": 9.16275670101532e-06,
+      "loss": 0.395,
+      "step": 991
+    },
+    {
+      "epoch": 0.8054127198917456,
+      "grad_norm": 0.26642531156539917,
+      "learning_rate": 9.160136305678167e-06,
+      "loss": 0.4016,
+      "step": 992
+    },
+    {
+      "epoch": 0.8062246278755074,
+      "grad_norm": 0.23850536346435547,
+      "learning_rate": 9.15751219202248e-06,
+      "loss": 0.3624,
+      "step": 993
+    },
+    {
+      "epoch": 0.8070365358592693,
+      "grad_norm": 0.24134017527103424,
+      "learning_rate": 9.15488436239369e-06,
+      "loss": 0.3729,
+      "step": 994
+    },
+    {
+      "epoch": 0.8078484438430311,
+      "grad_norm": 0.3051791489124298,
+      "learning_rate": 9.152252819140544e-06,
+      "loss": 0.3927,
+      "step": 995
+    },
+    {
+      "epoch": 0.808660351826793,
+      "grad_norm": 0.2775847017765045,
+      "learning_rate": 9.149617564615106e-06,
+      "loss": 0.3977,
+      "step": 996
+    },
+    {
+      "epoch": 0.8094722598105548,
+      "grad_norm": 0.25210246443748474,
+      "learning_rate": 9.146978601172761e-06,
+      "loss": 0.3856,
+      "step": 997
+    },
+    {
+      "epoch": 0.8102841677943167,
+      "grad_norm": 0.2751011848449707,
+      "learning_rate": 9.144335931172205e-06,
+      "loss": 0.3795,
+      "step": 998
+    },
+    {
+      "epoch": 0.8110960757780785,
+      "grad_norm": 0.2353699803352356,
+      "learning_rate": 9.141689556975453e-06,
+      "loss": 0.3735,
+      "step": 999
+    },
+    {
+      "epoch": 0.8119079837618404,
+      "grad_norm": 0.2807004153728485,
+      "learning_rate": 9.139039480947825e-06,
+      "loss": 0.3922,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8127198917456022,
+      "grad_norm": 0.26704707741737366,
+      "learning_rate": 9.136385705457951e-06,
+      "loss": 0.3893,
+      "step": 1001
+    },
+    {
+      "epoch": 0.813531799729364,
+      "grad_norm": 0.2574257254600525,
+      "learning_rate": 9.13372823287777e-06,
+      "loss": 0.3856,
+      "step": 1002
+    },
+    {
+      "epoch": 0.8143437077131258,
+      "grad_norm": 0.264737069606781,
+      "learning_rate": 9.131067065582521e-06,
+      "loss": 0.4073,
+      "step": 1003
+    },
+    {
+      "epoch": 0.8151556156968877,
+      "grad_norm": 0.24958108365535736,
+      "learning_rate": 9.128402205950751e-06,
+      "loss": 0.3622,
+      "step": 1004
+    },
+    {
+      "epoch": 0.8159675236806495,
+      "grad_norm": 0.2550022602081299,
+      "learning_rate": 9.125733656364304e-06,
+      "loss": 0.414,
+      "step": 1005
+    },
+    {
+      "epoch": 0.8167794316644114,
+      "grad_norm": 0.2698718309402466,
+      "learning_rate": 9.12306141920832e-06,
+      "loss": 0.3773,
+      "step": 1006
+    },
+    {
+      "epoch": 0.8175913396481732,
+      "grad_norm": 0.2695203423500061,
+      "learning_rate": 9.12038549687124e-06,
+      "loss": 0.4013,
+      "step": 1007
+    },
+    {
+      "epoch": 0.8184032476319351,
+      "grad_norm": 0.23414315283298492,
+      "learning_rate": 9.117705891744795e-06,
+      "loss": 0.3702,
+      "step": 1008
+    },
+    {
+      "epoch": 0.8192151556156969,
+      "grad_norm": 0.27482977509498596,
+      "learning_rate": 9.115022606224008e-06,
+      "loss": 0.3857,
+      "step": 1009
+    },
+    {
+      "epoch": 0.8200270635994588,
+      "grad_norm": 0.23628383874893188,
+      "learning_rate": 9.112335642707196e-06,
+      "loss": 0.3967,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8208389715832206,
+      "grad_norm": 0.2581934630870819,
+      "learning_rate": 9.109645003595954e-06,
+      "loss": 0.3962,
+      "step": 1011
+    },
+    {
+      "epoch": 0.8216508795669825,
+      "grad_norm": 0.23855005204677582,
+      "learning_rate": 9.106950691295172e-06,
+      "loss": 0.3996,
+      "step": 1012
+    },
+    {
+      "epoch": 0.8224627875507442,
+      "grad_norm": 0.27503228187561035,
+      "learning_rate": 9.104252708213018e-06,
+      "loss": 0.3886,
+      "step": 1013
+    },
+    {
+      "epoch": 0.823274695534506,
+      "grad_norm": 0.24632956087589264,
+      "learning_rate": 9.101551056760942e-06,
+      "loss": 0.3954,
+      "step": 1014
+    },
+    {
+      "epoch": 0.8240866035182679,
+      "grad_norm": 0.2645036578178406,
+      "learning_rate": 9.098845739353672e-06,
+      "loss": 0.3973,
+      "step": 1015
+    },
+    {
+      "epoch": 0.8248985115020298,
+      "grad_norm": 0.25859469175338745,
+      "learning_rate": 9.096136758409212e-06,
+      "loss": 0.4002,
+      "step": 1016
+    },
+    {
+      "epoch": 0.8257104194857916,
+      "grad_norm": 0.23207376897335052,
+      "learning_rate": 9.093424116348846e-06,
+      "loss": 0.3672,
+      "step": 1017
+    },
+    {
+      "epoch": 0.8265223274695535,
+      "grad_norm": 0.27344396710395813,
+      "learning_rate": 9.090707815597124e-06,
+      "loss": 0.4069,
+      "step": 1018
+    },
+    {
+      "epoch": 0.8273342354533153,
+      "grad_norm": 0.6669845581054688,
+      "learning_rate": 9.087987858581867e-06,
+      "loss": 0.3981,
+      "step": 1019
+    },
+    {
+      "epoch": 0.8281461434370772,
+      "grad_norm": 0.2469099462032318,
+      "learning_rate": 9.085264247734165e-06,
+      "loss": 0.3874,
+      "step": 1020
+    },
+    {
+      "epoch": 0.828958051420839,
+      "grad_norm": 0.2474839985370636,
+      "learning_rate": 9.082536985488377e-06,
+      "loss": 0.3925,
+      "step": 1021
+    },
+    {
+      "epoch": 0.8297699594046009,
+      "grad_norm": 0.2517271339893341,
+      "learning_rate": 9.07980607428212e-06,
+      "loss": 0.372,
+      "step": 1022
+    },
+    {
+      "epoch": 0.8305818673883627,
+      "grad_norm": 0.34447553753852844,
+      "learning_rate": 9.077071516556276e-06,
+      "loss": 0.3939,
+      "step": 1023
+    },
+    {
+      "epoch": 0.8313937753721244,
+      "grad_norm": 0.2683957517147064,
+      "learning_rate": 9.074333314754984e-06,
+      "loss": 0.3737,
+      "step": 1024
+    },
+    {
+      "epoch": 0.8322056833558863,
+      "grad_norm": 0.28619056940078735,
+      "learning_rate": 9.071591471325644e-06,
+      "loss": 0.3797,
+      "step": 1025
+    },
+    {
+      "epoch": 0.8330175913396481,
+      "grad_norm": 0.24576127529144287,
+      "learning_rate": 9.068845988718906e-06,
+      "loss": 0.4048,
+      "step": 1026
+    },
+    {
+      "epoch": 0.83382949932341,
+      "grad_norm": 0.23908191919326782,
+      "learning_rate": 9.066096869388674e-06,
+      "loss": 0.3839,
+      "step": 1027
+    },
+    {
+      "epoch": 0.8346414073071718,
+      "grad_norm": 0.24462969601154327,
+      "learning_rate": 9.063344115792108e-06,
+      "loss": 0.3808,
+      "step": 1028
+    },
+    {
+      "epoch": 0.8354533152909337,
+      "grad_norm": 0.2215019017457962,
+      "learning_rate": 9.060587730389606e-06,
+      "loss": 0.3802,
+      "step": 1029
+    },
+    {
+      "epoch": 0.8362652232746955,
+      "grad_norm": 0.25166958570480347,
+      "learning_rate": 9.05782771564482e-06,
+      "loss": 0.3894,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8370771312584574,
+      "grad_norm": 0.2516234219074249,
+      "learning_rate": 9.055064074024646e-06,
+      "loss": 0.4149,
+      "step": 1031
+    },
+    {
+      "epoch": 0.8378890392422192,
+      "grad_norm": 0.2796178460121155,
+      "learning_rate": 9.052296807999216e-06,
+      "loss": 0.422,
+      "step": 1032
+    },
+    {
+      "epoch": 0.8387009472259811,
+      "grad_norm": 0.2652052342891693,
+      "learning_rate": 9.049525920041909e-06,
+      "loss": 0.3783,
+      "step": 1033
+    },
+    {
+      "epoch": 0.839512855209743,
+      "grad_norm": 0.2469363808631897,
+      "learning_rate": 9.046751412629331e-06,
+      "loss": 0.3946,
+      "step": 1034
+    },
+    {
+      "epoch": 0.8403247631935047,
+      "grad_norm": 0.2655825614929199,
+      "learning_rate": 9.043973288241334e-06,
+      "loss": 0.3834,
+      "step": 1035
+    },
+    {
+      "epoch": 0.8411366711772665,
+      "grad_norm": 0.27901291847229004,
+      "learning_rate": 9.041191549360995e-06,
+      "loss": 0.3933,
+      "step": 1036
+    },
+    {
+      "epoch": 0.8419485791610284,
+      "grad_norm": 0.2793371081352234,
+      "learning_rate": 9.038406198474627e-06,
+      "loss": 0.4036,
+      "step": 1037
+    },
+    {
+      "epoch": 0.8427604871447902,
+      "grad_norm": 4.1689043045043945,
+      "learning_rate": 9.03561723807177e-06,
+      "loss": 0.4094,
+      "step": 1038
+    },
+    {
+      "epoch": 0.8435723951285521,
+      "grad_norm": 0.2881360352039337,
+      "learning_rate": 9.032824670645187e-06,
+      "loss": 0.3989,
+      "step": 1039
+    },
+    {
+      "epoch": 0.8443843031123139,
+      "grad_norm": 0.28218287229537964,
+      "learning_rate": 9.030028498690866e-06,
+      "loss": 0.3816,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8451962110960758,
+      "grad_norm": 0.2516503930091858,
+      "learning_rate": 9.027228724708022e-06,
+      "loss": 0.4054,
+      "step": 1041
+    },
+    {
+      "epoch": 0.8460081190798376,
+      "grad_norm": 0.264393150806427,
+      "learning_rate": 9.02442535119908e-06,
+      "loss": 0.3845,
+      "step": 1042
+    },
+    {
+      "epoch": 0.8468200270635995,
+      "grad_norm": 0.2540109157562256,
+      "learning_rate": 9.021618380669693e-06,
+      "loss": 0.3699,
+      "step": 1043
+    },
+    {
+      "epoch": 0.8476319350473613,
+      "grad_norm": 0.247217059135437,
+      "learning_rate": 9.018807815628721e-06,
+      "loss": 0.3918,
+      "step": 1044
+    },
+    {
+      "epoch": 0.8484438430311232,
+      "grad_norm": 0.26987186074256897,
+      "learning_rate": 9.01599365858824e-06,
+      "loss": 0.4109,
+      "step": 1045
+    },
+    {
+      "epoch": 0.8492557510148849,
+      "grad_norm": 0.26813167333602905,
+      "learning_rate": 9.013175912063534e-06,
+      "loss": 0.3852,
+      "step": 1046
+    },
+    {
+      "epoch": 0.8500676589986468,
+      "grad_norm": 0.24266992509365082,
+      "learning_rate": 9.0103545785731e-06,
+      "loss": 0.4158,
+      "step": 1047
+    },
+    {
+      "epoch": 0.8508795669824086,
+      "grad_norm": 0.2700927257537842,
+      "learning_rate": 9.007529660638637e-06,
+      "loss": 0.3822,
+      "step": 1048
+    },
+    {
+      "epoch": 0.8516914749661705,
+      "grad_norm": 0.2850741147994995,
+      "learning_rate": 9.004701160785047e-06,
+      "loss": 0.3941,
+      "step": 1049
+    },
+    {
+      "epoch": 0.8525033829499323,
+      "grad_norm": 0.27000558376312256,
+      "learning_rate": 9.00186908154044e-06,
+      "loss": 0.4118,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8533152909336942,
+      "grad_norm": 0.26864269375801086,
+      "learning_rate": 8.999033425436116e-06,
+      "loss": 0.3765,
+      "step": 1051
+    },
+    {
+      "epoch": 0.854127198917456,
+      "grad_norm": 0.2598525583744049,
+      "learning_rate": 8.996194195006582e-06,
+      "loss": 0.392,
+      "step": 1052
+    },
+    {
+      "epoch": 0.8549391069012179,
+      "grad_norm": 0.2682839334011078,
+      "learning_rate": 8.993351392789529e-06,
+      "loss": 0.3878,
+      "step": 1053
+    },
+    {
+      "epoch": 0.8557510148849797,
+      "grad_norm": 0.28347694873809814,
+      "learning_rate": 8.990505021325849e-06,
+      "loss": 0.3893,
+      "step": 1054
+    },
+    {
+      "epoch": 0.8565629228687416,
+      "grad_norm": 0.26080405712127686,
+      "learning_rate": 8.987655083159618e-06,
+      "loss": 0.3862,
+      "step": 1055
+    },
+    {
+      "epoch": 0.8573748308525034,
+      "grad_norm": 0.28711503744125366,
+      "learning_rate": 8.984801580838109e-06,
+      "loss": 0.3932,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8581867388362652,
+      "grad_norm": 0.24684882164001465,
+      "learning_rate": 8.98194451691177e-06,
+      "loss": 0.3894,
+      "step": 1057
+    },
+    {
+      "epoch": 0.858998646820027,
+      "grad_norm": 0.2611309587955475,
+      "learning_rate": 8.979083893934236e-06,
+      "loss": 0.4032,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8598105548037889,
+      "grad_norm": 0.3068278729915619,
+      "learning_rate": 8.976219714462326e-06,
+      "loss": 0.3903,
+      "step": 1059
+    },
+    {
+      "epoch": 0.8606224627875507,
+      "grad_norm": 0.28028175234794617,
+      "learning_rate": 8.973351981056037e-06,
+      "loss": 0.3894,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8614343707713126,
+      "grad_norm": 0.2739061117172241,
+      "learning_rate": 8.97048069627854e-06,
+      "loss": 0.3888,
+      "step": 1061
+    },
+    {
+      "epoch": 0.8622462787550744,
+      "grad_norm": 0.262588769197464,
+      "learning_rate": 8.967605862696183e-06,
+      "loss": 0.4018,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8630581867388363,
+      "grad_norm": 0.2792397439479828,
+      "learning_rate": 8.964727482878482e-06,
+      "loss": 0.3752,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8638700947225981,
+      "grad_norm": 0.26258668303489685,
+      "learning_rate": 8.961845559398125e-06,
+      "loss": 0.3928,
+      "step": 1064
+    },
+    {
+      "epoch": 0.86468200270636,
+      "grad_norm": 0.2593936622142792,
+      "learning_rate": 8.95896009483097e-06,
+      "loss": 0.3737,
+      "step": 1065
+    },
+    {
+      "epoch": 0.8654939106901218,
+      "grad_norm": 0.25506865978240967,
+      "learning_rate": 8.956071091756036e-06,
+      "loss": 0.402,
+      "step": 1066
+    },
+    {
+      "epoch": 0.8663058186738837,
+      "grad_norm": 0.2582107186317444,
+      "learning_rate": 8.953178552755506e-06,
+      "loss": 0.3957,
+      "step": 1067
+    },
+    {
+      "epoch": 0.8671177266576454,
+      "grad_norm": 0.2756621241569519,
+      "learning_rate": 8.950282480414723e-06,
+      "loss": 0.3664,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8679296346414073,
+      "grad_norm": 0.25939658284187317,
+      "learning_rate": 8.947382877322189e-06,
+      "loss": 0.3669,
+      "step": 1069
+    },
+    {
+      "epoch": 0.8687415426251691,
+      "grad_norm": 0.28070923686027527,
+      "learning_rate": 8.94447974606956e-06,
+      "loss": 0.3996,
+      "step": 1070
+    },
+    {
+      "epoch": 0.869553450608931,
+      "grad_norm": 0.23480218648910522,
+      "learning_rate": 8.941573089251652e-06,
+      "loss": 0.3968,
+      "step": 1071
+    },
+    {
+      "epoch": 0.8703653585926928,
+      "grad_norm": 0.26618343591690063,
+      "learning_rate": 8.93866290946642e-06,
+      "loss": 0.3781,
+      "step": 1072
+    },
+    {
+      "epoch": 0.8711772665764547,
+      "grad_norm": 0.2347402721643448,
+      "learning_rate": 8.935749209314981e-06,
+      "loss": 0.3828,
+      "step": 1073
+    },
+    {
+      "epoch": 0.8719891745602165,
+      "grad_norm": 0.26961296796798706,
+      "learning_rate": 8.932831991401587e-06,
+      "loss": 0.3832,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8728010825439784,
+      "grad_norm": 0.26282238960266113,
+      "learning_rate": 8.929911258333644e-06,
+      "loss": 0.3999,
+      "step": 1075
+    },
+    {
+      "epoch": 0.8736129905277402,
+      "grad_norm": 0.2269526869058609,
+      "learning_rate": 8.92698701272169e-06,
+      "loss": 0.4169,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8744248985115021,
+      "grad_norm": 0.26221898198127747,
+      "learning_rate": 8.924059257179414e-06,
+      "loss": 0.3758,
+      "step": 1077
+    },
+    {
+      "epoch": 0.8752368064952639,
+      "grad_norm": 0.28100213408470154,
+      "learning_rate": 8.92112799432363e-06,
+      "loss": 0.4029,
+      "step": 1078
+    },
+    {
+      "epoch": 0.8760487144790257,
+      "grad_norm": 0.23895332217216492,
+      "learning_rate": 8.9181932267743e-06,
+      "loss": 0.3791,
+      "step": 1079
+    },
+    {
+      "epoch": 0.8768606224627875,
+      "grad_norm": 0.24587878584861755,
+      "learning_rate": 8.915254957154501e-06,
+      "loss": 0.3825,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8776725304465494,
+      "grad_norm": 0.26791101694107056,
+      "learning_rate": 8.91231318809046e-06,
+      "loss": 0.3757,
+      "step": 1081
+    },
+    {
+      "epoch": 0.8784844384303112,
+      "grad_norm": 0.24598461389541626,
+      "learning_rate": 8.909367922211519e-06,
+      "loss": 0.3657,
+      "step": 1082
+    },
+    {
+      "epoch": 0.8792963464140731,
+      "grad_norm": 0.2681296765804291,
+      "learning_rate": 8.906419162150147e-06,
+      "loss": 0.3835,
+      "step": 1083
+    },
+    {
+      "epoch": 0.8801082543978349,
+      "grad_norm": 0.23437170684337616,
+      "learning_rate": 8.903466910541938e-06,
+      "loss": 0.3988,
+      "step": 1084
+    },
+    {
+      "epoch": 0.8809201623815968,
+      "grad_norm": 0.25690770149230957,
+      "learning_rate": 8.900511170025609e-06,
+      "loss": 0.3854,
+      "step": 1085
+    },
+    {
+      "epoch": 0.8817320703653586,
+      "grad_norm": 0.254558801651001,
+      "learning_rate": 8.89755194324299e-06,
+      "loss": 0.3891,
+      "step": 1086
+    },
+    {
+      "epoch": 0.8825439783491205,
+      "grad_norm": 0.2740063965320587,
+      "learning_rate": 8.89458923283903e-06,
+      "loss": 0.3825,
+      "step": 1087
+    },
+    {
+      "epoch": 0.8833558863328823,
+      "grad_norm": 0.24815283715724945,
+      "learning_rate": 8.891623041461794e-06,
+      "loss": 0.3874,
+      "step": 1088
+    },
+    {
+      "epoch": 0.8841677943166442,
+      "grad_norm": 0.2757580280303955,
+      "learning_rate": 8.888653371762454e-06,
+      "loss": 0.3799,
+      "step": 1089
+    },
+    {
+      "epoch": 0.8849797023004059,
+      "grad_norm": 0.2419026494026184,
+      "learning_rate": 8.885680226395295e-06,
+      "loss": 0.3922,
+      "step": 1090
+    },
+    {
+      "epoch": 0.8857916102841678,
+      "grad_norm": 0.25642162561416626,
+      "learning_rate": 8.882703608017702e-06,
+      "loss": 0.39,
+      "step": 1091
+    },
+    {
+      "epoch": 0.8866035182679296,
+      "grad_norm": 0.24010175466537476,
+      "learning_rate": 8.87972351929017e-06,
+      "loss": 0.3632,
+      "step": 1092
+    },
+    {
+      "epoch": 0.8874154262516915,
+      "grad_norm": 0.2585316598415375,
+      "learning_rate": 8.876739962876298e-06,
+      "loss": 0.4007,
+      "step": 1093
+    },
+    {
+      "epoch": 0.8882273342354533,
+      "grad_norm": 0.2339038997888565,
+      "learning_rate": 8.873752941442775e-06,
+      "loss": 0.3879,
+      "step": 1094
+    },
+    {
+      "epoch": 0.8890392422192152,
+      "grad_norm": 0.24977758526802063,
+      "learning_rate": 8.870762457659397e-06,
+      "loss": 0.3821,
+      "step": 1095
+    },
+    {
+      "epoch": 0.889851150202977,
+      "grad_norm": 0.2920962870121002,
+      "learning_rate": 8.867768514199046e-06,
+      "loss": 0.394,
+      "step": 1096
+    },
+    {
+      "epoch": 0.8906630581867389,
+      "grad_norm": 0.28696128726005554,
+      "learning_rate": 8.864771113737705e-06,
+      "loss": 0.4078,
+      "step": 1097
+    },
+    {
+      "epoch": 0.8914749661705007,
+      "grad_norm": 0.26207202672958374,
+      "learning_rate": 8.86177025895444e-06,
+      "loss": 0.3751,
+      "step": 1098
+    },
+    {
+      "epoch": 0.8922868741542626,
+      "grad_norm": 0.28151771426200867,
+      "learning_rate": 8.858765952531407e-06,
+      "loss": 0.3765,
+      "step": 1099
+    },
+    {
+      "epoch": 0.8930987821380244,
+      "grad_norm": 0.27616775035858154,
+      "learning_rate": 8.855758197153848e-06,
+      "loss": 0.4072,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8939106901217861,
+      "grad_norm": 0.26500093936920166,
+      "learning_rate": 8.852746995510084e-06,
+      "loss": 0.3906,
+      "step": 1101
+    },
+    {
+      "epoch": 0.894722598105548,
+      "grad_norm": 0.2739506661891937,
+      "learning_rate": 8.849732350291522e-06,
+      "loss": 0.3907,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8955345060893098,
+      "grad_norm": 0.26735809445381165,
+      "learning_rate": 8.846714264192642e-06,
+      "loss": 0.387,
+      "step": 1103
+    },
+    {
+      "epoch": 0.8963464140730717,
+      "grad_norm": 0.2642000615596771,
+      "learning_rate": 8.843692739911e-06,
+      "loss": 0.375,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8971583220568335,
+      "grad_norm": 0.2518375813961029,
+      "learning_rate": 8.84066778014723e-06,
+      "loss": 0.3839,
+      "step": 1105
+    },
+    {
+      "epoch": 0.8979702300405954,
+      "grad_norm": 0.2740698754787445,
+      "learning_rate": 8.837639387605031e-06,
+      "loss": 0.4147,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8987821380243572,
+      "grad_norm": 0.27368369698524475,
+      "learning_rate": 8.834607564991168e-06,
+      "loss": 0.397,
+      "step": 1107
+    },
+    {
+      "epoch": 0.8995940460081191,
+      "grad_norm": 0.26152658462524414,
+      "learning_rate": 8.831572315015484e-06,
+      "loss": 0.3745,
+      "step": 1108
+    },
+    {
+      "epoch": 0.900405953991881,
+      "grad_norm": 0.26505064964294434,
+      "learning_rate": 8.828533640390868e-06,
+      "loss": 0.3903,
+      "step": 1109
+    },
+    {
+      "epoch": 0.9012178619756428,
+      "grad_norm": 0.3119455575942993,
+      "learning_rate": 8.825491543833286e-06,
+      "loss": 0.3996,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9020297699594046,
+      "grad_norm": 0.26736173033714294,
+      "learning_rate": 8.822446028061752e-06,
+      "loss": 0.385,
+      "step": 1111
+    },
+    {
+      "epoch": 0.9028416779431664,
+      "grad_norm": 0.24869173765182495,
+      "learning_rate": 8.819397095798343e-06,
+      "loss": 0.3876,
+      "step": 1112
+    },
+    {
+      "epoch": 0.9036535859269282,
+      "grad_norm": 0.273971825838089,
+      "learning_rate": 8.816344749768184e-06,
+      "loss": 0.3734,
+      "step": 1113
+    },
+    {
+      "epoch": 0.9044654939106901,
+      "grad_norm": 0.27973851561546326,
+      "learning_rate": 8.813288992699454e-06,
+      "loss": 0.3966,
+      "step": 1114
+    },
+    {
+      "epoch": 0.9052774018944519,
+      "grad_norm": 0.25686103105545044,
+      "learning_rate": 8.810229827323378e-06,
+      "loss": 0.3955,
+      "step": 1115
+    },
+    {
+      "epoch": 0.9060893098782138,
+      "grad_norm": 0.26973727345466614,
+      "learning_rate": 8.807167256374234e-06,
+      "loss": 0.3813,
+      "step": 1116
+    },
+    {
+      "epoch": 0.9069012178619756,
+      "grad_norm": 0.2658529579639435,
+      "learning_rate": 8.804101282589338e-06,
+      "loss": 0.3817,
+      "step": 1117
+    },
+    {
+      "epoch": 0.9077131258457375,
+      "grad_norm": 0.2552364468574524,
+      "learning_rate": 8.801031908709047e-06,
+      "loss": 0.3672,
+      "step": 1118
+    },
+    {
+      "epoch": 0.9085250338294993,
+      "grad_norm": 0.28111568093299866,
+      "learning_rate": 8.797959137476763e-06,
+      "loss": 0.3896,
+      "step": 1119
+    },
+    {
+      "epoch": 0.9093369418132612,
+      "grad_norm": 0.26544660329818726,
+      "learning_rate": 8.794882971638916e-06,
+      "loss": 0.38,
+      "step": 1120
+    },
+    {
+      "epoch": 0.910148849797023,
+      "grad_norm": 0.26193979382514954,
+      "learning_rate": 8.791803413944979e-06,
+      "loss": 0.3903,
+      "step": 1121
+    },
+    {
+      "epoch": 0.9109607577807849,
+      "grad_norm": 0.29192590713500977,
+      "learning_rate": 8.78872046714745e-06,
+      "loss": 0.3885,
+      "step": 1122
+    },
+    {
+      "epoch": 0.9117726657645466,
+      "grad_norm": 0.29017141461372375,
+      "learning_rate": 8.785634134001856e-06,
+      "loss": 0.4202,
+      "step": 1123
+    },
+    {
+      "epoch": 0.9125845737483085,
+      "grad_norm": 0.28225943446159363,
+      "learning_rate": 8.782544417266758e-06,
+      "loss": 0.3912,
+      "step": 1124
+    },
+    {
+      "epoch": 0.9133964817320703,
+      "grad_norm": 0.2830917537212372,
+      "learning_rate": 8.779451319703735e-06,
+      "loss": 0.38,
+      "step": 1125
+    },
+    {
+      "epoch": 0.9142083897158322,
+      "grad_norm": 0.3120962381362915,
+      "learning_rate": 8.776354844077389e-06,
+      "loss": 0.3886,
+      "step": 1126
+    },
+    {
+      "epoch": 0.915020297699594,
+      "grad_norm": 0.27589544653892517,
+      "learning_rate": 8.77325499315534e-06,
+      "loss": 0.3921,
+      "step": 1127
+    },
+    {
+      "epoch": 0.9158322056833559,
+      "grad_norm": 0.2453465759754181,
+      "learning_rate": 8.770151769708226e-06,
+      "loss": 0.3979,
+      "step": 1128
+    },
+    {
+      "epoch": 0.9166441136671177,
+      "grad_norm": 0.306486040353775,
+      "learning_rate": 8.767045176509703e-06,
+      "loss": 0.3737,
+      "step": 1129
+    },
+    {
+      "epoch": 0.9174560216508796,
+      "grad_norm": 0.2780085504055023,
+      "learning_rate": 8.763935216336433e-06,
+      "loss": 0.4083,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9182679296346414,
+      "grad_norm": 0.2781031131744385,
+      "learning_rate": 8.76082189196809e-06,
+      "loss": 0.3927,
+      "step": 1131
+    },
+    {
+      "epoch": 0.9190798376184033,
+      "grad_norm": 0.26547306776046753,
+      "learning_rate": 8.757705206187357e-06,
+      "loss": 0.3765,
+      "step": 1132
+    },
+    {
+      "epoch": 0.9198917456021651,
+      "grad_norm": 0.2923036217689514,
+      "learning_rate": 8.754585161779916e-06,
+      "loss": 0.3755,
+      "step": 1133
+    },
+    {
+      "epoch": 0.9207036535859269,
+      "grad_norm": 0.2539641261100769,
+      "learning_rate": 8.751461761534457e-06,
+      "loss": 0.3859,
+      "step": 1134
+    },
+    {
+      "epoch": 0.9215155615696887,
+      "grad_norm": 0.28955936431884766,
+      "learning_rate": 8.748335008242667e-06,
+      "loss": 0.398,
+      "step": 1135
+    },
+    {
+      "epoch": 0.9223274695534506,
+      "grad_norm": 0.2661500871181488,
+      "learning_rate": 8.745204904699226e-06,
+      "loss": 0.3777,
+      "step": 1136
+    },
+    {
+      "epoch": 0.9231393775372124,
+      "grad_norm": 0.31144532561302185,
+      "learning_rate": 8.742071453701815e-06,
+      "loss": 0.3972,
+      "step": 1137
+    },
+    {
+      "epoch": 0.9239512855209743,
+      "grad_norm": 0.2706797420978546,
+      "learning_rate": 8.738934658051103e-06,
+      "loss": 0.365,
+      "step": 1138
+    },
+    {
+      "epoch": 0.9247631935047361,
+      "grad_norm": 0.2780349552631378,
+      "learning_rate": 8.73579452055075e-06,
+      "loss": 0.3914,
+      "step": 1139
+    },
+    {
+      "epoch": 0.925575101488498,
+      "grad_norm": 0.2859494686126709,
+      "learning_rate": 8.732651044007402e-06,
+      "loss": 0.4046,
+      "step": 1140
+    },
+    {
+      "epoch": 0.9263870094722598,
+      "grad_norm": 0.29466187953948975,
+      "learning_rate": 8.72950423123069e-06,
+      "loss": 0.3904,
+      "step": 1141
+    },
+    {
+      "epoch": 0.9271989174560217,
+      "grad_norm": 0.271281361579895,
+      "learning_rate": 8.726354085033227e-06,
+      "loss": 0.369,
+      "step": 1142
+    },
+    {
+      "epoch": 0.9280108254397835,
+      "grad_norm": 0.25415897369384766,
+      "learning_rate": 8.723200608230605e-06,
+      "loss": 0.3806,
+      "step": 1143
+    },
+    {
+      "epoch": 0.9288227334235454,
+      "grad_norm": 0.26181578636169434,
+      "learning_rate": 8.72004380364139e-06,
+      "loss": 0.3614,
+      "step": 1144
+    },
+    {
+      "epoch": 0.9296346414073072,
+      "grad_norm": 0.2862527370452881,
+      "learning_rate": 8.71688367408713e-06,
+      "loss": 0.3649,
+      "step": 1145
+    },
+    {
+      "epoch": 0.930446549391069,
+      "grad_norm": 0.28935208916664124,
+      "learning_rate": 8.713720222392338e-06,
+      "loss": 0.3688,
+      "step": 1146
+    },
+    {
+      "epoch": 0.9312584573748308,
+      "grad_norm": 0.2537871301174164,
+      "learning_rate": 8.710553451384499e-06,
+      "loss": 0.4009,
+      "step": 1147
+    },
+    {
+      "epoch": 0.9320703653585927,
+      "grad_norm": 0.27714279294013977,
+      "learning_rate": 8.707383363894064e-06,
+      "loss": 0.3993,
+      "step": 1148
+    },
+    {
+      "epoch": 0.9328822733423545,
+      "grad_norm": 0.253688246011734,
+      "learning_rate": 8.704209962754448e-06,
+      "loss": 0.3774,
+      "step": 1149
+    },
+    {
+      "epoch": 0.9336941813261164,
+      "grad_norm": 0.25840944051742554,
+      "learning_rate": 8.70103325080203e-06,
+      "loss": 0.373,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9345060893098782,
+      "grad_norm": 0.2704528272151947,
+      "learning_rate": 8.697853230876145e-06,
+      "loss": 0.4255,
+      "step": 1151
+    },
+    {
+      "epoch": 0.9353179972936401,
+      "grad_norm": 0.2830778956413269,
+      "learning_rate": 8.694669905819087e-06,
+      "loss": 0.3734,
+      "step": 1152
+    },
+    {
+      "epoch": 0.9361299052774019,
+      "grad_norm": 0.2583657205104828,
+      "learning_rate": 8.691483278476103e-06,
+      "loss": 0.3826,
+      "step": 1153
+    },
+    {
+      "epoch": 0.9369418132611638,
+      "grad_norm": 0.2756957709789276,
+      "learning_rate": 8.688293351695392e-06,
+      "loss": 0.3847,
+      "step": 1154
+    },
+    {
+      "epoch": 0.9377537212449256,
+      "grad_norm": 0.2781772315502167,
+      "learning_rate": 8.685100128328101e-06,
+      "loss": 0.4196,
+      "step": 1155
+    },
+    {
+      "epoch": 0.9385656292286875,
+      "grad_norm": 0.25186172127723694,
+      "learning_rate": 8.681903611228327e-06,
+      "loss": 0.3768,
+      "step": 1156
+    },
+    {
+      "epoch": 0.9393775372124492,
+      "grad_norm": 0.2761591374874115,
+      "learning_rate": 8.678703803253103e-06,
+      "loss": 0.4151,
+      "step": 1157
+    },
+    {
+      "epoch": 0.9401894451962111,
+      "grad_norm": 0.26894858479499817,
+      "learning_rate": 8.675500707262415e-06,
+      "loss": 0.3994,
+      "step": 1158
+    },
+    {
+      "epoch": 0.9410013531799729,
+      "grad_norm": 0.2716839611530304,
+      "learning_rate": 8.672294326119177e-06,
+      "loss": 0.4057,
+      "step": 1159
+    },
+    {
+      "epoch": 0.9418132611637348,
+      "grad_norm": 0.2808055281639099,
+      "learning_rate": 8.669084662689246e-06,
+      "loss": 0.3848,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9426251691474966,
+      "grad_norm": 0.26543232798576355,
+      "learning_rate": 8.66587171984141e-06,
+      "loss": 0.3866,
+      "step": 1161
+    },
+    {
+      "epoch": 0.9434370771312585,
+      "grad_norm": 0.27344685792922974,
+      "learning_rate": 8.662655500447388e-06,
+      "loss": 0.3709,
+      "step": 1162
+    },
+    {
+      "epoch": 0.9442489851150203,
+      "grad_norm": 0.26473289728164673,
+      "learning_rate": 8.65943600738183e-06,
+      "loss": 0.3812,
+      "step": 1163
+    },
+    {
+      "epoch": 0.9450608930987822,
+      "grad_norm": 0.27634522318840027,
+      "learning_rate": 8.656213243522307e-06,
+      "loss": 0.3814,
+      "step": 1164
+    },
+    {
+      "epoch": 0.945872801082544,
+      "grad_norm": 0.30250585079193115,
+      "learning_rate": 8.65298721174932e-06,
+      "loss": 0.3782,
+      "step": 1165
+    },
+    {
+      "epoch": 0.9466847090663059,
+      "grad_norm": 0.26937004923820496,
+      "learning_rate": 8.649757914946284e-06,
+      "loss": 0.4073,
+      "step": 1166
+    },
+    {
+      "epoch": 0.9474966170500677,
+      "grad_norm": 0.2870255410671234,
+      "learning_rate": 8.646525355999538e-06,
+      "loss": 0.3661,
+      "step": 1167
+    },
+    {
+      "epoch": 0.9483085250338295,
+      "grad_norm": 0.30247625708580017,
+      "learning_rate": 8.643289537798332e-06,
+      "loss": 0.3717,
+      "step": 1168
+    },
+    {
+      "epoch": 0.9491204330175913,
+      "grad_norm": 0.2880720794200897,
+      "learning_rate": 8.640050463234837e-06,
+      "loss": 0.3913,
+      "step": 1169
+    },
+    {
+      "epoch": 0.9499323410013532,
+      "grad_norm": 0.31085067987442017,
+      "learning_rate": 8.636808135204123e-06,
+      "loss": 0.3909,
+      "step": 1170
+    },
+    {
+      "epoch": 0.950744248985115,
+      "grad_norm": 0.28657767176628113,
+      "learning_rate": 8.633562556604177e-06,
+      "loss": 0.3971,
+      "step": 1171
+    },
+    {
+      "epoch": 0.9515561569688769,
+      "grad_norm": 0.3245067000389099,
+      "learning_rate": 8.630313730335888e-06,
+      "loss": 0.4022,
+      "step": 1172
+    },
+    {
+      "epoch": 0.9523680649526387,
+      "grad_norm": 0.2987765967845917,
+      "learning_rate": 8.627061659303047e-06,
+      "loss": 0.3831,
+      "step": 1173
+    },
+    {
+      "epoch": 0.9531799729364006,
+      "grad_norm": 0.2876870334148407,
+      "learning_rate": 8.623806346412349e-06,
+      "loss": 0.37,
+      "step": 1174
+    },
+    {
+      "epoch": 0.9539918809201624,
+      "grad_norm": 0.3034891188144684,
+      "learning_rate": 8.62054779457338e-06,
+      "loss": 0.3607,
+      "step": 1175
+    },
+    {
+      "epoch": 0.9548037889039243,
+      "grad_norm": 0.2737765312194824,
+      "learning_rate": 8.617286006698628e-06,
+      "loss": 0.411,
+      "step": 1176
+    },
+    {
+      "epoch": 0.9556156968876861,
+      "grad_norm": 0.2512311339378357,
+      "learning_rate": 8.614020985703468e-06,
+      "loss": 0.3877,
+      "step": 1177
+    },
+    {
+      "epoch": 0.956427604871448,
+      "grad_norm": 0.2541670799255371,
+      "learning_rate": 8.610752734506166e-06,
+      "loss": 0.4014,
+      "step": 1178
+    },
+    {
+      "epoch": 0.9572395128552097,
+      "grad_norm": 0.29216232895851135,
+      "learning_rate": 8.60748125602788e-06,
+      "loss": 0.3689,
+      "step": 1179
+    },
+    {
+      "epoch": 0.9580514208389715,
+      "grad_norm": 0.2463970184326172,
+      "learning_rate": 8.604206553192647e-06,
+      "loss": 0.391,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9588633288227334,
+      "grad_norm": 0.25518643856048584,
+      "learning_rate": 8.600928628927385e-06,
+      "loss": 0.4207,
+      "step": 1181
+    },
+    {
+      "epoch": 0.9596752368064952,
+      "grad_norm": 0.2640528678894043,
+      "learning_rate": 8.597647486161898e-06,
+      "loss": 0.3878,
+      "step": 1182
+    },
+    {
+      "epoch": 0.9604871447902571,
+      "grad_norm": 0.28012311458587646,
+      "learning_rate": 8.594363127828858e-06,
+      "loss": 0.3806,
+      "step": 1183
+    },
+    {
+      "epoch": 0.961299052774019,
+      "grad_norm": 0.24889670312404633,
+      "learning_rate": 8.591075556863818e-06,
+      "loss": 0.3735,
+      "step": 1184
+    },
+    {
+      "epoch": 0.9621109607577808,
+      "grad_norm": 0.2636782228946686,
+      "learning_rate": 8.5877847762052e-06,
+      "loss": 0.3937,
+      "step": 1185
+    },
+    {
+      "epoch": 0.9629228687415426,
+      "grad_norm": 0.39641010761260986,
+      "learning_rate": 8.584490788794296e-06,
+      "loss": 0.3831,
+      "step": 1186
+    },
+    {
+      "epoch": 0.9637347767253045,
+      "grad_norm": 0.2645554840564728,
+      "learning_rate": 8.581193597575259e-06,
+      "loss": 0.3771,
+      "step": 1187
+    },
+    {
+      "epoch": 0.9645466847090663,
+      "grad_norm": 0.2658025622367859,
+      "learning_rate": 8.577893205495113e-06,
+      "loss": 0.3755,
+      "step": 1188
+    },
+    {
+      "epoch": 0.9653585926928282,
+      "grad_norm": 0.27659785747528076,
+      "learning_rate": 8.574589615503739e-06,
+      "loss": 0.3751,
+      "step": 1189
+    },
+    {
+      "epoch": 0.9661705006765899,
+      "grad_norm": 0.24575616419315338,
+      "learning_rate": 8.571282830553875e-06,
+      "loss": 0.3939,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9669824086603518,
+      "grad_norm": 0.24864161014556885,
+      "learning_rate": 8.567972853601119e-06,
+      "loss": 0.4103,
+      "step": 1191
+    },
+    {
+      "epoch": 0.9677943166441136,
+      "grad_norm": 0.253998339176178,
+      "learning_rate": 8.564659687603917e-06,
+      "loss": 0.3921,
+      "step": 1192
+    },
+    {
+      "epoch": 0.9686062246278755,
+      "grad_norm": 0.27197936177253723,
+      "learning_rate": 8.56134333552357e-06,
+      "loss": 0.3756,
+      "step": 1193
+    },
+    {
+      "epoch": 0.9694181326116373,
+      "grad_norm": 0.2622896432876587,
+      "learning_rate": 8.558023800324223e-06,
+      "loss": 0.3887,
+      "step": 1194
+    },
+    {
+      "epoch": 0.9702300405953992,
+      "grad_norm": 0.28968143463134766,
+      "learning_rate": 8.554701084972869e-06,
+      "loss": 0.4039,
+      "step": 1195
+    },
+    {
+      "epoch": 0.971041948579161,
+      "grad_norm": 0.25985807180404663,
+      "learning_rate": 8.551375192439339e-06,
+      "loss": 0.3864,
+      "step": 1196
+    },
+    {
+      "epoch": 0.9718538565629229,
+      "grad_norm": 0.24716025590896606,
+      "learning_rate": 8.54804612569631e-06,
+      "loss": 0.3786,
+      "step": 1197
+    },
+    {
+      "epoch": 0.9726657645466847,
+      "grad_norm": 0.28425246477127075,
+      "learning_rate": 8.54471388771929e-06,
+      "loss": 0.391,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9734776725304466,
+      "grad_norm": 0.3173471689224243,
+      "learning_rate": 8.541378481486627e-06,
+      "loss": 0.3884,
+      "step": 1199
+    },
+    {
+      "epoch": 0.9742895805142084,
+      "grad_norm": 0.2839984893798828,
+      "learning_rate": 8.538039909979496e-06,
+      "loss": 0.4104,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9751014884979702,
+      "grad_norm": 0.2639375627040863,
+      "learning_rate": 8.534698176181904e-06,
+      "loss": 0.3894,
+      "step": 1201
+    },
+    {
+      "epoch": 0.975913396481732,
+      "grad_norm": 0.281556099653244,
+      "learning_rate": 8.531353283080683e-06,
+      "loss": 0.3671,
+      "step": 1202
+    },
+    {
+      "epoch": 0.9767253044654939,
+      "grad_norm": 0.28931108117103577,
+      "learning_rate": 8.528005233665491e-06,
+      "loss": 0.3861,
+      "step": 1203
+    },
+    {
+      "epoch": 0.9775372124492557,
+      "grad_norm": 0.2725782096385956,
+      "learning_rate": 8.524654030928802e-06,
+      "loss": 0.3952,
+      "step": 1204
+    },
+    {
+      "epoch": 0.9783491204330176,
+      "grad_norm": 0.26281148195266724,
+      "learning_rate": 8.521299677865915e-06,
+      "loss": 0.389,
+      "step": 1205
+    },
+    {
+      "epoch": 0.9791610284167794,
+      "grad_norm": 0.2440013885498047,
+      "learning_rate": 8.517942177474943e-06,
+      "loss": 0.3958,
+      "step": 1206
+    },
+    {
+      "epoch": 0.9799729364005413,
+      "grad_norm": 0.2817719578742981,
+      "learning_rate": 8.514581532756806e-06,
+      "loss": 0.3911,
+      "step": 1207
+    },
+    {
+      "epoch": 0.9807848443843031,
+      "grad_norm": 0.23647233843803406,
+      "learning_rate": 8.511217746715243e-06,
+      "loss": 0.3795,
+      "step": 1208
+    },
+    {
+      "epoch": 0.981596752368065,
+      "grad_norm": 0.2329581081867218,
+      "learning_rate": 8.507850822356796e-06,
+      "loss": 0.4069,
+      "step": 1209
+    },
+    {
+      "epoch": 0.9824086603518268,
+      "grad_norm": 0.24621187150478363,
+      "learning_rate": 8.50448076269081e-06,
+      "loss": 0.3699,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9832205683355887,
+      "grad_norm": 0.2368006557226181,
+      "learning_rate": 8.50110757072944e-06,
+      "loss": 0.374,
+      "step": 1211
+    },
+    {
+      "epoch": 0.9840324763193504,
+      "grad_norm": 0.24261754751205444,
+      "learning_rate": 8.497731249487632e-06,
+      "loss": 0.3779,
+      "step": 1212
+    },
+    {
+      "epoch": 0.9848443843031123,
+      "grad_norm": 0.2267138808965683,
+      "learning_rate": 8.494351801983135e-06,
+      "loss": 0.4018,
+      "step": 1213
+    },
+    {
+      "epoch": 0.9856562922868741,
+      "grad_norm": 0.2442365288734436,
+      "learning_rate": 8.490969231236487e-06,
+      "loss": 0.3975,
+      "step": 1214
+    },
+    {
+      "epoch": 0.986468200270636,
+      "grad_norm": 0.24782416224479675,
+      "learning_rate": 8.487583540271024e-06,
+      "loss": 0.3938,
+      "step": 1215
+    },
+    {
+      "epoch": 0.9872801082543978,
+      "grad_norm": 0.243353471159935,
+      "learning_rate": 8.484194732112866e-06,
+      "loss": 0.3865,
+      "step": 1216
+    },
+    {
+      "epoch": 0.9880920162381597,
+      "grad_norm": 0.24817124009132385,
+      "learning_rate": 8.480802809790922e-06,
+      "loss": 0.3842,
+      "step": 1217
+    },
+    {
+      "epoch": 0.9889039242219215,
+      "grad_norm": 0.23131495714187622,
+      "learning_rate": 8.47740777633688e-06,
+      "loss": 0.3954,
+      "step": 1218
+    },
+    {
+      "epoch": 0.9897158322056834,
+      "grad_norm": 0.25609683990478516,
+      "learning_rate": 8.474009634785214e-06,
+      "loss": 0.3845,
+      "step": 1219
+    },
+    {
+      "epoch": 0.9905277401894452,
+      "grad_norm": 0.2608782947063446,
+      "learning_rate": 8.47060838817317e-06,
+      "loss": 0.4005,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9913396481732071,
+      "grad_norm": 0.2761821448802948,
+      "learning_rate": 8.467204039540779e-06,
+      "loss": 0.3782,
+      "step": 1221
+    },
+    {
+      "epoch": 0.9921515561569689,
+      "grad_norm": 0.27413713932037354,
+      "learning_rate": 8.463796591930833e-06,
+      "loss": 0.3975,
+      "step": 1222
+    },
+    {
+      "epoch": 0.9929634641407307,
+      "grad_norm": 0.2304287701845169,
+      "learning_rate": 8.460386048388905e-06,
+      "loss": 0.376,
+      "step": 1223
+    },
+    {
+      "epoch": 0.9937753721244925,
+      "grad_norm": 0.27339884638786316,
+      "learning_rate": 8.456972411963323e-06,
+      "loss": 0.3901,
+      "step": 1224
+    },
+    {
+      "epoch": 0.9945872801082544,
+      "grad_norm": 0.23339788615703583,
+      "learning_rate": 8.453555685705192e-06,
+      "loss": 0.3751,
+      "step": 1225
+    },
+    {
+      "epoch": 0.9953991880920162,
+      "grad_norm": 0.255868136882782,
+      "learning_rate": 8.450135872668369e-06,
+      "loss": 0.4093,
+      "step": 1226
+    },
+    {
+      "epoch": 0.9962110960757781,
+      "grad_norm": 0.24789322912693024,
+      "learning_rate": 8.446712975909474e-06,
+      "loss": 0.3948,
+      "step": 1227
+    },
+    {
+      "epoch": 0.9970230040595399,
+      "grad_norm": 0.2643323242664337,
+      "learning_rate": 8.443286998487884e-06,
+      "loss": 0.3743,
+      "step": 1228
+    },
+    {
+      "epoch": 0.9978349120433018,
+      "grad_norm": 0.26547786593437195,
+      "learning_rate": 8.439857943465728e-06,
+      "loss": 0.3723,
+      "step": 1229
+    },
+    {
+      "epoch": 0.9986468200270636,
+      "grad_norm": 0.27032172679901123,
+      "learning_rate": 8.436425813907885e-06,
+      "loss": 0.3736,
+      "step": 1230
+    },
+    {
+      "epoch": 0.9994587280108255,
+      "grad_norm": 0.23936688899993896,
+      "learning_rate": 8.432990612881985e-06,
+      "loss": 0.4064,
+      "step": 1231
+    },
+    {
+      "epoch": 1.0002706359945872,
+      "grad_norm": 0.3786931037902832,
+      "learning_rate": 8.4295523434584e-06,
+      "loss": 0.4947,
+      "step": 1232
+    },
+    {
+      "epoch": 1.001082543978349,
+      "grad_norm": 0.26381340622901917,
+      "learning_rate": 8.426111008710245e-06,
+      "loss": 0.3468,
+      "step": 1233
+    },
+    {
+      "epoch": 1.001894451962111,
+      "grad_norm": 0.26769399642944336,
+      "learning_rate": 8.422666611713378e-06,
+      "loss": 0.3601,
+      "step": 1234
+    },
+    {
+      "epoch": 1.0027063599458728,
+      "grad_norm": 0.2532285153865814,
+      "learning_rate": 8.41921915554639e-06,
+      "loss": 0.3529,
+      "step": 1235
+    },
+    {
+      "epoch": 1.0035182679296346,
+      "grad_norm": 0.28053927421569824,
+      "learning_rate": 8.415768643290608e-06,
+      "loss": 0.3804,
+      "step": 1236
+    },
+    {
+      "epoch": 1.0043301759133965,
+      "grad_norm": 0.2574464976787567,
+      "learning_rate": 8.412315078030091e-06,
+      "loss": 0.3809,
+      "step": 1237
+    },
+    {
+      "epoch": 1.0051420838971583,
+      "grad_norm": 0.26367682218551636,
+      "learning_rate": 8.408858462851625e-06,
+      "loss": 0.3548,
+      "step": 1238
+    },
+    {
+      "epoch": 1.0059539918809202,
+      "grad_norm": 0.2688828706741333,
+      "learning_rate": 8.405398800844726e-06,
+      "loss": 0.3559,
+      "step": 1239
+    },
+    {
+      "epoch": 1.006765899864682,
+      "grad_norm": 0.2539481818675995,
+      "learning_rate": 8.401936095101629e-06,
+      "loss": 0.3794,
+      "step": 1240
+    },
+    {
+      "epoch": 1.0075778078484439,
+      "grad_norm": 0.2611037492752075,
+      "learning_rate": 8.398470348717288e-06,
+      "loss": 0.3681,
+      "step": 1241
+    },
+    {
+      "epoch": 1.0083897158322057,
+      "grad_norm": 0.25627532601356506,
+      "learning_rate": 8.395001564789385e-06,
+      "loss": 0.3583,
+      "step": 1242
+    },
+    {
+      "epoch": 1.0092016238159676,
+      "grad_norm": 0.2540230453014374,
+      "learning_rate": 8.391529746418303e-06,
+      "loss": 0.3654,
+      "step": 1243
+    },
+    {
+      "epoch": 1.0100135317997294,
+      "grad_norm": 0.3684833347797394,
+      "learning_rate": 8.388054896707146e-06,
+      "loss": 0.3781,
+      "step": 1244
+    },
+    {
+      "epoch": 1.0108254397834913,
+      "grad_norm": 0.2786010801792145,
+      "learning_rate": 8.384577018761725e-06,
+      "loss": 0.3896,
+      "step": 1245
+    },
+    {
+      "epoch": 1.0116373477672531,
+      "grad_norm": 0.23165570199489594,
+      "learning_rate": 8.38109611569056e-06,
+      "loss": 0.3492,
+      "step": 1246
+    },
+    {
+      "epoch": 1.012449255751015,
+      "grad_norm": 0.2300640195608139,
+      "learning_rate": 8.377612190604868e-06,
+      "loss": 0.3659,
+      "step": 1247
+    },
+    {
+      "epoch": 1.0132611637347768,
+      "grad_norm": 0.2658098638057709,
+      "learning_rate": 8.374125246618576e-06,
+      "loss": 0.3955,
+      "step": 1248
+    },
+    {
+      "epoch": 1.0140730717185387,
+      "grad_norm": 0.2428921014070511,
+      "learning_rate": 8.370635286848304e-06,
+      "loss": 0.3667,
+      "step": 1249
+    },
+    {
+      "epoch": 1.0148849797023005,
+      "grad_norm": 0.23254817724227905,
+      "learning_rate": 8.367142314413368e-06,
+      "loss": 0.3637,
+      "step": 1250
+    },
+    {
+      "epoch": 1.0156968876860621,
+      "grad_norm": 0.25003138184547424,
+      "learning_rate": 8.363646332435778e-06,
+      "loss": 0.3455,
+      "step": 1251
+    },
+    {
+      "epoch": 1.016508795669824,
+      "grad_norm": 0.2402811348438263,
+      "learning_rate": 8.360147344040233e-06,
+      "loss": 0.3229,
+      "step": 1252
+    },
+    {
+      "epoch": 1.0173207036535858,
+      "grad_norm": 0.24141517281532288,
+      "learning_rate": 8.356645352354117e-06,
+      "loss": 0.3683,
+      "step": 1253
+    },
+    {
+      "epoch": 1.0181326116373477,
+      "grad_norm": 0.24926507472991943,
+      "learning_rate": 8.353140360507506e-06,
+      "loss": 0.3555,
+      "step": 1254
+    },
+    {
+      "epoch": 1.0189445196211095,
+      "grad_norm": 0.31167128682136536,
+      "learning_rate": 8.349632371633146e-06,
+      "loss": 0.3273,
+      "step": 1255
+    },
+    {
+      "epoch": 1.0197564276048714,
+      "grad_norm": 0.24855254590511322,
+      "learning_rate": 8.346121388866472e-06,
+      "loss": 0.3899,
+      "step": 1256
+    },
+    {
+      "epoch": 1.0205683355886332,
+      "grad_norm": 0.21367809176445007,
+      "learning_rate": 8.34260741534559e-06,
+      "loss": 0.3067,
+      "step": 1257
+    },
+    {
+      "epoch": 1.021380243572395,
+      "grad_norm": 0.290089875459671,
+      "learning_rate": 8.339090454211278e-06,
+      "loss": 0.4129,
+      "step": 1258
+    },
+    {
+      "epoch": 1.022192151556157,
+      "grad_norm": 0.19848588109016418,
+      "learning_rate": 8.33557050860699e-06,
+      "loss": 0.2944,
+      "step": 1259
+    },
+    {
+      "epoch": 1.0230040595399188,
+      "grad_norm": 0.2751034200191498,
+      "learning_rate": 8.332047581678839e-06,
+      "loss": 0.3808,
+      "step": 1260
+    },
+    {
+      "epoch": 1.0238159675236806,
+      "grad_norm": 0.21761299669742584,
+      "learning_rate": 8.328521676575612e-06,
+      "loss": 0.3452,
+      "step": 1261
+    },
+    {
+      "epoch": 1.0246278755074425,
+      "grad_norm": 0.23547674715518951,
+      "learning_rate": 8.324992796448752e-06,
+      "loss": 0.3418,
+      "step": 1262
+    },
+    {
+      "epoch": 1.0254397834912043,
+      "grad_norm": 0.2683631181716919,
+      "learning_rate": 8.321460944452364e-06,
+      "loss": 0.3794,
+      "step": 1263
+    },
+    {
+      "epoch": 1.0262516914749662,
+      "grad_norm": 0.27912637591362,
+      "learning_rate": 8.317926123743202e-06,
+      "loss": 0.3885,
+      "step": 1264
+    },
+    {
+      "epoch": 1.027063599458728,
+      "grad_norm": 0.24314109981060028,
+      "learning_rate": 8.314388337480686e-06,
+      "loss": 0.3287,
+      "step": 1265
+    },
+    {
+      "epoch": 1.02787550744249,
+      "grad_norm": 0.2669338881969452,
+      "learning_rate": 8.310847588826876e-06,
+      "loss": 0.3661,
+      "step": 1266
+    },
+    {
+      "epoch": 1.0286874154262518,
+      "grad_norm": 0.23122340440750122,
+      "learning_rate": 8.307303880946485e-06,
+      "loss": 0.356,
+      "step": 1267
+    },
+    {
+      "epoch": 1.0294993234100136,
+      "grad_norm": 0.2574627101421356,
+      "learning_rate": 8.303757217006869e-06,
+      "loss": 0.3696,
+      "step": 1268
+    },
+    {
+      "epoch": 1.0303112313937755,
+      "grad_norm": 0.2548213303089142,
+      "learning_rate": 8.300207600178024e-06,
+      "loss": 0.3418,
+      "step": 1269
+    },
+    {
+      "epoch": 1.0311231393775373,
+      "grad_norm": 0.241521418094635,
+      "learning_rate": 8.296655033632592e-06,
+      "loss": 0.3414,
+      "step": 1270
+    },
+    {
+      "epoch": 1.0319350473612992,
+      "grad_norm": 0.25128719210624695,
+      "learning_rate": 8.293099520545845e-06,
+      "loss": 0.3624,
+      "step": 1271
+    },
+    {
+      "epoch": 1.0327469553450608,
+      "grad_norm": 0.303512305021286,
+      "learning_rate": 8.289541064095693e-06,
+      "loss": 0.376,
+      "step": 1272
+    },
+    {
+      "epoch": 1.0335588633288226,
+      "grad_norm": 0.2345477044582367,
+      "learning_rate": 8.285979667462671e-06,
+      "loss": 0.3562,
+      "step": 1273
+    },
+    {
+      "epoch": 1.0343707713125845,
+      "grad_norm": 0.2691783308982849,
+      "learning_rate": 8.28241533382995e-06,
+      "loss": 0.3609,
+      "step": 1274
+    },
+    {
+      "epoch": 1.0351826792963463,
+      "grad_norm": 0.24537056684494019,
+      "learning_rate": 8.27884806638332e-06,
+      "loss": 0.3436,
+      "step": 1275
+    },
+    {
+      "epoch": 1.0359945872801082,
+      "grad_norm": 0.2588464021682739,
+      "learning_rate": 8.275277868311193e-06,
+      "loss": 0.3569,
+      "step": 1276
+    },
+    {
+      "epoch": 1.03680649526387,
+      "grad_norm": 0.2704032063484192,
+      "learning_rate": 8.271704742804605e-06,
+      "loss": 0.3613,
+      "step": 1277
+    },
+    {
+      "epoch": 1.0376184032476319,
+      "grad_norm": 0.2431112378835678,
+      "learning_rate": 8.268128693057202e-06,
+      "loss": 0.3441,
+      "step": 1278
+    },
+    {
+      "epoch": 1.0384303112313937,
+      "grad_norm": 0.27162158489227295,
+      "learning_rate": 8.264549722265254e-06,
+      "loss": 0.3765,
+      "step": 1279
+    },
+    {
+      "epoch": 1.0392422192151556,
+      "grad_norm": 0.2841114401817322,
+      "learning_rate": 8.260967833627628e-06,
+      "loss": 0.3781,
+      "step": 1280
+    },
+    {
+      "epoch": 1.0400541271989174,
+      "grad_norm": 0.29869186878204346,
+      "learning_rate": 8.257383030345814e-06,
+      "loss": 0.3698,
+      "step": 1281
+    },
+    {
+      "epoch": 1.0408660351826793,
+      "grad_norm": 0.24000363051891327,
+      "learning_rate": 8.253795315623894e-06,
+      "loss": 0.3726,
+      "step": 1282
+    },
+    {
+      "epoch": 1.0416779431664411,
+      "grad_norm": 0.2628210783004761,
+      "learning_rate": 8.250204692668559e-06,
+      "loss": 0.3569,
+      "step": 1283
+    },
+    {
+      "epoch": 1.042489851150203,
+      "grad_norm": 0.2742999792098999,
+      "learning_rate": 8.246611164689097e-06,
+      "loss": 0.3805,
+      "step": 1284
+    },
+    {
+      "epoch": 1.0433017591339648,
+      "grad_norm": 0.24909597635269165,
+      "learning_rate": 8.243014734897397e-06,
+      "loss": 0.3825,
+      "step": 1285
+    },
+    {
+      "epoch": 1.0441136671177267,
+      "grad_norm": 0.27264365553855896,
+      "learning_rate": 8.239415406507934e-06,
+      "loss": 0.3647,
+      "step": 1286
+    },
+    {
+      "epoch": 1.0449255751014885,
+      "grad_norm": 0.28062430024147034,
+      "learning_rate": 8.235813182737782e-06,
+      "loss": 0.3874,
+      "step": 1287
+    },
+    {
+      "epoch": 1.0457374830852504,
+      "grad_norm": 0.23445986211299896,
+      "learning_rate": 8.232208066806596e-06,
+      "loss": 0.3408,
+      "step": 1288
+    },
+    {
+      "epoch": 1.0465493910690122,
+      "grad_norm": 0.26671406626701355,
+      "learning_rate": 8.228600061936617e-06,
+      "loss": 0.3792,
+      "step": 1289
+    },
+    {
+      "epoch": 1.047361299052774,
+      "grad_norm": 0.2500205636024475,
+      "learning_rate": 8.224989171352675e-06,
+      "loss": 0.3419,
+      "step": 1290
+    },
+    {
+      "epoch": 1.048173207036536,
+      "grad_norm": 0.2491471916437149,
+      "learning_rate": 8.221375398282172e-06,
+      "loss": 0.3636,
+      "step": 1291
+    },
+    {
+      "epoch": 1.0489851150202978,
+      "grad_norm": 0.24396269023418427,
+      "learning_rate": 8.217758745955087e-06,
+      "loss": 0.3583,
+      "step": 1292
+    },
+    {
+      "epoch": 1.0497970230040596,
+      "grad_norm": 0.24425946176052094,
+      "learning_rate": 8.214139217603975e-06,
+      "loss": 0.3659,
+      "step": 1293
+    },
+    {
+      "epoch": 1.0506089309878215,
+      "grad_norm": 0.2471698820590973,
+      "learning_rate": 8.210516816463961e-06,
+      "loss": 0.3896,
+      "step": 1294
+    },
+    {
+      "epoch": 1.0514208389715831,
+      "grad_norm": 0.25941702723503113,
+      "learning_rate": 8.206891545772738e-06,
+      "loss": 0.3886,
+      "step": 1295
+    },
+    {
+      "epoch": 1.052232746955345,
+      "grad_norm": 0.21585053205490112,
+      "learning_rate": 8.203263408770562e-06,
+      "loss": 0.3622,
+      "step": 1296
+    },
+    {
+      "epoch": 1.0530446549391068,
+      "grad_norm": 0.23950953781604767,
+      "learning_rate": 8.19963240870025e-06,
+      "loss": 0.3578,
+      "step": 1297
+    },
+    {
+      "epoch": 1.0538565629228687,
+      "grad_norm": 0.2693774998188019,
+      "learning_rate": 8.195998548807185e-06,
+      "loss": 0.397,
+      "step": 1298
+    },
+    {
+      "epoch": 1.0546684709066305,
+      "grad_norm": 0.2109154611825943,
+      "learning_rate": 8.192361832339299e-06,
+      "loss": 0.3161,
+      "step": 1299
+    },
+    {
+      "epoch": 1.0554803788903924,
+      "grad_norm": 0.26632025837898254,
+      "learning_rate": 8.188722262547078e-06,
+      "loss": 0.3991,
+      "step": 1300
+    },
+    {
+      "epoch": 1.0562922868741542,
+      "grad_norm": 0.25315964221954346,
+      "learning_rate": 8.185079842683558e-06,
+      "loss": 0.3599,
+      "step": 1301
+    },
+    {
+      "epoch": 1.057104194857916,
+      "grad_norm": 0.2375582456588745,
+      "learning_rate": 8.181434576004328e-06,
+      "loss": 0.3336,
+      "step": 1302
+    },
+    {
+      "epoch": 1.057916102841678,
+      "grad_norm": 0.26633065938949585,
+      "learning_rate": 8.177786465767517e-06,
+      "loss": 0.3743,
+      "step": 1303
+    },
+    {
+      "epoch": 1.0587280108254398,
+      "grad_norm": 0.2412031888961792,
+      "learning_rate": 8.174135515233792e-06,
+      "loss": 0.3377,
+      "step": 1304
+    },
+    {
+      "epoch": 1.0595399188092016,
+      "grad_norm": 0.2423148900270462,
+      "learning_rate": 8.170481727666366e-06,
+      "loss": 0.3704,
+      "step": 1305
+    },
+    {
+      "epoch": 1.0603518267929635,
+      "grad_norm": 0.2625727653503418,
+      "learning_rate": 8.166825106330985e-06,
+      "loss": 0.3331,
+      "step": 1306
+    },
+    {
+      "epoch": 1.0611637347767253,
+      "grad_norm": 0.28848201036453247,
+      "learning_rate": 8.163165654495925e-06,
+      "loss": 0.4138,
+      "step": 1307
+    },
+    {
+      "epoch": 1.0619756427604872,
+      "grad_norm": 0.2393466681241989,
+      "learning_rate": 8.159503375431996e-06,
+      "loss": 0.3387,
+      "step": 1308
+    },
+    {
+      "epoch": 1.062787550744249,
+      "grad_norm": 0.24995915591716766,
+      "learning_rate": 8.155838272412531e-06,
+      "loss": 0.3867,
+      "step": 1309
+    },
+    {
+      "epoch": 1.0635994587280109,
+      "grad_norm": 0.25035911798477173,
+      "learning_rate": 8.152170348713392e-06,
+      "loss": 0.3248,
+      "step": 1310
+    },
+    {
+      "epoch": 1.0644113667117727,
+      "grad_norm": 0.28035789728164673,
+      "learning_rate": 8.148499607612956e-06,
+      "loss": 0.3823,
+      "step": 1311
+    },
+    {
+      "epoch": 1.0652232746955346,
+      "grad_norm": 0.24067114293575287,
+      "learning_rate": 8.144826052392123e-06,
+      "loss": 0.3639,
+      "step": 1312
+    },
+    {
+      "epoch": 1.0660351826792964,
+      "grad_norm": 0.2965671122074127,
+      "learning_rate": 8.141149686334307e-06,
+      "loss": 0.3928,
+      "step": 1313
+    },
+    {
+      "epoch": 1.0668470906630583,
+      "grad_norm": 0.2135877162218094,
+      "learning_rate": 8.137470512725434e-06,
+      "loss": 0.3374,
+      "step": 1314
+    },
+    {
+      "epoch": 1.0676589986468201,
+      "grad_norm": 0.23596444725990295,
+      "learning_rate": 8.133788534853939e-06,
+      "loss": 0.4048,
+      "step": 1315
+    },
+    {
+      "epoch": 1.0684709066305818,
+      "grad_norm": 0.2528867721557617,
+      "learning_rate": 8.130103756010763e-06,
+      "loss": 0.3526,
+      "step": 1316
+    },
+    {
+      "epoch": 1.0692828146143438,
+      "grad_norm": 0.2211836725473404,
+      "learning_rate": 8.126416179489354e-06,
+      "loss": 0.3609,
+      "step": 1317
+    },
+    {
+      "epoch": 1.0700947225981055,
+      "grad_norm": 0.23112566769123077,
+      "learning_rate": 8.122725808585654e-06,
+      "loss": 0.3922,
+      "step": 1318
+    },
+    {
+      "epoch": 1.0709066305818673,
+      "grad_norm": 0.2539345920085907,
+      "learning_rate": 8.11903264659811e-06,
+      "loss": 0.3661,
+      "step": 1319
+    },
+    {
+      "epoch": 1.0717185385656292,
+      "grad_norm": 0.21523001790046692,
+      "learning_rate": 8.115336696827658e-06,
+      "loss": 0.349,
+      "step": 1320
+    },
+    {
+      "epoch": 1.072530446549391,
+      "grad_norm": 0.25778356194496155,
+      "learning_rate": 8.111637962577728e-06,
+      "loss": 0.3446,
+      "step": 1321
+    },
+    {
+      "epoch": 1.0733423545331529,
+      "grad_norm": 0.22282774746418,
+      "learning_rate": 8.10793644715424e-06,
+      "loss": 0.3127,
+      "step": 1322
+    },
+    {
+      "epoch": 1.0741542625169147,
+      "grad_norm": 0.24668002128601074,
+      "learning_rate": 8.104232153865597e-06,
+      "loss": 0.385,
+      "step": 1323
+    },
+    {
+      "epoch": 1.0749661705006766,
+      "grad_norm": 0.25453808903694153,
+      "learning_rate": 8.100525086022687e-06,
+      "loss": 0.3512,
+      "step": 1324
+    },
+    {
+      "epoch": 1.0757780784844384,
+      "grad_norm": 0.22986295819282532,
+      "learning_rate": 8.096815246938879e-06,
+      "loss": 0.3367,
+      "step": 1325
+    },
+    {
+      "epoch": 1.0765899864682003,
+      "grad_norm": 0.24788127839565277,
+      "learning_rate": 8.093102639930013e-06,
+      "loss": 0.3407,
+      "step": 1326
+    },
+    {
+      "epoch": 1.077401894451962,
+      "grad_norm": 0.2721409499645233,
+      "learning_rate": 8.089387268314408e-06,
+      "loss": 0.41,
+      "step": 1327
+    },
+    {
+      "epoch": 1.078213802435724,
+      "grad_norm": 0.21042361855506897,
+      "learning_rate": 8.085669135412856e-06,
+      "loss": 0.3555,
+      "step": 1328
+    },
+    {
+      "epoch": 1.0790257104194858,
+      "grad_norm": 0.2599429786205292,
+      "learning_rate": 8.081948244548615e-06,
+      "loss": 0.3785,
+      "step": 1329
+    },
+    {
+      "epoch": 1.0798376184032477,
+      "grad_norm": 0.23790892958641052,
+      "learning_rate": 8.0782245990474e-06,
+      "loss": 0.3562,
+      "step": 1330
+    },
+    {
+      "epoch": 1.0806495263870095,
+      "grad_norm": 0.2478892207145691,
+      "learning_rate": 8.074498202237402e-06,
+      "loss": 0.3532,
+      "step": 1331
+    },
+    {
+      "epoch": 1.0814614343707714,
+      "grad_norm": 0.23811066150665283,
+      "learning_rate": 8.070769057449262e-06,
+      "loss": 0.3734,
+      "step": 1332
+    },
+    {
+      "epoch": 1.0822733423545332,
+      "grad_norm": 0.27629661560058594,
+      "learning_rate": 8.067037168016079e-06,
+      "loss": 0.3974,
+      "step": 1333
+    },
+    {
+      "epoch": 1.083085250338295,
+      "grad_norm": 0.25024065375328064,
+      "learning_rate": 8.063302537273407e-06,
+      "loss": 0.3416,
+      "step": 1334
+    },
+    {
+      "epoch": 1.083897158322057,
+      "grad_norm": 0.23918890953063965,
+      "learning_rate": 8.059565168559247e-06,
+      "loss": 0.3678,
+      "step": 1335
+    },
+    {
+      "epoch": 1.0847090663058188,
+      "grad_norm": 0.26560482382774353,
+      "learning_rate": 8.05582506521405e-06,
+      "loss": 0.3631,
+      "step": 1336
+    },
+    {
+      "epoch": 1.0855209742895806,
+      "grad_norm": 0.23072859644889832,
+      "learning_rate": 8.052082230580711e-06,
+      "loss": 0.3553,
+      "step": 1337
+    },
+    {
+      "epoch": 1.0863328822733425,
+      "grad_norm": 0.23017700016498566,
+      "learning_rate": 8.048336668004564e-06,
+      "loss": 0.3782,
+      "step": 1338
+    },
+    {
+      "epoch": 1.087144790257104,
+      "grad_norm": 0.28096675872802734,
+      "learning_rate": 8.044588380833384e-06,
+      "loss": 0.3719,
+      "step": 1339
+    },
+    {
+      "epoch": 1.087956698240866,
+      "grad_norm": 0.2467406988143921,
+      "learning_rate": 8.040837372417379e-06,
+      "loss": 0.3525,
+      "step": 1340
+    },
+    {
+      "epoch": 1.0887686062246278,
+      "grad_norm": 0.25728631019592285,
+      "learning_rate": 8.03708364610919e-06,
+      "loss": 0.379,
+      "step": 1341
+    },
+    {
+      "epoch": 1.0895805142083896,
+      "grad_norm": 0.22535383701324463,
+      "learning_rate": 8.033327205263888e-06,
+      "loss": 0.3373,
+      "step": 1342
+    },
+    {
+      "epoch": 1.0903924221921515,
+      "grad_norm": 0.2512345612049103,
+      "learning_rate": 8.02956805323897e-06,
+      "loss": 0.383,
+      "step": 1343
+    },
+    {
+      "epoch": 1.0912043301759133,
+      "grad_norm": 0.24587014317512512,
+      "learning_rate": 8.025806193394354e-06,
+      "loss": 0.3522,
+      "step": 1344
+    },
+    {
+      "epoch": 1.0920162381596752,
+      "grad_norm": 0.23340022563934326,
+      "learning_rate": 8.022041629092382e-06,
+      "loss": 0.3525,
+      "step": 1345
+    },
+    {
+      "epoch": 1.092828146143437,
+      "grad_norm": 0.25500190258026123,
+      "learning_rate": 8.01827436369781e-06,
+      "loss": 0.3798,
+      "step": 1346
+    },
+    {
+      "epoch": 1.093640054127199,
+      "grad_norm": 0.2501123547554016,
+      "learning_rate": 8.01450440057781e-06,
+      "loss": 0.3851,
+      "step": 1347
+    },
+    {
+      "epoch": 1.0944519621109607,
+      "grad_norm": 0.26404014229774475,
+      "learning_rate": 8.010731743101967e-06,
+      "loss": 0.3584,
+      "step": 1348
+    },
+    {
+      "epoch": 1.0952638700947226,
+      "grad_norm": 0.22869935631752014,
+      "learning_rate": 8.006956394642268e-06,
+      "loss": 0.3761,
+      "step": 1349
+    },
+    {
+      "epoch": 1.0960757780784844,
+      "grad_norm": 0.21976928412914276,
+      "learning_rate": 8.003178358573112e-06,
+      "loss": 0.3294,
+      "step": 1350
+    },
+    {
+      "epoch": 1.0968876860622463,
+      "grad_norm": 0.28912416100502014,
+      "learning_rate": 7.999397638271298e-06,
+      "loss": 0.3934,
+      "step": 1351
+    },
+    {
+      "epoch": 1.0976995940460081,
+      "grad_norm": 0.24449172616004944,
+      "learning_rate": 7.99561423711602e-06,
+      "loss": 0.3258,
+      "step": 1352
+    },
+    {
+      "epoch": 1.09851150202977,
+      "grad_norm": 0.23645851016044617,
+      "learning_rate": 7.991828158488875e-06,
+      "loss": 0.3573,
+      "step": 1353
+    },
+    {
+      "epoch": 1.0993234100135318,
+      "grad_norm": 0.30291247367858887,
+      "learning_rate": 7.98803940577385e-06,
+      "loss": 0.399,
+      "step": 1354
+    },
+    {
+      "epoch": 1.1001353179972937,
+      "grad_norm": 0.2581310272216797,
+      "learning_rate": 7.984247982357317e-06,
+      "loss": 0.3743,
+      "step": 1355
+    },
+    {
+      "epoch": 1.1009472259810555,
+      "grad_norm": 0.2250755876302719,
+      "learning_rate": 7.980453891628047e-06,
+      "loss": 0.3429,
+      "step": 1356
+    },
+    {
+      "epoch": 1.1017591339648174,
+      "grad_norm": 0.25534266233444214,
+      "learning_rate": 7.976657136977181e-06,
+      "loss": 0.3725,
+      "step": 1357
+    },
+    {
+      "epoch": 1.1025710419485792,
+      "grad_norm": 0.25101548433303833,
+      "learning_rate": 7.972857721798254e-06,
+      "loss": 0.371,
+      "step": 1358
+    },
+    {
+      "epoch": 1.103382949932341,
+      "grad_norm": 0.2271679788827896,
+      "learning_rate": 7.96905564948717e-06,
+      "loss": 0.3572,
+      "step": 1359
+    },
+    {
+      "epoch": 1.104194857916103,
+      "grad_norm": 0.23516477644443512,
+      "learning_rate": 7.96525092344221e-06,
+      "loss": 0.3456,
+      "step": 1360
+    },
+    {
+      "epoch": 1.1050067658998648,
+      "grad_norm": 0.3003014624118805,
+      "learning_rate": 7.96144354706403e-06,
+      "loss": 0.3983,
+      "step": 1361
+    },
+    {
+      "epoch": 1.1058186738836264,
+      "grad_norm": 0.24628356099128723,
+      "learning_rate": 7.957633523755652e-06,
+      "loss": 0.3368,
+      "step": 1362
+    },
+    {
+      "epoch": 1.1066305818673883,
+      "grad_norm": 0.2230798900127411,
+      "learning_rate": 7.953820856922465e-06,
+      "loss": 0.3472,
+      "step": 1363
+    },
+    {
+      "epoch": 1.1074424898511501,
+      "grad_norm": 0.26453644037246704,
+      "learning_rate": 7.95000554997222e-06,
+      "loss": 0.3744,
+      "step": 1364
+    },
+    {
+      "epoch": 1.108254397834912,
+      "grad_norm": 0.2693396508693695,
+      "learning_rate": 7.946187606315026e-06,
+      "loss": 0.3952,
+      "step": 1365
+    },
+    {
+      "epoch": 1.1090663058186738,
+      "grad_norm": 0.25101929903030396,
+      "learning_rate": 7.942367029363351e-06,
+      "loss": 0.3834,
+      "step": 1366
+    },
+    {
+      "epoch": 1.1098782138024357,
+      "grad_norm": 0.2666069269180298,
+      "learning_rate": 7.938543822532019e-06,
+      "loss": 0.3491,
+      "step": 1367
+    },
+    {
+      "epoch": 1.1106901217861975,
+      "grad_norm": 0.20864960551261902,
+      "learning_rate": 7.9347179892382e-06,
+      "loss": 0.321,
+      "step": 1368
+    },
+    {
+      "epoch": 1.1115020297699594,
+      "grad_norm": 0.27392837405204773,
+      "learning_rate": 7.930889532901416e-06,
+      "loss": 0.3766,
+      "step": 1369
+    },
+    {
+      "epoch": 1.1123139377537212,
+      "grad_norm": 0.24697616696357727,
+      "learning_rate": 7.927058456943527e-06,
+      "loss": 0.3441,
+      "step": 1370
+    },
+    {
+      "epoch": 1.113125845737483,
+      "grad_norm": 0.2604011595249176,
+      "learning_rate": 7.92322476478874e-06,
+      "loss": 0.3838,
+      "step": 1371
+    },
+    {
+      "epoch": 1.113937753721245,
+      "grad_norm": 0.2400532066822052,
+      "learning_rate": 7.919388459863597e-06,
+      "loss": 0.3434,
+      "step": 1372
+    },
+    {
+      "epoch": 1.1147496617050068,
+      "grad_norm": 0.2242223620414734,
+      "learning_rate": 7.915549545596979e-06,
+      "loss": 0.3417,
+      "step": 1373
+    },
+    {
+      "epoch": 1.1155615696887686,
+      "grad_norm": 0.23429405689239502,
+      "learning_rate": 7.911708025420097e-06,
+      "loss": 0.3525,
+      "step": 1374
+    },
+    {
+      "epoch": 1.1163734776725305,
+      "grad_norm": 0.2680491507053375,
+      "learning_rate": 7.907863902766494e-06,
+      "loss": 0.3619,
+      "step": 1375
+    },
+    {
+      "epoch": 1.1171853856562923,
+      "grad_norm": 0.26044803857803345,
+      "learning_rate": 7.90401718107203e-06,
+      "loss": 0.3472,
+      "step": 1376
+    },
+    {
+      "epoch": 1.1179972936400542,
+      "grad_norm": 0.23161886632442474,
+      "learning_rate": 7.900167863774896e-06,
+      "loss": 0.3435,
+      "step": 1377
+    },
+    {
+      "epoch": 1.118809201623816,
+      "grad_norm": 0.25607553124427795,
+      "learning_rate": 7.896315954315608e-06,
+      "loss": 0.3754,
+      "step": 1378
+    },
+    {
+      "epoch": 1.1196211096075779,
+      "grad_norm": 0.229601189494133,
+      "learning_rate": 7.892461456136984e-06,
+      "loss": 0.3504,
+      "step": 1379
+    },
+    {
+      "epoch": 1.1204330175913397,
+      "grad_norm": 0.26143917441368103,
+      "learning_rate": 7.888604372684168e-06,
+      "loss": 0.3718,
+      "step": 1380
+    },
+    {
+      "epoch": 1.1212449255751016,
+      "grad_norm": 0.27604034543037415,
+      "learning_rate": 7.88474470740461e-06,
+      "loss": 0.3625,
+      "step": 1381
+    },
+    {
+      "epoch": 1.1220568335588634,
+      "grad_norm": 0.27955880761146545,
+      "learning_rate": 7.880882463748067e-06,
+      "loss": 0.3714,
+      "step": 1382
+    },
+    {
+      "epoch": 1.122868741542625,
+      "grad_norm": 0.25804170966148376,
+      "learning_rate": 7.877017645166606e-06,
+      "loss": 0.3321,
+      "step": 1383
+    },
+    {
+      "epoch": 1.123680649526387,
+      "grad_norm": 0.25660622119903564,
+      "learning_rate": 7.873150255114588e-06,
+      "loss": 0.3589,
+      "step": 1384
+    },
+    {
+      "epoch": 1.1244925575101488,
+      "grad_norm": 0.26388949155807495,
+      "learning_rate": 7.869280297048674e-06,
+      "loss": 0.354,
+      "step": 1385
+    },
+    {
+      "epoch": 1.1253044654939106,
+      "grad_norm": 0.2594480514526367,
+      "learning_rate": 7.865407774427828e-06,
+      "loss": 0.3599,
+      "step": 1386
+    },
+    {
+      "epoch": 1.1261163734776725,
+      "grad_norm": 0.25822290778160095,
+      "learning_rate": 7.861532690713293e-06,
+      "loss": 0.3621,
+      "step": 1387
+    },
+    {
+      "epoch": 1.1269282814614343,
+      "grad_norm": 0.27057790756225586,
+      "learning_rate": 7.857655049368614e-06,
+      "loss": 0.3494,
+      "step": 1388
+    },
+    {
+      "epoch": 1.1277401894451962,
+      "grad_norm": 0.2354855090379715,
+      "learning_rate": 7.853774853859612e-06,
+      "loss": 0.3619,
+      "step": 1389
+    },
+    {
+      "epoch": 1.128552097428958,
+      "grad_norm": 0.2514684200286865,
+      "learning_rate": 7.8498921076544e-06,
+      "loss": 0.3473,
+      "step": 1390
+    },
+    {
+      "epoch": 1.1293640054127199,
+      "grad_norm": 0.2539539635181427,
+      "learning_rate": 7.846006814223362e-06,
+      "loss": 0.3844,
+      "step": 1391
+    },
+    {
+      "epoch": 1.1301759133964817,
+      "grad_norm": 0.27979013323783875,
+      "learning_rate": 7.842118977039162e-06,
+      "loss": 0.4106,
+      "step": 1392
+    },
+    {
+      "epoch": 1.1309878213802436,
+      "grad_norm": 0.25499215722084045,
+      "learning_rate": 7.838228599576743e-06,
+      "loss": 0.3134,
+      "step": 1393
+    },
+    {
+      "epoch": 1.1317997293640054,
+      "grad_norm": 0.25103357434272766,
+      "learning_rate": 7.83433568531331e-06,
+      "loss": 0.3588,
+      "step": 1394
+    },
+    {
+      "epoch": 1.1326116373477673,
+      "grad_norm": 0.27248942852020264,
+      "learning_rate": 7.830440237728341e-06,
+      "loss": 0.3568,
+      "step": 1395
+    },
+    {
+      "epoch": 1.1334235453315291,
+      "grad_norm": 0.24771980941295624,
+      "learning_rate": 7.826542260303576e-06,
+      "loss": 0.3661,
+      "step": 1396
+    },
+    {
+      "epoch": 1.134235453315291,
+      "grad_norm": 0.2702447474002838,
+      "learning_rate": 7.822641756523015e-06,
+      "loss": 0.3536,
+      "step": 1397
+    },
+    {
+      "epoch": 1.1350473612990528,
+      "grad_norm": 0.259508341550827,
+      "learning_rate": 7.818738729872921e-06,
+      "loss": 0.356,
+      "step": 1398
+    },
+    {
+      "epoch": 1.1358592692828147,
+      "grad_norm": 0.24051053822040558,
+      "learning_rate": 7.814833183841805e-06,
+      "loss": 0.3738,
+      "step": 1399
+    },
+    {
+      "epoch": 1.1366711772665765,
+      "grad_norm": 0.25386369228363037,
+      "learning_rate": 7.810925121920437e-06,
+      "loss": 0.3644,
+      "step": 1400
+    },
+    {
+      "epoch": 1.1374830852503384,
+      "grad_norm": 0.2641783654689789,
+      "learning_rate": 7.80701454760183e-06,
+      "loss": 0.3414,
+      "step": 1401
+    },
+    {
+      "epoch": 1.1382949932341002,
+      "grad_norm": 0.25404879450798035,
+      "learning_rate": 7.803101464381244e-06,
+      "loss": 0.3447,
+      "step": 1402
+    },
+    {
+      "epoch": 1.139106901217862,
+      "grad_norm": 0.254201740026474,
+      "learning_rate": 7.799185875756184e-06,
+      "loss": 0.3422,
+      "step": 1403
+    },
+    {
+      "epoch": 1.1399188092016237,
+      "grad_norm": 0.2543894648551941,
+      "learning_rate": 7.795267785226392e-06,
+      "loss": 0.3598,
+      "step": 1404
+    },
+    {
+      "epoch": 1.1407307171853858,
+      "grad_norm": 0.23345771431922913,
+      "learning_rate": 7.791347196293846e-06,
+      "loss": 0.3242,
+      "step": 1405
+    },
+    {
+      "epoch": 1.1415426251691474,
+      "grad_norm": 0.2650417983531952,
+      "learning_rate": 7.787424112462758e-06,
+      "loss": 0.397,
+      "step": 1406
+    },
+    {
+      "epoch": 1.1423545331529092,
+      "grad_norm": 0.2438710778951645,
+      "learning_rate": 7.783498537239568e-06,
+      "loss": 0.3511,
+      "step": 1407
+    },
+    {
+      "epoch": 1.143166441136671,
+      "grad_norm": 0.2511480748653412,
+      "learning_rate": 7.779570474132949e-06,
+      "loss": 0.3417,
+      "step": 1408
+    },
+    {
+      "epoch": 1.143978349120433,
+      "grad_norm": 0.25840070843696594,
+      "learning_rate": 7.775639926653789e-06,
+      "loss": 0.39,
+      "step": 1409
+    },
+    {
+      "epoch": 1.1447902571041948,
+      "grad_norm": 0.24235279858112335,
+      "learning_rate": 7.771706898315204e-06,
+      "loss": 0.3235,
+      "step": 1410
+    },
+    {
+      "epoch": 1.1456021650879566,
+      "grad_norm": 0.2352583408355713,
+      "learning_rate": 7.767771392632518e-06,
+      "loss": 0.3425,
+      "step": 1411
+    },
+    {
+      "epoch": 1.1464140730717185,
+      "grad_norm": 0.2544795274734497,
+      "learning_rate": 7.763833413123281e-06,
+      "loss": 0.3785,
+      "step": 1412
+    },
+    {
+      "epoch": 1.1472259810554803,
+      "grad_norm": 0.2463865727186203,
+      "learning_rate": 7.759892963307249e-06,
+      "loss": 0.3699,
+      "step": 1413
+    },
+    {
+      "epoch": 1.1480378890392422,
+      "grad_norm": 0.2399873435497284,
+      "learning_rate": 7.75595004670638e-06,
+      "loss": 0.3638,
+      "step": 1414
+    },
+    {
+      "epoch": 1.148849797023004,
+      "grad_norm": 0.2529672682285309,
+      "learning_rate": 7.752004666844849e-06,
+      "loss": 0.3658,
+      "step": 1415
+    },
+    {
+      "epoch": 1.149661705006766,
+      "grad_norm": 0.23696070909500122,
+      "learning_rate": 7.748056827249018e-06,
+      "loss": 0.3587,
+      "step": 1416
+    },
+    {
+      "epoch": 1.1504736129905277,
+      "grad_norm": 0.2898157835006714,
+      "learning_rate": 7.744106531447462e-06,
+      "loss": 0.3544,
+      "step": 1417
+    },
+    {
+      "epoch": 1.1512855209742896,
+      "grad_norm": 0.253888338804245,
+      "learning_rate": 7.74015378297094e-06,
+      "loss": 0.318,
+      "step": 1418
+    },
+    {
+      "epoch": 1.1520974289580515,
+      "grad_norm": 0.24936243891716003,
+      "learning_rate": 7.736198585352411e-06,
+      "loss": 0.3612,
+      "step": 1419
+    },
+    {
+      "epoch": 1.1529093369418133,
+      "grad_norm": 0.27591177821159363,
+      "learning_rate": 7.732240942127018e-06,
+      "loss": 0.3935,
+      "step": 1420
+    },
+    {
+      "epoch": 1.1537212449255752,
+      "grad_norm": 0.24378764629364014,
+      "learning_rate": 7.728280856832094e-06,
+      "loss": 0.3261,
+      "step": 1421
+    },
+    {
+      "epoch": 1.154533152909337,
+      "grad_norm": 0.24467045068740845,
+      "learning_rate": 7.72431833300715e-06,
+      "loss": 0.3569,
+      "step": 1422
+    },
+    {
+      "epoch": 1.1553450608930989,
+      "grad_norm": 0.2420566827058792,
+      "learning_rate": 7.72035337419388e-06,
+      "loss": 0.352,
+      "step": 1423
+    },
+    {
+      "epoch": 1.1561569688768607,
+      "grad_norm": 0.2889433205127716,
+      "learning_rate": 7.716385983936154e-06,
+      "loss": 0.3721,
+      "step": 1424
+    },
+    {
+      "epoch": 1.1569688768606226,
+      "grad_norm": 0.2700004577636719,
+      "learning_rate": 7.712416165780014e-06,
+      "loss": 0.3785,
+      "step": 1425
+    },
+    {
+      "epoch": 1.1577807848443844,
+      "grad_norm": 0.24469533562660217,
+      "learning_rate": 7.708443923273671e-06,
+      "loss": 0.3273,
+      "step": 1426
+    },
+    {
+      "epoch": 1.158592692828146,
+      "grad_norm": 0.2802138030529022,
+      "learning_rate": 7.704469259967509e-06,
+      "loss": 0.3816,
+      "step": 1427
+    },
+    {
+      "epoch": 1.159404600811908,
+      "grad_norm": 0.2138456553220749,
+      "learning_rate": 7.700492179414067e-06,
+      "loss": 0.3497,
+      "step": 1428
+    },
+    {
+      "epoch": 1.1602165087956697,
+      "grad_norm": 0.257468581199646,
+      "learning_rate": 7.69651268516805e-06,
+      "loss": 0.3827,
+      "step": 1429
+    },
+    {
+      "epoch": 1.1610284167794316,
+      "grad_norm": 0.24826496839523315,
+      "learning_rate": 7.692530780786321e-06,
+      "loss": 0.3798,
+      "step": 1430
+    },
+    {
+      "epoch": 1.1618403247631934,
+      "grad_norm": 0.2858109474182129,
+      "learning_rate": 7.688546469827892e-06,
+      "loss": 0.3837,
+      "step": 1431
+    },
+    {
+      "epoch": 1.1626522327469553,
+      "grad_norm": 0.2474408894777298,
+      "learning_rate": 7.684559755853932e-06,
+      "loss": 0.3702,
+      "step": 1432
+    },
+    {
+      "epoch": 1.1634641407307171,
+      "grad_norm": 0.2348301112651825,
+      "learning_rate": 7.680570642427755e-06,
+      "loss": 0.3092,
+      "step": 1433
+    },
+    {
+      "epoch": 1.164276048714479,
+      "grad_norm": 0.2363719642162323,
+      "learning_rate": 7.676579133114819e-06,
+      "loss": 0.3567,
+      "step": 1434
+    },
+    {
+      "epoch": 1.1650879566982408,
+      "grad_norm": 0.2388361543416977,
+      "learning_rate": 7.672585231482723e-06,
+      "loss": 0.3659,
+      "step": 1435
+    },
+    {
+      "epoch": 1.1658998646820027,
+      "grad_norm": 0.24825267493724823,
+      "learning_rate": 7.668588941101208e-06,
+      "loss": 0.3774,
+      "step": 1436
+    },
+    {
+      "epoch": 1.1667117726657645,
+      "grad_norm": 0.2478463351726532,
+      "learning_rate": 7.664590265542144e-06,
+      "loss": 0.3721,
+      "step": 1437
+    },
+    {
+      "epoch": 1.1675236806495264,
+      "grad_norm": 0.24564437568187714,
+      "learning_rate": 7.66058920837954e-06,
+      "loss": 0.3773,
+      "step": 1438
+    },
+    {
+      "epoch": 1.1683355886332882,
+      "grad_norm": 0.2626170516014099,
+      "learning_rate": 7.65658577318953e-06,
+      "loss": 0.367,
+      "step": 1439
+    },
+    {
+      "epoch": 1.16914749661705,
+      "grad_norm": 0.22646848857402802,
+      "learning_rate": 7.65257996355037e-06,
+      "loss": 0.32,
+      "step": 1440
+    },
+    {
+      "epoch": 1.169959404600812,
+      "grad_norm": 0.2762152850627899,
+      "learning_rate": 7.648571783042445e-06,
+      "loss": 0.3818,
+      "step": 1441
+    },
+    {
+      "epoch": 1.1707713125845738,
+      "grad_norm": 0.25308188796043396,
+      "learning_rate": 7.644561235248254e-06,
+      "loss": 0.3563,
+      "step": 1442
+    },
+    {
+      "epoch": 1.1715832205683356,
+      "grad_norm": 0.2231668382883072,
+      "learning_rate": 7.640548323752412e-06,
+      "loss": 0.3362,
+      "step": 1443
+    },
+    {
+      "epoch": 1.1723951285520975,
+      "grad_norm": 0.25222721695899963,
+      "learning_rate": 7.636533052141653e-06,
+      "loss": 0.3651,
+      "step": 1444
+    },
+    {
+      "epoch": 1.1732070365358593,
+      "grad_norm": 0.2433440089225769,
+      "learning_rate": 7.632515424004813e-06,
+      "loss": 0.3714,
+      "step": 1445
+    },
+    {
+      "epoch": 1.1740189445196212,
+      "grad_norm": 0.22430385649204254,
+      "learning_rate": 7.628495442932838e-06,
+      "loss": 0.3351,
+      "step": 1446
+    },
+    {
+      "epoch": 1.174830852503383,
+      "grad_norm": 0.26103752851486206,
+      "learning_rate": 7.624473112518777e-06,
+      "loss": 0.3865,
+      "step": 1447
+    },
+    {
+      "epoch": 1.1756427604871447,
+      "grad_norm": 0.24525955319404602,
+      "learning_rate": 7.6204484363577756e-06,
+      "loss": 0.3408,
+      "step": 1448
+    },
+    {
+      "epoch": 1.1764546684709067,
+      "grad_norm": 0.22841687500476837,
+      "learning_rate": 7.61642141804708e-06,
+      "loss": 0.3441,
+      "step": 1449
+    },
+    {
+      "epoch": 1.1772665764546684,
+      "grad_norm": 0.2713543176651001,
+      "learning_rate": 7.612392061186027e-06,
+      "loss": 0.3914,
+      "step": 1450
+    },
+    {
+      "epoch": 1.1780784844384302,
+      "grad_norm": 0.22072599828243256,
+      "learning_rate": 7.608360369376047e-06,
+      "loss": 0.3461,
+      "step": 1451
+    },
+    {
+      "epoch": 1.178890392422192,
+      "grad_norm": 0.2488977611064911,
+      "learning_rate": 7.604326346220654e-06,
+      "loss": 0.3634,
+      "step": 1452
+    },
+    {
+      "epoch": 1.179702300405954,
+      "grad_norm": 0.24902021884918213,
+      "learning_rate": 7.600289995325446e-06,
+      "loss": 0.3789,
+      "step": 1453
+    },
+    {
+      "epoch": 1.1805142083897158,
+      "grad_norm": 0.2225618213415146,
+      "learning_rate": 7.596251320298105e-06,
+      "loss": 0.3593,
+      "step": 1454
+    },
+    {
+      "epoch": 1.1813261163734776,
+      "grad_norm": 0.25640109181404114,
+      "learning_rate": 7.592210324748384e-06,
+      "loss": 0.3896,
+      "step": 1455
+    },
+    {
+      "epoch": 1.1821380243572395,
+      "grad_norm": 0.20487511157989502,
+      "learning_rate": 7.588167012288117e-06,
+      "loss": 0.3378,
+      "step": 1456
+    },
+    {
+      "epoch": 1.1829499323410013,
+      "grad_norm": 0.262029767036438,
+      "learning_rate": 7.584121386531205e-06,
+      "loss": 0.3892,
+      "step": 1457
+    },
+    {
+      "epoch": 1.1837618403247632,
+      "grad_norm": 0.24068884551525116,
+      "learning_rate": 7.580073451093617e-06,
+      "loss": 0.3278,
+      "step": 1458
+    },
+    {
+      "epoch": 1.184573748308525,
+      "grad_norm": 0.2619876265525818,
+      "learning_rate": 7.576023209593386e-06,
+      "loss": 0.3676,
+      "step": 1459
+    },
+    {
+      "epoch": 1.1853856562922869,
+      "grad_norm": 0.23555003106594086,
+      "learning_rate": 7.571970665650607e-06,
+      "loss": 0.3657,
+      "step": 1460
+    },
+    {
+      "epoch": 1.1861975642760487,
+      "grad_norm": 0.23760762810707092,
+      "learning_rate": 7.567915822887434e-06,
+      "loss": 0.3421,
+      "step": 1461
+    },
+    {
+      "epoch": 1.1870094722598106,
+      "grad_norm": 0.2487548142671585,
+      "learning_rate": 7.563858684928074e-06,
+      "loss": 0.3428,
+      "step": 1462
+    },
+    {
+      "epoch": 1.1878213802435724,
+      "grad_norm": 0.24526885151863098,
+      "learning_rate": 7.559799255398785e-06,
+      "loss": 0.3713,
+      "step": 1463
+    },
+    {
+      "epoch": 1.1886332882273343,
+      "grad_norm": 0.22546742856502533,
+      "learning_rate": 7.555737537927875e-06,
+      "loss": 0.3845,
+      "step": 1464
+    },
+    {
+      "epoch": 1.1894451962110961,
+      "grad_norm": 0.2158726006746292,
+      "learning_rate": 7.551673536145695e-06,
+      "loss": 0.3312,
+      "step": 1465
+    },
+    {
+      "epoch": 1.190257104194858,
+      "grad_norm": 0.22490805387496948,
+      "learning_rate": 7.54760725368464e-06,
+      "loss": 0.3816,
+      "step": 1466
+    },
+    {
+      "epoch": 1.1910690121786198,
+      "grad_norm": 0.2474667727947235,
+      "learning_rate": 7.543538694179142e-06,
+      "loss": 0.3663,
+      "step": 1467
+    },
+    {
+      "epoch": 1.1918809201623817,
+      "grad_norm": 0.24773447215557098,
+      "learning_rate": 7.539467861265668e-06,
+      "loss": 0.3563,
+      "step": 1468
+    },
+    {
+      "epoch": 1.1926928281461435,
+      "grad_norm": 0.24624335765838623,
+      "learning_rate": 7.535394758582717e-06,
+      "loss": 0.3718,
+      "step": 1469
+    },
+    {
+      "epoch": 1.1935047361299054,
+      "grad_norm": 0.2261711061000824,
+      "learning_rate": 7.531319389770818e-06,
+      "loss": 0.3464,
+      "step": 1470
+    },
+    {
+      "epoch": 1.194316644113667,
+      "grad_norm": 0.22112810611724854,
+      "learning_rate": 7.527241758472525e-06,
+      "loss": 0.3091,
+      "step": 1471
+    },
+    {
+      "epoch": 1.195128552097429,
+      "grad_norm": 0.22846589982509613,
+      "learning_rate": 7.523161868332412e-06,
+      "loss": 0.3637,
+      "step": 1472
+    },
+    {
+      "epoch": 1.1959404600811907,
+      "grad_norm": 0.25148358941078186,
+      "learning_rate": 7.519079722997076e-06,
+      "loss": 0.3845,
+      "step": 1473
+    },
+    {
+      "epoch": 1.1967523680649526,
+      "grad_norm": 0.2260669320821762,
+      "learning_rate": 7.514995326115124e-06,
+      "loss": 0.3504,
+      "step": 1474
+    },
+    {
+      "epoch": 1.1975642760487144,
+      "grad_norm": 0.26081445813179016,
+      "learning_rate": 7.5109086813371835e-06,
+      "loss": 0.3779,
+      "step": 1475
+    },
+    {
+      "epoch": 1.1983761840324763,
+      "grad_norm": 0.25127705931663513,
+      "learning_rate": 7.506819792315883e-06,
+      "loss": 0.3578,
+      "step": 1476
+    },
+    {
+      "epoch": 1.199188092016238,
+      "grad_norm": 0.23205429315567017,
+      "learning_rate": 7.502728662705861e-06,
+      "loss": 0.3324,
+      "step": 1477
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.268428772687912,
+      "learning_rate": 7.498635296163758e-06,
+      "loss": 0.3596,
+      "step": 1478
+    },
+    {
+      "epoch": 1.2008119079837618,
+      "grad_norm": 0.23894675076007843,
+      "learning_rate": 7.494539696348215e-06,
+      "loss": 0.3641,
+      "step": 1479
+    },
+    {
+      "epoch": 1.2016238159675237,
+      "grad_norm": 0.2446397840976715,
+      "learning_rate": 7.490441866919869e-06,
+      "loss": 0.3547,
+      "step": 1480
+    },
+    {
+      "epoch": 1.2024357239512855,
+      "grad_norm": 0.2721719741821289,
+      "learning_rate": 7.486341811541345e-06,
+      "loss": 0.3839,
+      "step": 1481
+    },
+    {
+      "epoch": 1.2032476319350474,
+      "grad_norm": 0.2253265529870987,
+      "learning_rate": 7.482239533877264e-06,
+      "loss": 0.3383,
+      "step": 1482
+    },
+    {
+      "epoch": 1.2040595399188092,
+      "grad_norm": 0.26344582438468933,
+      "learning_rate": 7.4781350375942296e-06,
+      "loss": 0.3896,
+      "step": 1483
+    },
+    {
+      "epoch": 1.204871447902571,
+      "grad_norm": 0.2449388951063156,
+      "learning_rate": 7.4740283263608305e-06,
+      "loss": 0.3174,
+      "step": 1484
+    },
+    {
+      "epoch": 1.205683355886333,
+      "grad_norm": 0.2524270713329315,
+      "learning_rate": 7.469919403847633e-06,
+      "loss": 0.3767,
+      "step": 1485
+    },
+    {
+      "epoch": 1.2064952638700948,
+      "grad_norm": 0.24180158972740173,
+      "learning_rate": 7.465808273727182e-06,
+      "loss": 0.3967,
+      "step": 1486
+    },
+    {
+      "epoch": 1.2073071718538566,
+      "grad_norm": 0.23717573285102844,
+      "learning_rate": 7.461694939673992e-06,
+      "loss": 0.311,
+      "step": 1487
+    },
+    {
+      "epoch": 1.2081190798376185,
+      "grad_norm": 0.2801485061645508,
+      "learning_rate": 7.457579405364555e-06,
+      "loss": 0.3641,
+      "step": 1488
+    },
+    {
+      "epoch": 1.2089309878213803,
+      "grad_norm": 0.24097450077533722,
+      "learning_rate": 7.453461674477318e-06,
+      "loss": 0.3371,
+      "step": 1489
+    },
+    {
+      "epoch": 1.2097428958051422,
+      "grad_norm": 0.22208639979362488,
+      "learning_rate": 7.449341750692703e-06,
+      "loss": 0.3381,
+      "step": 1490
+    },
+    {
+      "epoch": 1.210554803788904,
+      "grad_norm": 0.2906947135925293,
+      "learning_rate": 7.445219637693082e-06,
+      "loss": 0.3541,
+      "step": 1491
+    },
+    {
+      "epoch": 1.2113667117726659,
+      "grad_norm": 0.2547138035297394,
+      "learning_rate": 7.441095339162791e-06,
+      "loss": 0.3535,
+      "step": 1492
+    },
+    {
+      "epoch": 1.2121786197564277,
+      "grad_norm": 0.242499440908432,
+      "learning_rate": 7.436968858788118e-06,
+      "loss": 0.3669,
+      "step": 1493
+    },
+    {
+      "epoch": 1.2129905277401893,
+      "grad_norm": 0.22720834612846375,
+      "learning_rate": 7.4328402002572985e-06,
+      "loss": 0.3071,
+      "step": 1494
+    },
+    {
+      "epoch": 1.2138024357239512,
+      "grad_norm": 0.24428679049015045,
+      "learning_rate": 7.428709367260516e-06,
+      "loss": 0.38,
+      "step": 1495
+    },
+    {
+      "epoch": 1.214614343707713,
+      "grad_norm": 0.2398340106010437,
+      "learning_rate": 7.424576363489897e-06,
+      "loss": 0.3492,
+      "step": 1496
+    },
+    {
+      "epoch": 1.215426251691475,
+      "grad_norm": 0.2369420975446701,
+      "learning_rate": 7.420441192639511e-06,
+      "loss": 0.3642,
+      "step": 1497
+    },
+    {
+      "epoch": 1.2162381596752367,
+      "grad_norm": 0.275720477104187,
+      "learning_rate": 7.416303858405363e-06,
+      "loss": 0.3715,
+      "step": 1498
+    },
+    {
+      "epoch": 1.2170500676589986,
+      "grad_norm": 0.2526825964450836,
+      "learning_rate": 7.412164364485388e-06,
+      "loss": 0.3662,
+      "step": 1499
+    },
+    {
+      "epoch": 1.2178619756427604,
+      "grad_norm": 0.2425920069217682,
+      "learning_rate": 7.408022714579457e-06,
+      "loss": 0.3407,
+      "step": 1500
+    },
+    {
+      "epoch": 1.2186738836265223,
+      "grad_norm": 0.251361608505249,
+      "learning_rate": 7.403878912389365e-06,
+      "loss": 0.3646,
+      "step": 1501
+    },
+    {
+      "epoch": 1.2194857916102841,
+      "grad_norm": 0.22192996740341187,
+      "learning_rate": 7.3997329616188305e-06,
+      "loss": 0.3406,
+      "step": 1502
+    },
+    {
+      "epoch": 1.220297699594046,
+      "grad_norm": 0.24959032237529755,
+      "learning_rate": 7.395584865973493e-06,
+      "loss": 0.3436,
+      "step": 1503
+    },
+    {
+      "epoch": 1.2211096075778078,
+      "grad_norm": 0.23616479337215424,
+      "learning_rate": 7.391434629160909e-06,
+      "loss": 0.356,
+      "step": 1504
+    },
+    {
+      "epoch": 1.2219215155615697,
+      "grad_norm": 0.2401830106973648,
+      "learning_rate": 7.38728225489055e-06,
+      "loss": 0.3823,
+      "step": 1505
+    },
+    {
+      "epoch": 1.2227334235453315,
+      "grad_norm": 0.252386212348938,
+      "learning_rate": 7.383127746873796e-06,
+      "loss": 0.3669,
+      "step": 1506
+    },
+    {
+      "epoch": 1.2235453315290934,
+      "grad_norm": 0.23745477199554443,
+      "learning_rate": 7.378971108823933e-06,
+      "loss": 0.3576,
+      "step": 1507
+    },
+    {
+      "epoch": 1.2243572395128552,
+      "grad_norm": 0.209089457988739,
+      "learning_rate": 7.374812344456157e-06,
+      "loss": 0.3535,
+      "step": 1508
+    },
+    {
+      "epoch": 1.225169147496617,
+      "grad_norm": 0.21981292963027954,
+      "learning_rate": 7.370651457487559e-06,
+      "loss": 0.3594,
+      "step": 1509
+    },
+    {
+      "epoch": 1.225981055480379,
+      "grad_norm": 0.22727377712726593,
+      "learning_rate": 7.366488451637126e-06,
+      "loss": 0.3242,
+      "step": 1510
+    },
+    {
+      "epoch": 1.2267929634641408,
+      "grad_norm": 0.26466935873031616,
+      "learning_rate": 7.362323330625744e-06,
+      "loss": 0.3719,
+      "step": 1511
+    },
+    {
+      "epoch": 1.2276048714479026,
+      "grad_norm": 0.23351871967315674,
+      "learning_rate": 7.358156098176185e-06,
+      "loss": 0.3373,
+      "step": 1512
+    },
+    {
+      "epoch": 1.2284167794316645,
+      "grad_norm": 0.24746312201023102,
+      "learning_rate": 7.353986758013112e-06,
+      "loss": 0.3574,
+      "step": 1513
+    },
+    {
+      "epoch": 1.2292286874154263,
+      "grad_norm": 0.2734043300151825,
+      "learning_rate": 7.349815313863068e-06,
+      "loss": 0.3581,
+      "step": 1514
+    },
+    {
+      "epoch": 1.230040595399188,
+      "grad_norm": 0.22939737141132355,
+      "learning_rate": 7.3456417694544804e-06,
+      "loss": 0.3719,
+      "step": 1515
+    },
+    {
+      "epoch": 1.23085250338295,
+      "grad_norm": 0.245289146900177,
+      "learning_rate": 7.341466128517649e-06,
+      "loss": 0.3606,
+      "step": 1516
+    },
+    {
+      "epoch": 1.2316644113667117,
+      "grad_norm": 0.24674183130264282,
+      "learning_rate": 7.337288394784754e-06,
+      "loss": 0.3984,
+      "step": 1517
+    },
+    {
+      "epoch": 1.2324763193504735,
+      "grad_norm": 0.22073611617088318,
+      "learning_rate": 7.3331085719898385e-06,
+      "loss": 0.3138,
+      "step": 1518
+    },
+    {
+      "epoch": 1.2332882273342354,
+      "grad_norm": 0.2932286560535431,
+      "learning_rate": 7.32892666386882e-06,
+      "loss": 0.3993,
+      "step": 1519
+    },
+    {
+      "epoch": 1.2341001353179972,
+      "grad_norm": 0.2377910166978836,
+      "learning_rate": 7.324742674159475e-06,
+      "loss": 0.3812,
+      "step": 1520
+    },
+    {
+      "epoch": 1.234912043301759,
+      "grad_norm": 0.22169746458530426,
+      "learning_rate": 7.3205566066014436e-06,
+      "loss": 0.3495,
+      "step": 1521
+    },
+    {
+      "epoch": 1.235723951285521,
+      "grad_norm": 0.22318966686725616,
+      "learning_rate": 7.316368464936219e-06,
+      "loss": 0.3544,
+      "step": 1522
+    },
+    {
+      "epoch": 1.2365358592692828,
+      "grad_norm": 0.24567922949790955,
+      "learning_rate": 7.312178252907154e-06,
+      "loss": 0.3399,
+      "step": 1523
+    },
+    {
+      "epoch": 1.2373477672530446,
+      "grad_norm": 0.23027023673057556,
+      "learning_rate": 7.307985974259445e-06,
+      "loss": 0.3844,
+      "step": 1524
+    },
+    {
+      "epoch": 1.2381596752368065,
+      "grad_norm": 0.23381324112415314,
+      "learning_rate": 7.303791632740143e-06,
+      "loss": 0.3571,
+      "step": 1525
+    },
+    {
+      "epoch": 1.2389715832205683,
+      "grad_norm": 0.2318311184644699,
+      "learning_rate": 7.2995952320981356e-06,
+      "loss": 0.3535,
+      "step": 1526
+    },
+    {
+      "epoch": 1.2397834912043302,
+      "grad_norm": 0.23061968386173248,
+      "learning_rate": 7.2953967760841585e-06,
+      "loss": 0.3707,
+      "step": 1527
+    },
+    {
+      "epoch": 1.240595399188092,
+      "grad_norm": 0.2705461382865906,
+      "learning_rate": 7.291196268450777e-06,
+      "loss": 0.3756,
+      "step": 1528
+    },
+    {
+      "epoch": 1.2414073071718539,
+      "grad_norm": 0.2447393536567688,
+      "learning_rate": 7.286993712952394e-06,
+      "loss": 0.341,
+      "step": 1529
+    },
+    {
+      "epoch": 1.2422192151556157,
+      "grad_norm": 0.2647668421268463,
+      "learning_rate": 7.282789113345242e-06,
+      "loss": 0.382,
+      "step": 1530
+    },
+    {
+      "epoch": 1.2430311231393776,
+      "grad_norm": 0.23277398943901062,
+      "learning_rate": 7.278582473387382e-06,
+      "loss": 0.3405,
+      "step": 1531
+    },
+    {
+      "epoch": 1.2438430311231394,
+      "grad_norm": 0.23825831711292267,
+      "learning_rate": 7.274373796838696e-06,
+      "loss": 0.3485,
+      "step": 1532
+    },
+    {
+      "epoch": 1.2446549391069013,
+      "grad_norm": 0.2280379831790924,
+      "learning_rate": 7.27016308746089e-06,
+      "loss": 0.3646,
+      "step": 1533
+    },
+    {
+      "epoch": 1.2454668470906631,
+      "grad_norm": 0.22873491048812866,
+      "learning_rate": 7.265950349017481e-06,
+      "loss": 0.3654,
+      "step": 1534
+    },
+    {
+      "epoch": 1.246278755074425,
+      "grad_norm": 0.22860011458396912,
+      "learning_rate": 7.261735585273804e-06,
+      "loss": 0.3751,
+      "step": 1535
+    },
+    {
+      "epoch": 1.2470906630581868,
+      "grad_norm": 0.23397758603096008,
+      "learning_rate": 7.257518799997007e-06,
+      "loss": 0.3371,
+      "step": 1536
+    },
+    {
+      "epoch": 1.2479025710419487,
+      "grad_norm": 0.24259443581104279,
+      "learning_rate": 7.253299996956038e-06,
+      "loss": 0.3502,
+      "step": 1537
+    },
+    {
+      "epoch": 1.2487144790257103,
+      "grad_norm": 0.24603794515132904,
+      "learning_rate": 7.249079179921652e-06,
+      "loss": 0.3815,
+      "step": 1538
+    },
+    {
+      "epoch": 1.2495263870094724,
+      "grad_norm": 0.2117844521999359,
+      "learning_rate": 7.244856352666404e-06,
+      "loss": 0.3302,
+      "step": 1539
+    },
+    {
+      "epoch": 1.250338294993234,
+      "grad_norm": 0.22870968282222748,
+      "learning_rate": 7.240631518964646e-06,
+      "loss": 0.3689,
+      "step": 1540
+    },
+    {
+      "epoch": 1.2511502029769959,
+      "grad_norm": 0.23572908341884613,
+      "learning_rate": 7.236404682592523e-06,
+      "loss": 0.3429,
+      "step": 1541
+    },
+    {
+      "epoch": 1.2519621109607577,
+      "grad_norm": 0.22629325091838837,
+      "learning_rate": 7.232175847327969e-06,
+      "loss": 0.3394,
+      "step": 1542
+    },
+    {
+      "epoch": 1.2527740189445196,
+      "grad_norm": 0.23883716762065887,
+      "learning_rate": 7.227945016950706e-06,
+      "loss": 0.3765,
+      "step": 1543
+    },
+    {
+      "epoch": 1.2535859269282814,
+      "grad_norm": 0.24776387214660645,
+      "learning_rate": 7.223712195242238e-06,
+      "loss": 0.3708,
+      "step": 1544
+    },
+    {
+      "epoch": 1.2543978349120433,
+      "grad_norm": 0.26760968565940857,
+      "learning_rate": 7.2194773859858495e-06,
+      "loss": 0.3451,
+      "step": 1545
+    },
+    {
+      "epoch": 1.2552097428958051,
+      "grad_norm": 0.2337782233953476,
+      "learning_rate": 7.215240592966603e-06,
+      "loss": 0.3696,
+      "step": 1546
+    },
+    {
+      "epoch": 1.256021650879567,
+      "grad_norm": 0.2598925828933716,
+      "learning_rate": 7.2110018199713325e-06,
+      "loss": 0.3602,
+      "step": 1547
+    },
+    {
+      "epoch": 1.2568335588633288,
+      "grad_norm": 0.2539239227771759,
+      "learning_rate": 7.206761070788641e-06,
+      "loss": 0.3335,
+      "step": 1548
+    },
+    {
+      "epoch": 1.2576454668470907,
+      "grad_norm": 0.2241261899471283,
+      "learning_rate": 7.202518349208898e-06,
+      "loss": 0.3649,
+      "step": 1549
+    },
+    {
+      "epoch": 1.2584573748308525,
+      "grad_norm": 0.24881933629512787,
+      "learning_rate": 7.1982736590242375e-06,
+      "loss": 0.3389,
+      "step": 1550
+    },
+    {
+      "epoch": 1.2592692828146144,
+      "grad_norm": 0.2607828378677368,
+      "learning_rate": 7.194027004028552e-06,
+      "loss": 0.3661,
+      "step": 1551
+    },
+    {
+      "epoch": 1.2600811907983762,
+      "grad_norm": 0.23785413801670074,
+      "learning_rate": 7.189778388017489e-06,
+      "loss": 0.374,
+      "step": 1552
+    },
+    {
+      "epoch": 1.260893098782138,
+      "grad_norm": 0.22073908150196075,
+      "learning_rate": 7.185527814788451e-06,
+      "loss": 0.3204,
+      "step": 1553
+    },
+    {
+      "epoch": 1.2617050067659,
+      "grad_norm": 0.26098576188087463,
+      "learning_rate": 7.18127528814059e-06,
+      "loss": 0.411,
+      "step": 1554
+    },
+    {
+      "epoch": 1.2625169147496618,
+      "grad_norm": 0.23809435963630676,
+      "learning_rate": 7.1770208118748e-06,
+      "loss": 0.3283,
+      "step": 1555
+    },
+    {
+      "epoch": 1.2633288227334236,
+      "grad_norm": 0.2629137337207794,
+      "learning_rate": 7.172764389793723e-06,
+      "loss": 0.3675,
+      "step": 1556
+    },
+    {
+      "epoch": 1.2641407307171855,
+      "grad_norm": 0.24093160033226013,
+      "learning_rate": 7.168506025701734e-06,
+      "loss": 0.3694,
+      "step": 1557
+    },
+    {
+      "epoch": 1.2649526387009473,
+      "grad_norm": 0.2612915635108948,
+      "learning_rate": 7.164245723404951e-06,
+      "loss": 0.3691,
+      "step": 1558
+    },
+    {
+      "epoch": 1.265764546684709,
+      "grad_norm": 0.22047817707061768,
+      "learning_rate": 7.159983486711219e-06,
+      "loss": 0.3395,
+      "step": 1559
+    },
+    {
+      "epoch": 1.266576454668471,
+      "grad_norm": 0.2716349959373474,
+      "learning_rate": 7.155719319430114e-06,
+      "loss": 0.4311,
+      "step": 1560
+    },
+    {
+      "epoch": 1.2673883626522326,
+      "grad_norm": 0.22551296651363373,
+      "learning_rate": 7.151453225372938e-06,
+      "loss": 0.318,
+      "step": 1561
+    },
+    {
+      "epoch": 1.2682002706359947,
+      "grad_norm": 0.21630807220935822,
+      "learning_rate": 7.147185208352712e-06,
+      "loss": 0.3197,
+      "step": 1562
+    },
+    {
+      "epoch": 1.2690121786197563,
+      "grad_norm": 0.25373777747154236,
+      "learning_rate": 7.142915272184179e-06,
+      "loss": 0.3836,
+      "step": 1563
+    },
+    {
+      "epoch": 1.2698240866035182,
+      "grad_norm": 0.24111326038837433,
+      "learning_rate": 7.138643420683795e-06,
+      "loss": 0.3908,
+      "step": 1564
+    },
+    {
+      "epoch": 1.27063599458728,
+      "grad_norm": 0.26475778222084045,
+      "learning_rate": 7.134369657669733e-06,
+      "loss": 0.3636,
+      "step": 1565
+    },
+    {
+      "epoch": 1.271447902571042,
+      "grad_norm": 0.28030601143836975,
+      "learning_rate": 7.130093986961868e-06,
+      "loss": 0.3564,
+      "step": 1566
+    },
+    {
+      "epoch": 1.2722598105548037,
+      "grad_norm": 0.2583480477333069,
+      "learning_rate": 7.1258164123817834e-06,
+      "loss": 0.3185,
+      "step": 1567
+    },
+    {
+      "epoch": 1.2730717185385656,
+      "grad_norm": 0.2874113619327545,
+      "learning_rate": 7.121536937752764e-06,
+      "loss": 0.3928,
+      "step": 1568
+    },
+    {
+      "epoch": 1.2738836265223275,
+      "grad_norm": 0.29563698172569275,
+      "learning_rate": 7.1172555668997925e-06,
+      "loss": 0.3978,
+      "step": 1569
+    },
+    {
+      "epoch": 1.2746955345060893,
+      "grad_norm": 0.21970148384571075,
+      "learning_rate": 7.112972303649546e-06,
+      "loss": 0.3252,
+      "step": 1570
+    },
+    {
+      "epoch": 1.2755074424898512,
+      "grad_norm": 0.2617684304714203,
+      "learning_rate": 7.108687151830394e-06,
+      "loss": 0.3677,
+      "step": 1571
+    },
+    {
+      "epoch": 1.276319350473613,
+      "grad_norm": 0.24680693447589874,
+      "learning_rate": 7.104400115272393e-06,
+      "loss": 0.3567,
+      "step": 1572
+    },
+    {
+      "epoch": 1.2771312584573749,
+      "grad_norm": 0.23597772419452667,
+      "learning_rate": 7.100111197807286e-06,
+      "loss": 0.3447,
+      "step": 1573
+    },
+    {
+      "epoch": 1.2779431664411367,
+      "grad_norm": 0.28573334217071533,
+      "learning_rate": 7.095820403268494e-06,
+      "loss": 0.3722,
+      "step": 1574
+    },
+    {
+      "epoch": 1.2787550744248986,
+      "grad_norm": 0.22354266047477722,
+      "learning_rate": 7.091527735491119e-06,
+      "loss": 0.3702,
+      "step": 1575
+    },
+    {
+      "epoch": 1.2795669824086604,
+      "grad_norm": 0.2540134787559509,
+      "learning_rate": 7.087233198311936e-06,
+      "loss": 0.3504,
+      "step": 1576
+    },
+    {
+      "epoch": 1.2803788903924223,
+      "grad_norm": 0.2594829797744751,
+      "learning_rate": 7.082936795569388e-06,
+      "loss": 0.3682,
+      "step": 1577
+    },
+    {
+      "epoch": 1.281190798376184,
+      "grad_norm": 0.2425624132156372,
+      "learning_rate": 7.078638531103592e-06,
+      "loss": 0.3702,
+      "step": 1578
+    },
+    {
+      "epoch": 1.282002706359946,
+      "grad_norm": 0.22648274898529053,
+      "learning_rate": 7.074338408756321e-06,
+      "loss": 0.3612,
+      "step": 1579
+    },
+    {
+      "epoch": 1.2828146143437076,
+      "grad_norm": 0.2166253626346588,
+      "learning_rate": 7.070036432371017e-06,
+      "loss": 0.3066,
+      "step": 1580
+    },
+    {
+      "epoch": 1.2836265223274697,
+      "grad_norm": 0.237893208861351,
+      "learning_rate": 7.065732605792771e-06,
+      "loss": 0.3738,
+      "step": 1581
+    },
+    {
+      "epoch": 1.2844384303112313,
+      "grad_norm": 0.23682132363319397,
+      "learning_rate": 7.061426932868334e-06,
+      "loss": 0.3699,
+      "step": 1582
+    },
+    {
+      "epoch": 1.2852503382949934,
+      "grad_norm": 0.22960659861564636,
+      "learning_rate": 7.057119417446102e-06,
+      "loss": 0.3618,
+      "step": 1583
+    },
+    {
+      "epoch": 1.286062246278755,
+      "grad_norm": 0.23040653765201569,
+      "learning_rate": 7.052810063376124e-06,
+      "loss": 0.3726,
+      "step": 1584
+    },
+    {
+      "epoch": 1.2868741542625168,
+      "grad_norm": 0.21522389352321625,
+      "learning_rate": 7.048498874510088e-06,
+      "loss": 0.3157,
+      "step": 1585
+    },
+    {
+      "epoch": 1.2876860622462787,
+      "grad_norm": 0.24170489609241486,
+      "learning_rate": 7.044185854701321e-06,
+      "loss": 0.3934,
+      "step": 1586
+    },
+    {
+      "epoch": 1.2884979702300405,
+      "grad_norm": 0.23357383906841278,
+      "learning_rate": 7.03987100780479e-06,
+      "loss": 0.3368,
+      "step": 1587
+    },
+    {
+      "epoch": 1.2893098782138024,
+      "grad_norm": 0.22911779582500458,
+      "learning_rate": 7.035554337677094e-06,
+      "loss": 0.3837,
+      "step": 1588
+    },
+    {
+      "epoch": 1.2901217861975642,
+      "grad_norm": 0.24219779670238495,
+      "learning_rate": 7.03123584817646e-06,
+      "loss": 0.3289,
+      "step": 1589
+    },
+    {
+      "epoch": 1.290933694181326,
+      "grad_norm": 0.23957718908786774,
+      "learning_rate": 7.026915543162741e-06,
+      "loss": 0.3785,
+      "step": 1590
+    },
+    {
+      "epoch": 1.291745602165088,
+      "grad_norm": 0.23434984683990479,
+      "learning_rate": 7.0225934264974175e-06,
+      "loss": 0.3407,
+      "step": 1591
+    },
+    {
+      "epoch": 1.2925575101488498,
+      "grad_norm": 0.24710939824581146,
+      "learning_rate": 7.018269502043581e-06,
+      "loss": 0.3775,
+      "step": 1592
+    },
+    {
+      "epoch": 1.2933694181326116,
+      "grad_norm": 0.24073292315006256,
+      "learning_rate": 7.013943773665947e-06,
+      "loss": 0.3853,
+      "step": 1593
+    },
+    {
+      "epoch": 1.2941813261163735,
+      "grad_norm": 0.20647989213466644,
+      "learning_rate": 7.009616245230838e-06,
+      "loss": 0.3282,
+      "step": 1594
+    },
+    {
+      "epoch": 1.2949932341001353,
+      "grad_norm": 0.22468654811382294,
+      "learning_rate": 7.005286920606188e-06,
+      "loss": 0.3565,
+      "step": 1595
+    },
+    {
+      "epoch": 1.2958051420838972,
+      "grad_norm": 0.21744848787784576,
+      "learning_rate": 7.0009558036615355e-06,
+      "loss": 0.3444,
+      "step": 1596
+    },
+    {
+      "epoch": 1.296617050067659,
+      "grad_norm": 0.2652900218963623,
+      "learning_rate": 6.996622898268021e-06,
+      "loss": 0.4008,
+      "step": 1597
+    },
+    {
+      "epoch": 1.2974289580514209,
+      "grad_norm": 0.2158915400505066,
+      "learning_rate": 6.992288208298383e-06,
+      "loss": 0.343,
+      "step": 1598
+    },
+    {
+      "epoch": 1.2982408660351827,
+      "grad_norm": 0.2695927023887634,
+      "learning_rate": 6.987951737626956e-06,
+      "loss": 0.383,
+      "step": 1599
+    },
+    {
+      "epoch": 1.2990527740189446,
+      "grad_norm": 0.23025627434253693,
+      "learning_rate": 6.983613490129666e-06,
+      "loss": 0.354,
+      "step": 1600
+    },
+    {
+      "epoch": 1.2998646820027064,
+      "grad_norm": 0.2505878806114197,
+      "learning_rate": 6.979273469684026e-06,
+      "loss": 0.3549,
+      "step": 1601
+    },
+    {
+      "epoch": 1.3006765899864683,
+      "grad_norm": 0.23811763525009155,
+      "learning_rate": 6.974931680169136e-06,
+      "loss": 0.3788,
+      "step": 1602
+    },
+    {
+      "epoch": 1.30148849797023,
+      "grad_norm": 0.232005774974823,
+      "learning_rate": 6.970588125465674e-06,
+      "loss": 0.3622,
+      "step": 1603
+    },
+    {
+      "epoch": 1.302300405953992,
+      "grad_norm": 0.22171185910701752,
+      "learning_rate": 6.966242809455899e-06,
+      "loss": 0.3489,
+      "step": 1604
+    },
+    {
+      "epoch": 1.3031123139377536,
+      "grad_norm": 0.22553686797618866,
+      "learning_rate": 6.961895736023641e-06,
+      "loss": 0.359,
+      "step": 1605
+    },
+    {
+      "epoch": 1.3039242219215157,
+      "grad_norm": 0.24445968866348267,
+      "learning_rate": 6.957546909054304e-06,
+      "loss": 0.3775,
+      "step": 1606
+    },
+    {
+      "epoch": 1.3047361299052773,
+      "grad_norm": 0.22749166190624237,
+      "learning_rate": 6.953196332434856e-06,
+      "loss": 0.3606,
+      "step": 1607
+    },
+    {
+      "epoch": 1.3055480378890392,
+      "grad_norm": 0.2186094969511032,
+      "learning_rate": 6.948844010053832e-06,
+      "loss": 0.3231,
+      "step": 1608
+    },
+    {
+      "epoch": 1.306359945872801,
+      "grad_norm": 0.26568809151649475,
+      "learning_rate": 6.944489945801326e-06,
+      "loss": 0.3867,
+      "step": 1609
+    },
+    {
+      "epoch": 1.3071718538565629,
+      "grad_norm": 0.22099466621875763,
+      "learning_rate": 6.940134143568987e-06,
+      "loss": 0.336,
+      "step": 1610
+    },
+    {
+      "epoch": 1.3079837618403247,
+      "grad_norm": 0.2374304085969925,
+      "learning_rate": 6.935776607250022e-06,
+      "loss": 0.354,
+      "step": 1611
+    },
+    {
+      "epoch": 1.3087956698240866,
+      "grad_norm": 0.23701415956020355,
+      "learning_rate": 6.931417340739183e-06,
+      "loss": 0.3466,
+      "step": 1612
+    },
+    {
+      "epoch": 1.3096075778078484,
+      "grad_norm": 0.2549269199371338,
+      "learning_rate": 6.927056347932772e-06,
+      "loss": 0.3917,
+      "step": 1613
+    },
+    {
+      "epoch": 1.3104194857916103,
+      "grad_norm": 0.2511473596096039,
+      "learning_rate": 6.92269363272863e-06,
+      "loss": 0.3467,
+      "step": 1614
+    },
+    {
+      "epoch": 1.3112313937753721,
+      "grad_norm": 0.22357183694839478,
+      "learning_rate": 6.918329199026143e-06,
+      "loss": 0.3701,
+      "step": 1615
+    },
+    {
+      "epoch": 1.312043301759134,
+      "grad_norm": 0.24373449385166168,
+      "learning_rate": 6.913963050726226e-06,
+      "loss": 0.341,
+      "step": 1616
+    },
+    {
+      "epoch": 1.3128552097428958,
+      "grad_norm": 0.23873990774154663,
+      "learning_rate": 6.909595191731337e-06,
+      "loss": 0.3442,
+      "step": 1617
+    },
+    {
+      "epoch": 1.3136671177266577,
+      "grad_norm": 0.21653050184249878,
+      "learning_rate": 6.905225625945451e-06,
+      "loss": 0.3702,
+      "step": 1618
+    },
+    {
+      "epoch": 1.3144790257104195,
+      "grad_norm": 0.25597596168518066,
+      "learning_rate": 6.900854357274075e-06,
+      "loss": 0.4113,
+      "step": 1619
+    },
+    {
+      "epoch": 1.3152909336941814,
+      "grad_norm": 0.22154821455478668,
+      "learning_rate": 6.896481389624239e-06,
+      "loss": 0.3236,
+      "step": 1620
+    },
+    {
+      "epoch": 1.3161028416779432,
+      "grad_norm": 0.23831982910633087,
+      "learning_rate": 6.892106726904486e-06,
+      "loss": 0.3889,
+      "step": 1621
+    },
+    {
+      "epoch": 1.316914749661705,
+      "grad_norm": 0.24241380393505096,
+      "learning_rate": 6.887730373024881e-06,
+      "loss": 0.3736,
+      "step": 1622
+    },
+    {
+      "epoch": 1.317726657645467,
+      "grad_norm": 0.24649596214294434,
+      "learning_rate": 6.883352331896998e-06,
+      "loss": 0.3766,
+      "step": 1623
+    },
+    {
+      "epoch": 1.3185385656292286,
+      "grad_norm": 0.2627733051776886,
+      "learning_rate": 6.878972607433915e-06,
+      "loss": 0.3505,
+      "step": 1624
+    },
+    {
+      "epoch": 1.3193504736129906,
+      "grad_norm": 0.2576459050178528,
+      "learning_rate": 6.8745912035502205e-06,
+      "loss": 0.3739,
+      "step": 1625
+    },
+    {
+      "epoch": 1.3201623815967523,
+      "grad_norm": 0.22633013129234314,
+      "learning_rate": 6.870208124161998e-06,
+      "loss": 0.3398,
+      "step": 1626
+    },
+    {
+      "epoch": 1.3209742895805143,
+      "grad_norm": 0.2224494218826294,
+      "learning_rate": 6.8658233731868355e-06,
+      "loss": 0.3428,
+      "step": 1627
+    },
+    {
+      "epoch": 1.321786197564276,
+      "grad_norm": 0.2283039689064026,
+      "learning_rate": 6.86143695454381e-06,
+      "loss": 0.3785,
+      "step": 1628
+    },
+    {
+      "epoch": 1.322598105548038,
+      "grad_norm": 0.24604202806949615,
+      "learning_rate": 6.857048872153491e-06,
+      "loss": 0.3297,
+      "step": 1629
+    },
+    {
+      "epoch": 1.3234100135317997,
+      "grad_norm": 0.25695744156837463,
+      "learning_rate": 6.852659129937936e-06,
+      "loss": 0.367,
+      "step": 1630
+    },
+    {
+      "epoch": 1.3242219215155615,
+      "grad_norm": 0.2513377070426941,
+      "learning_rate": 6.84826773182068e-06,
+      "loss": 0.3355,
+      "step": 1631
+    },
+    {
+      "epoch": 1.3250338294993234,
+      "grad_norm": 0.25853443145751953,
+      "learning_rate": 6.843874681726747e-06,
+      "loss": 0.3331,
+      "step": 1632
+    },
+    {
+      "epoch": 1.3258457374830852,
+      "grad_norm": 0.23811574280261993,
+      "learning_rate": 6.839479983582632e-06,
+      "loss": 0.3663,
+      "step": 1633
+    },
+    {
+      "epoch": 1.326657645466847,
+      "grad_norm": 0.23377200961112976,
+      "learning_rate": 6.835083641316304e-06,
+      "loss": 0.3818,
+      "step": 1634
+    },
+    {
+      "epoch": 1.327469553450609,
+      "grad_norm": 0.2548816502094269,
+      "learning_rate": 6.830685658857203e-06,
+      "loss": 0.3518,
+      "step": 1635
+    },
+    {
+      "epoch": 1.3282814614343708,
+      "grad_norm": 0.2500341832637787,
+      "learning_rate": 6.826286040136232e-06,
+      "loss": 0.3747,
+      "step": 1636
+    },
+    {
+      "epoch": 1.3290933694181326,
+      "grad_norm": 0.2460351288318634,
+      "learning_rate": 6.82188478908576e-06,
+      "loss": 0.3738,
+      "step": 1637
+    },
+    {
+      "epoch": 1.3299052774018945,
+      "grad_norm": 0.2237626016139984,
+      "learning_rate": 6.817481909639611e-06,
+      "loss": 0.3293,
+      "step": 1638
+    },
+    {
+      "epoch": 1.3307171853856563,
+      "grad_norm": 0.26320114731788635,
+      "learning_rate": 6.813077405733068e-06,
+      "loss": 0.3699,
+      "step": 1639
+    },
+    {
+      "epoch": 1.3315290933694182,
+      "grad_norm": 0.3213578760623932,
+      "learning_rate": 6.808671281302865e-06,
+      "loss": 0.3534,
+      "step": 1640
+    },
+    {
+      "epoch": 1.33234100135318,
+      "grad_norm": 0.22325892746448517,
+      "learning_rate": 6.804263540287184e-06,
+      "loss": 0.3623,
+      "step": 1641
+    },
+    {
+      "epoch": 1.3331529093369419,
+      "grad_norm": 0.26711389422416687,
+      "learning_rate": 6.799854186625651e-06,
+      "loss": 0.3825,
+      "step": 1642
+    },
+    {
+      "epoch": 1.3339648173207037,
+      "grad_norm": 0.25444886088371277,
+      "learning_rate": 6.795443224259335e-06,
+      "loss": 0.3545,
+      "step": 1643
+    },
+    {
+      "epoch": 1.3347767253044656,
+      "grad_norm": 0.25449466705322266,
+      "learning_rate": 6.7910306571307425e-06,
+      "loss": 0.359,
+      "step": 1644
+    },
+    {
+      "epoch": 1.3355886332882274,
+      "grad_norm": 0.23088806867599487,
+      "learning_rate": 6.786616489183814e-06,
+      "loss": 0.348,
+      "step": 1645
+    },
+    {
+      "epoch": 1.3364005412719893,
+      "grad_norm": 0.2614940106868744,
+      "learning_rate": 6.78220072436392e-06,
+      "loss": 0.3846,
+      "step": 1646
+    },
+    {
+      "epoch": 1.337212449255751,
+      "grad_norm": 0.2244085669517517,
+      "learning_rate": 6.7777833666178594e-06,
+      "loss": 0.3264,
+      "step": 1647
+    },
+    {
+      "epoch": 1.338024357239513,
+      "grad_norm": 0.2499360591173172,
+      "learning_rate": 6.773364419893854e-06,
+      "loss": 0.3453,
+      "step": 1648
+    },
+    {
+      "epoch": 1.3388362652232746,
+      "grad_norm": 0.25482529401779175,
+      "learning_rate": 6.768943888141548e-06,
+      "loss": 0.3588,
+      "step": 1649
+    },
+    {
+      "epoch": 1.3396481732070367,
+      "grad_norm": 0.2248770147562027,
+      "learning_rate": 6.764521775312e-06,
+      "loss": 0.3672,
+      "step": 1650
+    },
+    {
+      "epoch": 1.3404600811907983,
+      "grad_norm": 0.22947247326374054,
+      "learning_rate": 6.760098085357681e-06,
+      "loss": 0.3328,
+      "step": 1651
+    },
+    {
+      "epoch": 1.3412719891745601,
+      "grad_norm": 0.24114727973937988,
+      "learning_rate": 6.755672822232475e-06,
+      "loss": 0.346,
+      "step": 1652
+    },
+    {
+      "epoch": 1.342083897158322,
+      "grad_norm": 0.2311500757932663,
+      "learning_rate": 6.7512459898916704e-06,
+      "loss": 0.3482,
+      "step": 1653
+    },
+    {
+      "epoch": 1.3428958051420838,
+      "grad_norm": 0.23954668641090393,
+      "learning_rate": 6.746817592291957e-06,
+      "loss": 0.3853,
+      "step": 1654
+    },
+    {
+      "epoch": 1.3437077131258457,
+      "grad_norm": 0.23173227906227112,
+      "learning_rate": 6.742387633391423e-06,
+      "loss": 0.4052,
+      "step": 1655
+    },
+    {
+      "epoch": 1.3445196211096075,
+      "grad_norm": 0.22577990591526031,
+      "learning_rate": 6.737956117149555e-06,
+      "loss": 0.3278,
+      "step": 1656
+    },
+    {
+      "epoch": 1.3453315290933694,
+      "grad_norm": 0.2270469218492508,
+      "learning_rate": 6.733523047527231e-06,
+      "loss": 0.362,
+      "step": 1657
+    },
+    {
+      "epoch": 1.3461434370771312,
+      "grad_norm": 0.2364434003829956,
+      "learning_rate": 6.729088428486717e-06,
+      "loss": 0.3346,
+      "step": 1658
+    },
+    {
+      "epoch": 1.346955345060893,
+      "grad_norm": 0.22551614046096802,
+      "learning_rate": 6.72465226399166e-06,
+      "loss": 0.348,
+      "step": 1659
+    },
+    {
+      "epoch": 1.347767253044655,
+      "grad_norm": 0.2582223117351532,
+      "learning_rate": 6.720214558007094e-06,
+      "loss": 0.3861,
+      "step": 1660
+    },
+    {
+      "epoch": 1.3485791610284168,
+      "grad_norm": 0.2598671019077301,
+      "learning_rate": 6.715775314499431e-06,
+      "loss": 0.3748,
+      "step": 1661
+    },
+    {
+      "epoch": 1.3493910690121786,
+      "grad_norm": 0.21023385226726532,
+      "learning_rate": 6.711334537436451e-06,
+      "loss": 0.3378,
+      "step": 1662
+    },
+    {
+      "epoch": 1.3502029769959405,
+      "grad_norm": 0.2525627315044403,
+      "learning_rate": 6.7068922307873104e-06,
+      "loss": 0.3809,
+      "step": 1663
+    },
+    {
+      "epoch": 1.3510148849797023,
+      "grad_norm": 0.23106525838375092,
+      "learning_rate": 6.7024483985225295e-06,
+      "loss": 0.3672,
+      "step": 1664
+    },
+    {
+      "epoch": 1.3518267929634642,
+      "grad_norm": 0.26395919919013977,
+      "learning_rate": 6.698003044613997e-06,
+      "loss": 0.4016,
+      "step": 1665
+    },
+    {
+      "epoch": 1.352638700947226,
+      "grad_norm": 0.23417459428310394,
+      "learning_rate": 6.693556173034953e-06,
+      "loss": 0.3578,
+      "step": 1666
+    },
+    {
+      "epoch": 1.353450608930988,
+      "grad_norm": 0.24902866780757904,
+      "learning_rate": 6.689107787760002e-06,
+      "loss": 0.3506,
+      "step": 1667
+    },
+    {
+      "epoch": 1.3542625169147495,
+      "grad_norm": 0.2178947478532791,
+      "learning_rate": 6.6846578927650985e-06,
+      "loss": 0.3596,
+      "step": 1668
+    },
+    {
+      "epoch": 1.3550744248985116,
+      "grad_norm": 0.2619500756263733,
+      "learning_rate": 6.6802064920275475e-06,
+      "loss": 0.3701,
+      "step": 1669
+    },
+    {
+      "epoch": 1.3558863328822732,
+      "grad_norm": 0.2681584358215332,
+      "learning_rate": 6.675753589525996e-06,
+      "loss": 0.3491,
+      "step": 1670
+    },
+    {
+      "epoch": 1.3566982408660353,
+      "grad_norm": 0.1995447427034378,
+      "learning_rate": 6.671299189240439e-06,
+      "loss": 0.3176,
+      "step": 1671
+    },
+    {
+      "epoch": 1.357510148849797,
+      "grad_norm": 0.24435283243656158,
+      "learning_rate": 6.666843295152207e-06,
+      "loss": 0.4062,
+      "step": 1672
+    },
+    {
+      "epoch": 1.358322056833559,
+      "grad_norm": 0.25931650400161743,
+      "learning_rate": 6.662385911243965e-06,
+      "loss": 0.3773,
+      "step": 1673
+    },
+    {
+      "epoch": 1.3591339648173206,
+      "grad_norm": 0.22480809688568115,
+      "learning_rate": 6.657927041499711e-06,
+      "loss": 0.3649,
+      "step": 1674
+    },
+    {
+      "epoch": 1.3599458728010825,
+      "grad_norm": 0.2147214561700821,
+      "learning_rate": 6.653466689904771e-06,
+      "loss": 0.3459,
+      "step": 1675
+    },
+    {
+      "epoch": 1.3607577807848443,
+      "grad_norm": 0.24678853154182434,
+      "learning_rate": 6.649004860445795e-06,
+      "loss": 0.3941,
+      "step": 1676
+    },
+    {
+      "epoch": 1.3615696887686062,
+      "grad_norm": 0.21983982622623444,
+      "learning_rate": 6.644541557110753e-06,
+      "loss": 0.3445,
+      "step": 1677
+    },
+    {
+      "epoch": 1.362381596752368,
+      "grad_norm": 0.2349981665611267,
+      "learning_rate": 6.640076783888934e-06,
+      "loss": 0.3735,
+      "step": 1678
+    },
+    {
+      "epoch": 1.3631935047361299,
+      "grad_norm": 0.2262328714132309,
+      "learning_rate": 6.63561054477094e-06,
+      "loss": 0.3548,
+      "step": 1679
+    },
+    {
+      "epoch": 1.3640054127198917,
+      "grad_norm": 0.24358214437961578,
+      "learning_rate": 6.631142843748686e-06,
+      "loss": 0.3441,
+      "step": 1680
+    },
+    {
+      "epoch": 1.3648173207036536,
+      "grad_norm": 0.21898531913757324,
+      "learning_rate": 6.626673684815389e-06,
+      "loss": 0.3435,
+      "step": 1681
+    },
+    {
+      "epoch": 1.3656292286874154,
+      "grad_norm": 0.22444714605808258,
+      "learning_rate": 6.62220307196557e-06,
+      "loss": 0.3522,
+      "step": 1682
+    },
+    {
+      "epoch": 1.3664411366711773,
+      "grad_norm": 0.2602396607398987,
+      "learning_rate": 6.617731009195052e-06,
+      "loss": 0.3763,
+      "step": 1683
+    },
+    {
+      "epoch": 1.3672530446549391,
+      "grad_norm": 0.21283552050590515,
+      "learning_rate": 6.613257500500952e-06,
+      "loss": 0.3287,
+      "step": 1684
+    },
+    {
+      "epoch": 1.368064952638701,
+      "grad_norm": 0.2262137085199356,
+      "learning_rate": 6.60878254988168e-06,
+      "loss": 0.3824,
+      "step": 1685
+    },
+    {
+      "epoch": 1.3688768606224628,
+      "grad_norm": 0.23716458678245544,
+      "learning_rate": 6.6043061613369356e-06,
+      "loss": 0.3689,
+      "step": 1686
+    },
+    {
+      "epoch": 1.3696887686062247,
+      "grad_norm": 0.22525520622730255,
+      "learning_rate": 6.5998283388677005e-06,
+      "loss": 0.3431,
+      "step": 1687
+    },
+    {
+      "epoch": 1.3705006765899865,
+      "grad_norm": 0.22924190759658813,
+      "learning_rate": 6.595349086476244e-06,
+      "loss": 0.3755,
+      "step": 1688
+    },
+    {
+      "epoch": 1.3713125845737484,
+      "grad_norm": 0.21128986775875092,
+      "learning_rate": 6.590868408166105e-06,
+      "loss": 0.3433,
+      "step": 1689
+    },
+    {
+      "epoch": 1.3721244925575102,
+      "grad_norm": 0.2427276372909546,
+      "learning_rate": 6.586386307942105e-06,
+      "loss": 0.3671,
+      "step": 1690
+    },
+    {
+      "epoch": 1.3729364005412719,
+      "grad_norm": 0.22183404862880707,
+      "learning_rate": 6.581902789810333e-06,
+      "loss": 0.3323,
+      "step": 1691
+    },
+    {
+      "epoch": 1.373748308525034,
+      "grad_norm": 0.24643155932426453,
+      "learning_rate": 6.577417857778142e-06,
+      "loss": 0.3951,
+      "step": 1692
+    },
+    {
+      "epoch": 1.3745602165087956,
+      "grad_norm": 0.2246680110692978,
+      "learning_rate": 6.572931515854158e-06,
+      "loss": 0.3334,
+      "step": 1693
+    },
+    {
+      "epoch": 1.3753721244925576,
+      "grad_norm": 0.23663026094436646,
+      "learning_rate": 6.568443768048254e-06,
+      "loss": 0.3485,
+      "step": 1694
+    },
+    {
+      "epoch": 1.3761840324763193,
+      "grad_norm": 0.2429826259613037,
+      "learning_rate": 6.563954618371573e-06,
+      "loss": 0.3744,
+      "step": 1695
+    },
+    {
+      "epoch": 1.3769959404600811,
+      "grad_norm": 0.25151705741882324,
+      "learning_rate": 6.559464070836501e-06,
+      "loss": 0.3877,
+      "step": 1696
+    },
+    {
+      "epoch": 1.377807848443843,
+      "grad_norm": 0.24555322527885437,
+      "learning_rate": 6.5549721294566794e-06,
+      "loss": 0.3248,
+      "step": 1697
+    },
+    {
+      "epoch": 1.3786197564276048,
+      "grad_norm": 0.28208625316619873,
+      "learning_rate": 6.550478798246991e-06,
+      "loss": 0.3935,
+      "step": 1698
+    },
+    {
+      "epoch": 1.3794316644113667,
+      "grad_norm": 0.22733859717845917,
+      "learning_rate": 6.545984081223564e-06,
+      "loss": 0.329,
+      "step": 1699
+    },
+    {
+      "epoch": 1.3802435723951285,
+      "grad_norm": 0.2508426308631897,
+      "learning_rate": 6.541487982403767e-06,
+      "loss": 0.363,
+      "step": 1700
+    },
+    {
+      "epoch": 1.3810554803788904,
+      "grad_norm": 0.23861393332481384,
+      "learning_rate": 6.536990505806197e-06,
+      "loss": 0.329,
+      "step": 1701
+    },
+    {
+      "epoch": 1.3818673883626522,
+      "grad_norm": 0.22381877899169922,
+      "learning_rate": 6.532491655450688e-06,
+      "loss": 0.3901,
+      "step": 1702
+    },
+    {
+      "epoch": 1.382679296346414,
+      "grad_norm": 0.21469658613204956,
+      "learning_rate": 6.5279914353582996e-06,
+      "loss": 0.3245,
+      "step": 1703
+    },
+    {
+      "epoch": 1.383491204330176,
+      "grad_norm": 0.24692672491073608,
+      "learning_rate": 6.523489849551316e-06,
+      "loss": 0.3812,
+      "step": 1704
+    },
+    {
+      "epoch": 1.3843031123139378,
+      "grad_norm": 0.22466197609901428,
+      "learning_rate": 6.518986902053244e-06,
+      "loss": 0.3662,
+      "step": 1705
+    },
+    {
+      "epoch": 1.3851150202976996,
+      "grad_norm": 0.23736517131328583,
+      "learning_rate": 6.514482596888807e-06,
+      "loss": 0.3667,
+      "step": 1706
+    },
+    {
+      "epoch": 1.3859269282814615,
+      "grad_norm": 0.23065081238746643,
+      "learning_rate": 6.509976938083936e-06,
+      "loss": 0.3539,
+      "step": 1707
+    },
+    {
+      "epoch": 1.3867388362652233,
+      "grad_norm": 0.2519584000110626,
+      "learning_rate": 6.505469929665781e-06,
+      "loss": 0.355,
+      "step": 1708
+    },
+    {
+      "epoch": 1.3875507442489852,
+      "grad_norm": 0.28415682911872864,
+      "learning_rate": 6.500961575662691e-06,
+      "loss": 0.3901,
+      "step": 1709
+    },
+    {
+      "epoch": 1.388362652232747,
+      "grad_norm": 0.21060457825660706,
+      "learning_rate": 6.496451880104222e-06,
+      "loss": 0.3259,
+      "step": 1710
+    },
+    {
+      "epoch": 1.3891745602165089,
+      "grad_norm": 0.2778107225894928,
+      "learning_rate": 6.491940847021128e-06,
+      "loss": 0.4246,
+      "step": 1711
+    },
+    {
+      "epoch": 1.3899864682002705,
+      "grad_norm": 0.23779264092445374,
+      "learning_rate": 6.487428480445357e-06,
+      "loss": 0.3386,
+      "step": 1712
+    },
+    {
+      "epoch": 1.3907983761840326,
+      "grad_norm": 0.2581709921360016,
+      "learning_rate": 6.48291478441005e-06,
+      "loss": 0.3817,
+      "step": 1713
+    },
+    {
+      "epoch": 1.3916102841677942,
+      "grad_norm": 0.2421477884054184,
+      "learning_rate": 6.478399762949537e-06,
+      "loss": 0.3497,
+      "step": 1714
+    },
+    {
+      "epoch": 1.3924221921515563,
+      "grad_norm": 0.22824953496456146,
+      "learning_rate": 6.47388342009933e-06,
+      "loss": 0.3789,
+      "step": 1715
+    },
+    {
+      "epoch": 1.393234100135318,
+      "grad_norm": 0.23937761783599854,
+      "learning_rate": 6.469365759896128e-06,
+      "loss": 0.3509,
+      "step": 1716
+    },
+    {
+      "epoch": 1.39404600811908,
+      "grad_norm": 0.2388204038143158,
+      "learning_rate": 6.464846786377801e-06,
+      "loss": 0.3519,
+      "step": 1717
+    },
+    {
+      "epoch": 1.3948579161028416,
+      "grad_norm": 0.28960391879081726,
+      "learning_rate": 6.460326503583394e-06,
+      "loss": 0.3851,
+      "step": 1718
+    },
+    {
+      "epoch": 1.3956698240866034,
+      "grad_norm": 0.20418748259544373,
+      "learning_rate": 6.4558049155531255e-06,
+      "loss": 0.3172,
+      "step": 1719
+    },
+    {
+      "epoch": 1.3964817320703653,
+      "grad_norm": 0.22102616727352142,
+      "learning_rate": 6.451282026328377e-06,
+      "loss": 0.3303,
+      "step": 1720
+    },
+    {
+      "epoch": 1.3972936400541272,
+      "grad_norm": 0.23094037175178528,
+      "learning_rate": 6.4467578399516965e-06,
+      "loss": 0.3513,
+      "step": 1721
+    },
+    {
+      "epoch": 1.398105548037889,
+      "grad_norm": 0.2748175859451294,
+      "learning_rate": 6.442232360466789e-06,
+      "loss": 0.3747,
+      "step": 1722
+    },
+    {
+      "epoch": 1.3989174560216509,
+      "grad_norm": 0.23592063784599304,
+      "learning_rate": 6.437705591918514e-06,
+      "loss": 0.3667,
+      "step": 1723
+    },
+    {
+      "epoch": 1.3997293640054127,
+      "grad_norm": 0.24202480912208557,
+      "learning_rate": 6.433177538352887e-06,
+      "loss": 0.3855,
+      "step": 1724
+    },
+    {
+      "epoch": 1.4005412719891746,
+      "grad_norm": 0.2489633560180664,
+      "learning_rate": 6.428648203817069e-06,
+      "loss": 0.3901,
+      "step": 1725
+    },
+    {
+      "epoch": 1.4013531799729364,
+      "grad_norm": 0.24372032284736633,
+      "learning_rate": 6.424117592359367e-06,
+      "loss": 0.3288,
+      "step": 1726
+    },
+    {
+      "epoch": 1.4021650879566983,
+      "grad_norm": 0.26702484488487244,
+      "learning_rate": 6.419585708029229e-06,
+      "loss": 0.3429,
+      "step": 1727
+    },
+    {
+      "epoch": 1.40297699594046,
+      "grad_norm": 0.25901302695274353,
+      "learning_rate": 6.415052554877241e-06,
+      "loss": 0.3527,
+      "step": 1728
+    },
+    {
+      "epoch": 1.403788903924222,
+      "grad_norm": 0.22453421354293823,
+      "learning_rate": 6.410518136955124e-06,
+      "loss": 0.3369,
+      "step": 1729
+    },
+    {
+      "epoch": 1.4046008119079838,
+      "grad_norm": 0.2638757526874542,
+      "learning_rate": 6.405982458315727e-06,
+      "loss": 0.372,
+      "step": 1730
+    },
+    {
+      "epoch": 1.4054127198917457,
+      "grad_norm": 0.26225829124450684,
+      "learning_rate": 6.401445523013028e-06,
+      "loss": 0.3445,
+      "step": 1731
+    },
+    {
+      "epoch": 1.4062246278755075,
+      "grad_norm": 0.2849729359149933,
+      "learning_rate": 6.396907335102127e-06,
+      "loss": 0.401,
+      "step": 1732
+    },
+    {
+      "epoch": 1.4070365358592694,
+      "grad_norm": 0.24043384194374084,
+      "learning_rate": 6.392367898639245e-06,
+      "loss": 0.3539,
+      "step": 1733
+    },
+    {
+      "epoch": 1.4078484438430312,
+      "grad_norm": 0.2139657735824585,
+      "learning_rate": 6.387827217681717e-06,
+      "loss": 0.3142,
+      "step": 1734
+    },
+    {
+      "epoch": 1.4086603518267928,
+      "grad_norm": 0.2653227746486664,
+      "learning_rate": 6.383285296287991e-06,
+      "loss": 0.3794,
+      "step": 1735
+    },
+    {
+      "epoch": 1.409472259810555,
+      "grad_norm": 0.2645275890827179,
+      "learning_rate": 6.378742138517624e-06,
+      "loss": 0.3772,
+      "step": 1736
+    },
+    {
+      "epoch": 1.4102841677943165,
+      "grad_norm": 0.22779908776283264,
+      "learning_rate": 6.37419774843128e-06,
+      "loss": 0.3532,
+      "step": 1737
+    },
+    {
+      "epoch": 1.4110960757780786,
+      "grad_norm": 0.24808387458324432,
+      "learning_rate": 6.3696521300907214e-06,
+      "loss": 0.3655,
+      "step": 1738
+    },
+    {
+      "epoch": 1.4119079837618402,
+      "grad_norm": 0.2539712190628052,
+      "learning_rate": 6.3651052875588085e-06,
+      "loss": 0.3242,
+      "step": 1739
+    },
+    {
+      "epoch": 1.412719891745602,
+      "grad_norm": 0.3312995731830597,
+      "learning_rate": 6.360557224899496e-06,
+      "loss": 0.4296,
+      "step": 1740
+    },
+    {
+      "epoch": 1.413531799729364,
+      "grad_norm": 0.24024130403995514,
+      "learning_rate": 6.356007946177833e-06,
+      "loss": 0.3487,
+      "step": 1741
+    },
+    {
+      "epoch": 1.4143437077131258,
+      "grad_norm": 0.25608792901039124,
+      "learning_rate": 6.351457455459953e-06,
+      "loss": 0.3677,
+      "step": 1742
+    },
+    {
+      "epoch": 1.4151556156968876,
+      "grad_norm": 0.26171669363975525,
+      "learning_rate": 6.346905756813069e-06,
+      "loss": 0.3598,
+      "step": 1743
+    },
+    {
+      "epoch": 1.4159675236806495,
+      "grad_norm": 0.2326306849718094,
+      "learning_rate": 6.342352854305477e-06,
+      "loss": 0.3433,
+      "step": 1744
+    },
+    {
+      "epoch": 1.4167794316644113,
+      "grad_norm": 0.26165810227394104,
+      "learning_rate": 6.337798752006551e-06,
+      "loss": 0.3788,
+      "step": 1745
+    },
+    {
+      "epoch": 1.4175913396481732,
+      "grad_norm": 0.2557559907436371,
+      "learning_rate": 6.333243453986734e-06,
+      "loss": 0.3622,
+      "step": 1746
+    },
+    {
+      "epoch": 1.418403247631935,
+      "grad_norm": 0.22170065343379974,
+      "learning_rate": 6.32868696431754e-06,
+      "loss": 0.3384,
+      "step": 1747
+    },
+    {
+      "epoch": 1.4192151556156969,
+      "grad_norm": 0.2760625183582306,
+      "learning_rate": 6.324129287071546e-06,
+      "loss": 0.3679,
+      "step": 1748
+    },
+    {
+      "epoch": 1.4200270635994587,
+      "grad_norm": 0.23983579874038696,
+      "learning_rate": 6.3195704263223914e-06,
+      "loss": 0.3636,
+      "step": 1749
+    },
+    {
+      "epoch": 1.4208389715832206,
+      "grad_norm": 0.2324240654706955,
+      "learning_rate": 6.315010386144776e-06,
+      "loss": 0.3595,
+      "step": 1750
+    },
+    {
+      "epoch": 1.4216508795669824,
+      "grad_norm": 0.23462934792041779,
+      "learning_rate": 6.310449170614446e-06,
+      "loss": 0.34,
+      "step": 1751
+    },
+    {
+      "epoch": 1.4224627875507443,
+      "grad_norm": 0.2503843307495117,
+      "learning_rate": 6.305886783808209e-06,
+      "loss": 0.3727,
+      "step": 1752
+    },
+    {
+      "epoch": 1.4232746955345061,
+      "grad_norm": 0.22547759115695953,
+      "learning_rate": 6.301323229803913e-06,
+      "loss": 0.3236,
+      "step": 1753
+    },
+    {
+      "epoch": 1.424086603518268,
+      "grad_norm": 0.24910621345043182,
+      "learning_rate": 6.296758512680448e-06,
+      "loss": 0.3843,
+      "step": 1754
+    },
+    {
+      "epoch": 1.4248985115020298,
+      "grad_norm": 0.23730558156967163,
+      "learning_rate": 6.292192636517744e-06,
+      "loss": 0.3559,
+      "step": 1755
+    },
+    {
+      "epoch": 1.4257104194857917,
+      "grad_norm": 0.2331499606370926,
+      "learning_rate": 6.287625605396774e-06,
+      "loss": 0.351,
+      "step": 1756
+    },
+    {
+      "epoch": 1.4265223274695535,
+      "grad_norm": 0.23323003947734833,
+      "learning_rate": 6.2830574233995356e-06,
+      "loss": 0.3806,
+      "step": 1757
+    },
+    {
+      "epoch": 1.4273342354533152,
+      "grad_norm": 0.21901674568653107,
+      "learning_rate": 6.278488094609057e-06,
+      "loss": 0.3325,
+      "step": 1758
+    },
+    {
+      "epoch": 1.4281461434370772,
+      "grad_norm": 0.20840182900428772,
+      "learning_rate": 6.273917623109392e-06,
+      "loss": 0.3382,
+      "step": 1759
+    },
+    {
+      "epoch": 1.4289580514208389,
+      "grad_norm": 0.27469485998153687,
+      "learning_rate": 6.269346012985617e-06,
+      "loss": 0.3732,
+      "step": 1760
+    },
+    {
+      "epoch": 1.429769959404601,
+      "grad_norm": 0.2309303879737854,
+      "learning_rate": 6.264773268323821e-06,
+      "loss": 0.37,
+      "step": 1761
+    },
+    {
+      "epoch": 1.4305818673883626,
+      "grad_norm": 0.23009029030799866,
+      "learning_rate": 6.260199393211115e-06,
+      "loss": 0.3683,
+      "step": 1762
+    },
+    {
+      "epoch": 1.4313937753721244,
+      "grad_norm": 0.24633707106113434,
+      "learning_rate": 6.255624391735614e-06,
+      "loss": 0.3719,
+      "step": 1763
+    },
+    {
+      "epoch": 1.4322056833558863,
+      "grad_norm": 0.27767467498779297,
+      "learning_rate": 6.2510482679864425e-06,
+      "loss": 0.3738,
+      "step": 1764
+    },
+    {
+      "epoch": 1.4330175913396481,
+      "grad_norm": 0.21657705307006836,
+      "learning_rate": 6.246471026053727e-06,
+      "loss": 0.3546,
+      "step": 1765
+    },
+    {
+      "epoch": 1.43382949932341,
+      "grad_norm": 0.2344108521938324,
+      "learning_rate": 6.241892670028595e-06,
+      "loss": 0.3655,
+      "step": 1766
+    },
+    {
+      "epoch": 1.4346414073071718,
+      "grad_norm": 0.22900553047657013,
+      "learning_rate": 6.237313204003167e-06,
+      "loss": 0.3627,
+      "step": 1767
+    },
+    {
+      "epoch": 1.4354533152909337,
+      "grad_norm": 0.26144254207611084,
+      "learning_rate": 6.23273263207056e-06,
+      "loss": 0.4231,
+      "step": 1768
+    },
+    {
+      "epoch": 1.4362652232746955,
+      "grad_norm": 0.2172008603811264,
+      "learning_rate": 6.228150958324872e-06,
+      "loss": 0.3297,
+      "step": 1769
+    },
+    {
+      "epoch": 1.4370771312584574,
+      "grad_norm": 0.21712778508663177,
+      "learning_rate": 6.223568186861195e-06,
+      "loss": 0.3382,
+      "step": 1770
+    },
+    {
+      "epoch": 1.4378890392422192,
+      "grad_norm": 0.240541473031044,
+      "learning_rate": 6.218984321775596e-06,
+      "loss": 0.3891,
+      "step": 1771
+    },
+    {
+      "epoch": 1.438700947225981,
+      "grad_norm": 0.23823702335357666,
+      "learning_rate": 6.214399367165122e-06,
+      "loss": 0.354,
+      "step": 1772
+    },
+    {
+      "epoch": 1.439512855209743,
+      "grad_norm": 0.2326749563217163,
+      "learning_rate": 6.209813327127792e-06,
+      "loss": 0.3411,
+      "step": 1773
+    },
+    {
+      "epoch": 1.4403247631935048,
+      "grad_norm": 0.23649263381958008,
+      "learning_rate": 6.205226205762595e-06,
+      "loss": 0.3946,
+      "step": 1774
+    },
+    {
+      "epoch": 1.4411366711772666,
+      "grad_norm": 0.21656061708927155,
+      "learning_rate": 6.20063800716949e-06,
+      "loss": 0.3476,
+      "step": 1775
+    },
+    {
+      "epoch": 1.4419485791610285,
+      "grad_norm": 0.22107309103012085,
+      "learning_rate": 6.196048735449396e-06,
+      "loss": 0.354,
+      "step": 1776
+    },
+    {
+      "epoch": 1.4427604871447903,
+      "grad_norm": 0.22435951232910156,
+      "learning_rate": 6.1914583947041906e-06,
+      "loss": 0.38,
+      "step": 1777
+    },
+    {
+      "epoch": 1.4435723951285522,
+      "grad_norm": 0.22486943006515503,
+      "learning_rate": 6.186866989036708e-06,
+      "loss": 0.3371,
+      "step": 1778
+    },
+    {
+      "epoch": 1.4443843031123138,
+      "grad_norm": 0.23842036724090576,
+      "learning_rate": 6.1822745225507355e-06,
+      "loss": 0.3612,
+      "step": 1779
+    },
+    {
+      "epoch": 1.4451962110960759,
+      "grad_norm": 0.22367502748966217,
+      "learning_rate": 6.177680999351006e-06,
+      "loss": 0.3439,
+      "step": 1780
+    },
+    {
+      "epoch": 1.4460081190798375,
+      "grad_norm": 0.24618986248970032,
+      "learning_rate": 6.173086423543198e-06,
+      "loss": 0.3898,
+      "step": 1781
+    },
+    {
+      "epoch": 1.4468200270635996,
+      "grad_norm": 0.2130477875471115,
+      "learning_rate": 6.168490799233931e-06,
+      "loss": 0.3256,
+      "step": 1782
+    },
+    {
+      "epoch": 1.4476319350473612,
+      "grad_norm": 0.22934691607952118,
+      "learning_rate": 6.163894130530761e-06,
+      "loss": 0.3348,
+      "step": 1783
+    },
+    {
+      "epoch": 1.448443843031123,
+      "grad_norm": 0.236383855342865,
+      "learning_rate": 6.159296421542178e-06,
+      "loss": 0.3714,
+      "step": 1784
+    },
+    {
+      "epoch": 1.449255751014885,
+      "grad_norm": 0.2144795060157776,
+      "learning_rate": 6.154697676377605e-06,
+      "loss": 0.3487,
+      "step": 1785
+    },
+    {
+      "epoch": 1.4500676589986468,
+      "grad_norm": 0.2192850112915039,
+      "learning_rate": 6.150097899147384e-06,
+      "loss": 0.342,
+      "step": 1786
+    },
+    {
+      "epoch": 1.4508795669824086,
+      "grad_norm": 0.24613067507743835,
+      "learning_rate": 6.1454970939627845e-06,
+      "loss": 0.3416,
+      "step": 1787
+    },
+    {
+      "epoch": 1.4516914749661705,
+      "grad_norm": 0.23278966546058655,
+      "learning_rate": 6.140895264935993e-06,
+      "loss": 0.3692,
+      "step": 1788
+    },
+    {
+      "epoch": 1.4525033829499323,
+      "grad_norm": 0.23425890505313873,
+      "learning_rate": 6.136292416180114e-06,
+      "loss": 0.3723,
+      "step": 1789
+    },
+    {
+      "epoch": 1.4533152909336942,
+      "grad_norm": 0.23012934625148773,
+      "learning_rate": 6.131688551809161e-06,
+      "loss": 0.3696,
+      "step": 1790
+    },
+    {
+      "epoch": 1.454127198917456,
+      "grad_norm": 0.26892438530921936,
+      "learning_rate": 6.127083675938053e-06,
+      "loss": 0.3626,
+      "step": 1791
+    },
+    {
+      "epoch": 1.4549391069012179,
+      "grad_norm": 0.24430692195892334,
+      "learning_rate": 6.122477792682616e-06,
+      "loss": 0.3628,
+      "step": 1792
+    },
+    {
+      "epoch": 1.4557510148849797,
+      "grad_norm": 0.23383599519729614,
+      "learning_rate": 6.117870906159577e-06,
+      "loss": 0.3447,
+      "step": 1793
+    },
+    {
+      "epoch": 1.4565629228687416,
+      "grad_norm": 0.24481017887592316,
+      "learning_rate": 6.113263020486559e-06,
+      "loss": 0.3443,
+      "step": 1794
+    },
+    {
+      "epoch": 1.4573748308525034,
+      "grad_norm": 0.25695857405662537,
+      "learning_rate": 6.1086541397820785e-06,
+      "loss": 0.384,
+      "step": 1795
+    },
+    {
+      "epoch": 1.4581867388362653,
+      "grad_norm": 0.25220805406570435,
+      "learning_rate": 6.104044268165539e-06,
+      "loss": 0.3606,
+      "step": 1796
+    },
+    {
+      "epoch": 1.4589986468200271,
+      "grad_norm": 0.220266655087471,
+      "learning_rate": 6.0994334097572305e-06,
+      "loss": 0.382,
+      "step": 1797
+    },
+    {
+      "epoch": 1.459810554803789,
+      "grad_norm": 0.2436649054288864,
+      "learning_rate": 6.094821568678328e-06,
+      "loss": 0.3563,
+      "step": 1798
+    },
+    {
+      "epoch": 1.4606224627875508,
+      "grad_norm": 0.24222387373447418,
+      "learning_rate": 6.090208749050881e-06,
+      "loss": 0.3443,
+      "step": 1799
+    },
+    {
+      "epoch": 1.4614343707713127,
+      "grad_norm": 0.24605031311511993,
+      "learning_rate": 6.085594954997814e-06,
+      "loss": 0.3837,
+      "step": 1800
+    },
+    {
+      "epoch": 1.4622462787550745,
+      "grad_norm": 0.23830477893352509,
+      "learning_rate": 6.080980190642927e-06,
+      "loss": 0.3439,
+      "step": 1801
+    },
+    {
+      "epoch": 1.4630581867388361,
+      "grad_norm": 0.24046534299850464,
+      "learning_rate": 6.076364460110881e-06,
+      "loss": 0.3754,
+      "step": 1802
+    },
+    {
+      "epoch": 1.4638700947225982,
+      "grad_norm": 0.2770979106426239,
+      "learning_rate": 6.071747767527204e-06,
+      "loss": 0.3998,
+      "step": 1803
+    },
+    {
+      "epoch": 1.4646820027063598,
+      "grad_norm": 0.20743681490421295,
+      "learning_rate": 6.067130117018284e-06,
+      "loss": 0.3108,
+      "step": 1804
+    },
+    {
+      "epoch": 1.465493910690122,
+      "grad_norm": 0.24249324202537537,
+      "learning_rate": 6.062511512711362e-06,
+      "loss": 0.3578,
+      "step": 1805
+    },
+    {
+      "epoch": 1.4663058186738835,
+      "grad_norm": 0.2467086911201477,
+      "learning_rate": 6.057891958734538e-06,
+      "loss": 0.37,
+      "step": 1806
+    },
+    {
+      "epoch": 1.4671177266576454,
+      "grad_norm": 0.21984650194644928,
+      "learning_rate": 6.053271459216755e-06,
+      "loss": 0.3466,
+      "step": 1807
+    },
+    {
+      "epoch": 1.4679296346414072,
+      "grad_norm": 0.2430969625711441,
+      "learning_rate": 6.048650018287803e-06,
+      "loss": 0.3667,
+      "step": 1808
+    },
+    {
+      "epoch": 1.468741542625169,
+      "grad_norm": 0.2241179496049881,
+      "learning_rate": 6.044027640078312e-06,
+      "loss": 0.3592,
+      "step": 1809
+    },
+    {
+      "epoch": 1.469553450608931,
+      "grad_norm": 0.22605526447296143,
+      "learning_rate": 6.0394043287197524e-06,
+      "loss": 0.3375,
+      "step": 1810
+    },
+    {
+      "epoch": 1.4703653585926928,
+      "grad_norm": 0.21936161816120148,
+      "learning_rate": 6.034780088344427e-06,
+      "loss": 0.3426,
+      "step": 1811
+    },
+    {
+      "epoch": 1.4711772665764546,
+      "grad_norm": 0.23898614943027496,
+      "learning_rate": 6.030154923085469e-06,
+      "loss": 0.3779,
+      "step": 1812
+    },
+    {
+      "epoch": 1.4719891745602165,
+      "grad_norm": 0.22187158465385437,
+      "learning_rate": 6.025528837076841e-06,
+      "loss": 0.3359,
+      "step": 1813
+    },
+    {
+      "epoch": 1.4728010825439783,
+      "grad_norm": 0.2343793660402298,
+      "learning_rate": 6.020901834453322e-06,
+      "loss": 0.3283,
+      "step": 1814
+    },
+    {
+      "epoch": 1.4736129905277402,
+      "grad_norm": 0.22249604761600494,
+      "learning_rate": 6.016273919350517e-06,
+      "loss": 0.3786,
+      "step": 1815
+    },
+    {
+      "epoch": 1.474424898511502,
+      "grad_norm": 0.22781893610954285,
+      "learning_rate": 6.011645095904845e-06,
+      "loss": 0.3369,
+      "step": 1816
+    },
+    {
+      "epoch": 1.475236806495264,
+      "grad_norm": 0.2724662721157074,
+      "learning_rate": 6.0070153682535315e-06,
+      "loss": 0.3681,
+      "step": 1817
+    },
+    {
+      "epoch": 1.4760487144790257,
+      "grad_norm": 0.22135840356349945,
+      "learning_rate": 6.002384740534619e-06,
+      "loss": 0.3539,
+      "step": 1818
+    },
+    {
+      "epoch": 1.4768606224627876,
+      "grad_norm": 0.22987376153469086,
+      "learning_rate": 5.997753216886948e-06,
+      "loss": 0.3372,
+      "step": 1819
+    },
+    {
+      "epoch": 1.4776725304465494,
+      "grad_norm": 0.23796653747558594,
+      "learning_rate": 5.993120801450162e-06,
+      "loss": 0.3475,
+      "step": 1820
+    },
+    {
+      "epoch": 1.4784844384303113,
+      "grad_norm": 0.22833077609539032,
+      "learning_rate": 5.9884874983647e-06,
+      "loss": 0.3678,
+      "step": 1821
+    },
+    {
+      "epoch": 1.4792963464140731,
+      "grad_norm": 0.24801380932331085,
+      "learning_rate": 5.983853311771799e-06,
+      "loss": 0.3719,
+      "step": 1822
+    },
+    {
+      "epoch": 1.4801082543978348,
+      "grad_norm": 0.23287218809127808,
+      "learning_rate": 5.979218245813478e-06,
+      "loss": 0.3661,
+      "step": 1823
+    },
+    {
+      "epoch": 1.4809201623815968,
+      "grad_norm": 0.19757086038589478,
+      "learning_rate": 5.974582304632551e-06,
+      "loss": 0.352,
+      "step": 1824
+    },
+    {
+      "epoch": 1.4817320703653585,
+      "grad_norm": 0.2344127595424652,
+      "learning_rate": 5.969945492372606e-06,
+      "loss": 0.3655,
+      "step": 1825
+    },
+    {
+      "epoch": 1.4825439783491206,
+      "grad_norm": 0.26045721769332886,
+      "learning_rate": 5.965307813178015e-06,
+      "loss": 0.378,
+      "step": 1826
+    },
+    {
+      "epoch": 1.4833558863328822,
+      "grad_norm": 0.2116851508617401,
+      "learning_rate": 5.960669271193922e-06,
+      "loss": 0.3641,
+      "step": 1827
+    },
+    {
+      "epoch": 1.4841677943166443,
+      "grad_norm": 0.20913740992546082,
+      "learning_rate": 5.9560298705662444e-06,
+      "loss": 0.339,
+      "step": 1828
+    },
+    {
+      "epoch": 1.4849797023004059,
+      "grad_norm": 0.22162680327892303,
+      "learning_rate": 5.951389615441666e-06,
+      "loss": 0.376,
+      "step": 1829
+    },
+    {
+      "epoch": 1.4857916102841677,
+      "grad_norm": 0.2362152487039566,
+      "learning_rate": 5.9467485099676325e-06,
+      "loss": 0.3722,
+      "step": 1830
+    },
+    {
+      "epoch": 1.4866035182679296,
+      "grad_norm": 0.19379474222660065,
+      "learning_rate": 5.942106558292352e-06,
+      "loss": 0.3197,
+      "step": 1831
+    },
+    {
+      "epoch": 1.4874154262516914,
+      "grad_norm": 0.23799993097782135,
+      "learning_rate": 5.93746376456479e-06,
+      "loss": 0.3862,
+      "step": 1832
+    },
+    {
+      "epoch": 1.4882273342354533,
+      "grad_norm": 0.23614934086799622,
+      "learning_rate": 5.932820132934661e-06,
+      "loss": 0.3833,
+      "step": 1833
+    },
+    {
+      "epoch": 1.4890392422192151,
+      "grad_norm": 0.2417062669992447,
+      "learning_rate": 5.928175667552432e-06,
+      "loss": 0.3615,
+      "step": 1834
+    },
+    {
+      "epoch": 1.489851150202977,
+      "grad_norm": 0.23723454773426056,
+      "learning_rate": 5.923530372569312e-06,
+      "loss": 0.3519,
+      "step": 1835
+    },
+    {
+      "epoch": 1.4906630581867388,
+      "grad_norm": 0.23173874616622925,
+      "learning_rate": 5.9188842521372535e-06,
+      "loss": 0.3705,
+      "step": 1836
+    },
+    {
+      "epoch": 1.4914749661705007,
+      "grad_norm": 0.23151184618473053,
+      "learning_rate": 5.914237310408947e-06,
+      "loss": 0.3479,
+      "step": 1837
+    },
+    {
+      "epoch": 1.4922868741542625,
+      "grad_norm": 0.261500746011734,
+      "learning_rate": 5.909589551537814e-06,
+      "loss": 0.3541,
+      "step": 1838
+    },
+    {
+      "epoch": 1.4930987821380244,
+      "grad_norm": 0.23359160125255585,
+      "learning_rate": 5.904940979678012e-06,
+      "loss": 0.3399,
+      "step": 1839
+    },
+    {
+      "epoch": 1.4939106901217862,
+      "grad_norm": 0.2311207503080368,
+      "learning_rate": 5.90029159898442e-06,
+      "loss": 0.3738,
+      "step": 1840
+    },
+    {
+      "epoch": 1.494722598105548,
+      "grad_norm": 0.24646176397800446,
+      "learning_rate": 5.895641413612643e-06,
+      "loss": 0.3558,
+      "step": 1841
+    },
+    {
+      "epoch": 1.49553450608931,
+      "grad_norm": 0.25415724515914917,
+      "learning_rate": 5.890990427719003e-06,
+      "loss": 0.3486,
+      "step": 1842
+    },
+    {
+      "epoch": 1.4963464140730718,
+      "grad_norm": 0.24593311548233032,
+      "learning_rate": 5.886338645460539e-06,
+      "loss": 0.3417,
+      "step": 1843
+    },
+    {
+      "epoch": 1.4971583220568336,
+      "grad_norm": 0.26894286274909973,
+      "learning_rate": 5.881686070995001e-06,
+      "loss": 0.3984,
+      "step": 1844
+    },
+    {
+      "epoch": 1.4979702300405955,
+      "grad_norm": 0.25415679812431335,
+      "learning_rate": 5.877032708480847e-06,
+      "loss": 0.3666,
+      "step": 1845
+    },
+    {
+      "epoch": 1.4987821380243571,
+      "grad_norm": 0.23997807502746582,
+      "learning_rate": 5.872378562077241e-06,
+      "loss": 0.3458,
+      "step": 1846
+    },
+    {
+      "epoch": 1.4995940460081192,
+      "grad_norm": 0.24879159033298492,
+      "learning_rate": 5.867723635944047e-06,
+      "loss": 0.3694,
+      "step": 1847
+    },
+    {
+      "epoch": 1.5004059539918808,
+      "grad_norm": 0.22614943981170654,
+      "learning_rate": 5.863067934241823e-06,
+      "loss": 0.3297,
+      "step": 1848
+    },
+    {
+      "epoch": 1.5012178619756429,
+      "grad_norm": 0.24048499763011932,
+      "learning_rate": 5.8584114611318225e-06,
+      "loss": 0.373,
+      "step": 1849
+    },
+    {
+      "epoch": 1.5020297699594045,
+      "grad_norm": 0.2668588161468506,
+      "learning_rate": 5.853754220775992e-06,
+      "loss": 0.3909,
+      "step": 1850
+    },
+    {
+      "epoch": 1.5028416779431666,
+      "grad_norm": 0.2006106823682785,
+      "learning_rate": 5.849096217336956e-06,
+      "loss": 0.3195,
+      "step": 1851
+    },
+    {
+      "epoch": 1.5036535859269282,
+      "grad_norm": 0.24283482134342194,
+      "learning_rate": 5.844437454978029e-06,
+      "loss": 0.3978,
+      "step": 1852
+    },
+    {
+      "epoch": 1.5044654939106903,
+      "grad_norm": 0.22241604328155518,
+      "learning_rate": 5.8397779378631955e-06,
+      "loss": 0.3406,
+      "step": 1853
+    },
+    {
+      "epoch": 1.505277401894452,
+      "grad_norm": 0.207333043217659,
+      "learning_rate": 5.835117670157126e-06,
+      "loss": 0.346,
+      "step": 1854
+    },
+    {
+      "epoch": 1.5060893098782138,
+      "grad_norm": 0.24109458923339844,
+      "learning_rate": 5.83045665602515e-06,
+      "loss": 0.3625,
+      "step": 1855
+    },
+    {
+      "epoch": 1.5069012178619756,
+      "grad_norm": 0.2081637680530548,
+      "learning_rate": 5.82579489963327e-06,
+      "loss": 0.3725,
+      "step": 1856
+    },
+    {
+      "epoch": 1.5077131258457375,
+      "grad_norm": 0.24687153100967407,
+      "learning_rate": 5.8211324051481534e-06,
+      "loss": 0.4074,
+      "step": 1857
+    },
+    {
+      "epoch": 1.5085250338294993,
+      "grad_norm": 0.23652689158916473,
+      "learning_rate": 5.816469176737122e-06,
+      "loss": 0.3507,
+      "step": 1858
+    },
+    {
+      "epoch": 1.5093369418132612,
+      "grad_norm": 0.2190493792295456,
+      "learning_rate": 5.8118052185681584e-06,
+      "loss": 0.3825,
+      "step": 1859
+    },
+    {
+      "epoch": 1.510148849797023,
+      "grad_norm": 0.23038357496261597,
+      "learning_rate": 5.807140534809896e-06,
+      "loss": 0.376,
+      "step": 1860
+    },
+    {
+      "epoch": 1.5109607577807849,
+      "grad_norm": 0.22267024219036102,
+      "learning_rate": 5.802475129631616e-06,
+      "loss": 0.3427,
+      "step": 1861
+    },
+    {
+      "epoch": 1.5117726657645467,
+      "grad_norm": 0.23588265478610992,
+      "learning_rate": 5.797809007203245e-06,
+      "loss": 0.3618,
+      "step": 1862
+    },
+    {
+      "epoch": 1.5125845737483086,
+      "grad_norm": 0.2542931139469147,
+      "learning_rate": 5.79314217169535e-06,
+      "loss": 0.383,
+      "step": 1863
+    },
+    {
+      "epoch": 1.5133964817320704,
+      "grad_norm": 0.22999854385852814,
+      "learning_rate": 5.788474627279137e-06,
+      "loss": 0.3533,
+      "step": 1864
+    },
+    {
+      "epoch": 1.514208389715832,
+      "grad_norm": 0.22740691900253296,
+      "learning_rate": 5.783806378126443e-06,
+      "loss": 0.338,
+      "step": 1865
+    },
+    {
+      "epoch": 1.5150202976995941,
+      "grad_norm": 0.25408798456192017,
+      "learning_rate": 5.779137428409738e-06,
+      "loss": 0.3343,
+      "step": 1866
+    },
+    {
+      "epoch": 1.5158322056833557,
+      "grad_norm": 0.3036448359489441,
+      "learning_rate": 5.774467782302115e-06,
+      "loss": 0.42,
+      "step": 1867
+    },
+    {
+      "epoch": 1.5166441136671178,
+      "grad_norm": 0.21863307058811188,
+      "learning_rate": 5.7697974439772934e-06,
+      "loss": 0.3298,
+      "step": 1868
+    },
+    {
+      "epoch": 1.5174560216508794,
+      "grad_norm": 0.2676413357257843,
+      "learning_rate": 5.765126417609606e-06,
+      "loss": 0.3897,
+      "step": 1869
+    },
+    {
+      "epoch": 1.5182679296346415,
+      "grad_norm": 0.27003639936447144,
+      "learning_rate": 5.760454707374004e-06,
+      "loss": 0.3476,
+      "step": 1870
+    },
+    {
+      "epoch": 1.5190798376184032,
+      "grad_norm": 0.23248445987701416,
+      "learning_rate": 5.755782317446051e-06,
+      "loss": 0.3459,
+      "step": 1871
+    },
+    {
+      "epoch": 1.5198917456021652,
+      "grad_norm": 0.2276916801929474,
+      "learning_rate": 5.751109252001916e-06,
+      "loss": 0.3429,
+      "step": 1872
+    },
+    {
+      "epoch": 1.5207036535859269,
+      "grad_norm": 0.2247655987739563,
+      "learning_rate": 5.746435515218373e-06,
+      "loss": 0.388,
+      "step": 1873
+    },
+    {
+      "epoch": 1.521515561569689,
+      "grad_norm": 0.20654261112213135,
+      "learning_rate": 5.741761111272792e-06,
+      "loss": 0.3409,
+      "step": 1874
+    },
+    {
+      "epoch": 1.5223274695534506,
+      "grad_norm": 0.21401123702526093,
+      "learning_rate": 5.737086044343145e-06,
+      "loss": 0.3551,
+      "step": 1875
+    },
+    {
+      "epoch": 1.5231393775372124,
+      "grad_norm": 0.23796486854553223,
+      "learning_rate": 5.732410318607993e-06,
+      "loss": 0.3514,
+      "step": 1876
+    },
+    {
+      "epoch": 1.5239512855209743,
+      "grad_norm": 0.2316002994775772,
+      "learning_rate": 5.72773393824649e-06,
+      "loss": 0.3707,
+      "step": 1877
+    },
+    {
+      "epoch": 1.524763193504736,
+      "grad_norm": 0.21801616251468658,
+      "learning_rate": 5.723056907438369e-06,
+      "loss": 0.3614,
+      "step": 1878
+    },
+    {
+      "epoch": 1.525575101488498,
+      "grad_norm": 0.240801140666008,
+      "learning_rate": 5.718379230363948e-06,
+      "loss": 0.3515,
+      "step": 1879
+    },
+    {
+      "epoch": 1.5263870094722598,
+      "grad_norm": 0.242412269115448,
+      "learning_rate": 5.713700911204121e-06,
+      "loss": 0.3396,
+      "step": 1880
+    },
+    {
+      "epoch": 1.5271989174560217,
+      "grad_norm": 0.24924618005752563,
+      "learning_rate": 5.709021954140359e-06,
+      "loss": 0.3881,
+      "step": 1881
+    },
+    {
+      "epoch": 1.5280108254397835,
+      "grad_norm": 0.20775189995765686,
+      "learning_rate": 5.704342363354701e-06,
+      "loss": 0.3135,
+      "step": 1882
+    },
+    {
+      "epoch": 1.5288227334235454,
+      "grad_norm": 0.2666032314300537,
+      "learning_rate": 5.6996621430297514e-06,
+      "loss": 0.4087,
+      "step": 1883
+    },
+    {
+      "epoch": 1.5296346414073072,
+      "grad_norm": 0.21502672135829926,
+      "learning_rate": 5.6949812973486785e-06,
+      "loss": 0.3495,
+      "step": 1884
+    },
+    {
+      "epoch": 1.530446549391069,
+      "grad_norm": 0.2398320734500885,
+      "learning_rate": 5.690299830495211e-06,
+      "loss": 0.3432,
+      "step": 1885
+    },
+    {
+      "epoch": 1.531258457374831,
+      "grad_norm": 0.26820045709609985,
+      "learning_rate": 5.685617746653629e-06,
+      "loss": 0.3347,
+      "step": 1886
+    },
+    {
+      "epoch": 1.5320703653585928,
+      "grad_norm": 0.2375265657901764,
+      "learning_rate": 5.680935050008767e-06,
+      "loss": 0.3582,
+      "step": 1887
+    },
+    {
+      "epoch": 1.5328822733423544,
+      "grad_norm": 0.26426970958709717,
+      "learning_rate": 5.676251744746008e-06,
+      "loss": 0.3592,
+      "step": 1888
+    },
+    {
+      "epoch": 1.5336941813261165,
+      "grad_norm": 0.26119476556777954,
+      "learning_rate": 5.671567835051274e-06,
+      "loss": 0.3322,
+      "step": 1889
+    },
+    {
+      "epoch": 1.534506089309878,
+      "grad_norm": 0.22735695540905,
+      "learning_rate": 5.666883325111035e-06,
+      "loss": 0.3436,
+      "step": 1890
+    },
+    {
+      "epoch": 1.5353179972936402,
+      "grad_norm": 0.24559371173381805,
+      "learning_rate": 5.662198219112291e-06,
+      "loss": 0.3662,
+      "step": 1891
+    },
+    {
+      "epoch": 1.5361299052774018,
+      "grad_norm": 0.23068661987781525,
+      "learning_rate": 5.657512521242579e-06,
+      "loss": 0.3528,
+      "step": 1892
+    },
+    {
+      "epoch": 1.5369418132611639,
+      "grad_norm": 0.2137380987405777,
+      "learning_rate": 5.652826235689961e-06,
+      "loss": 0.3448,
+      "step": 1893
+    },
+    {
+      "epoch": 1.5377537212449255,
+      "grad_norm": 0.24910862743854523,
+      "learning_rate": 5.648139366643026e-06,
+      "loss": 0.4084,
+      "step": 1894
+    },
+    {
+      "epoch": 1.5385656292286876,
+      "grad_norm": 0.22860826551914215,
+      "learning_rate": 5.643451918290885e-06,
+      "loss": 0.3518,
+      "step": 1895
+    },
+    {
+      "epoch": 1.5393775372124492,
+      "grad_norm": 0.2173006236553192,
+      "learning_rate": 5.638763894823169e-06,
+      "loss": 0.354,
+      "step": 1896
+    },
+    {
+      "epoch": 1.5401894451962113,
+      "grad_norm": 0.23718596994876862,
+      "learning_rate": 5.634075300430017e-06,
+      "loss": 0.3489,
+      "step": 1897
+    },
+    {
+      "epoch": 1.5410013531799729,
+      "grad_norm": 0.26992347836494446,
+      "learning_rate": 5.6293861393020855e-06,
+      "loss": 0.3829,
+      "step": 1898
+    },
+    {
+      "epoch": 1.5418132611637347,
+      "grad_norm": 0.2111724466085434,
+      "learning_rate": 5.624696415630532e-06,
+      "loss": 0.3022,
+      "step": 1899
+    },
+    {
+      "epoch": 1.5426251691474966,
+      "grad_norm": 0.25430551171302795,
+      "learning_rate": 5.620006133607017e-06,
+      "loss": 0.3821,
+      "step": 1900
+    },
+    {
+      "epoch": 1.5434370771312584,
+      "grad_norm": 0.23348163068294525,
+      "learning_rate": 5.615315297423707e-06,
+      "loss": 0.3556,
+      "step": 1901
+    },
+    {
+      "epoch": 1.5442489851150203,
+      "grad_norm": 0.21112586557865143,
+      "learning_rate": 5.610623911273254e-06,
+      "loss": 0.3229,
+      "step": 1902
+    },
+    {
+      "epoch": 1.5450608930987821,
+      "grad_norm": 0.22651159763336182,
+      "learning_rate": 5.605931979348807e-06,
+      "loss": 0.3126,
+      "step": 1903
+    },
+    {
+      "epoch": 1.545872801082544,
+      "grad_norm": 0.2410609871149063,
+      "learning_rate": 5.601239505844005e-06,
+      "loss": 0.3709,
+      "step": 1904
+    },
+    {
+      "epoch": 1.5466847090663058,
+      "grad_norm": 0.21337734162807465,
+      "learning_rate": 5.596546494952965e-06,
+      "loss": 0.3385,
+      "step": 1905
+    },
+    {
+      "epoch": 1.5474966170500677,
+      "grad_norm": 0.2314242124557495,
+      "learning_rate": 5.591852950870287e-06,
+      "loss": 0.3627,
+      "step": 1906
+    },
+    {
+      "epoch": 1.5483085250338295,
+      "grad_norm": 0.22739017009735107,
+      "learning_rate": 5.58715887779105e-06,
+      "loss": 0.3602,
+      "step": 1907
+    },
+    {
+      "epoch": 1.5491204330175914,
+      "grad_norm": 0.2496870905160904,
+      "learning_rate": 5.582464279910802e-06,
+      "loss": 0.3662,
+      "step": 1908
+    },
+    {
+      "epoch": 1.549932341001353,
+      "grad_norm": 0.238424614071846,
+      "learning_rate": 5.577769161425563e-06,
+      "loss": 0.3705,
+      "step": 1909
+    },
+    {
+      "epoch": 1.550744248985115,
+      "grad_norm": 0.2425748109817505,
+      "learning_rate": 5.573073526531818e-06,
+      "loss": 0.3452,
+      "step": 1910
+    },
+    {
+      "epoch": 1.5515561569688767,
+      "grad_norm": 0.2436589002609253,
+      "learning_rate": 5.568377379426511e-06,
+      "loss": 0.389,
+      "step": 1911
+    },
+    {
+      "epoch": 1.5523680649526388,
+      "grad_norm": 0.2247631996870041,
+      "learning_rate": 5.563680724307046e-06,
+      "loss": 0.3477,
+      "step": 1912
+    },
+    {
+      "epoch": 1.5531799729364004,
+      "grad_norm": 0.23485568165779114,
+      "learning_rate": 5.558983565371281e-06,
+      "loss": 0.361,
+      "step": 1913
+    },
+    {
+      "epoch": 1.5539918809201625,
+      "grad_norm": 0.22930237650871277,
+      "learning_rate": 5.554285906817524e-06,
+      "loss": 0.3582,
+      "step": 1914
+    },
+    {
+      "epoch": 1.5548037889039241,
+      "grad_norm": 0.23389457166194916,
+      "learning_rate": 5.549587752844529e-06,
+      "loss": 0.3611,
+      "step": 1915
+    },
+    {
+      "epoch": 1.5556156968876862,
+      "grad_norm": 0.22899988293647766,
+      "learning_rate": 5.5448891076514925e-06,
+      "loss": 0.3684,
+      "step": 1916
+    },
+    {
+      "epoch": 1.5564276048714478,
+      "grad_norm": 0.23988580703735352,
+      "learning_rate": 5.540189975438053e-06,
+      "loss": 0.3708,
+      "step": 1917
+    },
+    {
+      "epoch": 1.55723951285521,
+      "grad_norm": 0.2213442325592041,
+      "learning_rate": 5.53549036040428e-06,
+      "loss": 0.3124,
+      "step": 1918
+    },
+    {
+      "epoch": 1.5580514208389715,
+      "grad_norm": 0.23449601233005524,
+      "learning_rate": 5.530790266750675e-06,
+      "loss": 0.3643,
+      "step": 1919
+    },
+    {
+      "epoch": 1.5588633288227334,
+      "grad_norm": 0.2198362648487091,
+      "learning_rate": 5.526089698678174e-06,
+      "loss": 0.3407,
+      "step": 1920
+    },
+    {
+      "epoch": 1.5596752368064952,
+      "grad_norm": 0.2280084490776062,
+      "learning_rate": 5.521388660388129e-06,
+      "loss": 0.3647,
+      "step": 1921
+    },
+    {
+      "epoch": 1.560487144790257,
+      "grad_norm": 0.23780398070812225,
+      "learning_rate": 5.516687156082317e-06,
+      "loss": 0.3875,
+      "step": 1922
+    },
+    {
+      "epoch": 1.561299052774019,
+      "grad_norm": 0.21463750302791595,
+      "learning_rate": 5.511985189962926e-06,
+      "loss": 0.359,
+      "step": 1923
+    },
+    {
+      "epoch": 1.5621109607577808,
+      "grad_norm": 0.2423272281885147,
+      "learning_rate": 5.507282766232565e-06,
+      "loss": 0.349,
+      "step": 1924
+    },
+    {
+      "epoch": 1.5629228687415426,
+      "grad_norm": 0.23174826800823212,
+      "learning_rate": 5.502579889094244e-06,
+      "loss": 0.3423,
+      "step": 1925
+    },
+    {
+      "epoch": 1.5637347767253045,
+      "grad_norm": 0.22575487196445465,
+      "learning_rate": 5.497876562751384e-06,
+      "loss": 0.344,
+      "step": 1926
+    },
+    {
+      "epoch": 1.5645466847090663,
+      "grad_norm": 0.2561476230621338,
+      "learning_rate": 5.493172791407803e-06,
+      "loss": 0.3769,
+      "step": 1927
+    },
+    {
+      "epoch": 1.5653585926928282,
+      "grad_norm": 0.2439117729663849,
+      "learning_rate": 5.488468579267721e-06,
+      "loss": 0.3345,
+      "step": 1928
+    },
+    {
+      "epoch": 1.56617050067659,
+      "grad_norm": 0.22771796584129333,
+      "learning_rate": 5.483763930535752e-06,
+      "loss": 0.3492,
+      "step": 1929
+    },
+    {
+      "epoch": 1.5669824086603519,
+      "grad_norm": 0.2597627341747284,
+      "learning_rate": 5.479058849416894e-06,
+      "loss": 0.3677,
+      "step": 1930
+    },
+    {
+      "epoch": 1.5677943166441137,
+      "grad_norm": 0.2201060950756073,
+      "learning_rate": 5.474353340116538e-06,
+      "loss": 0.388,
+      "step": 1931
+    },
+    {
+      "epoch": 1.5686062246278754,
+      "grad_norm": 0.2273837774991989,
+      "learning_rate": 5.469647406840456e-06,
+      "loss": 0.3649,
+      "step": 1932
+    },
+    {
+      "epoch": 1.5694181326116374,
+      "grad_norm": 0.21880699694156647,
+      "learning_rate": 5.464941053794797e-06,
+      "loss": 0.3414,
+      "step": 1933
+    },
+    {
+      "epoch": 1.570230040595399,
+      "grad_norm": 0.22775837779045105,
+      "learning_rate": 5.460234285186087e-06,
+      "loss": 0.362,
+      "step": 1934
+    },
+    {
+      "epoch": 1.5710419485791611,
+      "grad_norm": 0.23724433779716492,
+      "learning_rate": 5.455527105221222e-06,
+      "loss": 0.4117,
+      "step": 1935
+    },
+    {
+      "epoch": 1.5718538565629228,
+      "grad_norm": 0.2293754667043686,
+      "learning_rate": 5.450819518107468e-06,
+      "loss": 0.3217,
+      "step": 1936
+    },
+    {
+      "epoch": 1.5726657645466848,
+      "grad_norm": 0.2310798019170761,
+      "learning_rate": 5.446111528052455e-06,
+      "loss": 0.351,
+      "step": 1937
+    },
+    {
+      "epoch": 1.5734776725304465,
+      "grad_norm": 0.22842593491077423,
+      "learning_rate": 5.441403139264169e-06,
+      "loss": 0.3223,
+      "step": 1938
+    },
+    {
+      "epoch": 1.5742895805142085,
+      "grad_norm": 0.24148593842983246,
+      "learning_rate": 5.436694355950955e-06,
+      "loss": 0.3685,
+      "step": 1939
+    },
+    {
+      "epoch": 1.5751014884979702,
+      "grad_norm": 0.23123513162136078,
+      "learning_rate": 5.431985182321513e-06,
+      "loss": 0.3742,
+      "step": 1940
+    },
+    {
+      "epoch": 1.5759133964817322,
+      "grad_norm": 0.20439419150352478,
+      "learning_rate": 5.427275622584888e-06,
+      "loss": 0.3217,
+      "step": 1941
+    },
+    {
+      "epoch": 1.5767253044654939,
+      "grad_norm": 0.23474393784999847,
+      "learning_rate": 5.422565680950472e-06,
+      "loss": 0.3595,
+      "step": 1942
+    },
+    {
+      "epoch": 1.5775372124492557,
+      "grad_norm": 0.25419148802757263,
+      "learning_rate": 5.417855361627998e-06,
+      "loss": 0.3768,
+      "step": 1943
+    },
+    {
+      "epoch": 1.5783491204330176,
+      "grad_norm": 0.22208000719547272,
+      "learning_rate": 5.413144668827538e-06,
+      "loss": 0.354,
+      "step": 1944
+    },
+    {
+      "epoch": 1.5791610284167794,
+      "grad_norm": 0.21728238463401794,
+      "learning_rate": 5.4084336067594945e-06,
+      "loss": 0.3786,
+      "step": 1945
+    },
+    {
+      "epoch": 1.5799729364005413,
+      "grad_norm": 0.24146491289138794,
+      "learning_rate": 5.403722179634602e-06,
+      "loss": 0.3625,
+      "step": 1946
+    },
+    {
+      "epoch": 1.5807848443843031,
+      "grad_norm": 0.20292335748672485,
+      "learning_rate": 5.399010391663923e-06,
+      "loss": 0.323,
+      "step": 1947
+    },
+    {
+      "epoch": 1.581596752368065,
+      "grad_norm": 0.24149513244628906,
+      "learning_rate": 5.3942982470588386e-06,
+      "loss": 0.388,
+      "step": 1948
+    },
+    {
+      "epoch": 1.5824086603518268,
+      "grad_norm": 0.2230261266231537,
+      "learning_rate": 5.389585750031053e-06,
+      "loss": 0.3646,
+      "step": 1949
+    },
+    {
+      "epoch": 1.5832205683355887,
+      "grad_norm": 0.2176000326871872,
+      "learning_rate": 5.384872904792583e-06,
+      "loss": 0.3585,
+      "step": 1950
+    },
+    {
+      "epoch": 1.5840324763193505,
+      "grad_norm": 0.23575378954410553,
+      "learning_rate": 5.380159715555755e-06,
+      "loss": 0.3612,
+      "step": 1951
+    },
+    {
+      "epoch": 1.5848443843031124,
+      "grad_norm": 0.23089317977428436,
+      "learning_rate": 5.375446186533206e-06,
+      "loss": 0.3237,
+      "step": 1952
+    },
+    {
+      "epoch": 1.585656292286874,
+      "grad_norm": 0.250468909740448,
+      "learning_rate": 5.370732321937876e-06,
+      "loss": 0.3901,
+      "step": 1953
+    },
+    {
+      "epoch": 1.586468200270636,
+      "grad_norm": 0.20545685291290283,
+      "learning_rate": 5.3660181259830024e-06,
+      "loss": 0.3468,
+      "step": 1954
+    },
+    {
+      "epoch": 1.5872801082543977,
+      "grad_norm": 0.233934223651886,
+      "learning_rate": 5.361303602882123e-06,
+      "loss": 0.3986,
+      "step": 1955
+    },
+    {
+      "epoch": 1.5880920162381598,
+      "grad_norm": 0.23152099549770355,
+      "learning_rate": 5.356588756849065e-06,
+      "loss": 0.3365,
+      "step": 1956
+    },
+    {
+      "epoch": 1.5889039242219214,
+      "grad_norm": 0.21001078188419342,
+      "learning_rate": 5.3518735920979436e-06,
+      "loss": 0.3598,
+      "step": 1957
+    },
+    {
+      "epoch": 1.5897158322056835,
+      "grad_norm": 0.22550509870052338,
+      "learning_rate": 5.347158112843163e-06,
+      "loss": 0.3588,
+      "step": 1958
+    },
+    {
+      "epoch": 1.590527740189445,
+      "grad_norm": 0.2458970993757248,
+      "learning_rate": 5.342442323299403e-06,
+      "loss": 0.3206,
+      "step": 1959
+    },
+    {
+      "epoch": 1.5913396481732072,
+      "grad_norm": 0.23145334422588348,
+      "learning_rate": 5.3377262276816245e-06,
+      "loss": 0.3325,
+      "step": 1960
+    },
+    {
+      "epoch": 1.5921515561569688,
+      "grad_norm": 0.21277864277362823,
+      "learning_rate": 5.3330098302050615e-06,
+      "loss": 0.3701,
+      "step": 1961
+    },
+    {
+      "epoch": 1.5929634641407309,
+      "grad_norm": 0.24452944099903107,
+      "learning_rate": 5.328293135085215e-06,
+      "loss": 0.3881,
+      "step": 1962
+    },
+    {
+      "epoch": 1.5937753721244925,
+      "grad_norm": 0.21212197840213776,
+      "learning_rate": 5.323576146537858e-06,
+      "loss": 0.3496,
+      "step": 1963
+    },
+    {
+      "epoch": 1.5945872801082543,
+      "grad_norm": 0.2445046603679657,
+      "learning_rate": 5.31885886877902e-06,
+      "loss": 0.392,
+      "step": 1964
+    },
+    {
+      "epoch": 1.5953991880920162,
+      "grad_norm": 0.21388094127178192,
+      "learning_rate": 5.314141306024991e-06,
+      "loss": 0.3137,
+      "step": 1965
+    },
+    {
+      "epoch": 1.596211096075778,
+      "grad_norm": 0.25753089785575867,
+      "learning_rate": 5.309423462492314e-06,
+      "loss": 0.382,
+      "step": 1966
+    },
+    {
+      "epoch": 1.59702300405954,
+      "grad_norm": 0.20615671575069427,
+      "learning_rate": 5.30470534239779e-06,
+      "loss": 0.3482,
+      "step": 1967
+    },
+    {
+      "epoch": 1.5978349120433017,
+      "grad_norm": 0.2596169114112854,
+      "learning_rate": 5.299986949958457e-06,
+      "loss": 0.38,
+      "step": 1968
+    },
+    {
+      "epoch": 1.5986468200270636,
+      "grad_norm": 0.2279483526945114,
+      "learning_rate": 5.295268289391603e-06,
+      "loss": 0.3718,
+      "step": 1969
+    },
+    {
+      "epoch": 1.5994587280108254,
+      "grad_norm": 0.21201787889003754,
+      "learning_rate": 5.290549364914754e-06,
+      "loss": 0.3365,
+      "step": 1970
+    },
+    {
+      "epoch": 1.6002706359945873,
+      "grad_norm": 0.2298806756734848,
+      "learning_rate": 5.285830180745673e-06,
+      "loss": 0.3774,
+      "step": 1971
+    },
+    {
+      "epoch": 1.6010825439783491,
+      "grad_norm": 0.20095713436603546,
+      "learning_rate": 5.281110741102352e-06,
+      "loss": 0.3424,
+      "step": 1972
+    },
+    {
+      "epoch": 1.601894451962111,
+      "grad_norm": 0.222127765417099,
+      "learning_rate": 5.276391050203015e-06,
+      "loss": 0.3268,
+      "step": 1973
+    },
+    {
+      "epoch": 1.6027063599458728,
+      "grad_norm": 0.23903107643127441,
+      "learning_rate": 5.2716711122661065e-06,
+      "loss": 0.3774,
+      "step": 1974
+    },
+    {
+      "epoch": 1.6035182679296347,
+      "grad_norm": 0.23499740660190582,
+      "learning_rate": 5.266950931510296e-06,
+      "loss": 0.3862,
+      "step": 1975
+    },
+    {
+      "epoch": 1.6043301759133963,
+      "grad_norm": 0.21363723278045654,
+      "learning_rate": 5.262230512154469e-06,
+      "loss": 0.3137,
+      "step": 1976
+    },
+    {
+      "epoch": 1.6051420838971584,
+      "grad_norm": 0.21218273043632507,
+      "learning_rate": 5.257509858417723e-06,
+      "loss": 0.3421,
+      "step": 1977
+    },
+    {
+      "epoch": 1.60595399188092,
+      "grad_norm": 0.24108603596687317,
+      "learning_rate": 5.252788974519364e-06,
+      "loss": 0.4026,
+      "step": 1978
+    },
+    {
+      "epoch": 1.606765899864682,
+      "grad_norm": 0.20261381566524506,
+      "learning_rate": 5.248067864678906e-06,
+      "loss": 0.3245,
+      "step": 1979
+    },
+    {
+      "epoch": 1.6075778078484437,
+      "grad_norm": 0.22024303674697876,
+      "learning_rate": 5.243346533116065e-06,
+      "loss": 0.3649,
+      "step": 1980
+    },
+    {
+      "epoch": 1.6083897158322058,
+      "grad_norm": 0.23742438852787018,
+      "learning_rate": 5.238624984050754e-06,
+      "loss": 0.3541,
+      "step": 1981
+    },
+    {
+      "epoch": 1.6092016238159674,
+      "grad_norm": 0.21429529786109924,
+      "learning_rate": 5.233903221703079e-06,
+      "loss": 0.3802,
+      "step": 1982
+    },
+    {
+      "epoch": 1.6100135317997295,
+      "grad_norm": 0.21690510213375092,
+      "learning_rate": 5.229181250293341e-06,
+      "loss": 0.3444,
+      "step": 1983
+    },
+    {
+      "epoch": 1.6108254397834911,
+      "grad_norm": 0.22316835820674896,
+      "learning_rate": 5.224459074042023e-06,
+      "loss": 0.3424,
+      "step": 1984
+    },
+    {
+      "epoch": 1.6116373477672532,
+      "grad_norm": 0.22063319385051727,
+      "learning_rate": 5.219736697169795e-06,
+      "loss": 0.3535,
+      "step": 1985
+    },
+    {
+      "epoch": 1.6124492557510148,
+      "grad_norm": 0.21610133349895477,
+      "learning_rate": 5.215014123897504e-06,
+      "loss": 0.3727,
+      "step": 1986
+    },
+    {
+      "epoch": 1.6132611637347767,
+      "grad_norm": 0.19234685599803925,
+      "learning_rate": 5.210291358446173e-06,
+      "loss": 0.3283,
+      "step": 1987
+    },
+    {
+      "epoch": 1.6140730717185385,
+      "grad_norm": 0.24211633205413818,
+      "learning_rate": 5.205568405036996e-06,
+      "loss": 0.3515,
+      "step": 1988
+    },
+    {
+      "epoch": 1.6148849797023004,
+      "grad_norm": 0.23750005662441254,
+      "learning_rate": 5.200845267891338e-06,
+      "loss": 0.3488,
+      "step": 1989
+    },
+    {
+      "epoch": 1.6156968876860622,
+      "grad_norm": 0.25909870862960815,
+      "learning_rate": 5.196121951230726e-06,
+      "loss": 0.3894,
+      "step": 1990
+    },
+    {
+      "epoch": 1.616508795669824,
+      "grad_norm": 0.22888711094856262,
+      "learning_rate": 5.1913984592768455e-06,
+      "loss": 0.3507,
+      "step": 1991
+    },
+    {
+      "epoch": 1.617320703653586,
+      "grad_norm": 0.2067672163248062,
+      "learning_rate": 5.186674796251543e-06,
+      "loss": 0.3295,
+      "step": 1992
+    },
+    {
+      "epoch": 1.6181326116373478,
+      "grad_norm": 0.2428632378578186,
+      "learning_rate": 5.181950966376814e-06,
+      "loss": 0.3869,
+      "step": 1993
+    },
+    {
+      "epoch": 1.6189445196211096,
+      "grad_norm": 0.2573719620704651,
+      "learning_rate": 5.177226973874805e-06,
+      "loss": 0.3718,
+      "step": 1994
+    },
+    {
+      "epoch": 1.6197564276048715,
+      "grad_norm": 0.22836843132972717,
+      "learning_rate": 5.172502822967808e-06,
+      "loss": 0.3544,
+      "step": 1995
+    },
+    {
+      "epoch": 1.6205683355886333,
+      "grad_norm": 0.22402632236480713,
+      "learning_rate": 5.167778517878258e-06,
+      "loss": 0.3561,
+      "step": 1996
+    },
+    {
+      "epoch": 1.621380243572395,
+      "grad_norm": 0.22760261595249176,
+      "learning_rate": 5.163054062828724e-06,
+      "loss": 0.3635,
+      "step": 1997
+    },
+    {
+      "epoch": 1.622192151556157,
+      "grad_norm": 0.2096368372440338,
+      "learning_rate": 5.158329462041911e-06,
+      "loss": 0.3514,
+      "step": 1998
+    },
+    {
+      "epoch": 1.6230040595399187,
+      "grad_norm": 0.21171718835830688,
+      "learning_rate": 5.153604719740655e-06,
+      "loss": 0.3668,
+      "step": 1999
+    },
+    {
+      "epoch": 1.6238159675236807,
+      "grad_norm": 0.22099575400352478,
+      "learning_rate": 5.148879840147918e-06,
+      "loss": 0.3586,
+      "step": 2000
+    },
+    {
+      "epoch": 1.6246278755074424,
+      "grad_norm": 0.24632349610328674,
+      "learning_rate": 5.1441548274867834e-06,
+      "loss": 0.3679,
+      "step": 2001
+    },
+    {
+      "epoch": 1.6254397834912044,
+      "grad_norm": 0.23499879240989685,
+      "learning_rate": 5.139429685980457e-06,
+      "loss": 0.3429,
+      "step": 2002
+    },
+    {
+      "epoch": 1.626251691474966,
+      "grad_norm": 0.2227485328912735,
+      "learning_rate": 5.134704419852253e-06,
+      "loss": 0.3476,
+      "step": 2003
+    },
+    {
+      "epoch": 1.6270635994587281,
+      "grad_norm": 0.20957009494304657,
+      "learning_rate": 5.129979033325608e-06,
+      "loss": 0.3732,
+      "step": 2004
+    },
+    {
+      "epoch": 1.6278755074424898,
+      "grad_norm": 0.2279195487499237,
+      "learning_rate": 5.125253530624054e-06,
+      "loss": 0.3596,
+      "step": 2005
+    },
+    {
+      "epoch": 1.6286874154262518,
+      "grad_norm": 0.2311449646949768,
+      "learning_rate": 5.120527915971235e-06,
+      "loss": 0.349,
+      "step": 2006
+    },
+    {
+      "epoch": 1.6294993234100135,
+      "grad_norm": 0.2110489010810852,
+      "learning_rate": 5.115802193590893e-06,
+      "loss": 0.3498,
+      "step": 2007
+    },
+    {
+      "epoch": 1.6303112313937753,
+      "grad_norm": 0.21867816150188446,
+      "learning_rate": 5.111076367706864e-06,
+      "loss": 0.3682,
+      "step": 2008
+    },
+    {
+      "epoch": 1.6311231393775372,
+      "grad_norm": 0.22895948588848114,
+      "learning_rate": 5.10635044254308e-06,
+      "loss": 0.3594,
+      "step": 2009
+    },
+    {
+      "epoch": 1.631935047361299,
+      "grad_norm": 0.22730299830436707,
+      "learning_rate": 5.1016244223235576e-06,
+      "loss": 0.369,
+      "step": 2010
+    },
+    {
+      "epoch": 1.6327469553450609,
+      "grad_norm": 0.2319604456424713,
+      "learning_rate": 5.096898311272404e-06,
+      "loss": 0.3439,
+      "step": 2011
+    },
+    {
+      "epoch": 1.6335588633288227,
+      "grad_norm": 0.24227219820022583,
+      "learning_rate": 5.0921721136138025e-06,
+      "loss": 0.3973,
+      "step": 2012
+    },
+    {
+      "epoch": 1.6343707713125846,
+      "grad_norm": 0.21264171600341797,
+      "learning_rate": 5.087445833572017e-06,
+      "loss": 0.3493,
+      "step": 2013
+    },
+    {
+      "epoch": 1.6351826792963464,
+      "grad_norm": 0.22045813500881195,
+      "learning_rate": 5.082719475371382e-06,
+      "loss": 0.3467,
+      "step": 2014
+    },
+    {
+      "epoch": 1.6359945872801083,
+      "grad_norm": 0.2252490073442459,
+      "learning_rate": 5.077993043236306e-06,
+      "loss": 0.3423,
+      "step": 2015
+    },
+    {
+      "epoch": 1.6368064952638701,
+      "grad_norm": 0.3103943467140198,
+      "learning_rate": 5.073266541391259e-06,
+      "loss": 0.3314,
+      "step": 2016
+    },
+    {
+      "epoch": 1.637618403247632,
+      "grad_norm": 0.22662578523159027,
+      "learning_rate": 5.068539974060776e-06,
+      "loss": 0.3657,
+      "step": 2017
+    },
+    {
+      "epoch": 1.6384303112313938,
+      "grad_norm": 0.24624420702457428,
+      "learning_rate": 5.06381334546945e-06,
+      "loss": 0.3991,
+      "step": 2018
+    },
+    {
+      "epoch": 1.6392422192151557,
+      "grad_norm": 0.21409954130649567,
+      "learning_rate": 5.05908665984193e-06,
+      "loss": 0.368,
+      "step": 2019
+    },
+    {
+      "epoch": 1.6400541271989173,
+      "grad_norm": 0.23742777109146118,
+      "learning_rate": 5.054359921402914e-06,
+      "loss": 0.3585,
+      "step": 2020
+    },
+    {
+      "epoch": 1.6408660351826794,
+      "grad_norm": 0.22447805106639862,
+      "learning_rate": 5.049633134377147e-06,
+      "loss": 0.3735,
+      "step": 2021
+    },
+    {
+      "epoch": 1.641677943166441,
+      "grad_norm": 0.23393981158733368,
+      "learning_rate": 5.04490630298942e-06,
+      "loss": 0.3588,
+      "step": 2022
+    },
+    {
+      "epoch": 1.642489851150203,
+      "grad_norm": 0.22307370603084564,
+      "learning_rate": 5.04017943146456e-06,
+      "loss": 0.3354,
+      "step": 2023
+    },
+    {
+      "epoch": 1.6433017591339647,
+      "grad_norm": 0.2660270631313324,
+      "learning_rate": 5.035452524027436e-06,
+      "loss": 0.3799,
+      "step": 2024
+    },
+    {
+      "epoch": 1.6441136671177268,
+      "grad_norm": 0.2283170372247696,
+      "learning_rate": 5.03072558490294e-06,
+      "loss": 0.361,
+      "step": 2025
+    },
+    {
+      "epoch": 1.6449255751014884,
+      "grad_norm": 0.24613089859485626,
+      "learning_rate": 5.0259986183160006e-06,
+      "loss": 0.3667,
+      "step": 2026
+    },
+    {
+      "epoch": 1.6457374830852505,
+      "grad_norm": 0.23177827894687653,
+      "learning_rate": 5.021271628491566e-06,
+      "loss": 0.3769,
+      "step": 2027
+    },
+    {
+      "epoch": 1.646549391069012,
+      "grad_norm": 0.22954992949962616,
+      "learning_rate": 5.016544619654609e-06,
+      "loss": 0.3496,
+      "step": 2028
+    },
+    {
+      "epoch": 1.6473612990527742,
+      "grad_norm": 0.23398730158805847,
+      "learning_rate": 5.011817596030115e-06,
+      "loss": 0.3407,
+      "step": 2029
+    },
+    {
+      "epoch": 1.6481732070365358,
+      "grad_norm": 0.26466625928878784,
+      "learning_rate": 5.0070905618430865e-06,
+      "loss": 0.3669,
+      "step": 2030
+    },
+    {
+      "epoch": 1.6489851150202977,
+      "grad_norm": 0.20944908261299133,
+      "learning_rate": 5.0023635213185325e-06,
+      "loss": 0.3223,
+      "step": 2031
+    },
+    {
+      "epoch": 1.6497970230040595,
+      "grad_norm": 0.22746600210666656,
+      "learning_rate": 4.997636478681468e-06,
+      "loss": 0.3841,
+      "step": 2032
+    },
+    {
+      "epoch": 1.6506089309878214,
+      "grad_norm": 0.22021806240081787,
+      "learning_rate": 4.992909438156915e-06,
+      "loss": 0.3501,
+      "step": 2033
+    },
+    {
+      "epoch": 1.6514208389715832,
+      "grad_norm": 0.22996674478054047,
+      "learning_rate": 4.988182403969886e-06,
+      "loss": 0.3432,
+      "step": 2034
+    },
+    {
+      "epoch": 1.652232746955345,
+      "grad_norm": 0.21422921121120453,
+      "learning_rate": 4.983455380345392e-06,
+      "loss": 0.3519,
+      "step": 2035
+    },
+    {
+      "epoch": 1.653044654939107,
+      "grad_norm": 0.2215646654367447,
+      "learning_rate": 4.978728371508433e-06,
+      "loss": 0.3562,
+      "step": 2036
+    },
+    {
+      "epoch": 1.6538565629228688,
+      "grad_norm": 0.23107053339481354,
+      "learning_rate": 4.974001381684e-06,
+      "loss": 0.3627,
+      "step": 2037
+    },
+    {
+      "epoch": 1.6546684709066306,
+      "grad_norm": 0.21991296112537384,
+      "learning_rate": 4.96927441509706e-06,
+      "loss": 0.3293,
+      "step": 2038
+    },
+    {
+      "epoch": 1.6554803788903925,
+      "grad_norm": 0.2307509034872055,
+      "learning_rate": 4.964547475972565e-06,
+      "loss": 0.3783,
+      "step": 2039
+    },
+    {
+      "epoch": 1.6562922868741543,
+      "grad_norm": 0.229376882314682,
+      "learning_rate": 4.9598205685354404e-06,
+      "loss": 0.351,
+      "step": 2040
+    },
+    {
+      "epoch": 1.6571041948579162,
+      "grad_norm": 0.22200506925582886,
+      "learning_rate": 4.955093697010581e-06,
+      "loss": 0.3753,
+      "step": 2041
+    },
+    {
+      "epoch": 1.657916102841678,
+      "grad_norm": 0.2260511815547943,
+      "learning_rate": 4.950366865622854e-06,
+      "loss": 0.339,
+      "step": 2042
+    },
+    {
+      "epoch": 1.6587280108254396,
+      "grad_norm": 0.24783973395824432,
+      "learning_rate": 4.9456400785970875e-06,
+      "loss": 0.3674,
+      "step": 2043
+    },
+    {
+      "epoch": 1.6595399188092017,
+      "grad_norm": 0.250405490398407,
+      "learning_rate": 4.940913340158072e-06,
+      "loss": 0.3675,
+      "step": 2044
+    },
+    {
+      "epoch": 1.6603518267929633,
+      "grad_norm": 0.23701797425746918,
+      "learning_rate": 4.936186654530551e-06,
+      "loss": 0.3275,
+      "step": 2045
+    },
+    {
+      "epoch": 1.6611637347767254,
+      "grad_norm": 0.23613330721855164,
+      "learning_rate": 4.931460025939226e-06,
+      "loss": 0.3395,
+      "step": 2046
+    },
+    {
+      "epoch": 1.661975642760487,
+      "grad_norm": 0.23792241513729095,
+      "learning_rate": 4.926733458608744e-06,
+      "loss": 0.394,
+      "step": 2047
+    },
+    {
+      "epoch": 1.662787550744249,
+      "grad_norm": 0.21871714293956757,
+      "learning_rate": 4.922006956763697e-06,
+      "loss": 0.3315,
+      "step": 2048
+    },
+    {
+      "epoch": 1.6635994587280107,
+      "grad_norm": 0.23582173883914948,
+      "learning_rate": 4.917280524628619e-06,
+      "loss": 0.3688,
+      "step": 2049
+    },
+    {
+      "epoch": 1.6644113667117728,
+      "grad_norm": 0.2296038568019867,
+      "learning_rate": 4.912554166427985e-06,
+      "loss": 0.3545,
+      "step": 2050
+    },
+    {
+      "epoch": 1.6652232746955344,
+      "grad_norm": 0.2224859595298767,
+      "learning_rate": 4.907827886386199e-06,
+      "loss": 0.342,
+      "step": 2051
+    },
+    {
+      "epoch": 1.6660351826792965,
+      "grad_norm": 0.208472341299057,
+      "learning_rate": 4.903101688727598e-06,
+      "loss": 0.3201,
+      "step": 2052
+    },
+    {
+      "epoch": 1.6668470906630581,
+      "grad_norm": 0.24549537897109985,
+      "learning_rate": 4.898375577676444e-06,
+      "loss": 0.3976,
+      "step": 2053
+    },
+    {
+      "epoch": 1.66765899864682,
+      "grad_norm": 0.2230387181043625,
+      "learning_rate": 4.893649557456922e-06,
+      "loss": 0.3611,
+      "step": 2054
+    },
+    {
+      "epoch": 1.6684709066305818,
+      "grad_norm": 0.2504335343837738,
+      "learning_rate": 4.888923632293137e-06,
+      "loss": 0.3321,
+      "step": 2055
+    },
+    {
+      "epoch": 1.6692828146143437,
+      "grad_norm": 0.2363443225622177,
+      "learning_rate": 4.884197806409109e-06,
+      "loss": 0.3433,
+      "step": 2056
+    },
+    {
+      "epoch": 1.6700947225981055,
+      "grad_norm": 0.22222736477851868,
+      "learning_rate": 4.8794720840287665e-06,
+      "loss": 0.353,
+      "step": 2057
+    },
+    {
+      "epoch": 1.6709066305818674,
+      "grad_norm": 0.23795849084854126,
+      "learning_rate": 4.874746469375947e-06,
+      "loss": 0.358,
+      "step": 2058
+    },
+    {
+      "epoch": 1.6717185385656292,
+      "grad_norm": 0.21926665306091309,
+      "learning_rate": 4.870020966674394e-06,
+      "loss": 0.3582,
+      "step": 2059
+    },
+    {
+      "epoch": 1.672530446549391,
+      "grad_norm": 0.24244748055934906,
+      "learning_rate": 4.865295580147747e-06,
+      "loss": 0.4043,
+      "step": 2060
+    },
+    {
+      "epoch": 1.673342354533153,
+      "grad_norm": 0.2079298347234726,
+      "learning_rate": 4.860570314019546e-06,
+      "loss": 0.3358,
+      "step": 2061
+    },
+    {
+      "epoch": 1.6741542625169148,
+      "grad_norm": 0.2525360882282257,
+      "learning_rate": 4.855845172513219e-06,
+      "loss": 0.3735,
+      "step": 2062
+    },
+    {
+      "epoch": 1.6749661705006766,
+      "grad_norm": 0.22536899149417877,
+      "learning_rate": 4.851120159852085e-06,
+      "loss": 0.3468,
+      "step": 2063
+    },
+    {
+      "epoch": 1.6757780784844383,
+      "grad_norm": 0.23605121672153473,
+      "learning_rate": 4.846395280259348e-06,
+      "loss": 0.3493,
+      "step": 2064
+    },
+    {
+      "epoch": 1.6765899864682003,
+      "grad_norm": 0.24195890128612518,
+      "learning_rate": 4.841670537958092e-06,
+      "loss": 0.3441,
+      "step": 2065
+    },
+    {
+      "epoch": 1.677401894451962,
+      "grad_norm": 0.2338932454586029,
+      "learning_rate": 4.836945937171279e-06,
+      "loss": 0.3586,
+      "step": 2066
+    },
+    {
+      "epoch": 1.678213802435724,
+      "grad_norm": 0.43127959966659546,
+      "learning_rate": 4.832221482121745e-06,
+      "loss": 0.3326,
+      "step": 2067
+    },
+    {
+      "epoch": 1.6790257104194857,
+      "grad_norm": 0.2522030174732208,
+      "learning_rate": 4.827497177032193e-06,
+      "loss": 0.3643,
+      "step": 2068
+    },
+    {
+      "epoch": 1.6798376184032477,
+      "grad_norm": 0.22540558874607086,
+      "learning_rate": 4.822773026125197e-06,
+      "loss": 0.3649,
+      "step": 2069
+    },
+    {
+      "epoch": 1.6806495263870094,
+      "grad_norm": 0.2500484585762024,
+      "learning_rate": 4.8180490336231865e-06,
+      "loss": 0.3971,
+      "step": 2070
+    },
+    {
+      "epoch": 1.6814614343707714,
+      "grad_norm": 0.22149935364723206,
+      "learning_rate": 4.813325203748458e-06,
+      "loss": 0.3103,
+      "step": 2071
+    },
+    {
+      "epoch": 1.682273342354533,
+      "grad_norm": 0.2178916037082672,
+      "learning_rate": 4.8086015407231545e-06,
+      "loss": 0.3412,
+      "step": 2072
+    },
+    {
+      "epoch": 1.6830852503382951,
+      "grad_norm": 0.24394278228282928,
+      "learning_rate": 4.803878048769275e-06,
+      "loss": 0.3995,
+      "step": 2073
+    },
+    {
+      "epoch": 1.6838971583220568,
+      "grad_norm": 0.22522945702075958,
+      "learning_rate": 4.799154732108662e-06,
+      "loss": 0.3222,
+      "step": 2074
+    },
+    {
+      "epoch": 1.6847090663058186,
+      "grad_norm": 0.22441378235816956,
+      "learning_rate": 4.794431594963004e-06,
+      "loss": 0.3829,
+      "step": 2075
+    },
+    {
+      "epoch": 1.6855209742895805,
+      "grad_norm": 0.2116931825876236,
+      "learning_rate": 4.789708641553828e-06,
+      "loss": 0.3364,
+      "step": 2076
+    },
+    {
+      "epoch": 1.6863328822733423,
+      "grad_norm": 0.24719005823135376,
+      "learning_rate": 4.784985876102498e-06,
+      "loss": 0.3269,
+      "step": 2077
+    },
+    {
+      "epoch": 1.6871447902571042,
+      "grad_norm": 0.23803463578224182,
+      "learning_rate": 4.7802633028302065e-06,
+      "loss": 0.3857,
+      "step": 2078
+    },
+    {
+      "epoch": 1.687956698240866,
+      "grad_norm": 0.26219871640205383,
+      "learning_rate": 4.7755409259579786e-06,
+      "loss": 0.374,
+      "step": 2079
+    },
+    {
+      "epoch": 1.6887686062246279,
+      "grad_norm": 0.2567477226257324,
+      "learning_rate": 4.7708187497066595e-06,
+      "loss": 0.366,
+      "step": 2080
+    },
+    {
+      "epoch": 1.6895805142083897,
+      "grad_norm": 0.2233663946390152,
+      "learning_rate": 4.766096778296922e-06,
+      "loss": 0.343,
+      "step": 2081
+    },
+    {
+      "epoch": 1.6903924221921516,
+      "grad_norm": 0.2194279581308365,
+      "learning_rate": 4.761375015949248e-06,
+      "loss": 0.385,
+      "step": 2082
+    },
+    {
+      "epoch": 1.6912043301759134,
+      "grad_norm": 0.19891737401485443,
+      "learning_rate": 4.756653466883937e-06,
+      "loss": 0.3186,
+      "step": 2083
+    },
+    {
+      "epoch": 1.6920162381596753,
+      "grad_norm": 0.22539734840393066,
+      "learning_rate": 4.751932135321095e-06,
+      "loss": 0.3533,
+      "step": 2084
+    },
+    {
+      "epoch": 1.6928281461434371,
+      "grad_norm": 0.24812281131744385,
+      "learning_rate": 4.7472110254806375e-06,
+      "loss": 0.3801,
+      "step": 2085
+    },
+    {
+      "epoch": 1.693640054127199,
+      "grad_norm": 0.2293502390384674,
+      "learning_rate": 4.742490141582279e-06,
+      "loss": 0.3443,
+      "step": 2086
+    },
+    {
+      "epoch": 1.6944519621109606,
+      "grad_norm": 0.22224600613117218,
+      "learning_rate": 4.737769487845532e-06,
+      "loss": 0.3634,
+      "step": 2087
+    },
+    {
+      "epoch": 1.6952638700947227,
+      "grad_norm": 0.22039075195789337,
+      "learning_rate": 4.733049068489704e-06,
+      "loss": 0.3493,
+      "step": 2088
+    },
+    {
+      "epoch": 1.6960757780784843,
+      "grad_norm": 0.2321782112121582,
+      "learning_rate": 4.728328887733894e-06,
+      "loss": 0.3542,
+      "step": 2089
+    },
+    {
+      "epoch": 1.6968876860622464,
+      "grad_norm": 0.23392166197299957,
+      "learning_rate": 4.723608949796987e-06,
+      "loss": 0.388,
+      "step": 2090
+    },
+    {
+      "epoch": 1.697699594046008,
+      "grad_norm": 0.23521338403224945,
+      "learning_rate": 4.7188892588976496e-06,
+      "loss": 0.3603,
+      "step": 2091
+    },
+    {
+      "epoch": 1.69851150202977,
+      "grad_norm": 0.2185409963130951,
+      "learning_rate": 4.714169819254328e-06,
+      "loss": 0.3648,
+      "step": 2092
+    },
+    {
+      "epoch": 1.6993234100135317,
+      "grad_norm": 0.23572126030921936,
+      "learning_rate": 4.709450635085246e-06,
+      "loss": 0.3596,
+      "step": 2093
+    },
+    {
+      "epoch": 1.7001353179972938,
+      "grad_norm": 0.2501370310783386,
+      "learning_rate": 4.704731710608398e-06,
+      "loss": 0.3553,
+      "step": 2094
+    },
+    {
+      "epoch": 1.7009472259810554,
+      "grad_norm": 0.22917629778385162,
+      "learning_rate": 4.700013050041544e-06,
+      "loss": 0.3479,
+      "step": 2095
+    },
+    {
+      "epoch": 1.7017591339648175,
+      "grad_norm": 0.2135496586561203,
+      "learning_rate": 4.695294657602212e-06,
+      "loss": 0.3519,
+      "step": 2096
+    },
+    {
+      "epoch": 1.7025710419485791,
+      "grad_norm": 0.235653817653656,
+      "learning_rate": 4.6905765375076865e-06,
+      "loss": 0.3746,
+      "step": 2097
+    },
+    {
+      "epoch": 1.703382949932341,
+      "grad_norm": 0.2271241694688797,
+      "learning_rate": 4.685858693975011e-06,
+      "loss": 0.3528,
+      "step": 2098
+    },
+    {
+      "epoch": 1.7041948579161028,
+      "grad_norm": 0.204807847738266,
+      "learning_rate": 4.681141131220982e-06,
+      "loss": 0.3469,
+      "step": 2099
+    },
+    {
+      "epoch": 1.7050067658998647,
+      "grad_norm": 0.23420177400112152,
+      "learning_rate": 4.6764238534621436e-06,
+      "loss": 0.362,
+      "step": 2100
+    },
+    {
+      "epoch": 1.7058186738836265,
+      "grad_norm": 0.24142847955226898,
+      "learning_rate": 4.671706864914786e-06,
+      "loss": 0.3952,
+      "step": 2101
+    },
+    {
+      "epoch": 1.7066305818673884,
+      "grad_norm": 0.19491349160671234,
+      "learning_rate": 4.666990169794942e-06,
+      "loss": 0.323,
+      "step": 2102
+    },
+    {
+      "epoch": 1.7074424898511502,
+      "grad_norm": 0.2200826108455658,
+      "learning_rate": 4.662273772318378e-06,
+      "loss": 0.3414,
+      "step": 2103
+    },
+    {
+      "epoch": 1.708254397834912,
+      "grad_norm": 0.2446584850549698,
+      "learning_rate": 4.6575576767006e-06,
+      "loss": 0.3491,
+      "step": 2104
+    },
+    {
+      "epoch": 1.709066305818674,
+      "grad_norm": 0.22027452290058136,
+      "learning_rate": 4.652841887156841e-06,
+      "loss": 0.3446,
+      "step": 2105
+    },
+    {
+      "epoch": 1.7098782138024358,
+      "grad_norm": 0.22442978620529175,
+      "learning_rate": 4.648126407902058e-06,
+      "loss": 0.3744,
+      "step": 2106
+    },
+    {
+      "epoch": 1.7106901217861976,
+      "grad_norm": 0.22898204624652863,
+      "learning_rate": 4.6434112431509376e-06,
+      "loss": 0.3459,
+      "step": 2107
+    },
+    {
+      "epoch": 1.7115020297699592,
+      "grad_norm": 0.22009025514125824,
+      "learning_rate": 4.638696397117877e-06,
+      "loss": 0.3352,
+      "step": 2108
+    },
+    {
+      "epoch": 1.7123139377537213,
+      "grad_norm": 0.24266277253627777,
+      "learning_rate": 4.6339818740169975e-06,
+      "loss": 0.3658,
+      "step": 2109
+    },
+    {
+      "epoch": 1.713125845737483,
+      "grad_norm": 0.2355370670557022,
+      "learning_rate": 4.6292676780621245e-06,
+      "loss": 0.321,
+      "step": 2110
+    },
+    {
+      "epoch": 1.713937753721245,
+      "grad_norm": 0.20500701665878296,
+      "learning_rate": 4.624553813466794e-06,
+      "loss": 0.3701,
+      "step": 2111
+    },
+    {
+      "epoch": 1.7147496617050066,
+      "grad_norm": 0.2113361358642578,
+      "learning_rate": 4.619840284444245e-06,
+      "loss": 0.3669,
+      "step": 2112
+    },
+    {
+      "epoch": 1.7155615696887687,
+      "grad_norm": 0.21919837594032288,
+      "learning_rate": 4.615127095207418e-06,
+      "loss": 0.3757,
+      "step": 2113
+    },
+    {
+      "epoch": 1.7163734776725303,
+      "grad_norm": 0.21524176001548767,
+      "learning_rate": 4.610414249968947e-06,
+      "loss": 0.3545,
+      "step": 2114
+    },
+    {
+      "epoch": 1.7171853856562924,
+      "grad_norm": 0.23822060227394104,
+      "learning_rate": 4.605701752941161e-06,
+      "loss": 0.3701,
+      "step": 2115
+    },
+    {
+      "epoch": 1.717997293640054,
+      "grad_norm": 0.19620686769485474,
+      "learning_rate": 4.600989608336078e-06,
+      "loss": 0.3407,
+      "step": 2116
+    },
+    {
+      "epoch": 1.7188092016238161,
+      "grad_norm": 0.24326026439666748,
+      "learning_rate": 4.596277820365398e-06,
+      "loss": 0.3358,
+      "step": 2117
+    },
+    {
+      "epoch": 1.7196211096075777,
+      "grad_norm": 0.2433251589536667,
+      "learning_rate": 4.591566393240507e-06,
+      "loss": 0.3691,
+      "step": 2118
+    },
+    {
+      "epoch": 1.7204330175913396,
+      "grad_norm": 0.2223580777645111,
+      "learning_rate": 4.5868553311724635e-06,
+      "loss": 0.3931,
+      "step": 2119
+    },
+    {
+      "epoch": 1.7212449255751014,
+      "grad_norm": 0.23850855231285095,
+      "learning_rate": 4.5821446383720025e-06,
+      "loss": 0.3456,
+      "step": 2120
+    },
+    {
+      "epoch": 1.7220568335588633,
+      "grad_norm": 0.24507680535316467,
+      "learning_rate": 4.577434319049529e-06,
+      "loss": 0.3704,
+      "step": 2121
+    },
+    {
+      "epoch": 1.7228687415426251,
+      "grad_norm": 0.2253115177154541,
+      "learning_rate": 4.572724377415114e-06,
+      "loss": 0.3853,
+      "step": 2122
+    },
+    {
+      "epoch": 1.723680649526387,
+      "grad_norm": 0.20647768676280975,
+      "learning_rate": 4.568014817678488e-06,
+      "loss": 0.3483,
+      "step": 2123
+    },
+    {
+      "epoch": 1.7244925575101488,
+      "grad_norm": 0.26571783423423767,
+      "learning_rate": 4.5633056440490455e-06,
+      "loss": 0.3816,
+      "step": 2124
+    },
+    {
+      "epoch": 1.7253044654939107,
+      "grad_norm": 0.2427477389574051,
+      "learning_rate": 4.558596860735833e-06,
+      "loss": 0.3953,
+      "step": 2125
+    },
+    {
+      "epoch": 1.7261163734776725,
+      "grad_norm": 0.22305655479431152,
+      "learning_rate": 4.553888471947546e-06,
+      "loss": 0.3321,
+      "step": 2126
+    },
+    {
+      "epoch": 1.7269282814614344,
+      "grad_norm": 0.2457035332918167,
+      "learning_rate": 4.549180481892532e-06,
+      "loss": 0.3468,
+      "step": 2127
+    },
+    {
+      "epoch": 1.7277401894451963,
+      "grad_norm": 0.2372967153787613,
+      "learning_rate": 4.544472894778779e-06,
+      "loss": 0.3719,
+      "step": 2128
+    },
+    {
+      "epoch": 1.728552097428958,
+      "grad_norm": 0.22203095257282257,
+      "learning_rate": 4.539765714813915e-06,
+      "loss": 0.351,
+      "step": 2129
+    },
+    {
+      "epoch": 1.72936400541272,
+      "grad_norm": 0.2441115826368332,
+      "learning_rate": 4.535058946205205e-06,
+      "loss": 0.3861,
+      "step": 2130
+    },
+    {
+      "epoch": 1.7301759133964816,
+      "grad_norm": 0.23017102479934692,
+      "learning_rate": 4.530352593159546e-06,
+      "loss": 0.3373,
+      "step": 2131
+    },
+    {
+      "epoch": 1.7309878213802437,
+      "grad_norm": 0.22350594401359558,
+      "learning_rate": 4.525646659883463e-06,
+      "loss": 0.3714,
+      "step": 2132
+    },
+    {
+      "epoch": 1.7317997293640053,
+      "grad_norm": 0.20636244118213654,
+      "learning_rate": 4.5209411505831066e-06,
+      "loss": 0.3458,
+      "step": 2133
+    },
+    {
+      "epoch": 1.7326116373477674,
+      "grad_norm": 0.24830478429794312,
+      "learning_rate": 4.51623606946425e-06,
+      "loss": 0.3756,
+      "step": 2134
+    },
+    {
+      "epoch": 1.733423545331529,
+      "grad_norm": 0.22760944068431854,
+      "learning_rate": 4.5115314207322794e-06,
+      "loss": 0.3575,
+      "step": 2135
+    },
+    {
+      "epoch": 1.734235453315291,
+      "grad_norm": 0.2360357642173767,
+      "learning_rate": 4.506827208592198e-06,
+      "loss": 0.3884,
+      "step": 2136
+    },
+    {
+      "epoch": 1.7350473612990527,
+      "grad_norm": 0.22573520243167877,
+      "learning_rate": 4.502123437248619e-06,
+      "loss": 0.324,
+      "step": 2137
+    },
+    {
+      "epoch": 1.7358592692828148,
+      "grad_norm": 0.23045194149017334,
+      "learning_rate": 4.497420110905758e-06,
+      "loss": 0.3828,
+      "step": 2138
+    },
+    {
+      "epoch": 1.7366711772665764,
+      "grad_norm": 0.2092134654521942,
+      "learning_rate": 4.492717233767438e-06,
+      "loss": 0.3391,
+      "step": 2139
+    },
+    {
+      "epoch": 1.7374830852503385,
+      "grad_norm": 0.22113755345344543,
+      "learning_rate": 4.4880148100370765e-06,
+      "loss": 0.3583,
+      "step": 2140
+    },
+    {
+      "epoch": 1.7382949932341,
+      "grad_norm": 0.23820653557777405,
+      "learning_rate": 4.483312843917685e-06,
+      "loss": 0.3751,
+      "step": 2141
+    },
+    {
+      "epoch": 1.739106901217862,
+      "grad_norm": 0.21718120574951172,
+      "learning_rate": 4.4786113396118715e-06,
+      "loss": 0.3215,
+      "step": 2142
+    },
+    {
+      "epoch": 1.7399188092016238,
+      "grad_norm": 0.25260668992996216,
+      "learning_rate": 4.473910301321827e-06,
+      "loss": 0.4002,
+      "step": 2143
+    },
+    {
+      "epoch": 1.7407307171853856,
+      "grad_norm": 0.22052276134490967,
+      "learning_rate": 4.469209733249326e-06,
+      "loss": 0.3667,
+      "step": 2144
+    },
+    {
+      "epoch": 1.7415426251691475,
+      "grad_norm": 0.21894143521785736,
+      "learning_rate": 4.4645096395957235e-06,
+      "loss": 0.355,
+      "step": 2145
+    },
+    {
+      "epoch": 1.7423545331529093,
+      "grad_norm": 0.21630899608135223,
+      "learning_rate": 4.4598100245619505e-06,
+      "loss": 0.374,
+      "step": 2146
+    },
+    {
+      "epoch": 1.7431664411366712,
+      "grad_norm": 0.24362193048000336,
+      "learning_rate": 4.455110892348508e-06,
+      "loss": 0.3755,
+      "step": 2147
+    },
+    {
+      "epoch": 1.743978349120433,
+      "grad_norm": 0.2260470986366272,
+      "learning_rate": 4.450412247155473e-06,
+      "loss": 0.3277,
+      "step": 2148
+    },
+    {
+      "epoch": 1.7447902571041949,
+      "grad_norm": 0.2684026062488556,
+      "learning_rate": 4.445714093182477e-06,
+      "loss": 0.3731,
+      "step": 2149
+    },
+    {
+      "epoch": 1.7456021650879567,
+      "grad_norm": 0.25399261713027954,
+      "learning_rate": 4.441016434628719e-06,
+      "loss": 0.3858,
+      "step": 2150
+    },
+    {
+      "epoch": 1.7464140730717186,
+      "grad_norm": 0.22806675732135773,
+      "learning_rate": 4.436319275692954e-06,
+      "loss": 0.361,
+      "step": 2151
+    },
+    {
+      "epoch": 1.7472259810554802,
+      "grad_norm": 0.2798563539981842,
+      "learning_rate": 4.43162262057349e-06,
+      "loss": 0.3607,
+      "step": 2152
+    },
+    {
+      "epoch": 1.7480378890392423,
+      "grad_norm": 0.22141033411026,
+      "learning_rate": 4.426926473468183e-06,
+      "loss": 0.352,
+      "step": 2153
+    },
+    {
+      "epoch": 1.748849797023004,
+      "grad_norm": 0.22465194761753082,
+      "learning_rate": 4.4222308385744375e-06,
+      "loss": 0.3743,
+      "step": 2154
+    },
+    {
+      "epoch": 1.749661705006766,
+      "grad_norm": 0.22997747361660004,
+      "learning_rate": 4.417535720089199e-06,
+      "loss": 0.3612,
+      "step": 2155
+    },
+    {
+      "epoch": 1.7504736129905276,
+      "grad_norm": 0.24141564965248108,
+      "learning_rate": 4.412841122208951e-06,
+      "loss": 0.3704,
+      "step": 2156
+    },
+    {
+      "epoch": 1.7512855209742897,
+      "grad_norm": 0.22115804255008698,
+      "learning_rate": 4.408147049129715e-06,
+      "loss": 0.3356,
+      "step": 2157
+    },
+    {
+      "epoch": 1.7520974289580513,
+      "grad_norm": 0.23318108916282654,
+      "learning_rate": 4.403453505047037e-06,
+      "loss": 0.3372,
+      "step": 2158
+    },
+    {
+      "epoch": 1.7529093369418134,
+      "grad_norm": 0.22440466284751892,
+      "learning_rate": 4.3987604941559966e-06,
+      "loss": 0.3608,
+      "step": 2159
+    },
+    {
+      "epoch": 1.753721244925575,
+      "grad_norm": 0.24263817071914673,
+      "learning_rate": 4.3940680206511935e-06,
+      "loss": 0.3805,
+      "step": 2160
+    },
+    {
+      "epoch": 1.754533152909337,
+      "grad_norm": 0.2338429093360901,
+      "learning_rate": 4.389376088726747e-06,
+      "loss": 0.3105,
+      "step": 2161
+    },
+    {
+      "epoch": 1.7553450608930987,
+      "grad_norm": 0.27486127614974976,
+      "learning_rate": 4.3846847025762944e-06,
+      "loss": 0.3962,
+      "step": 2162
+    },
+    {
+      "epoch": 1.7561569688768606,
+      "grad_norm": 0.209196999669075,
+      "learning_rate": 4.3799938663929835e-06,
+      "loss": 0.3293,
+      "step": 2163
+    },
+    {
+      "epoch": 1.7569688768606224,
+      "grad_norm": 0.24542148411273956,
+      "learning_rate": 4.37530358436947e-06,
+      "loss": 0.3824,
+      "step": 2164
+    },
+    {
+      "epoch": 1.7577807848443843,
+      "grad_norm": 0.21622587740421295,
+      "learning_rate": 4.370613860697916e-06,
+      "loss": 0.3584,
+      "step": 2165
+    },
+    {
+      "epoch": 1.7585926928281461,
+      "grad_norm": 0.2066020369529724,
+      "learning_rate": 4.3659246995699845e-06,
+      "loss": 0.3238,
+      "step": 2166
+    },
+    {
+      "epoch": 1.759404600811908,
+      "grad_norm": 0.22473132610321045,
+      "learning_rate": 4.361236105176832e-06,
+      "loss": 0.3506,
+      "step": 2167
+    },
+    {
+      "epoch": 1.7602165087956698,
+      "grad_norm": 0.21264219284057617,
+      "learning_rate": 4.356548081709116e-06,
+      "loss": 0.3734,
+      "step": 2168
+    },
+    {
+      "epoch": 1.7610284167794317,
+      "grad_norm": 0.20075322687625885,
+      "learning_rate": 4.351860633356976e-06,
+      "loss": 0.3206,
+      "step": 2169
+    },
+    {
+      "epoch": 1.7618403247631935,
+      "grad_norm": 0.2389746606349945,
+      "learning_rate": 4.347173764310041e-06,
+      "loss": 0.3794,
+      "step": 2170
+    },
+    {
+      "epoch": 1.7626522327469554,
+      "grad_norm": 0.22590701282024384,
+      "learning_rate": 4.342487478757422e-06,
+      "loss": 0.3648,
+      "step": 2171
+    },
+    {
+      "epoch": 1.7634641407307172,
+      "grad_norm": 0.1987086832523346,
+      "learning_rate": 4.3378017808877095e-06,
+      "loss": 0.3128,
+      "step": 2172
+    },
+    {
+      "epoch": 1.764276048714479,
+      "grad_norm": 0.2341879904270172,
+      "learning_rate": 4.333116674888966e-06,
+      "loss": 0.3635,
+      "step": 2173
+    },
+    {
+      "epoch": 1.765087956698241,
+      "grad_norm": 0.21438905596733093,
+      "learning_rate": 4.328432164948727e-06,
+      "loss": 0.3725,
+      "step": 2174
+    },
+    {
+      "epoch": 1.7658998646820026,
+      "grad_norm": 0.19643835723400116,
+      "learning_rate": 4.323748255253995e-06,
+      "loss": 0.3482,
+      "step": 2175
+    },
+    {
+      "epoch": 1.7667117726657646,
+      "grad_norm": 0.2233283817768097,
+      "learning_rate": 4.319064949991236e-06,
+      "loss": 0.3428,
+      "step": 2176
+    },
+    {
+      "epoch": 1.7675236806495263,
+      "grad_norm": 0.2417917251586914,
+      "learning_rate": 4.314382253346374e-06,
+      "loss": 0.3819,
+      "step": 2177
+    },
+    {
+      "epoch": 1.7683355886332883,
+      "grad_norm": 0.20478829741477966,
+      "learning_rate": 4.309700169504792e-06,
+      "loss": 0.357,
+      "step": 2178
+    },
+    {
+      "epoch": 1.76914749661705,
+      "grad_norm": 0.23109965026378632,
+      "learning_rate": 4.305018702651323e-06,
+      "loss": 0.3747,
+      "step": 2179
+    },
+    {
+      "epoch": 1.769959404600812,
+      "grad_norm": 0.23093485832214355,
+      "learning_rate": 4.300337856970251e-06,
+      "loss": 0.337,
+      "step": 2180
+    },
+    {
+      "epoch": 1.7707713125845737,
+      "grad_norm": 0.21209514141082764,
+      "learning_rate": 4.295657636645302e-06,
+      "loss": 0.357,
+      "step": 2181
+    },
+    {
+      "epoch": 1.7715832205683357,
+      "grad_norm": 0.2146962434053421,
+      "learning_rate": 4.290978045859643e-06,
+      "loss": 0.3486,
+      "step": 2182
+    },
+    {
+      "epoch": 1.7723951285520974,
+      "grad_norm": 0.20830734074115753,
+      "learning_rate": 4.2862990887958815e-06,
+      "loss": 0.3876,
+      "step": 2183
+    },
+    {
+      "epoch": 1.7732070365358594,
+      "grad_norm": 0.2245628833770752,
+      "learning_rate": 4.281620769636055e-06,
+      "loss": 0.3534,
+      "step": 2184
+    },
+    {
+      "epoch": 1.774018944519621,
+      "grad_norm": 0.2165120542049408,
+      "learning_rate": 4.276943092561632e-06,
+      "loss": 0.3354,
+      "step": 2185
+    },
+    {
+      "epoch": 1.774830852503383,
+      "grad_norm": 0.21295399963855743,
+      "learning_rate": 4.2722660617535105e-06,
+      "loss": 0.3784,
+      "step": 2186
+    },
+    {
+      "epoch": 1.7756427604871448,
+      "grad_norm": 0.20074157416820526,
+      "learning_rate": 4.267589681392006e-06,
+      "loss": 0.3225,
+      "step": 2187
+    },
+    {
+      "epoch": 1.7764546684709066,
+      "grad_norm": 0.23749874532222748,
+      "learning_rate": 4.262913955656856e-06,
+      "loss": 0.3489,
+      "step": 2188
+    },
+    {
+      "epoch": 1.7772665764546685,
+      "grad_norm": 0.21890892088413239,
+      "learning_rate": 4.258238888727209e-06,
+      "loss": 0.3495,
+      "step": 2189
+    },
+    {
+      "epoch": 1.7780784844384303,
+      "grad_norm": 0.25863462686538696,
+      "learning_rate": 4.253564484781629e-06,
+      "loss": 0.374,
+      "step": 2190
+    },
+    {
+      "epoch": 1.7788903924221922,
+      "grad_norm": 0.21450206637382507,
+      "learning_rate": 4.248890747998085e-06,
+      "loss": 0.3619,
+      "step": 2191
+    },
+    {
+      "epoch": 1.779702300405954,
+      "grad_norm": 0.21328723430633545,
+      "learning_rate": 4.24421768255395e-06,
+      "loss": 0.3182,
+      "step": 2192
+    },
+    {
+      "epoch": 1.7805142083897159,
+      "grad_norm": 0.24588316679000854,
+      "learning_rate": 4.2395452926259965e-06,
+      "loss": 0.3528,
+      "step": 2193
+    },
+    {
+      "epoch": 1.7813261163734777,
+      "grad_norm": 0.22966864705085754,
+      "learning_rate": 4.234873582390395e-06,
+      "loss": 0.3602,
+      "step": 2194
+    },
+    {
+      "epoch": 1.7821380243572396,
+      "grad_norm": 0.21149961650371552,
+      "learning_rate": 4.230202556022708e-06,
+      "loss": 0.3517,
+      "step": 2195
+    },
+    {
+      "epoch": 1.7829499323410012,
+      "grad_norm": 0.21298307180404663,
+      "learning_rate": 4.225532217697886e-06,
+      "loss": 0.3352,
+      "step": 2196
+    },
+    {
+      "epoch": 1.7837618403247633,
+      "grad_norm": 0.23622529208660126,
+      "learning_rate": 4.220862571590264e-06,
+      "loss": 0.3738,
+      "step": 2197
+    },
+    {
+      "epoch": 1.7845737483085249,
+      "grad_norm": 0.22376419603824615,
+      "learning_rate": 4.216193621873559e-06,
+      "loss": 0.3391,
+      "step": 2198
+    },
+    {
+      "epoch": 1.785385656292287,
+      "grad_norm": 0.21622315049171448,
+      "learning_rate": 4.211525372720865e-06,
+      "loss": 0.3881,
+      "step": 2199
+    },
+    {
+      "epoch": 1.7861975642760486,
+      "grad_norm": 0.2302180528640747,
+      "learning_rate": 4.206857828304651e-06,
+      "loss": 0.3201,
+      "step": 2200
+    },
+    {
+      "epoch": 1.7870094722598107,
+      "grad_norm": 0.23346953094005585,
+      "learning_rate": 4.202190992796756e-06,
+      "loss": 0.3383,
+      "step": 2201
+    },
+    {
+      "epoch": 1.7878213802435723,
+      "grad_norm": 0.24025745689868927,
+      "learning_rate": 4.1975248703683855e-06,
+      "loss": 0.3663,
+      "step": 2202
+    },
+    {
+      "epoch": 1.7886332882273344,
+      "grad_norm": 0.22679300606250763,
+      "learning_rate": 4.1928594651901055e-06,
+      "loss": 0.3659,
+      "step": 2203
+    },
+    {
+      "epoch": 1.789445196211096,
+      "grad_norm": 0.21890240907669067,
+      "learning_rate": 4.188194781431843e-06,
+      "loss": 0.3476,
+      "step": 2204
+    },
+    {
+      "epoch": 1.790257104194858,
+      "grad_norm": 0.20694489777088165,
+      "learning_rate": 4.18353082326288e-06,
+      "loss": 0.324,
+      "step": 2205
+    },
+    {
+      "epoch": 1.7910690121786197,
+      "grad_norm": 0.23832136392593384,
+      "learning_rate": 4.178867594851849e-06,
+      "loss": 0.3771,
+      "step": 2206
+    },
+    {
+      "epoch": 1.7918809201623818,
+      "grad_norm": 0.208040252327919,
+      "learning_rate": 4.1742051003667315e-06,
+      "loss": 0.3265,
+      "step": 2207
+    },
+    {
+      "epoch": 1.7926928281461434,
+      "grad_norm": 0.2409697026014328,
+      "learning_rate": 4.169543343974852e-06,
+      "loss": 0.3281,
+      "step": 2208
+    },
+    {
+      "epoch": 1.7935047361299052,
+      "grad_norm": 0.23674654960632324,
+      "learning_rate": 4.164882329842876e-06,
+      "loss": 0.3964,
+      "step": 2209
+    },
+    {
+      "epoch": 1.794316644113667,
+      "grad_norm": 0.22860224545001984,
+      "learning_rate": 4.160222062136805e-06,
+      "loss": 0.4014,
+      "step": 2210
+    },
+    {
+      "epoch": 1.795128552097429,
+      "grad_norm": 0.1965637356042862,
+      "learning_rate": 4.1555625450219725e-06,
+      "loss": 0.3775,
+      "step": 2211
+    },
+    {
+      "epoch": 1.7959404600811908,
+      "grad_norm": 0.21507620811462402,
+      "learning_rate": 4.150903782663045e-06,
+      "loss": 0.3048,
+      "step": 2212
+    },
+    {
+      "epoch": 1.7967523680649526,
+      "grad_norm": 0.24203522503376007,
+      "learning_rate": 4.146245779224011e-06,
+      "loss": 0.3964,
+      "step": 2213
+    },
+    {
+      "epoch": 1.7975642760487145,
+      "grad_norm": 0.21548448503017426,
+      "learning_rate": 4.141588538868179e-06,
+      "loss": 0.3468,
+      "step": 2214
+    },
+    {
+      "epoch": 1.7983761840324763,
+      "grad_norm": 0.19690640270709991,
+      "learning_rate": 4.13693206575818e-06,
+      "loss": 0.3366,
+      "step": 2215
+    },
+    {
+      "epoch": 1.7991880920162382,
+      "grad_norm": 0.22393326461315155,
+      "learning_rate": 4.132276364055956e-06,
+      "loss": 0.3753,
+      "step": 2216
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2627128064632416,
+      "learning_rate": 4.127621437922761e-06,
+      "loss": 0.3742,
+      "step": 2217
+    },
+    {
+      "epoch": 1.800811907983762,
+      "grad_norm": 0.2122894674539566,
+      "learning_rate": 4.122967291519155e-06,
+      "loss": 0.3469,
+      "step": 2218
+    },
+    {
+      "epoch": 1.8016238159675235,
+      "grad_norm": 0.1910112053155899,
+      "learning_rate": 4.118313929005002e-06,
+      "loss": 0.3584,
+      "step": 2219
+    },
+    {
+      "epoch": 1.8024357239512856,
+      "grad_norm": 0.22993305325508118,
+      "learning_rate": 4.113661354539464e-06,
+      "loss": 0.3696,
+      "step": 2220
+    },
+    {
+      "epoch": 1.8032476319350472,
+      "grad_norm": 0.22561129927635193,
+      "learning_rate": 4.109009572280999e-06,
+      "loss": 0.3728,
+      "step": 2221
+    },
+    {
+      "epoch": 1.8040595399188093,
+      "grad_norm": 0.24680252373218536,
+      "learning_rate": 4.1043585863873595e-06,
+      "loss": 0.3931,
+      "step": 2222
+    },
+    {
+      "epoch": 1.804871447902571,
+      "grad_norm": 0.2240302711725235,
+      "learning_rate": 4.09970840101558e-06,
+      "loss": 0.3393,
+      "step": 2223
+    },
+    {
+      "epoch": 1.805683355886333,
+      "grad_norm": 0.22259902954101562,
+      "learning_rate": 4.095059020321988e-06,
+      "loss": 0.3388,
+      "step": 2224
+    },
+    {
+      "epoch": 1.8064952638700946,
+      "grad_norm": 0.2160307765007019,
+      "learning_rate": 4.090410448462186e-06,
+      "loss": 0.3439,
+      "step": 2225
+    },
+    {
+      "epoch": 1.8073071718538567,
+      "grad_norm": 0.22399340569972992,
+      "learning_rate": 4.085762689591054e-06,
+      "loss": 0.331,
+      "step": 2226
+    },
+    {
+      "epoch": 1.8081190798376183,
+      "grad_norm": 0.23935823142528534,
+      "learning_rate": 4.081115747862747e-06,
+      "loss": 0.3715,
+      "step": 2227
+    },
+    {
+      "epoch": 1.8089309878213804,
+      "grad_norm": 0.23225374519824982,
+      "learning_rate": 4.076469627430689e-06,
+      "loss": 0.3808,
+      "step": 2228
+    },
+    {
+      "epoch": 1.809742895805142,
+      "grad_norm": 0.2082996964454651,
+      "learning_rate": 4.071824332447568e-06,
+      "loss": 0.3496,
+      "step": 2229
+    },
+    {
+      "epoch": 1.8105548037889039,
+      "grad_norm": 0.24279405176639557,
+      "learning_rate": 4.067179867065339e-06,
+      "loss": 0.3508,
+      "step": 2230
+    },
+    {
+      "epoch": 1.8113667117726657,
+      "grad_norm": 0.22993583977222443,
+      "learning_rate": 4.0625362354352105e-06,
+      "loss": 0.3406,
+      "step": 2231
+    },
+    {
+      "epoch": 1.8121786197564276,
+      "grad_norm": 0.22104690968990326,
+      "learning_rate": 4.0578934417076486e-06,
+      "loss": 0.3474,
+      "step": 2232
+    },
+    {
+      "epoch": 1.8129905277401894,
+      "grad_norm": 0.2338695526123047,
+      "learning_rate": 4.053251490032368e-06,
+      "loss": 0.3793,
+      "step": 2233
+    },
+    {
+      "epoch": 1.8138024357239513,
+      "grad_norm": 0.22883456945419312,
+      "learning_rate": 4.048610384558335e-06,
+      "loss": 0.369,
+      "step": 2234
+    },
+    {
+      "epoch": 1.8146143437077131,
+      "grad_norm": 0.22854796051979065,
+      "learning_rate": 4.043970129433757e-06,
+      "loss": 0.3936,
+      "step": 2235
+    },
+    {
+      "epoch": 1.815426251691475,
+      "grad_norm": 0.20791934430599213,
+      "learning_rate": 4.039330728806079e-06,
+      "loss": 0.3409,
+      "step": 2236
+    },
+    {
+      "epoch": 1.8162381596752368,
+      "grad_norm": 0.21706539392471313,
+      "learning_rate": 4.034692186821986e-06,
+      "loss": 0.3424,
+      "step": 2237
+    },
+    {
+      "epoch": 1.8170500676589987,
+      "grad_norm": 0.22956059873104095,
+      "learning_rate": 4.030054507627395e-06,
+      "loss": 0.3583,
+      "step": 2238
+    },
+    {
+      "epoch": 1.8178619756427605,
+      "grad_norm": 0.22232778370380402,
+      "learning_rate": 4.0254176953674505e-06,
+      "loss": 0.3295,
+      "step": 2239
+    },
+    {
+      "epoch": 1.8186738836265224,
+      "grad_norm": 0.2492658793926239,
+      "learning_rate": 4.020781754186523e-06,
+      "loss": 0.3549,
+      "step": 2240
+    },
+    {
+      "epoch": 1.8194857916102842,
+      "grad_norm": 0.2638021409511566,
+      "learning_rate": 4.016146688228203e-06,
+      "loss": 0.3712,
+      "step": 2241
+    },
+    {
+      "epoch": 1.8202976995940459,
+      "grad_norm": 0.2248041182756424,
+      "learning_rate": 4.011512501635301e-06,
+      "loss": 0.3214,
+      "step": 2242
+    },
+    {
+      "epoch": 1.821109607577808,
+      "grad_norm": 0.2544277310371399,
+      "learning_rate": 4.00687919854984e-06,
+      "loss": 0.3832,
+      "step": 2243
+    },
+    {
+      "epoch": 1.8219215155615696,
+      "grad_norm": 0.24470742046833038,
+      "learning_rate": 4.002246783113054e-06,
+      "loss": 0.3706,
+      "step": 2244
+    },
+    {
+      "epoch": 1.8227334235453316,
+      "grad_norm": 0.20484289526939392,
+      "learning_rate": 3.997615259465382e-06,
+      "loss": 0.319,
+      "step": 2245
+    },
+    {
+      "epoch": 1.8235453315290933,
+      "grad_norm": 0.21204587817192078,
+      "learning_rate": 3.992984631746469e-06,
+      "loss": 0.3637,
+      "step": 2246
+    },
+    {
+      "epoch": 1.8243572395128553,
+      "grad_norm": 0.22727376222610474,
+      "learning_rate": 3.988354904095157e-06,
+      "loss": 0.3203,
+      "step": 2247
+    },
+    {
+      "epoch": 1.825169147496617,
+      "grad_norm": 0.2591909170150757,
+      "learning_rate": 3.983726080649484e-06,
+      "loss": 0.3725,
+      "step": 2248
+    },
+    {
+      "epoch": 1.825981055480379,
+      "grad_norm": 0.23672711849212646,
+      "learning_rate": 3.979098165546679e-06,
+      "loss": 0.3655,
+      "step": 2249
+    },
+    {
+      "epoch": 1.8267929634641407,
+      "grad_norm": 0.5230373740196228,
+      "learning_rate": 3.974471162923162e-06,
+      "loss": 0.3552,
+      "step": 2250
+    },
+    {
+      "epoch": 1.8276048714479027,
+      "grad_norm": 0.25595054030418396,
+      "learning_rate": 3.969845076914532e-06,
+      "loss": 0.3314,
+      "step": 2251
+    },
+    {
+      "epoch": 1.8284167794316644,
+      "grad_norm": 0.22569827735424042,
+      "learning_rate": 3.965219911655575e-06,
+      "loss": 0.3578,
+      "step": 2252
+    },
+    {
+      "epoch": 1.8292286874154262,
+      "grad_norm": 0.22836364805698395,
+      "learning_rate": 3.96059567128025e-06,
+      "loss": 0.3414,
+      "step": 2253
+    },
+    {
+      "epoch": 1.830040595399188,
+      "grad_norm": 0.2391643077135086,
+      "learning_rate": 3.9559723599216905e-06,
+      "loss": 0.3503,
+      "step": 2254
+    },
+    {
+      "epoch": 1.83085250338295,
+      "grad_norm": 0.24605396389961243,
+      "learning_rate": 3.951349981712199e-06,
+      "loss": 0.3385,
+      "step": 2255
+    },
+    {
+      "epoch": 1.8316644113667118,
+      "grad_norm": 0.2345791608095169,
+      "learning_rate": 3.946728540783247e-06,
+      "loss": 0.3688,
+      "step": 2256
+    },
+    {
+      "epoch": 1.8324763193504736,
+      "grad_norm": 0.23470529913902283,
+      "learning_rate": 3.942108041265463e-06,
+      "loss": 0.3501,
+      "step": 2257
+    },
+    {
+      "epoch": 1.8332882273342355,
+      "grad_norm": 0.24500544369220734,
+      "learning_rate": 3.9374884872886394e-06,
+      "loss": 0.4068,
+      "step": 2258
+    },
+    {
+      "epoch": 1.8341001353179973,
+      "grad_norm": 0.23498539626598358,
+      "learning_rate": 3.9328698829817195e-06,
+      "loss": 0.3642,
+      "step": 2259
+    },
+    {
+      "epoch": 1.8349120433017592,
+      "grad_norm": 0.21185819804668427,
+      "learning_rate": 3.928252232472799e-06,
+      "loss": 0.3359,
+      "step": 2260
+    },
+    {
+      "epoch": 1.835723951285521,
+      "grad_norm": 0.2537451684474945,
+      "learning_rate": 3.923635539889121e-06,
+      "loss": 0.3606,
+      "step": 2261
+    },
+    {
+      "epoch": 1.8365358592692829,
+      "grad_norm": 0.23583146929740906,
+      "learning_rate": 3.919019809357075e-06,
+      "loss": 0.33,
+      "step": 2262
+    },
+    {
+      "epoch": 1.8373477672530445,
+      "grad_norm": 0.22639107704162598,
+      "learning_rate": 3.914405045002186e-06,
+      "loss": 0.3403,
+      "step": 2263
+    },
+    {
+      "epoch": 1.8381596752368066,
+      "grad_norm": 0.214708611369133,
+      "learning_rate": 3.9097912509491195e-06,
+      "loss": 0.3543,
+      "step": 2264
+    },
+    {
+      "epoch": 1.8389715832205682,
+      "grad_norm": 0.2307552695274353,
+      "learning_rate": 3.905178431321672e-06,
+      "loss": 0.3887,
+      "step": 2265
+    },
+    {
+      "epoch": 1.8397834912043303,
+      "grad_norm": 0.21002420783042908,
+      "learning_rate": 3.9005665902427695e-06,
+      "loss": 0.3403,
+      "step": 2266
+    },
+    {
+      "epoch": 1.840595399188092,
+      "grad_norm": 0.21478334069252014,
+      "learning_rate": 3.895955731834461e-06,
+      "loss": 0.3558,
+      "step": 2267
+    },
+    {
+      "epoch": 1.841407307171854,
+      "grad_norm": 0.2038838118314743,
+      "learning_rate": 3.8913458602179215e-06,
+      "loss": 0.3429,
+      "step": 2268
+    },
+    {
+      "epoch": 1.8422192151556156,
+      "grad_norm": 0.23084427416324615,
+      "learning_rate": 3.88673697951344e-06,
+      "loss": 0.385,
+      "step": 2269
+    },
+    {
+      "epoch": 1.8430311231393777,
+      "grad_norm": 0.21960021555423737,
+      "learning_rate": 3.882129093840423e-06,
+      "loss": 0.3493,
+      "step": 2270
+    },
+    {
+      "epoch": 1.8438430311231393,
+      "grad_norm": 0.23820167779922485,
+      "learning_rate": 3.8775222073173844e-06,
+      "loss": 0.3766,
+      "step": 2271
+    },
+    {
+      "epoch": 1.8446549391069014,
+      "grad_norm": 0.21245934069156647,
+      "learning_rate": 3.872916324061949e-06,
+      "loss": 0.3255,
+      "step": 2272
+    },
+    {
+      "epoch": 1.845466847090663,
+      "grad_norm": 0.2356433868408203,
+      "learning_rate": 3.8683114481908415e-06,
+      "loss": 0.3735,
+      "step": 2273
+    },
+    {
+      "epoch": 1.8462787550744248,
+      "grad_norm": 0.2275618463754654,
+      "learning_rate": 3.8637075838198864e-06,
+      "loss": 0.3353,
+      "step": 2274
+    },
+    {
+      "epoch": 1.8470906630581867,
+      "grad_norm": 0.19161613285541534,
+      "learning_rate": 3.859104735064008e-06,
+      "loss": 0.321,
+      "step": 2275
+    },
+    {
+      "epoch": 1.8479025710419485,
+      "grad_norm": 0.22274288535118103,
+      "learning_rate": 3.854502906037217e-06,
+      "loss": 0.3917,
+      "step": 2276
+    },
+    {
+      "epoch": 1.8487144790257104,
+      "grad_norm": 0.2100948691368103,
+      "learning_rate": 3.849902100852618e-06,
+      "loss": 0.3431,
+      "step": 2277
+    },
+    {
+      "epoch": 1.8495263870094723,
+      "grad_norm": 0.23240548372268677,
+      "learning_rate": 3.845302323622398e-06,
+      "loss": 0.3799,
+      "step": 2278
+    },
+    {
+      "epoch": 1.850338294993234,
+      "grad_norm": 0.1893819272518158,
+      "learning_rate": 3.8407035784578226e-06,
+      "loss": 0.309,
+      "step": 2279
+    },
+    {
+      "epoch": 1.851150202976996,
+      "grad_norm": 0.24205568432807922,
+      "learning_rate": 3.8361058694692414e-06,
+      "loss": 0.3784,
+      "step": 2280
+    },
+    {
+      "epoch": 1.8519621109607578,
+      "grad_norm": 0.22100558876991272,
+      "learning_rate": 3.83150920076607e-06,
+      "loss": 0.3558,
+      "step": 2281
+    },
+    {
+      "epoch": 1.8527740189445197,
+      "grad_norm": 0.20714813470840454,
+      "learning_rate": 3.826913576456803e-06,
+      "loss": 0.3783,
+      "step": 2282
+    },
+    {
+      "epoch": 1.8535859269282815,
+      "grad_norm": 0.2086331695318222,
+      "learning_rate": 3.822319000648995e-06,
+      "loss": 0.3565,
+      "step": 2283
+    },
+    {
+      "epoch": 1.8543978349120434,
+      "grad_norm": 0.22080780565738678,
+      "learning_rate": 3.817725477449265e-06,
+      "loss": 0.3621,
+      "step": 2284
+    },
+    {
+      "epoch": 1.8552097428958052,
+      "grad_norm": 0.19861924648284912,
+      "learning_rate": 3.8131330109632925e-06,
+      "loss": 0.3509,
+      "step": 2285
+    },
+    {
+      "epoch": 1.8560216508795668,
+      "grad_norm": 0.20991556346416473,
+      "learning_rate": 3.8085416052958107e-06,
+      "loss": 0.3752,
+      "step": 2286
+    },
+    {
+      "epoch": 1.856833558863329,
+      "grad_norm": 0.21120166778564453,
+      "learning_rate": 3.8039512645506055e-06,
+      "loss": 0.356,
+      "step": 2287
+    },
+    {
+      "epoch": 1.8576454668470905,
+      "grad_norm": 0.21484355628490448,
+      "learning_rate": 3.7993619928305113e-06,
+      "loss": 0.3528,
+      "step": 2288
+    },
+    {
+      "epoch": 1.8584573748308526,
+      "grad_norm": 0.20564588904380798,
+      "learning_rate": 3.7947737942374065e-06,
+      "loss": 0.3389,
+      "step": 2289
+    },
+    {
+      "epoch": 1.8592692828146142,
+      "grad_norm": 0.20879849791526794,
+      "learning_rate": 3.7901866728722102e-06,
+      "loss": 0.3662,
+      "step": 2290
+    },
+    {
+      "epoch": 1.8600811907983763,
+      "grad_norm": 0.2058820277452469,
+      "learning_rate": 3.7856006328348805e-06,
+      "loss": 0.354,
+      "step": 2291
+    },
+    {
+      "epoch": 1.860893098782138,
+      "grad_norm": 0.20740748941898346,
+      "learning_rate": 3.7810156782244055e-06,
+      "loss": 0.3406,
+      "step": 2292
+    },
+    {
+      "epoch": 1.8617050067659,
+      "grad_norm": 0.21630573272705078,
+      "learning_rate": 3.7764318131388067e-06,
+      "loss": 0.3666,
+      "step": 2293
+    },
+    {
+      "epoch": 1.8625169147496616,
+      "grad_norm": 0.21548645198345184,
+      "learning_rate": 3.7718490416751297e-06,
+      "loss": 0.3421,
+      "step": 2294
+    },
+    {
+      "epoch": 1.8633288227334237,
+      "grad_norm": 0.2062557488679886,
+      "learning_rate": 3.7672673679294436e-06,
+      "loss": 0.347,
+      "step": 2295
+    },
+    {
+      "epoch": 1.8641407307171853,
+      "grad_norm": 0.22425541281700134,
+      "learning_rate": 3.762686795996835e-06,
+      "loss": 0.3693,
+      "step": 2296
+    },
+    {
+      "epoch": 1.8649526387009472,
+      "grad_norm": 0.21054182946681976,
+      "learning_rate": 3.7581073299714077e-06,
+      "loss": 0.3597,
+      "step": 2297
+    },
+    {
+      "epoch": 1.865764546684709,
+      "grad_norm": 0.20903700590133667,
+      "learning_rate": 3.7535289739462757e-06,
+      "loss": 0.3825,
+      "step": 2298
+    },
+    {
+      "epoch": 1.8665764546684709,
+      "grad_norm": 0.19771204888820648,
+      "learning_rate": 3.7489517320135583e-06,
+      "loss": 0.339,
+      "step": 2299
+    },
+    {
+      "epoch": 1.8673883626522327,
+      "grad_norm": 0.2012229710817337,
+      "learning_rate": 3.7443756082643866e-06,
+      "loss": 0.3178,
+      "step": 2300
+    },
+    {
+      "epoch": 1.8682002706359946,
+      "grad_norm": 0.23623476922512054,
+      "learning_rate": 3.739800606788886e-06,
+      "loss": 0.395,
+      "step": 2301
+    },
+    {
+      "epoch": 1.8690121786197564,
+      "grad_norm": 0.18302761018276215,
+      "learning_rate": 3.735226731676179e-06,
+      "loss": 0.308,
+      "step": 2302
+    },
+    {
+      "epoch": 1.8698240866035183,
+      "grad_norm": 0.21334075927734375,
+      "learning_rate": 3.7306539870143853e-06,
+      "loss": 0.3974,
+      "step": 2303
+    },
+    {
+      "epoch": 1.8706359945872801,
+      "grad_norm": 0.21827538311481476,
+      "learning_rate": 3.7260823768906096e-06,
+      "loss": 0.3338,
+      "step": 2304
+    },
+    {
+      "epoch": 1.871447902571042,
+      "grad_norm": 0.19939731061458588,
+      "learning_rate": 3.7215119053909444e-06,
+      "loss": 0.3276,
+      "step": 2305
+    },
+    {
+      "epoch": 1.8722598105548038,
+      "grad_norm": 0.20477169752120972,
+      "learning_rate": 3.7169425766004653e-06,
+      "loss": 0.3499,
+      "step": 2306
+    },
+    {
+      "epoch": 1.8730717185385655,
+      "grad_norm": 0.22647078335285187,
+      "learning_rate": 3.7123743946032263e-06,
+      "loss": 0.3521,
+      "step": 2307
+    },
+    {
+      "epoch": 1.8738836265223275,
+      "grad_norm": 0.22722691297531128,
+      "learning_rate": 3.7078073634822565e-06,
+      "loss": 0.3937,
+      "step": 2308
+    },
+    {
+      "epoch": 1.8746955345060892,
+      "grad_norm": 0.20447053015232086,
+      "learning_rate": 3.7032414873195544e-06,
+      "loss": 0.3096,
+      "step": 2309
+    },
+    {
+      "epoch": 1.8755074424898512,
+      "grad_norm": 0.21931712329387665,
+      "learning_rate": 3.698676770196089e-06,
+      "loss": 0.3447,
+      "step": 2310
+    },
+    {
+      "epoch": 1.8763193504736129,
+      "grad_norm": 0.2144707590341568,
+      "learning_rate": 3.6941132161917916e-06,
+      "loss": 0.3577,
+      "step": 2311
+    },
+    {
+      "epoch": 1.877131258457375,
+      "grad_norm": 0.2142912596464157,
+      "learning_rate": 3.6895508293855546e-06,
+      "loss": 0.3425,
+      "step": 2312
+    },
+    {
+      "epoch": 1.8779431664411366,
+      "grad_norm": 0.21552814543247223,
+      "learning_rate": 3.6849896138552265e-06,
+      "loss": 0.3471,
+      "step": 2313
+    },
+    {
+      "epoch": 1.8787550744248986,
+      "grad_norm": 0.2169177532196045,
+      "learning_rate": 3.68042957367761e-06,
+      "loss": 0.3364,
+      "step": 2314
+    },
+    {
+      "epoch": 1.8795669824086603,
+      "grad_norm": 0.20932993292808533,
+      "learning_rate": 3.6758707129284556e-06,
+      "loss": 0.368,
+      "step": 2315
+    },
+    {
+      "epoch": 1.8803788903924223,
+      "grad_norm": 0.218265563249588,
+      "learning_rate": 3.6713130356824615e-06,
+      "loss": 0.326,
+      "step": 2316
+    },
+    {
+      "epoch": 1.881190798376184,
+      "grad_norm": 0.2042979747056961,
+      "learning_rate": 3.666756546013267e-06,
+      "loss": 0.3361,
+      "step": 2317
+    },
+    {
+      "epoch": 1.8820027063599458,
+      "grad_norm": 0.236994668841362,
+      "learning_rate": 3.6622012479934505e-06,
+      "loss": 0.3787,
+      "step": 2318
+    },
+    {
+      "epoch": 1.8828146143437077,
+      "grad_norm": 0.19558244943618774,
+      "learning_rate": 3.6576471456945244e-06,
+      "loss": 0.3658,
+      "step": 2319
+    },
+    {
+      "epoch": 1.8836265223274695,
+      "grad_norm": 0.1938365250825882,
+      "learning_rate": 3.653094243186933e-06,
+      "loss": 0.3395,
+      "step": 2320
+    },
+    {
+      "epoch": 1.8844384303112314,
+      "grad_norm": 0.21914272010326385,
+      "learning_rate": 3.648542544540049e-06,
+      "loss": 0.3499,
+      "step": 2321
+    },
+    {
+      "epoch": 1.8852503382949932,
+      "grad_norm": 0.21327003836631775,
+      "learning_rate": 3.6439920538221675e-06,
+      "loss": 0.3427,
+      "step": 2322
+    },
+    {
+      "epoch": 1.886062246278755,
+      "grad_norm": 0.2073378711938858,
+      "learning_rate": 3.639442775100505e-06,
+      "loss": 0.3964,
+      "step": 2323
+    },
+    {
+      "epoch": 1.886874154262517,
+      "grad_norm": 0.20922596752643585,
+      "learning_rate": 3.634894712441193e-06,
+      "loss": 0.3592,
+      "step": 2324
+    },
+    {
+      "epoch": 1.8876860622462788,
+      "grad_norm": 0.1993296593427658,
+      "learning_rate": 3.630347869909281e-06,
+      "loss": 0.3279,
+      "step": 2325
+    },
+    {
+      "epoch": 1.8884979702300406,
+      "grad_norm": 0.21011541783809662,
+      "learning_rate": 3.6258022515687215e-06,
+      "loss": 0.3612,
+      "step": 2326
+    },
+    {
+      "epoch": 1.8893098782138025,
+      "grad_norm": 0.21431997418403625,
+      "learning_rate": 3.6212578614823767e-06,
+      "loss": 0.3558,
+      "step": 2327
+    },
+    {
+      "epoch": 1.8901217861975643,
+      "grad_norm": 0.21559761464595795,
+      "learning_rate": 3.616714703712011e-06,
+      "loss": 0.3215,
+      "step": 2328
+    },
+    {
+      "epoch": 1.8909336941813262,
+      "grad_norm": 0.20969070494174957,
+      "learning_rate": 3.6121727823182854e-06,
+      "loss": 0.3954,
+      "step": 2329
+    },
+    {
+      "epoch": 1.8917456021650878,
+      "grad_norm": 0.21176838874816895,
+      "learning_rate": 3.6076321013607575e-06,
+      "loss": 0.3864,
+      "step": 2330
+    },
+    {
+      "epoch": 1.8925575101488499,
+      "grad_norm": 0.2141079306602478,
+      "learning_rate": 3.6030926648978748e-06,
+      "loss": 0.3105,
+      "step": 2331
+    },
+    {
+      "epoch": 1.8933694181326115,
+      "grad_norm": 0.2425554245710373,
+      "learning_rate": 3.598554476986974e-06,
+      "loss": 0.3839,
+      "step": 2332
+    },
+    {
+      "epoch": 1.8941813261163736,
+      "grad_norm": 0.18451888859272003,
+      "learning_rate": 3.594017541684275e-06,
+      "loss": 0.3138,
+      "step": 2333
+    },
+    {
+      "epoch": 1.8949932341001352,
+      "grad_norm": 0.22819893062114716,
+      "learning_rate": 3.589481863044878e-06,
+      "loss": 0.3546,
+      "step": 2334
+    },
+    {
+      "epoch": 1.8958051420838973,
+      "grad_norm": 0.21030019223690033,
+      "learning_rate": 3.5849474451227607e-06,
+      "loss": 0.3626,
+      "step": 2335
+    },
+    {
+      "epoch": 1.896617050067659,
+      "grad_norm": 0.2025078684091568,
+      "learning_rate": 3.5804142919707734e-06,
+      "loss": 0.3508,
+      "step": 2336
+    },
+    {
+      "epoch": 1.897428958051421,
+      "grad_norm": 0.20038406550884247,
+      "learning_rate": 3.5758824076406358e-06,
+      "loss": 0.334,
+      "step": 2337
+    },
+    {
+      "epoch": 1.8982408660351826,
+      "grad_norm": 0.24745076894760132,
+      "learning_rate": 3.5713517961829314e-06,
+      "loss": 0.4073,
+      "step": 2338
+    },
+    {
+      "epoch": 1.8990527740189447,
+      "grad_norm": 0.23628047108650208,
+      "learning_rate": 3.5668224616471136e-06,
+      "loss": 0.3431,
+      "step": 2339
+    },
+    {
+      "epoch": 1.8998646820027063,
+      "grad_norm": 0.23678430914878845,
+      "learning_rate": 3.562294408081486e-06,
+      "loss": 0.3407,
+      "step": 2340
+    },
+    {
+      "epoch": 1.9006765899864682,
+      "grad_norm": 0.19295133650302887,
+      "learning_rate": 3.5577676395332116e-06,
+      "loss": 0.31,
+      "step": 2341
+    },
+    {
+      "epoch": 1.90148849797023,
+      "grad_norm": 0.23327584564685822,
+      "learning_rate": 3.5532421600483035e-06,
+      "loss": 0.3889,
+      "step": 2342
+    },
+    {
+      "epoch": 1.9023004059539919,
+      "grad_norm": 0.22439667582511902,
+      "learning_rate": 3.5487179736716233e-06,
+      "loss": 0.3644,
+      "step": 2343
+    },
+    {
+      "epoch": 1.9031123139377537,
+      "grad_norm": 0.21219919621944427,
+      "learning_rate": 3.5441950844468754e-06,
+      "loss": 0.3395,
+      "step": 2344
+    },
+    {
+      "epoch": 1.9039242219215156,
+      "grad_norm": 0.23101094365119934,
+      "learning_rate": 3.539673496416607e-06,
+      "loss": 0.426,
+      "step": 2345
+    },
+    {
+      "epoch": 1.9047361299052774,
+      "grad_norm": 0.2137201428413391,
+      "learning_rate": 3.5351532136222012e-06,
+      "loss": 0.3748,
+      "step": 2346
+    },
+    {
+      "epoch": 1.9055480378890393,
+      "grad_norm": 0.21368184685707092,
+      "learning_rate": 3.5306342401038727e-06,
+      "loss": 0.35,
+      "step": 2347
+    },
+    {
+      "epoch": 1.906359945872801,
+      "grad_norm": 0.18339461088180542,
+      "learning_rate": 3.52611657990067e-06,
+      "loss": 0.2873,
+      "step": 2348
+    },
+    {
+      "epoch": 1.907171853856563,
+      "grad_norm": 0.22756202518939972,
+      "learning_rate": 3.5216002370504656e-06,
+      "loss": 0.3952,
+      "step": 2349
+    },
+    {
+      "epoch": 1.9079837618403248,
+      "grad_norm": 0.20742341876029968,
+      "learning_rate": 3.5170852155899517e-06,
+      "loss": 0.3262,
+      "step": 2350
+    },
+    {
+      "epoch": 1.9087956698240864,
+      "grad_norm": 0.2005053609609604,
+      "learning_rate": 3.512571519554645e-06,
+      "loss": 0.3497,
+      "step": 2351
+    },
+    {
+      "epoch": 1.9096075778078485,
+      "grad_norm": 0.21324387192726135,
+      "learning_rate": 3.5080591529788736e-06,
+      "loss": 0.3269,
+      "step": 2352
+    },
+    {
+      "epoch": 1.9104194857916101,
+      "grad_norm": 0.20170335471630096,
+      "learning_rate": 3.5035481198957794e-06,
+      "loss": 0.3797,
+      "step": 2353
+    },
+    {
+      "epoch": 1.9112313937753722,
+      "grad_norm": 0.22454455494880676,
+      "learning_rate": 3.4990384243373105e-06,
+      "loss": 0.339,
+      "step": 2354
+    },
+    {
+      "epoch": 1.9120433017591338,
+      "grad_norm": 0.22518198192119598,
+      "learning_rate": 3.494530070334221e-06,
+      "loss": 0.3669,
+      "step": 2355
+    },
+    {
+      "epoch": 1.912855209742896,
+      "grad_norm": 0.21244165301322937,
+      "learning_rate": 3.490023061916065e-06,
+      "loss": 0.372,
+      "step": 2356
+    },
+    {
+      "epoch": 1.9136671177266575,
+      "grad_norm": 0.212235227227211,
+      "learning_rate": 3.4855174031111947e-06,
+      "loss": 0.3387,
+      "step": 2357
+    },
+    {
+      "epoch": 1.9144790257104196,
+      "grad_norm": 0.21469572186470032,
+      "learning_rate": 3.481013097946756e-06,
+      "loss": 0.3516,
+      "step": 2358
+    },
+    {
+      "epoch": 1.9152909336941812,
+      "grad_norm": 0.21541748940944672,
+      "learning_rate": 3.4765101504486842e-06,
+      "loss": 0.3622,
+      "step": 2359
+    },
+    {
+      "epoch": 1.9161028416779433,
+      "grad_norm": 0.20494012534618378,
+      "learning_rate": 3.4720085646417025e-06,
+      "loss": 0.3858,
+      "step": 2360
+    },
+    {
+      "epoch": 1.916914749661705,
+      "grad_norm": 0.21332043409347534,
+      "learning_rate": 3.467508344549314e-06,
+      "loss": 0.3506,
+      "step": 2361
+    },
+    {
+      "epoch": 1.9177266576454668,
+      "grad_norm": 0.2103058397769928,
+      "learning_rate": 3.4630094941938054e-06,
+      "loss": 0.3503,
+      "step": 2362
+    },
+    {
+      "epoch": 1.9185385656292286,
+      "grad_norm": 0.20831827819347382,
+      "learning_rate": 3.458512017596235e-06,
+      "loss": 0.3483,
+      "step": 2363
+    },
+    {
+      "epoch": 1.9193504736129905,
+      "grad_norm": 0.20172883570194244,
+      "learning_rate": 3.454015918776437e-06,
+      "loss": 0.3638,
+      "step": 2364
+    },
+    {
+      "epoch": 1.9201623815967523,
+      "grad_norm": 0.20724555850028992,
+      "learning_rate": 3.449521201753011e-06,
+      "loss": 0.3712,
+      "step": 2365
+    },
+    {
+      "epoch": 1.9209742895805142,
+      "grad_norm": 0.2172815352678299,
+      "learning_rate": 3.445027870543323e-06,
+      "loss": 0.3301,
+      "step": 2366
+    },
+    {
+      "epoch": 1.921786197564276,
+      "grad_norm": 0.22412952780723572,
+      "learning_rate": 3.4405359291635008e-06,
+      "loss": 0.347,
+      "step": 2367
+    },
+    {
+      "epoch": 1.922598105548038,
+      "grad_norm": 0.2065112441778183,
+      "learning_rate": 3.436045381628429e-06,
+      "loss": 0.3415,
+      "step": 2368
+    },
+    {
+      "epoch": 1.9234100135317997,
+      "grad_norm": 0.2040354460477829,
+      "learning_rate": 3.431556231951747e-06,
+      "loss": 0.3148,
+      "step": 2369
+    },
+    {
+      "epoch": 1.9242219215155616,
+      "grad_norm": 0.1993257999420166,
+      "learning_rate": 3.4270684841458445e-06,
+      "loss": 0.3371,
+      "step": 2370
+    },
+    {
+      "epoch": 1.9250338294993234,
+      "grad_norm": 0.22844244539737701,
+      "learning_rate": 3.4225821422218585e-06,
+      "loss": 0.4007,
+      "step": 2371
+    },
+    {
+      "epoch": 1.9258457374830853,
+      "grad_norm": 0.20063476264476776,
+      "learning_rate": 3.418097210189669e-06,
+      "loss": 0.3405,
+      "step": 2372
+    },
+    {
+      "epoch": 1.9266576454668471,
+      "grad_norm": 0.2524063289165497,
+      "learning_rate": 3.4136136920578967e-06,
+      "loss": 0.394,
+      "step": 2373
+    },
+    {
+      "epoch": 1.9274695534506088,
+      "grad_norm": 0.2083909958600998,
+      "learning_rate": 3.4091315918338968e-06,
+      "loss": 0.3392,
+      "step": 2374
+    },
+    {
+      "epoch": 1.9282814614343708,
+      "grad_norm": 0.2099296748638153,
+      "learning_rate": 3.4046509135237593e-06,
+      "loss": 0.3492,
+      "step": 2375
+    },
+    {
+      "epoch": 1.9290933694181325,
+      "grad_norm": 0.20975758135318756,
+      "learning_rate": 3.400171661132299e-06,
+      "loss": 0.3576,
+      "step": 2376
+    },
+    {
+      "epoch": 1.9299052774018945,
+      "grad_norm": 0.24611753225326538,
+      "learning_rate": 3.395693838663065e-06,
+      "loss": 0.3741,
+      "step": 2377
+    },
+    {
+      "epoch": 1.9307171853856562,
+      "grad_norm": 0.2186264544725418,
+      "learning_rate": 3.3912174501183206e-06,
+      "loss": 0.3294,
+      "step": 2378
+    },
+    {
+      "epoch": 1.9315290933694182,
+      "grad_norm": 0.2212553173303604,
+      "learning_rate": 3.3867424994990483e-06,
+      "loss": 0.3549,
+      "step": 2379
+    },
+    {
+      "epoch": 1.9323410013531799,
+      "grad_norm": 0.20149734616279602,
+      "learning_rate": 3.3822689908049487e-06,
+      "loss": 0.3588,
+      "step": 2380
+    },
+    {
+      "epoch": 1.933152909336942,
+      "grad_norm": 0.2309603989124298,
+      "learning_rate": 3.37779692803443e-06,
+      "loss": 0.3671,
+      "step": 2381
+    },
+    {
+      "epoch": 1.9339648173207036,
+      "grad_norm": 0.22916977107524872,
+      "learning_rate": 3.373326315184612e-06,
+      "loss": 0.3583,
+      "step": 2382
+    },
+    {
+      "epoch": 1.9347767253044656,
+      "grad_norm": 0.20013582706451416,
+      "learning_rate": 3.368857156251314e-06,
+      "loss": 0.3601,
+      "step": 2383
+    },
+    {
+      "epoch": 1.9355886332882273,
+      "grad_norm": 0.19291295111179352,
+      "learning_rate": 3.364389455229059e-06,
+      "loss": 0.3535,
+      "step": 2384
+    },
+    {
+      "epoch": 1.9364005412719891,
+      "grad_norm": 0.2049330174922943,
+      "learning_rate": 3.359923216111067e-06,
+      "loss": 0.3361,
+      "step": 2385
+    },
+    {
+      "epoch": 1.937212449255751,
+      "grad_norm": 0.20975331962108612,
+      "learning_rate": 3.3554584428892488e-06,
+      "loss": 0.3512,
+      "step": 2386
+    },
+    {
+      "epoch": 1.9380243572395128,
+      "grad_norm": 0.20831237733364105,
+      "learning_rate": 3.3509951395542073e-06,
+      "loss": 0.3535,
+      "step": 2387
+    },
+    {
+      "epoch": 1.9388362652232747,
+      "grad_norm": 0.21285778284072876,
+      "learning_rate": 3.3465333100952306e-06,
+      "loss": 0.3417,
+      "step": 2388
+    },
+    {
+      "epoch": 1.9396481732070365,
+      "grad_norm": 0.1937015801668167,
+      "learning_rate": 3.342072958500291e-06,
+      "loss": 0.331,
+      "step": 2389
+    },
+    {
+      "epoch": 1.9404600811907984,
+      "grad_norm": 0.2231116145849228,
+      "learning_rate": 3.337614088756037e-06,
+      "loss": 0.3838,
+      "step": 2390
+    },
+    {
+      "epoch": 1.9412719891745602,
+      "grad_norm": 0.19913934171199799,
+      "learning_rate": 3.3331567048477946e-06,
+      "loss": 0.3545,
+      "step": 2391
+    },
+    {
+      "epoch": 1.942083897158322,
+      "grad_norm": 0.21456439793109894,
+      "learning_rate": 3.328700810759562e-06,
+      "loss": 0.3276,
+      "step": 2392
+    },
+    {
+      "epoch": 1.942895805142084,
+      "grad_norm": 0.23801031708717346,
+      "learning_rate": 3.324246410474006e-06,
+      "loss": 0.3253,
+      "step": 2393
+    },
+    {
+      "epoch": 1.9437077131258458,
+      "grad_norm": 0.23404885828495026,
+      "learning_rate": 3.319793507972454e-06,
+      "loss": 0.3754,
+      "step": 2394
+    },
+    {
+      "epoch": 1.9445196211096076,
+      "grad_norm": 0.22845235466957092,
+      "learning_rate": 3.3153421072349024e-06,
+      "loss": 0.3589,
+      "step": 2395
+    },
+    {
+      "epoch": 1.9453315290933695,
+      "grad_norm": 0.2114274799823761,
+      "learning_rate": 3.310892212239999e-06,
+      "loss": 0.3451,
+      "step": 2396
+    },
+    {
+      "epoch": 1.946143437077131,
+      "grad_norm": 0.20808006823062897,
+      "learning_rate": 3.306443826965049e-06,
+      "loss": 0.353,
+      "step": 2397
+    },
+    {
+      "epoch": 1.9469553450608932,
+      "grad_norm": 0.23001490533351898,
+      "learning_rate": 3.3019969553860056e-06,
+      "loss": 0.3671,
+      "step": 2398
+    },
+    {
+      "epoch": 1.9477672530446548,
+      "grad_norm": 0.23892471194267273,
+      "learning_rate": 3.2975516014774713e-06,
+      "loss": 0.3936,
+      "step": 2399
+    },
+    {
+      "epoch": 1.9485791610284169,
+      "grad_norm": 0.18842612206935883,
+      "learning_rate": 3.2931077692126912e-06,
+      "loss": 0.2913,
+      "step": 2400
+    },
+    {
+      "epoch": 1.9493910690121785,
+      "grad_norm": 0.21449343860149384,
+      "learning_rate": 3.2886654625635507e-06,
+      "loss": 0.4028,
+      "step": 2401
+    },
+    {
+      "epoch": 1.9502029769959406,
+      "grad_norm": 0.22380705177783966,
+      "learning_rate": 3.284224685500571e-06,
+      "loss": 0.3707,
+      "step": 2402
+    },
+    {
+      "epoch": 1.9510148849797022,
+      "grad_norm": 0.2277042120695114,
+      "learning_rate": 3.2797854419929064e-06,
+      "loss": 0.3692,
+      "step": 2403
+    },
+    {
+      "epoch": 1.9518267929634643,
+      "grad_norm": 0.21464522182941437,
+      "learning_rate": 3.2753477360083417e-06,
+      "loss": 0.358,
+      "step": 2404
+    },
+    {
+      "epoch": 1.952638700947226,
+      "grad_norm": 0.2019171416759491,
+      "learning_rate": 3.2709115715132868e-06,
+      "loss": 0.3211,
+      "step": 2405
+    },
+    {
+      "epoch": 1.953450608930988,
+      "grad_norm": 0.19943921267986298,
+      "learning_rate": 3.2664769524727712e-06,
+      "loss": 0.3727,
+      "step": 2406
+    },
+    {
+      "epoch": 1.9542625169147496,
+      "grad_norm": 0.2094951719045639,
+      "learning_rate": 3.2620438828504473e-06,
+      "loss": 0.3646,
+      "step": 2407
+    },
+    {
+      "epoch": 1.9550744248985115,
+      "grad_norm": 0.20186595618724823,
+      "learning_rate": 3.25761236660858e-06,
+      "loss": 0.3608,
+      "step": 2408
+    },
+    {
+      "epoch": 1.9558863328822733,
+      "grad_norm": 0.21997885406017303,
+      "learning_rate": 3.2531824077080466e-06,
+      "loss": 0.3555,
+      "step": 2409
+    },
+    {
+      "epoch": 1.9566982408660352,
+      "grad_norm": 0.23163960874080658,
+      "learning_rate": 3.2487540101083325e-06,
+      "loss": 0.3244,
+      "step": 2410
+    },
+    {
+      "epoch": 1.957510148849797,
+      "grad_norm": 0.24625183641910553,
+      "learning_rate": 3.244327177767527e-06,
+      "loss": 0.3912,
+      "step": 2411
+    },
+    {
+      "epoch": 1.9583220568335589,
+      "grad_norm": 0.1968807876110077,
+      "learning_rate": 3.23990191464232e-06,
+      "loss": 0.3694,
+      "step": 2412
+    },
+    {
+      "epoch": 1.9591339648173207,
+      "grad_norm": 0.18264277279376984,
+      "learning_rate": 3.2354782246880016e-06,
+      "loss": 0.3109,
+      "step": 2413
+    },
+    {
+      "epoch": 1.9599458728010826,
+      "grad_norm": 0.1957288682460785,
+      "learning_rate": 3.2310561118584526e-06,
+      "loss": 0.3321,
+      "step": 2414
+    },
+    {
+      "epoch": 1.9607577807848444,
+      "grad_norm": 0.21746955811977386,
+      "learning_rate": 3.226635580106147e-06,
+      "loss": 0.3763,
+      "step": 2415
+    },
+    {
+      "epoch": 1.9615696887686063,
+      "grad_norm": 0.22992181777954102,
+      "learning_rate": 3.222216633382142e-06,
+      "loss": 0.4158,
+      "step": 2416
+    },
+    {
+      "epoch": 1.9623815967523681,
+      "grad_norm": 0.20123282074928284,
+      "learning_rate": 3.2177992756360814e-06,
+      "loss": 0.3313,
+      "step": 2417
+    },
+    {
+      "epoch": 1.9631935047361297,
+      "grad_norm": 0.2200862318277359,
+      "learning_rate": 3.2133835108161875e-06,
+      "loss": 0.3271,
+      "step": 2418
+    },
+    {
+      "epoch": 1.9640054127198918,
+      "grad_norm": 0.22300367057323456,
+      "learning_rate": 3.2089693428692587e-06,
+      "loss": 0.3702,
+      "step": 2419
+    },
+    {
+      "epoch": 1.9648173207036534,
+      "grad_norm": 0.22419105470180511,
+      "learning_rate": 3.204556775740665e-06,
+      "loss": 0.3849,
+      "step": 2420
+    },
+    {
+      "epoch": 1.9656292286874155,
+      "grad_norm": 0.24026374518871307,
+      "learning_rate": 3.200145813374349e-06,
+      "loss": 0.3504,
+      "step": 2421
+    },
+    {
+      "epoch": 1.9664411366711771,
+      "grad_norm": 0.21357578039169312,
+      "learning_rate": 3.195736459712817e-06,
+      "loss": 0.3338,
+      "step": 2422
+    },
+    {
+      "epoch": 1.9672530446549392,
+      "grad_norm": 0.20934058725833893,
+      "learning_rate": 3.1913287186971353e-06,
+      "loss": 0.3286,
+      "step": 2423
+    },
+    {
+      "epoch": 1.9680649526387008,
+      "grad_norm": 0.2045539915561676,
+      "learning_rate": 3.186922594266932e-06,
+      "loss": 0.3405,
+      "step": 2424
+    },
+    {
+      "epoch": 1.968876860622463,
+      "grad_norm": 0.2566884458065033,
+      "learning_rate": 3.1825180903603903e-06,
+      "loss": 0.3881,
+      "step": 2425
+    },
+    {
+      "epoch": 1.9696887686062245,
+      "grad_norm": 0.2191406935453415,
+      "learning_rate": 3.178115210914242e-06,
+      "loss": 0.3725,
+      "step": 2426
+    },
+    {
+      "epoch": 1.9705006765899866,
+      "grad_norm": 0.19938130676746368,
+      "learning_rate": 3.1737139598637688e-06,
+      "loss": 0.3437,
+      "step": 2427
+    },
+    {
+      "epoch": 1.9713125845737482,
+      "grad_norm": 0.2257574200630188,
+      "learning_rate": 3.1693143411427984e-06,
+      "loss": 0.3564,
+      "step": 2428
+    },
+    {
+      "epoch": 1.97212449255751,
+      "grad_norm": 0.21032379567623138,
+      "learning_rate": 3.164916358683697e-06,
+      "loss": 0.3485,
+      "step": 2429
+    },
+    {
+      "epoch": 1.972936400541272,
+      "grad_norm": 0.20796795189380646,
+      "learning_rate": 3.160520016417369e-06,
+      "loss": 0.3465,
+      "step": 2430
+    },
+    {
+      "epoch": 1.9737483085250338,
+      "grad_norm": 0.21004754304885864,
+      "learning_rate": 3.1561253182732544e-06,
+      "loss": 0.3503,
+      "step": 2431
+    },
+    {
+      "epoch": 1.9745602165087957,
+      "grad_norm": 0.21836450695991516,
+      "learning_rate": 3.1517322681793215e-06,
+      "loss": 0.3645,
+      "step": 2432
+    },
+    {
+      "epoch": 1.9753721244925575,
+      "grad_norm": 0.21118243038654327,
+      "learning_rate": 3.1473408700620665e-06,
+      "loss": 0.3484,
+      "step": 2433
+    },
+    {
+      "epoch": 1.9761840324763194,
+      "grad_norm": 0.19537122547626495,
+      "learning_rate": 3.1429511278465104e-06,
+      "loss": 0.3451,
+      "step": 2434
+    },
+    {
+      "epoch": 1.9769959404600812,
+      "grad_norm": 0.21631085872650146,
+      "learning_rate": 3.138563045456191e-06,
+      "loss": 0.3408,
+      "step": 2435
+    },
+    {
+      "epoch": 1.977807848443843,
+      "grad_norm": 0.23328594863414764,
+      "learning_rate": 3.1341766268131657e-06,
+      "loss": 0.381,
+      "step": 2436
+    },
+    {
+      "epoch": 1.978619756427605,
+      "grad_norm": 0.2099274843931198,
+      "learning_rate": 3.1297918758380035e-06,
+      "loss": 0.3341,
+      "step": 2437
+    },
+    {
+      "epoch": 1.9794316644113668,
+      "grad_norm": 0.19503405690193176,
+      "learning_rate": 3.1254087964497816e-06,
+      "loss": 0.3427,
+      "step": 2438
+    },
+    {
+      "epoch": 1.9802435723951286,
+      "grad_norm": 0.21517369151115417,
+      "learning_rate": 3.121027392566086e-06,
+      "loss": 0.3701,
+      "step": 2439
+    },
+    {
+      "epoch": 1.9810554803788905,
+      "grad_norm": 0.1956460326910019,
+      "learning_rate": 3.1166476681030035e-06,
+      "loss": 0.3217,
+      "step": 2440
+    },
+    {
+      "epoch": 1.981867388362652,
+      "grad_norm": 0.19480697810649872,
+      "learning_rate": 3.1122696269751196e-06,
+      "loss": 0.3536,
+      "step": 2441
+    },
+    {
+      "epoch": 1.9826792963464142,
+      "grad_norm": 0.22723007202148438,
+      "learning_rate": 3.1078932730955152e-06,
+      "loss": 0.3778,
+      "step": 2442
+    },
+    {
+      "epoch": 1.9834912043301758,
+      "grad_norm": 0.2107202708721161,
+      "learning_rate": 3.103518610375764e-06,
+      "loss": 0.3552,
+      "step": 2443
+    },
+    {
+      "epoch": 1.9843031123139379,
+      "grad_norm": 0.20562255382537842,
+      "learning_rate": 3.099145642725927e-06,
+      "loss": 0.3618,
+      "step": 2444
+    },
+    {
+      "epoch": 1.9851150202976995,
+      "grad_norm": 0.2056446075439453,
+      "learning_rate": 3.0947743740545523e-06,
+      "loss": 0.3481,
+      "step": 2445
+    },
+    {
+      "epoch": 1.9859269282814616,
+      "grad_norm": 0.21181541681289673,
+      "learning_rate": 3.0904048082686655e-06,
+      "loss": 0.3522,
+      "step": 2446
+    },
+    {
+      "epoch": 1.9867388362652232,
+      "grad_norm": 0.22390051186084747,
+      "learning_rate": 3.0860369492737753e-06,
+      "loss": 0.3643,
+      "step": 2447
+    },
+    {
+      "epoch": 1.9875507442489853,
+      "grad_norm": 0.21579238772392273,
+      "learning_rate": 3.0816708009738605e-06,
+      "loss": 0.3559,
+      "step": 2448
+    },
+    {
+      "epoch": 1.9883626522327469,
+      "grad_norm": 0.1935131996870041,
+      "learning_rate": 3.0773063672713725e-06,
+      "loss": 0.3182,
+      "step": 2449
+    },
+    {
+      "epoch": 1.989174560216509,
+      "grad_norm": 0.19716985523700714,
+      "learning_rate": 3.0729436520672318e-06,
+      "loss": 0.3344,
+      "step": 2450
+    },
+    {
+      "epoch": 1.9899864682002706,
+      "grad_norm": 0.21238993108272552,
+      "learning_rate": 3.0685826592608193e-06,
+      "loss": 0.3782,
+      "step": 2451
+    },
+    {
+      "epoch": 1.9907983761840324,
+      "grad_norm": 0.227983757853508,
+      "learning_rate": 3.0642233927499786e-06,
+      "loss": 0.3624,
+      "step": 2452
+    },
+    {
+      "epoch": 1.9916102841677943,
+      "grad_norm": 0.19452013075351715,
+      "learning_rate": 3.0598658564310122e-06,
+      "loss": 0.3403,
+      "step": 2453
+    },
+    {
+      "epoch": 1.9924221921515561,
+      "grad_norm": 0.20309583842754364,
+      "learning_rate": 3.055510054198675e-06,
+      "loss": 0.3985,
+      "step": 2454
+    },
+    {
+      "epoch": 1.993234100135318,
+      "grad_norm": 0.19820711016654968,
+      "learning_rate": 3.0511559899461684e-06,
+      "loss": 0.3457,
+      "step": 2455
+    },
+    {
+      "epoch": 1.9940460081190798,
+      "grad_norm": 0.19590263068675995,
+      "learning_rate": 3.0468036675651447e-06,
+      "loss": 0.3273,
+      "step": 2456
+    },
+    {
+      "epoch": 1.9948579161028417,
+      "grad_norm": 0.21249519288539886,
+      "learning_rate": 3.0424530909456974e-06,
+      "loss": 0.3874,
+      "step": 2457
+    },
+    {
+      "epoch": 1.9956698240866035,
+      "grad_norm": 0.20713287591934204,
+      "learning_rate": 3.03810426397636e-06,
+      "loss": 0.3604,
+      "step": 2458
+    },
+    {
+      "epoch": 1.9964817320703654,
+      "grad_norm": 0.21494632959365845,
+      "learning_rate": 3.033757190544102e-06,
+      "loss": 0.3678,
+      "step": 2459
+    },
+    {
+      "epoch": 1.9972936400541272,
+      "grad_norm": 0.20102201402187347,
+      "learning_rate": 3.029411874534327e-06,
+      "loss": 0.3148,
+      "step": 2460
+    },
+    {
+      "epoch": 1.998105548037889,
+      "grad_norm": 0.2322881817817688,
+      "learning_rate": 3.0250683198308656e-06,
+      "loss": 0.3607,
+      "step": 2461
+    },
+    {
+      "epoch": 1.9989174560216507,
+      "grad_norm": 0.21304436028003693,
+      "learning_rate": 3.020726530315975e-06,
+      "loss": 0.361,
+      "step": 2462
+    },
+    {
+      "epoch": 1.9997293640054128,
+      "grad_norm": 0.2798072099685669,
+      "learning_rate": 3.016386509870335e-06,
+      "loss": 0.3745,
+      "step": 2463
+    },
+    {
+      "epoch": 2.0005412719891744,
+      "grad_norm": 0.2860136926174164,
+      "learning_rate": 3.0120482623730452e-06,
+      "loss": 0.4044,
+      "step": 2464
+    },
+    {
+      "epoch": 2.0013531799729365,
+      "grad_norm": 0.20581486821174622,
+      "learning_rate": 3.0077117917016187e-06,
+      "loss": 0.3228,
+      "step": 2465
+    },
+    {
+      "epoch": 2.002165087956698,
+      "grad_norm": 0.2026837319135666,
+      "learning_rate": 3.00337710173198e-06,
+      "loss": 0.32,
+      "step": 2466
+    },
+    {
+      "epoch": 2.00297699594046,
+      "grad_norm": 0.21703356504440308,
+      "learning_rate": 2.9990441963384657e-06,
+      "loss": 0.3174,
+      "step": 2467
+    },
+    {
+      "epoch": 2.003788903924222,
+      "grad_norm": 0.24868090450763702,
+      "learning_rate": 2.9947130793938127e-06,
+      "loss": 0.3408,
+      "step": 2468
+    },
+    {
+      "epoch": 2.004600811907984,
+      "grad_norm": 0.22960342466831207,
+      "learning_rate": 2.9903837547691634e-06,
+      "loss": 0.3628,
+      "step": 2469
+    },
+    {
+      "epoch": 2.0054127198917455,
+      "grad_norm": 0.2056204080581665,
+      "learning_rate": 2.986056226334054e-06,
+      "loss": 0.3007,
+      "step": 2470
+    },
+    {
+      "epoch": 2.0062246278755076,
+      "grad_norm": 0.21412457525730133,
+      "learning_rate": 2.9817304979564206e-06,
+      "loss": 0.3083,
+      "step": 2471
+    },
+    {
+      "epoch": 2.007036535859269,
+      "grad_norm": 0.215728297829628,
+      "learning_rate": 2.977406573502585e-06,
+      "loss": 0.3431,
+      "step": 2472
+    },
+    {
+      "epoch": 2.0078484438430313,
+      "grad_norm": 0.23913609981536865,
+      "learning_rate": 2.9730844568372598e-06,
+      "loss": 0.3245,
+      "step": 2473
+    },
+    {
+      "epoch": 2.008660351826793,
+      "grad_norm": 0.2301327884197235,
+      "learning_rate": 2.968764151823542e-06,
+      "loss": 0.3615,
+      "step": 2474
+    },
+    {
+      "epoch": 2.009472259810555,
+      "grad_norm": 0.2228858768939972,
+      "learning_rate": 2.964445662322908e-06,
+      "loss": 0.3106,
+      "step": 2475
+    },
+    {
+      "epoch": 2.0102841677943166,
+      "grad_norm": 0.21926817297935486,
+      "learning_rate": 2.960128992195211e-06,
+      "loss": 0.3437,
+      "step": 2476
+    },
+    {
+      "epoch": 2.0110960757780787,
+      "grad_norm": 0.2225901186466217,
+      "learning_rate": 2.95581414529868e-06,
+      "loss": 0.3745,
+      "step": 2477
+    },
+    {
+      "epoch": 2.0119079837618403,
+      "grad_norm": 0.22040916979312897,
+      "learning_rate": 2.951501125489914e-06,
+      "loss": 0.3079,
+      "step": 2478
+    },
+    {
+      "epoch": 2.012719891745602,
+      "grad_norm": 0.23161236941814423,
+      "learning_rate": 2.9471899366238777e-06,
+      "loss": 0.3438,
+      "step": 2479
+    },
+    {
+      "epoch": 2.013531799729364,
+      "grad_norm": 0.20193779468536377,
+      "learning_rate": 2.9428805825538993e-06,
+      "loss": 0.3131,
+      "step": 2480
+    },
+    {
+      "epoch": 2.0143437077131257,
+      "grad_norm": 0.21600410342216492,
+      "learning_rate": 2.9385730671316687e-06,
+      "loss": 0.3339,
+      "step": 2481
+    },
+    {
+      "epoch": 2.0151556156968877,
+      "grad_norm": 0.19607877731323242,
+      "learning_rate": 2.934267394207231e-06,
+      "loss": 0.3348,
+      "step": 2482
+    },
+    {
+      "epoch": 2.0159675236806494,
+      "grad_norm": 0.20735116302967072,
+      "learning_rate": 2.929963567628985e-06,
+      "loss": 0.3273,
+      "step": 2483
+    },
+    {
+      "epoch": 2.0167794316644114,
+      "grad_norm": 0.22218379378318787,
+      "learning_rate": 2.9256615912436803e-06,
+      "loss": 0.3612,
+      "step": 2484
+    },
+    {
+      "epoch": 2.017591339648173,
+      "grad_norm": 0.2026771754026413,
+      "learning_rate": 2.9213614688964104e-06,
+      "loss": 0.32,
+      "step": 2485
+    },
+    {
+      "epoch": 2.018403247631935,
+      "grad_norm": 0.19235950708389282,
+      "learning_rate": 2.9170632044306137e-06,
+      "loss": 0.3317,
+      "step": 2486
+    },
+    {
+      "epoch": 2.0192151556156968,
+      "grad_norm": 0.17794857919216156,
+      "learning_rate": 2.9127668016880674e-06,
+      "loss": 0.3236,
+      "step": 2487
+    },
+    {
+      "epoch": 2.020027063599459,
+      "grad_norm": 0.18587854504585266,
+      "learning_rate": 2.9084722645088825e-06,
+      "loss": 0.2966,
+      "step": 2488
+    },
+    {
+      "epoch": 2.0208389715832205,
+      "grad_norm": 0.21765445172786713,
+      "learning_rate": 2.904179596731509e-06,
+      "loss": 0.3563,
+      "step": 2489
+    },
+    {
+      "epoch": 2.0216508795669825,
+      "grad_norm": 0.1933736652135849,
+      "learning_rate": 2.8998888021927146e-06,
+      "loss": 0.2998,
+      "step": 2490
+    },
+    {
+      "epoch": 2.022462787550744,
+      "grad_norm": 0.20295892655849457,
+      "learning_rate": 2.895599884727608e-06,
+      "loss": 0.3552,
+      "step": 2491
+    },
+    {
+      "epoch": 2.0232746955345062,
+      "grad_norm": 0.19951565563678741,
+      "learning_rate": 2.8913128481696067e-06,
+      "loss": 0.3716,
+      "step": 2492
+    },
+    {
+      "epoch": 2.024086603518268,
+      "grad_norm": 0.195914164185524,
+      "learning_rate": 2.8870276963504562e-06,
+      "loss": 0.295,
+      "step": 2493
+    },
+    {
+      "epoch": 2.02489851150203,
+      "grad_norm": 0.21008634567260742,
+      "learning_rate": 2.8827444331002087e-06,
+      "loss": 0.3154,
+      "step": 2494
+    },
+    {
+      "epoch": 2.0257104194857916,
+      "grad_norm": 0.20718730986118317,
+      "learning_rate": 2.8784630622472376e-06,
+      "loss": 0.3435,
+      "step": 2495
+    },
+    {
+      "epoch": 2.0265223274695536,
+      "grad_norm": 0.18462742865085602,
+      "learning_rate": 2.8741835876182165e-06,
+      "loss": 0.3066,
+      "step": 2496
+    },
+    {
+      "epoch": 2.0273342354533153,
+      "grad_norm": 0.22178871929645538,
+      "learning_rate": 2.8699060130381334e-06,
+      "loss": 0.364,
+      "step": 2497
+    },
+    {
+      "epoch": 2.0281461434370773,
+      "grad_norm": 0.22039838135242462,
+      "learning_rate": 2.8656303423302678e-06,
+      "loss": 0.35,
+      "step": 2498
+    },
+    {
+      "epoch": 2.028958051420839,
+      "grad_norm": 0.21203839778900146,
+      "learning_rate": 2.861356579316204e-06,
+      "loss": 0.3644,
+      "step": 2499
+    },
+    {
+      "epoch": 2.029769959404601,
+      "grad_norm": 0.1858183592557907,
+      "learning_rate": 2.857084727815823e-06,
+      "loss": 0.3071,
+      "step": 2500
+    },
+    {
+      "epoch": 2.0305818673883627,
+      "grad_norm": 0.2414398044347763,
+      "learning_rate": 2.8528147916472895e-06,
+      "loss": 0.3649,
+      "step": 2501
+    },
+    {
+      "epoch": 2.0313937753721243,
+      "grad_norm": 0.2168034166097641,
+      "learning_rate": 2.848546774627064e-06,
+      "loss": 0.3154,
+      "step": 2502
+    },
+    {
+      "epoch": 2.0322056833558864,
+      "grad_norm": 0.1821415275335312,
+      "learning_rate": 2.8442806805698857e-06,
+      "loss": 0.2662,
+      "step": 2503
+    },
+    {
+      "epoch": 2.033017591339648,
+      "grad_norm": 0.21955081820487976,
+      "learning_rate": 2.840016513288782e-06,
+      "loss": 0.3585,
+      "step": 2504
+    },
+    {
+      "epoch": 2.03382949932341,
+      "grad_norm": 0.1974199265241623,
+      "learning_rate": 2.8357542765950493e-06,
+      "loss": 0.3149,
+      "step": 2505
+    },
+    {
+      "epoch": 2.0346414073071717,
+      "grad_norm": 0.2319677770137787,
+      "learning_rate": 2.8314939742982673e-06,
+      "loss": 0.3557,
+      "step": 2506
+    },
+    {
+      "epoch": 2.0354533152909338,
+      "grad_norm": 0.21963539719581604,
+      "learning_rate": 2.8272356102062792e-06,
+      "loss": 0.3124,
+      "step": 2507
+    },
+    {
+      "epoch": 2.0362652232746954,
+      "grad_norm": 0.20201516151428223,
+      "learning_rate": 2.8229791881252023e-06,
+      "loss": 0.321,
+      "step": 2508
+    },
+    {
+      "epoch": 2.0370771312584575,
+      "grad_norm": 0.216538205742836,
+      "learning_rate": 2.818724711859412e-06,
+      "loss": 0.3417,
+      "step": 2509
+    },
+    {
+      "epoch": 2.037889039242219,
+      "grad_norm": 0.2016521692276001,
+      "learning_rate": 2.814472185211551e-06,
+      "loss": 0.3118,
+      "step": 2510
+    },
+    {
+      "epoch": 2.038700947225981,
+      "grad_norm": 0.1894722729921341,
+      "learning_rate": 2.810221611982512e-06,
+      "loss": 0.298,
+      "step": 2511
+    },
+    {
+      "epoch": 2.039512855209743,
+      "grad_norm": 0.22017988562583923,
+      "learning_rate": 2.805972995971451e-06,
+      "loss": 0.3632,
+      "step": 2512
+    },
+    {
+      "epoch": 2.040324763193505,
+      "grad_norm": 0.188715860247612,
+      "learning_rate": 2.8017263409757646e-06,
+      "loss": 0.2974,
+      "step": 2513
+    },
+    {
+      "epoch": 2.0411366711772665,
+      "grad_norm": 0.19456727802753448,
+      "learning_rate": 2.7974816507911047e-06,
+      "loss": 0.3427,
+      "step": 2514
+    },
+    {
+      "epoch": 2.0419485791610286,
+      "grad_norm": 0.20121699571609497,
+      "learning_rate": 2.7932389292113604e-06,
+      "loss": 0.3137,
+      "step": 2515
+    },
+    {
+      "epoch": 2.04276048714479,
+      "grad_norm": 0.21681542694568634,
+      "learning_rate": 2.788998180028668e-06,
+      "loss": 0.306,
+      "step": 2516
+    },
+    {
+      "epoch": 2.0435723951285523,
+      "grad_norm": 0.21026195585727692,
+      "learning_rate": 2.784759407033398e-06,
+      "loss": 0.3533,
+      "step": 2517
+    },
+    {
+      "epoch": 2.044384303112314,
+      "grad_norm": 0.2107834815979004,
+      "learning_rate": 2.7805226140141505e-06,
+      "loss": 0.3496,
+      "step": 2518
+    },
+    {
+      "epoch": 2.045196211096076,
+      "grad_norm": 0.20738348364830017,
+      "learning_rate": 2.776287804757765e-06,
+      "loss": 0.342,
+      "step": 2519
+    },
+    {
+      "epoch": 2.0460081190798376,
+      "grad_norm": 0.19592416286468506,
+      "learning_rate": 2.772054983049296e-06,
+      "loss": 0.3277,
+      "step": 2520
+    },
+    {
+      "epoch": 2.0468200270635997,
+      "grad_norm": 0.20149774849414825,
+      "learning_rate": 2.7678241526720338e-06,
+      "loss": 0.3272,
+      "step": 2521
+    },
+    {
+      "epoch": 2.0476319350473613,
+      "grad_norm": 0.20924854278564453,
+      "learning_rate": 2.7635953174074787e-06,
+      "loss": 0.3309,
+      "step": 2522
+    },
+    {
+      "epoch": 2.048443843031123,
+      "grad_norm": 0.21309807896614075,
+      "learning_rate": 2.759368481035356e-06,
+      "loss": 0.3586,
+      "step": 2523
+    },
+    {
+      "epoch": 2.049255751014885,
+      "grad_norm": 0.19526107609272003,
+      "learning_rate": 2.755143647333597e-06,
+      "loss": 0.3372,
+      "step": 2524
+    },
+    {
+      "epoch": 2.0500676589986466,
+      "grad_norm": 0.22154748439788818,
+      "learning_rate": 2.7509208200783505e-06,
+      "loss": 0.3458,
+      "step": 2525
+    },
+    {
+      "epoch": 2.0508795669824087,
+      "grad_norm": 0.2063003033399582,
+      "learning_rate": 2.746700003043964e-06,
+      "loss": 0.297,
+      "step": 2526
+    },
+    {
+      "epoch": 2.0516914749661703,
+      "grad_norm": 0.1975739747285843,
+      "learning_rate": 2.7424812000029955e-06,
+      "loss": 0.3298,
+      "step": 2527
+    },
+    {
+      "epoch": 2.0525033829499324,
+      "grad_norm": 0.20413611829280853,
+      "learning_rate": 2.738264414726196e-06,
+      "loss": 0.3241,
+      "step": 2528
+    },
+    {
+      "epoch": 2.053315290933694,
+      "grad_norm": 0.20195543766021729,
+      "learning_rate": 2.73404965098252e-06,
+      "loss": 0.3062,
+      "step": 2529
+    },
+    {
+      "epoch": 2.054127198917456,
+      "grad_norm": 0.2174825817346573,
+      "learning_rate": 2.7298369125391112e-06,
+      "loss": 0.3184,
+      "step": 2530
+    },
+    {
+      "epoch": 2.0549391069012177,
+      "grad_norm": 0.19972918927669525,
+      "learning_rate": 2.7256262031613044e-06,
+      "loss": 0.3023,
+      "step": 2531
+    },
+    {
+      "epoch": 2.05575101488498,
+      "grad_norm": 0.21753333508968353,
+      "learning_rate": 2.7214175266126184e-06,
+      "loss": 0.3439,
+      "step": 2532
+    },
+    {
+      "epoch": 2.0565629228687414,
+      "grad_norm": 0.18228372931480408,
+      "learning_rate": 2.7172108866547593e-06,
+      "loss": 0.2977,
+      "step": 2533
+    },
+    {
+      "epoch": 2.0573748308525035,
+      "grad_norm": 0.21457816660404205,
+      "learning_rate": 2.7130062870476072e-06,
+      "loss": 0.34,
+      "step": 2534
+    },
+    {
+      "epoch": 2.058186738836265,
+      "grad_norm": 0.20791196823120117,
+      "learning_rate": 2.7088037315492254e-06,
+      "loss": 0.2975,
+      "step": 2535
+    },
+    {
+      "epoch": 2.058998646820027,
+      "grad_norm": 0.20045089721679688,
+      "learning_rate": 2.7046032239158428e-06,
+      "loss": 0.323,
+      "step": 2536
+    },
+    {
+      "epoch": 2.059810554803789,
+      "grad_norm": 0.19426392018795013,
+      "learning_rate": 2.7004047679018652e-06,
+      "loss": 0.3522,
+      "step": 2537
+    },
+    {
+      "epoch": 2.060622462787551,
+      "grad_norm": 0.1906808465719223,
+      "learning_rate": 2.6962083672598584e-06,
+      "loss": 0.3544,
+      "step": 2538
+    },
+    {
+      "epoch": 2.0614343707713125,
+      "grad_norm": 0.19458819925785065,
+      "learning_rate": 2.692014025740557e-06,
+      "loss": 0.3115,
+      "step": 2539
+    },
+    {
+      "epoch": 2.0622462787550746,
+      "grad_norm": 0.19927465915679932,
+      "learning_rate": 2.687821747092848e-06,
+      "loss": 0.3223,
+      "step": 2540
+    },
+    {
+      "epoch": 2.0630581867388362,
+      "grad_norm": 0.21004174649715424,
+      "learning_rate": 2.683631535063783e-06,
+      "loss": 0.3078,
+      "step": 2541
+    },
+    {
+      "epoch": 2.0638700947225983,
+      "grad_norm": 0.20895028114318848,
+      "learning_rate": 2.679443393398558e-06,
+      "loss": 0.343,
+      "step": 2542
+    },
+    {
+      "epoch": 2.06468200270636,
+      "grad_norm": 0.21156232059001923,
+      "learning_rate": 2.6752573258405245e-06,
+      "loss": 0.3328,
+      "step": 2543
+    },
+    {
+      "epoch": 2.0654939106901216,
+      "grad_norm": 0.17583221197128296,
+      "learning_rate": 2.6710733361311812e-06,
+      "loss": 0.2754,
+      "step": 2544
+    },
+    {
+      "epoch": 2.0663058186738836,
+      "grad_norm": 0.19910255074501038,
+      "learning_rate": 2.6668914280101615e-06,
+      "loss": 0.3172,
+      "step": 2545
+    },
+    {
+      "epoch": 2.0671177266576453,
+      "grad_norm": 0.20044097304344177,
+      "learning_rate": 2.662711605215248e-06,
+      "loss": 0.3595,
+      "step": 2546
+    },
+    {
+      "epoch": 2.0679296346414073,
+      "grad_norm": 0.19012795388698578,
+      "learning_rate": 2.658533871482351e-06,
+      "loss": 0.3183,
+      "step": 2547
+    },
+    {
+      "epoch": 2.068741542625169,
+      "grad_norm": 0.2012583315372467,
+      "learning_rate": 2.6543582305455217e-06,
+      "loss": 0.3361,
+      "step": 2548
+    },
+    {
+      "epoch": 2.069553450608931,
+      "grad_norm": 0.18782205879688263,
+      "learning_rate": 2.650184686136932e-06,
+      "loss": 0.2924,
+      "step": 2549
+    },
+    {
+      "epoch": 2.0703653585926927,
+      "grad_norm": 0.19488516449928284,
+      "learning_rate": 2.6460132419868893e-06,
+      "loss": 0.3223,
+      "step": 2550
+    },
+    {
+      "epoch": 2.0711772665764547,
+      "grad_norm": 0.20608069002628326,
+      "learning_rate": 2.641843901823815e-06,
+      "loss": 0.3083,
+      "step": 2551
+    },
+    {
+      "epoch": 2.0719891745602164,
+      "grad_norm": 0.19479815661907196,
+      "learning_rate": 2.637676669374258e-06,
+      "loss": 0.3023,
+      "step": 2552
+    },
+    {
+      "epoch": 2.0728010825439784,
+      "grad_norm": 0.19623591005802155,
+      "learning_rate": 2.6335115483628747e-06,
+      "loss": 0.3502,
+      "step": 2553
+    },
+    {
+      "epoch": 2.07361299052774,
+      "grad_norm": 0.20341069996356964,
+      "learning_rate": 2.629348542512443e-06,
+      "loss": 0.3513,
+      "step": 2554
+    },
+    {
+      "epoch": 2.074424898511502,
+      "grad_norm": 0.189675435423851,
+      "learning_rate": 2.625187655543844e-06,
+      "loss": 0.2978,
+      "step": 2555
+    },
+    {
+      "epoch": 2.0752368064952638,
+      "grad_norm": 0.2008848339319229,
+      "learning_rate": 2.6210288911760684e-06,
+      "loss": 0.3581,
+      "step": 2556
+    },
+    {
+      "epoch": 2.076048714479026,
+      "grad_norm": 0.20676665008068085,
+      "learning_rate": 2.616872253126207e-06,
+      "loss": 0.3185,
+      "step": 2557
+    },
+    {
+      "epoch": 2.0768606224627875,
+      "grad_norm": 0.1911642700433731,
+      "learning_rate": 2.612717745109453e-06,
+      "loss": 0.2947,
+      "step": 2558
+    },
+    {
+      "epoch": 2.0776725304465495,
+      "grad_norm": 0.19723407924175262,
+      "learning_rate": 2.6085653708390936e-06,
+      "loss": 0.3302,
+      "step": 2559
+    },
+    {
+      "epoch": 2.078484438430311,
+      "grad_norm": 0.18586239218711853,
+      "learning_rate": 2.604415134026509e-06,
+      "loss": 0.3298,
+      "step": 2560
+    },
+    {
+      "epoch": 2.0792963464140732,
+      "grad_norm": 0.22774268686771393,
+      "learning_rate": 2.6002670383811724e-06,
+      "loss": 0.3769,
+      "step": 2561
+    },
+    {
+      "epoch": 2.080108254397835,
+      "grad_norm": 0.20576432347297668,
+      "learning_rate": 2.596121087610636e-06,
+      "loss": 0.3238,
+      "step": 2562
+    },
+    {
+      "epoch": 2.080920162381597,
+      "grad_norm": 0.20354606211185455,
+      "learning_rate": 2.591977285420545e-06,
+      "loss": 0.3061,
+      "step": 2563
+    },
+    {
+      "epoch": 2.0817320703653586,
+      "grad_norm": 0.19074790179729462,
+      "learning_rate": 2.587835635514613e-06,
+      "loss": 0.3164,
+      "step": 2564
+    },
+    {
+      "epoch": 2.0825439783491206,
+      "grad_norm": 0.2213759869337082,
+      "learning_rate": 2.5836961415946394e-06,
+      "loss": 0.3799,
+      "step": 2565
+    },
+    {
+      "epoch": 2.0833558863328823,
+      "grad_norm": 0.2077367603778839,
+      "learning_rate": 2.579558807360489e-06,
+      "loss": 0.3266,
+      "step": 2566
+    },
+    {
+      "epoch": 2.0841677943166443,
+      "grad_norm": 0.24216191470623016,
+      "learning_rate": 2.5754236365101024e-06,
+      "loss": 0.3578,
+      "step": 2567
+    },
+    {
+      "epoch": 2.084979702300406,
+      "grad_norm": 0.19593918323516846,
+      "learning_rate": 2.5712906327394836e-06,
+      "loss": 0.3057,
+      "step": 2568
+    },
+    {
+      "epoch": 2.0857916102841676,
+      "grad_norm": 0.2040565311908722,
+      "learning_rate": 2.5671597997427023e-06,
+      "loss": 0.333,
+      "step": 2569
+    },
+    {
+      "epoch": 2.0866035182679297,
+      "grad_norm": 0.2209572196006775,
+      "learning_rate": 2.5630311412118815e-06,
+      "loss": 0.3489,
+      "step": 2570
+    },
+    {
+      "epoch": 2.0874154262516913,
+      "grad_norm": 0.20549562573432922,
+      "learning_rate": 2.558904660837209e-06,
+      "loss": 0.3125,
+      "step": 2571
+    },
+    {
+      "epoch": 2.0882273342354534,
+      "grad_norm": 0.21414828300476074,
+      "learning_rate": 2.554780362306918e-06,
+      "loss": 0.3461,
+      "step": 2572
+    },
+    {
+      "epoch": 2.089039242219215,
+      "grad_norm": 0.19350294768810272,
+      "learning_rate": 2.5506582493072994e-06,
+      "loss": 0.3304,
+      "step": 2573
+    },
+    {
+      "epoch": 2.089851150202977,
+      "grad_norm": 0.18959267437458038,
+      "learning_rate": 2.546538325522682e-06,
+      "loss": 0.332,
+      "step": 2574
+    },
+    {
+      "epoch": 2.0906630581867387,
+      "grad_norm": 0.18942318856716156,
+      "learning_rate": 2.5424205946354475e-06,
+      "loss": 0.3411,
+      "step": 2575
+    },
+    {
+      "epoch": 2.0914749661705008,
+      "grad_norm": 0.1887650191783905,
+      "learning_rate": 2.5383050603260074e-06,
+      "loss": 0.2951,
+      "step": 2576
+    },
+    {
+      "epoch": 2.0922868741542624,
+      "grad_norm": 0.20228976011276245,
+      "learning_rate": 2.53419172627282e-06,
+      "loss": 0.3134,
+      "step": 2577
+    },
+    {
+      "epoch": 2.0930987821380245,
+      "grad_norm": 0.2510368227958679,
+      "learning_rate": 2.5300805961523677e-06,
+      "loss": 0.3463,
+      "step": 2578
+    },
+    {
+      "epoch": 2.093910690121786,
+      "grad_norm": 0.20003686845302582,
+      "learning_rate": 2.525971673639172e-06,
+      "loss": 0.3346,
+      "step": 2579
+    },
+    {
+      "epoch": 2.094722598105548,
+      "grad_norm": 0.1906794011592865,
+      "learning_rate": 2.5218649624057717e-06,
+      "loss": 0.351,
+      "step": 2580
+    },
+    {
+      "epoch": 2.09553450608931,
+      "grad_norm": 0.2002934217453003,
+      "learning_rate": 2.517760466122739e-06,
+      "loss": 0.3042,
+      "step": 2581
+    },
+    {
+      "epoch": 2.096346414073072,
+      "grad_norm": 0.23740451037883759,
+      "learning_rate": 2.513658188458657e-06,
+      "loss": 0.3463,
+      "step": 2582
+    },
+    {
+      "epoch": 2.0971583220568335,
+      "grad_norm": 0.22402381896972656,
+      "learning_rate": 2.5095581330801345e-06,
+      "loss": 0.3213,
+      "step": 2583
+    },
+    {
+      "epoch": 2.0979702300405956,
+      "grad_norm": 0.22374382615089417,
+      "learning_rate": 2.5054603036517856e-06,
+      "loss": 0.3571,
+      "step": 2584
+    },
+    {
+      "epoch": 2.098782138024357,
+      "grad_norm": 0.18976548314094543,
+      "learning_rate": 2.5013647038362437e-06,
+      "loss": 0.327,
+      "step": 2585
+    },
+    {
+      "epoch": 2.0995940460081193,
+      "grad_norm": 0.2133355438709259,
+      "learning_rate": 2.4972713372941406e-06,
+      "loss": 0.3545,
+      "step": 2586
+    },
+    {
+      "epoch": 2.100405953991881,
+      "grad_norm": 0.2056700587272644,
+      "learning_rate": 2.4931802076841176e-06,
+      "loss": 0.341,
+      "step": 2587
+    },
+    {
+      "epoch": 2.101217861975643,
+      "grad_norm": 0.19280005991458893,
+      "learning_rate": 2.4890913186628178e-06,
+      "loss": 0.2964,
+      "step": 2588
+    },
+    {
+      "epoch": 2.1020297699594046,
+      "grad_norm": 0.19891755282878876,
+      "learning_rate": 2.4850046738848752e-06,
+      "loss": 0.2945,
+      "step": 2589
+    },
+    {
+      "epoch": 2.1028416779431662,
+      "grad_norm": 0.20743916928768158,
+      "learning_rate": 2.480920277002926e-06,
+      "loss": 0.3251,
+      "step": 2590
+    },
+    {
+      "epoch": 2.1036535859269283,
+      "grad_norm": 0.1988065242767334,
+      "learning_rate": 2.4768381316675883e-06,
+      "loss": 0.3303,
+      "step": 2591
+    },
+    {
+      "epoch": 2.10446549391069,
+      "grad_norm": 0.21556751430034637,
+      "learning_rate": 2.472758241527477e-06,
+      "loss": 0.3652,
+      "step": 2592
+    },
+    {
+      "epoch": 2.105277401894452,
+      "grad_norm": 0.1915753185749054,
+      "learning_rate": 2.468680610229182e-06,
+      "loss": 0.2919,
+      "step": 2593
+    },
+    {
+      "epoch": 2.1060893098782136,
+      "grad_norm": 0.20534710586071014,
+      "learning_rate": 2.4646052414172844e-06,
+      "loss": 0.3345,
+      "step": 2594
+    },
+    {
+      "epoch": 2.1069012178619757,
+      "grad_norm": 0.22919823229312897,
+      "learning_rate": 2.4605321387343334e-06,
+      "loss": 0.3682,
+      "step": 2595
+    },
+    {
+      "epoch": 2.1077131258457373,
+      "grad_norm": 0.18596819043159485,
+      "learning_rate": 2.4564613058208603e-06,
+      "loss": 0.3109,
+      "step": 2596
+    },
+    {
+      "epoch": 2.1085250338294994,
+      "grad_norm": 0.20546583831310272,
+      "learning_rate": 2.452392746315361e-06,
+      "loss": 0.3359,
+      "step": 2597
+    },
+    {
+      "epoch": 2.109336941813261,
+      "grad_norm": 0.22032728791236877,
+      "learning_rate": 2.4483264638543076e-06,
+      "loss": 0.3499,
+      "step": 2598
+    },
+    {
+      "epoch": 2.110148849797023,
+      "grad_norm": 0.19730329513549805,
+      "learning_rate": 2.4442624620721274e-06,
+      "loss": 0.3102,
+      "step": 2599
+    },
+    {
+      "epoch": 2.1109607577807847,
+      "grad_norm": 0.22564752399921417,
+      "learning_rate": 2.440200744601218e-06,
+      "loss": 0.3199,
+      "step": 2600
+    },
+    {
+      "epoch": 2.111772665764547,
+      "grad_norm": 0.20735275745391846,
+      "learning_rate": 2.4361413150719287e-06,
+      "loss": 0.3149,
+      "step": 2601
+    },
+    {
+      "epoch": 2.1125845737483084,
+      "grad_norm": 0.1938306838274002,
+      "learning_rate": 2.432084177112569e-06,
+      "loss": 0.3245,
+      "step": 2602
+    },
+    {
+      "epoch": 2.1133964817320705,
+      "grad_norm": 0.18469147384166718,
+      "learning_rate": 2.4280293343493954e-06,
+      "loss": 0.3078,
+      "step": 2603
+    },
+    {
+      "epoch": 2.114208389715832,
+      "grad_norm": 0.20468854904174805,
+      "learning_rate": 2.423976790406616e-06,
+      "loss": 0.3393,
+      "step": 2604
+    },
+    {
+      "epoch": 2.115020297699594,
+      "grad_norm": 0.2222663015127182,
+      "learning_rate": 2.4199265489063846e-06,
+      "loss": 0.3789,
+      "step": 2605
+    },
+    {
+      "epoch": 2.115832205683356,
+      "grad_norm": 0.22328872978687286,
+      "learning_rate": 2.4158786134687966e-06,
+      "loss": 0.3152,
+      "step": 2606
+    },
+    {
+      "epoch": 2.116644113667118,
+      "grad_norm": 0.1808295100927353,
+      "learning_rate": 2.4118329877118836e-06,
+      "loss": 0.332,
+      "step": 2607
+    },
+    {
+      "epoch": 2.1174560216508795,
+      "grad_norm": 0.18614795804023743,
+      "learning_rate": 2.407789675251617e-06,
+      "loss": 0.3677,
+      "step": 2608
+    },
+    {
+      "epoch": 2.1182679296346416,
+      "grad_norm": 0.17931774258613586,
+      "learning_rate": 2.403748679701896e-06,
+      "loss": 0.2959,
+      "step": 2609
+    },
+    {
+      "epoch": 2.1190798376184032,
+      "grad_norm": 0.1960575431585312,
+      "learning_rate": 2.3997100046745554e-06,
+      "loss": 0.3357,
+      "step": 2610
+    },
+    {
+      "epoch": 2.119891745602165,
+      "grad_norm": 0.20216582715511322,
+      "learning_rate": 2.3956736537793468e-06,
+      "loss": 0.3268,
+      "step": 2611
+    },
+    {
+      "epoch": 2.120703653585927,
+      "grad_norm": 0.21501852571964264,
+      "learning_rate": 2.3916396306239525e-06,
+      "loss": 0.3389,
+      "step": 2612
+    },
+    {
+      "epoch": 2.1215155615696886,
+      "grad_norm": 0.1912708878517151,
+      "learning_rate": 2.3876079388139733e-06,
+      "loss": 0.3039,
+      "step": 2613
+    },
+    {
+      "epoch": 2.1223274695534506,
+      "grad_norm": 0.2036161869764328,
+      "learning_rate": 2.38357858195292e-06,
+      "loss": 0.3527,
+      "step": 2614
+    },
+    {
+      "epoch": 2.1231393775372123,
+      "grad_norm": 0.20998884737491608,
+      "learning_rate": 2.3795515636422257e-06,
+      "loss": 0.3192,
+      "step": 2615
+    },
+    {
+      "epoch": 2.1239512855209743,
+      "grad_norm": 0.2021828293800354,
+      "learning_rate": 2.3755268874812237e-06,
+      "loss": 0.3402,
+      "step": 2616
+    },
+    {
+      "epoch": 2.124763193504736,
+      "grad_norm": 0.18152351677417755,
+      "learning_rate": 2.371504557067163e-06,
+      "loss": 0.2977,
+      "step": 2617
+    },
+    {
+      "epoch": 2.125575101488498,
+      "grad_norm": 0.20884640514850616,
+      "learning_rate": 2.367484575995187e-06,
+      "loss": 0.3449,
+      "step": 2618
+    },
+    {
+      "epoch": 2.1263870094722597,
+      "grad_norm": 0.19364672899246216,
+      "learning_rate": 2.3634669478583484e-06,
+      "loss": 0.33,
+      "step": 2619
+    },
+    {
+      "epoch": 2.1271989174560217,
+      "grad_norm": 0.20248380303382874,
+      "learning_rate": 2.3594516762475882e-06,
+      "loss": 0.3167,
+      "step": 2620
+    },
+    {
+      "epoch": 2.1280108254397834,
+      "grad_norm": 0.19341282546520233,
+      "learning_rate": 2.3554387647517497e-06,
+      "loss": 0.304,
+      "step": 2621
+    },
+    {
+      "epoch": 2.1288227334235454,
+      "grad_norm": 0.2365405112504959,
+      "learning_rate": 2.3514282169575575e-06,
+      "loss": 0.3396,
+      "step": 2622
+    },
+    {
+      "epoch": 2.129634641407307,
+      "grad_norm": 0.1930687129497528,
+      "learning_rate": 2.3474200364496326e-06,
+      "loss": 0.3147,
+      "step": 2623
+    },
+    {
+      "epoch": 2.130446549391069,
+      "grad_norm": 0.19267767667770386,
+      "learning_rate": 2.343414226810472e-06,
+      "loss": 0.285,
+      "step": 2624
+    },
+    {
+      "epoch": 2.1312584573748308,
+      "grad_norm": 0.20868007838726044,
+      "learning_rate": 2.3394107916204617e-06,
+      "loss": 0.3434,
+      "step": 2625
+    },
+    {
+      "epoch": 2.132070365358593,
+      "grad_norm": 0.2028043270111084,
+      "learning_rate": 2.3354097344578565e-06,
+      "loss": 0.3773,
+      "step": 2626
+    },
+    {
+      "epoch": 2.1328822733423545,
+      "grad_norm": 0.20307624340057373,
+      "learning_rate": 2.3314110588987946e-06,
+      "loss": 0.3004,
+      "step": 2627
+    },
+    {
+      "epoch": 2.1336941813261165,
+      "grad_norm": 0.21297332644462585,
+      "learning_rate": 2.327414768517278e-06,
+      "loss": 0.2993,
+      "step": 2628
+    },
+    {
+      "epoch": 2.134506089309878,
+      "grad_norm": 0.21407069265842438,
+      "learning_rate": 2.3234208668851843e-06,
+      "loss": 0.3697,
+      "step": 2629
+    },
+    {
+      "epoch": 2.1353179972936402,
+      "grad_norm": 0.2335767298936844,
+      "learning_rate": 2.3194293575722464e-06,
+      "loss": 0.3313,
+      "step": 2630
+    },
+    {
+      "epoch": 2.136129905277402,
+      "grad_norm": 0.20621147751808167,
+      "learning_rate": 2.3154402441460685e-06,
+      "loss": 0.3323,
+      "step": 2631
+    },
+    {
+      "epoch": 2.1369418132611635,
+      "grad_norm": 0.21469604969024658,
+      "learning_rate": 2.311453530172109e-06,
+      "loss": 0.3201,
+      "step": 2632
+    },
+    {
+      "epoch": 2.1377537212449256,
+      "grad_norm": 0.19321385025978088,
+      "learning_rate": 2.3074692192136803e-06,
+      "loss": 0.3361,
+      "step": 2633
+    },
+    {
+      "epoch": 2.1385656292286876,
+      "grad_norm": 0.21296092867851257,
+      "learning_rate": 2.303487314831952e-06,
+      "loss": 0.3312,
+      "step": 2634
+    },
+    {
+      "epoch": 2.1393775372124493,
+      "grad_norm": 0.1964598149061203,
+      "learning_rate": 2.2995078205859344e-06,
+      "loss": 0.3224,
+      "step": 2635
+    },
+    {
+      "epoch": 2.140189445196211,
+      "grad_norm": 0.20847280323505402,
+      "learning_rate": 2.295530740032494e-06,
+      "loss": 0.3054,
+      "step": 2636
+    },
+    {
+      "epoch": 2.141001353179973,
+      "grad_norm": 0.22070710361003876,
+      "learning_rate": 2.2915560767263296e-06,
+      "loss": 0.3325,
+      "step": 2637
+    },
+    {
+      "epoch": 2.1418132611637346,
+      "grad_norm": 0.2008322924375534,
+      "learning_rate": 2.287583834219989e-06,
+      "loss": 0.3052,
+      "step": 2638
+    },
+    {
+      "epoch": 2.1426251691474967,
+      "grad_norm": 0.18989551067352295,
+      "learning_rate": 2.2836140160638477e-06,
+      "loss": 0.3193,
+      "step": 2639
+    },
+    {
+      "epoch": 2.1434370771312583,
+      "grad_norm": 0.20419514179229736,
+      "learning_rate": 2.2796466258061223e-06,
+      "loss": 0.3531,
+      "step": 2640
+    },
+    {
+      "epoch": 2.1442489851150204,
+      "grad_norm": 0.201409712433815,
+      "learning_rate": 2.2756816669928516e-06,
+      "loss": 0.3094,
+      "step": 2641
+    },
+    {
+      "epoch": 2.145060893098782,
+      "grad_norm": 0.21053028106689453,
+      "learning_rate": 2.2717191431679085e-06,
+      "loss": 0.3134,
+      "step": 2642
+    },
+    {
+      "epoch": 2.145872801082544,
+      "grad_norm": 0.20614908635616302,
+      "learning_rate": 2.2677590578729815e-06,
+      "loss": 0.3081,
+      "step": 2643
+    },
+    {
+      "epoch": 2.1466847090663057,
+      "grad_norm": 0.19644257426261902,
+      "learning_rate": 2.2638014146475894e-06,
+      "loss": 0.3468,
+      "step": 2644
+    },
+    {
+      "epoch": 2.147496617050068,
+      "grad_norm": 0.21588250994682312,
+      "learning_rate": 2.2598462170290598e-06,
+      "loss": 0.3756,
+      "step": 2645
+    },
+    {
+      "epoch": 2.1483085250338294,
+      "grad_norm": 0.18895165622234344,
+      "learning_rate": 2.25589346855254e-06,
+      "loss": 0.3154,
+      "step": 2646
+    },
+    {
+      "epoch": 2.1491204330175915,
+      "grad_norm": 0.21070504188537598,
+      "learning_rate": 2.2519431727509823e-06,
+      "loss": 0.3387,
+      "step": 2647
+    },
+    {
+      "epoch": 2.149932341001353,
+      "grad_norm": 0.19214271008968353,
+      "learning_rate": 2.2479953331551534e-06,
+      "loss": 0.326,
+      "step": 2648
+    },
+    {
+      "epoch": 2.150744248985115,
+      "grad_norm": 0.20506441593170166,
+      "learning_rate": 2.2440499532936196e-06,
+      "loss": 0.328,
+      "step": 2649
+    },
+    {
+      "epoch": 2.151556156968877,
+      "grad_norm": 0.20923295617103577,
+      "learning_rate": 2.240107036692753e-06,
+      "loss": 0.3318,
+      "step": 2650
+    },
+    {
+      "epoch": 2.152368064952639,
+      "grad_norm": 0.18997439742088318,
+      "learning_rate": 2.2361665868767185e-06,
+      "loss": 0.3209,
+      "step": 2651
+    },
+    {
+      "epoch": 2.1531799729364005,
+      "grad_norm": 0.20215849578380585,
+      "learning_rate": 2.2322286073674832e-06,
+      "loss": 0.3434,
+      "step": 2652
+    },
+    {
+      "epoch": 2.1539918809201626,
+      "grad_norm": 0.19955042004585266,
+      "learning_rate": 2.228293101684799e-06,
+      "loss": 0.355,
+      "step": 2653
+    },
+    {
+      "epoch": 2.154803788903924,
+      "grad_norm": 0.21945920586585999,
+      "learning_rate": 2.224360073346213e-06,
+      "loss": 0.3337,
+      "step": 2654
+    },
+    {
+      "epoch": 2.1556156968876863,
+      "grad_norm": 0.19662538170814514,
+      "learning_rate": 2.2204295258670524e-06,
+      "loss": 0.337,
+      "step": 2655
+    },
+    {
+      "epoch": 2.156427604871448,
+      "grad_norm": 0.20688027143478394,
+      "learning_rate": 2.2165014627604313e-06,
+      "loss": 0.3479,
+      "step": 2656
+    },
+    {
+      "epoch": 2.1572395128552095,
+      "grad_norm": 0.18472614884376526,
+      "learning_rate": 2.212575887537244e-06,
+      "loss": 0.2859,
+      "step": 2657
+    },
+    {
+      "epoch": 2.1580514208389716,
+      "grad_norm": 0.2119249552488327,
+      "learning_rate": 2.208652803706155e-06,
+      "loss": 0.3408,
+      "step": 2658
+    },
+    {
+      "epoch": 2.1588633288227332,
+      "grad_norm": 0.19319657981395721,
+      "learning_rate": 2.2047322147736096e-06,
+      "loss": 0.3033,
+      "step": 2659
+    },
+    {
+      "epoch": 2.1596752368064953,
+      "grad_norm": 0.20557701587677002,
+      "learning_rate": 2.2008141242438165e-06,
+      "loss": 0.302,
+      "step": 2660
+    },
+    {
+      "epoch": 2.160487144790257,
+      "grad_norm": 0.19780895113945007,
+      "learning_rate": 2.196898535618757e-06,
+      "loss": 0.304,
+      "step": 2661
+    },
+    {
+      "epoch": 2.161299052774019,
+      "grad_norm": 0.20874595642089844,
+      "learning_rate": 2.192985452398171e-06,
+      "loss": 0.3152,
+      "step": 2662
+    },
+    {
+      "epoch": 2.1621109607577806,
+      "grad_norm": 0.21763932704925537,
+      "learning_rate": 2.1890748780795644e-06,
+      "loss": 0.3432,
+      "step": 2663
+    },
+    {
+      "epoch": 2.1629228687415427,
+      "grad_norm": 0.20360839366912842,
+      "learning_rate": 2.1851668161581946e-06,
+      "loss": 0.3117,
+      "step": 2664
+    },
+    {
+      "epoch": 2.1637347767253043,
+      "grad_norm": 0.22165033221244812,
+      "learning_rate": 2.181261270127081e-06,
+      "loss": 0.3181,
+      "step": 2665
+    },
+    {
+      "epoch": 2.1645466847090664,
+      "grad_norm": 0.29684194922447205,
+      "learning_rate": 2.1773582434769854e-06,
+      "loss": 0.3017,
+      "step": 2666
+    },
+    {
+      "epoch": 2.165358592692828,
+      "grad_norm": 0.22399559617042542,
+      "learning_rate": 2.1734577396964264e-06,
+      "loss": 0.3212,
+      "step": 2667
+    },
+    {
+      "epoch": 2.16617050067659,
+      "grad_norm": 0.19824543595314026,
+      "learning_rate": 2.1695597622716603e-06,
+      "loss": 0.3154,
+      "step": 2668
+    },
+    {
+      "epoch": 2.1669824086603517,
+      "grad_norm": 0.2087014615535736,
+      "learning_rate": 2.165664314686692e-06,
+      "loss": 0.3383,
+      "step": 2669
+    },
+    {
+      "epoch": 2.167794316644114,
+      "grad_norm": 0.20376873016357422,
+      "learning_rate": 2.1617714004232588e-06,
+      "loss": 0.3409,
+      "step": 2670
+    },
+    {
+      "epoch": 2.1686062246278754,
+      "grad_norm": 0.19039559364318848,
+      "learning_rate": 2.15788102296084e-06,
+      "loss": 0.2971,
+      "step": 2671
+    },
+    {
+      "epoch": 2.1694181326116375,
+      "grad_norm": 0.18953335285186768,
+      "learning_rate": 2.153993185776641e-06,
+      "loss": 0.3372,
+      "step": 2672
+    },
+    {
+      "epoch": 2.170230040595399,
+      "grad_norm": 0.2060459703207016,
+      "learning_rate": 2.150107892345602e-06,
+      "loss": 0.3573,
+      "step": 2673
+    },
+    {
+      "epoch": 2.171041948579161,
+      "grad_norm": 0.20267577469348907,
+      "learning_rate": 2.14622514614039e-06,
+      "loss": 0.3417,
+      "step": 2674
+    },
+    {
+      "epoch": 2.171853856562923,
+      "grad_norm": 0.21609437465667725,
+      "learning_rate": 2.142344950631388e-06,
+      "loss": 0.3331,
+      "step": 2675
+    },
+    {
+      "epoch": 2.172665764546685,
+      "grad_norm": 0.23034100234508514,
+      "learning_rate": 2.138467309286709e-06,
+      "loss": 0.3401,
+      "step": 2676
+    },
+    {
+      "epoch": 2.1734776725304465,
+      "grad_norm": 0.19266951084136963,
+      "learning_rate": 2.1345922255721745e-06,
+      "loss": 0.3189,
+      "step": 2677
+    },
+    {
+      "epoch": 2.174289580514208,
+      "grad_norm": 0.19394086301326752,
+      "learning_rate": 2.1307197029513276e-06,
+      "loss": 0.3325,
+      "step": 2678
+    },
+    {
+      "epoch": 2.1751014884979702,
+      "grad_norm": 0.19365593791007996,
+      "learning_rate": 2.1268497448854143e-06,
+      "loss": 0.3351,
+      "step": 2679
+    },
+    {
+      "epoch": 2.175913396481732,
+      "grad_norm": 0.19103801250457764,
+      "learning_rate": 2.1229823548333965e-06,
+      "loss": 0.3026,
+      "step": 2680
+    },
+    {
+      "epoch": 2.176725304465494,
+      "grad_norm": 0.20074136555194855,
+      "learning_rate": 2.1191175362519316e-06,
+      "loss": 0.3211,
+      "step": 2681
+    },
+    {
+      "epoch": 2.1775372124492556,
+      "grad_norm": 0.2082689255475998,
+      "learning_rate": 2.115255292595391e-06,
+      "loss": 0.3365,
+      "step": 2682
+    },
+    {
+      "epoch": 2.1783491204330176,
+      "grad_norm": 0.19937241077423096,
+      "learning_rate": 2.111395627315832e-06,
+      "loss": 0.3367,
+      "step": 2683
+    },
+    {
+      "epoch": 2.1791610284167793,
+      "grad_norm": 0.20308703184127808,
+      "learning_rate": 2.1075385438630168e-06,
+      "loss": 0.3205,
+      "step": 2684
+    },
+    {
+      "epoch": 2.1799729364005414,
+      "grad_norm": 0.22848749160766602,
+      "learning_rate": 2.103684045684393e-06,
+      "loss": 0.346,
+      "step": 2685
+    },
+    {
+      "epoch": 2.180784844384303,
+      "grad_norm": 0.19911277294158936,
+      "learning_rate": 2.0998321362251036e-06,
+      "loss": 0.2974,
+      "step": 2686
+    },
+    {
+      "epoch": 2.181596752368065,
+      "grad_norm": 0.19660820066928864,
+      "learning_rate": 2.0959828189279712e-06,
+      "loss": 0.3458,
+      "step": 2687
+    },
+    {
+      "epoch": 2.1824086603518267,
+      "grad_norm": 0.21418237686157227,
+      "learning_rate": 2.0921360972335086e-06,
+      "loss": 0.3194,
+      "step": 2688
+    },
+    {
+      "epoch": 2.1832205683355888,
+      "grad_norm": 0.20105479657649994,
+      "learning_rate": 2.0882919745799023e-06,
+      "loss": 0.319,
+      "step": 2689
+    },
+    {
+      "epoch": 2.1840324763193504,
+      "grad_norm": 0.19248653948307037,
+      "learning_rate": 2.084450454403022e-06,
+      "loss": 0.3193,
+      "step": 2690
+    },
+    {
+      "epoch": 2.1848443843031125,
+      "grad_norm": 0.21864183247089386,
+      "learning_rate": 2.0806115401364036e-06,
+      "loss": 0.3548,
+      "step": 2691
+    },
+    {
+      "epoch": 2.185656292286874,
+      "grad_norm": 0.20074015855789185,
+      "learning_rate": 2.076775235211263e-06,
+      "loss": 0.3314,
+      "step": 2692
+    },
+    {
+      "epoch": 2.186468200270636,
+      "grad_norm": 0.20773935317993164,
+      "learning_rate": 2.0729415430564748e-06,
+      "loss": 0.3292,
+      "step": 2693
+    },
+    {
+      "epoch": 2.187280108254398,
+      "grad_norm": 0.20151840150356293,
+      "learning_rate": 2.0691104670985868e-06,
+      "loss": 0.3208,
+      "step": 2694
+    },
+    {
+      "epoch": 2.18809201623816,
+      "grad_norm": 0.1962594836950302,
+      "learning_rate": 2.0652820107618e-06,
+      "loss": 0.3315,
+      "step": 2695
+    },
+    {
+      "epoch": 2.1889039242219215,
+      "grad_norm": 0.21160277724266052,
+      "learning_rate": 2.0614561774679826e-06,
+      "loss": 0.3433,
+      "step": 2696
+    },
+    {
+      "epoch": 2.1897158322056836,
+      "grad_norm": 0.19455872476100922,
+      "learning_rate": 2.0576329706366494e-06,
+      "loss": 0.3164,
+      "step": 2697
+    },
+    {
+      "epoch": 2.190527740189445,
+      "grad_norm": 0.1972394585609436,
+      "learning_rate": 2.053812393684977e-06,
+      "loss": 0.3567,
+      "step": 2698
+    },
+    {
+      "epoch": 2.191339648173207,
+      "grad_norm": 0.193943589925766,
+      "learning_rate": 2.0499944500277825e-06,
+      "loss": 0.3295,
+      "step": 2699
+    },
+    {
+      "epoch": 2.192151556156969,
+      "grad_norm": 0.20639540255069733,
+      "learning_rate": 2.046179143077535e-06,
+      "loss": 0.3315,
+      "step": 2700
+    },
+    {
+      "epoch": 2.1929634641407305,
+      "grad_norm": 0.22367596626281738,
+      "learning_rate": 2.0423664762443483e-06,
+      "loss": 0.3614,
+      "step": 2701
+    },
+    {
+      "epoch": 2.1937753721244926,
+      "grad_norm": 0.2091004103422165,
+      "learning_rate": 2.0385564529359696e-06,
+      "loss": 0.3827,
+      "step": 2702
+    },
+    {
+      "epoch": 2.194587280108254,
+      "grad_norm": 0.1892516314983368,
+      "learning_rate": 2.034749076557791e-06,
+      "loss": 0.3046,
+      "step": 2703
+    },
+    {
+      "epoch": 2.1953991880920163,
+      "grad_norm": 0.20649179816246033,
+      "learning_rate": 2.030944350512831e-06,
+      "loss": 0.344,
+      "step": 2704
+    },
+    {
+      "epoch": 2.196211096075778,
+      "grad_norm": 0.21210359036922455,
+      "learning_rate": 2.027142278201748e-06,
+      "loss": 0.3047,
+      "step": 2705
+    },
+    {
+      "epoch": 2.19702300405954,
+      "grad_norm": 0.19034118950366974,
+      "learning_rate": 2.023342863022819e-06,
+      "loss": 0.2983,
+      "step": 2706
+    },
+    {
+      "epoch": 2.1978349120433016,
+      "grad_norm": 0.223609060049057,
+      "learning_rate": 2.0195461083719557e-06,
+      "loss": 0.3315,
+      "step": 2707
+    },
+    {
+      "epoch": 2.1986468200270637,
+      "grad_norm": 0.2061319351196289,
+      "learning_rate": 2.015752017642684e-06,
+      "loss": 0.3315,
+      "step": 2708
+    },
+    {
+      "epoch": 2.1994587280108253,
+      "grad_norm": 0.18590758740901947,
+      "learning_rate": 2.0119605942261535e-06,
+      "loss": 0.2899,
+      "step": 2709
+    },
+    {
+      "epoch": 2.2002706359945874,
+      "grad_norm": 0.22332528233528137,
+      "learning_rate": 2.0081718415111263e-06,
+      "loss": 0.3737,
+      "step": 2710
+    },
+    {
+      "epoch": 2.201082543978349,
+      "grad_norm": 0.24580521881580353,
+      "learning_rate": 2.004385762883982e-06,
+      "loss": 0.3316,
+      "step": 2711
+    },
+    {
+      "epoch": 2.201894451962111,
+      "grad_norm": 0.2166924774646759,
+      "learning_rate": 2.000602361728704e-06,
+      "loss": 0.3073,
+      "step": 2712
+    },
+    {
+      "epoch": 2.2027063599458727,
+      "grad_norm": 0.1968022584915161,
+      "learning_rate": 1.99682164142689e-06,
+      "loss": 0.3259,
+      "step": 2713
+    },
+    {
+      "epoch": 2.203518267929635,
+      "grad_norm": 0.18310172855854034,
+      "learning_rate": 1.993043605357733e-06,
+      "loss": 0.3386,
+      "step": 2714
+    },
+    {
+      "epoch": 2.2043301759133964,
+      "grad_norm": 0.20416781306266785,
+      "learning_rate": 1.989268256898036e-06,
+      "loss": 0.3394,
+      "step": 2715
+    },
+    {
+      "epoch": 2.2051420838971585,
+      "grad_norm": 0.2932181656360626,
+      "learning_rate": 1.985495599422191e-06,
+      "loss": 0.3411,
+      "step": 2716
+    },
+    {
+      "epoch": 2.20595399188092,
+      "grad_norm": 0.17935824394226074,
+      "learning_rate": 1.9817256363021907e-06,
+      "loss": 0.3093,
+      "step": 2717
+    },
+    {
+      "epoch": 2.206765899864682,
+      "grad_norm": 0.21472738683223724,
+      "learning_rate": 1.9779583709076207e-06,
+      "loss": 0.3134,
+      "step": 2718
+    },
+    {
+      "epoch": 2.207577807848444,
+      "grad_norm": 0.19384266436100006,
+      "learning_rate": 1.9741938066056475e-06,
+      "loss": 0.3321,
+      "step": 2719
+    },
+    {
+      "epoch": 2.208389715832206,
+      "grad_norm": 0.20872431993484497,
+      "learning_rate": 1.970431946761031e-06,
+      "loss": 0.3456,
+      "step": 2720
+    },
+    {
+      "epoch": 2.2092016238159675,
+      "grad_norm": 0.20243212580680847,
+      "learning_rate": 1.9666727947361135e-06,
+      "loss": 0.3298,
+      "step": 2721
+    },
+    {
+      "epoch": 2.2100135317997296,
+      "grad_norm": 0.20460738241672516,
+      "learning_rate": 1.962916353890811e-06,
+      "loss": 0.3266,
+      "step": 2722
+    },
+    {
+      "epoch": 2.210825439783491,
+      "grad_norm": 0.19484153389930725,
+      "learning_rate": 1.959162627582623e-06,
+      "loss": 0.3156,
+      "step": 2723
+    },
+    {
+      "epoch": 2.211637347767253,
+      "grad_norm": 0.23214007914066315,
+      "learning_rate": 1.955411619166617e-06,
+      "loss": 0.3396,
+      "step": 2724
+    },
+    {
+      "epoch": 2.212449255751015,
+      "grad_norm": 0.18918152153491974,
+      "learning_rate": 1.951663331995436e-06,
+      "loss": 0.2997,
+      "step": 2725
+    },
+    {
+      "epoch": 2.2132611637347765,
+      "grad_norm": 0.1921251267194748,
+      "learning_rate": 1.94791776941929e-06,
+      "loss": 0.3441,
+      "step": 2726
+    },
+    {
+      "epoch": 2.2140730717185386,
+      "grad_norm": 0.19895407557487488,
+      "learning_rate": 1.9441749347859503e-06,
+      "loss": 0.3042,
+      "step": 2727
+    },
+    {
+      "epoch": 2.2148849797023002,
+      "grad_norm": 0.18331022560596466,
+      "learning_rate": 1.9404348314407544e-06,
+      "loss": 0.3235,
+      "step": 2728
+    },
+    {
+      "epoch": 2.2156968876860623,
+      "grad_norm": 0.22145287692546844,
+      "learning_rate": 1.9366974627265945e-06,
+      "loss": 0.3843,
+      "step": 2729
+    },
+    {
+      "epoch": 2.216508795669824,
+      "grad_norm": 0.1870713233947754,
+      "learning_rate": 1.9329628319839227e-06,
+      "loss": 0.2915,
+      "step": 2730
+    },
+    {
+      "epoch": 2.217320703653586,
+      "grad_norm": 0.212056964635849,
+      "learning_rate": 1.929230942550739e-06,
+      "loss": 0.322,
+      "step": 2731
+    },
+    {
+      "epoch": 2.2181326116373477,
+      "grad_norm": 0.197708860039711,
+      "learning_rate": 1.9255017977625995e-06,
+      "loss": 0.3237,
+      "step": 2732
+    },
+    {
+      "epoch": 2.2189445196211097,
+      "grad_norm": 0.2124820053577423,
+      "learning_rate": 1.9217754009526e-06,
+      "loss": 0.3297,
+      "step": 2733
+    },
+    {
+      "epoch": 2.2197564276048714,
+      "grad_norm": 0.20985586941242218,
+      "learning_rate": 1.9180517554513884e-06,
+      "loss": 0.3223,
+      "step": 2734
+    },
+    {
+      "epoch": 2.2205683355886334,
+      "grad_norm": 0.18615496158599854,
+      "learning_rate": 1.914330864587144e-06,
+      "loss": 0.3218,
+      "step": 2735
+    },
+    {
+      "epoch": 2.221380243572395,
+      "grad_norm": 0.18996688723564148,
+      "learning_rate": 1.910612731685593e-06,
+      "loss": 0.3226,
+      "step": 2736
+    },
+    {
+      "epoch": 2.222192151556157,
+      "grad_norm": 0.19439645111560822,
+      "learning_rate": 1.9068973600699887e-06,
+      "loss": 0.3521,
+      "step": 2737
+    },
+    {
+      "epoch": 2.2230040595399188,
+      "grad_norm": 0.19310873746871948,
+      "learning_rate": 1.9031847530611237e-06,
+      "loss": 0.3018,
+      "step": 2738
+    },
+    {
+      "epoch": 2.223815967523681,
+      "grad_norm": 0.20303863286972046,
+      "learning_rate": 1.8994749139773133e-06,
+      "loss": 0.3429,
+      "step": 2739
+    },
+    {
+      "epoch": 2.2246278755074425,
+      "grad_norm": 0.19008158147335052,
+      "learning_rate": 1.8957678461344048e-06,
+      "loss": 0.2937,
+      "step": 2740
+    },
+    {
+      "epoch": 2.2254397834912045,
+      "grad_norm": 0.17495815455913544,
+      "learning_rate": 1.892063552845761e-06,
+      "loss": 0.3084,
+      "step": 2741
+    },
+    {
+      "epoch": 2.226251691474966,
+      "grad_norm": 0.1992001086473465,
+      "learning_rate": 1.888362037422274e-06,
+      "loss": 0.3512,
+      "step": 2742
+    },
+    {
+      "epoch": 2.2270635994587282,
+      "grad_norm": 0.18336911499500275,
+      "learning_rate": 1.8846633031723438e-06,
+      "loss": 0.2986,
+      "step": 2743
+    },
+    {
+      "epoch": 2.22787550744249,
+      "grad_norm": 0.20365765690803528,
+      "learning_rate": 1.8809673534018908e-06,
+      "loss": 0.3346,
+      "step": 2744
+    },
+    {
+      "epoch": 2.2286874154262515,
+      "grad_norm": 0.20219756662845612,
+      "learning_rate": 1.8772741914143472e-06,
+      "loss": 0.348,
+      "step": 2745
+    },
+    {
+      "epoch": 2.2294993234100136,
+      "grad_norm": 0.19700740277767181,
+      "learning_rate": 1.873583820510647e-06,
+      "loss": 0.3358,
+      "step": 2746
+    },
+    {
+      "epoch": 2.230311231393775,
+      "grad_norm": 0.20072822272777557,
+      "learning_rate": 1.8698962439892383e-06,
+      "loss": 0.3527,
+      "step": 2747
+    },
+    {
+      "epoch": 2.2311231393775373,
+      "grad_norm": 0.1896645724773407,
+      "learning_rate": 1.8662114651460617e-06,
+      "loss": 0.3067,
+      "step": 2748
+    },
+    {
+      "epoch": 2.231935047361299,
+      "grad_norm": 0.194441556930542,
+      "learning_rate": 1.8625294872745675e-06,
+      "loss": 0.3544,
+      "step": 2749
+    },
+    {
+      "epoch": 2.232746955345061,
+      "grad_norm": 0.2042710781097412,
+      "learning_rate": 1.8588503136656938e-06,
+      "loss": 0.331,
+      "step": 2750
+    },
+    {
+      "epoch": 2.2335588633288226,
+      "grad_norm": 0.19736871123313904,
+      "learning_rate": 1.8551739476078784e-06,
+      "loss": 0.3179,
+      "step": 2751
+    },
+    {
+      "epoch": 2.2343707713125847,
+      "grad_norm": 0.20771640539169312,
+      "learning_rate": 1.851500392387045e-06,
+      "loss": 0.332,
+      "step": 2752
+    },
+    {
+      "epoch": 2.2351826792963463,
+      "grad_norm": 0.19622039794921875,
+      "learning_rate": 1.8478296512866105e-06,
+      "loss": 0.3118,
+      "step": 2753
+    },
+    {
+      "epoch": 2.2359945872801084,
+      "grad_norm": 0.18444985151290894,
+      "learning_rate": 1.8441617275874701e-06,
+      "loss": 0.3473,
+      "step": 2754
+    },
+    {
+      "epoch": 2.23680649526387,
+      "grad_norm": 0.2314058244228363,
+      "learning_rate": 1.8404966245680068e-06,
+      "loss": 0.3383,
+      "step": 2755
+    },
+    {
+      "epoch": 2.237618403247632,
+      "grad_norm": 0.19246596097946167,
+      "learning_rate": 1.836834345504076e-06,
+      "loss": 0.3084,
+      "step": 2756
+    },
+    {
+      "epoch": 2.2384303112313937,
+      "grad_norm": 0.21328933537006378,
+      "learning_rate": 1.8331748936690175e-06,
+      "loss": 0.3734,
+      "step": 2757
+    },
+    {
+      "epoch": 2.2392422192151558,
+      "grad_norm": 0.18760091066360474,
+      "learning_rate": 1.829518272333633e-06,
+      "loss": 0.311,
+      "step": 2758
+    },
+    {
+      "epoch": 2.2400541271989174,
+      "grad_norm": 0.19140098989009857,
+      "learning_rate": 1.8258644847662087e-06,
+      "loss": 0.3177,
+      "step": 2759
+    },
+    {
+      "epoch": 2.2408660351826795,
+      "grad_norm": 0.23402103781700134,
+      "learning_rate": 1.8222135342324844e-06,
+      "loss": 0.3134,
+      "step": 2760
+    },
+    {
+      "epoch": 2.241677943166441,
+      "grad_norm": 0.19101478159427643,
+      "learning_rate": 1.8185654239956723e-06,
+      "loss": 0.339,
+      "step": 2761
+    },
+    {
+      "epoch": 2.242489851150203,
+      "grad_norm": 0.20588520169258118,
+      "learning_rate": 1.8149201573164415e-06,
+      "loss": 0.3194,
+      "step": 2762
+    },
+    {
+      "epoch": 2.243301759133965,
+      "grad_norm": 0.19614145159721375,
+      "learning_rate": 1.811277737452924e-06,
+      "loss": 0.3257,
+      "step": 2763
+    },
+    {
+      "epoch": 2.244113667117727,
+      "grad_norm": 0.18740014731884003,
+      "learning_rate": 1.807638167660702e-06,
+      "loss": 0.3374,
+      "step": 2764
+    },
+    {
+      "epoch": 2.2449255751014885,
+      "grad_norm": 0.22859041392803192,
+      "learning_rate": 1.8040014511928155e-06,
+      "loss": 0.3112,
+      "step": 2765
+    },
+    {
+      "epoch": 2.24573748308525,
+      "grad_norm": 0.1751399040222168,
+      "learning_rate": 1.8003675912997487e-06,
+      "loss": 0.3139,
+      "step": 2766
+    },
+    {
+      "epoch": 2.246549391069012,
+      "grad_norm": 0.2003294825553894,
+      "learning_rate": 1.7967365912294398e-06,
+      "loss": 0.334,
+      "step": 2767
+    },
+    {
+      "epoch": 2.247361299052774,
+      "grad_norm": 0.19750528037548065,
+      "learning_rate": 1.7931084542272632e-06,
+      "loss": 0.3187,
+      "step": 2768
+    },
+    {
+      "epoch": 2.248173207036536,
+      "grad_norm": 0.2001194804906845,
+      "learning_rate": 1.7894831835360387e-06,
+      "loss": 0.3617,
+      "step": 2769
+    },
+    {
+      "epoch": 2.2489851150202975,
+      "grad_norm": 0.18170025944709778,
+      "learning_rate": 1.7858607823960262e-06,
+      "loss": 0.2995,
+      "step": 2770
+    },
+    {
+      "epoch": 2.2497970230040596,
+      "grad_norm": 0.1911078840494156,
+      "learning_rate": 1.7822412540449141e-06,
+      "loss": 0.3077,
+      "step": 2771
+    },
+    {
+      "epoch": 2.250608930987821,
+      "grad_norm": 0.1887860745191574,
+      "learning_rate": 1.77862460171783e-06,
+      "loss": 0.3348,
+      "step": 2772
+    },
+    {
+      "epoch": 2.2514208389715833,
+      "grad_norm": 0.19924630224704742,
+      "learning_rate": 1.7750108286473255e-06,
+      "loss": 0.3414,
+      "step": 2773
+    },
+    {
+      "epoch": 2.252232746955345,
+      "grad_norm": 0.1870625913143158,
+      "learning_rate": 1.771399938063384e-06,
+      "loss": 0.3445,
+      "step": 2774
+    },
+    {
+      "epoch": 2.253044654939107,
+      "grad_norm": 0.1835235208272934,
+      "learning_rate": 1.7677919331934057e-06,
+      "loss": 0.3222,
+      "step": 2775
+    },
+    {
+      "epoch": 2.2538565629228686,
+      "grad_norm": 0.19843874871730804,
+      "learning_rate": 1.7641868172622201e-06,
+      "loss": 0.3532,
+      "step": 2776
+    },
+    {
+      "epoch": 2.2546684709066307,
+      "grad_norm": 0.20303508639335632,
+      "learning_rate": 1.7605845934920663e-06,
+      "loss": 0.3117,
+      "step": 2777
+    },
+    {
+      "epoch": 2.2554803788903923,
+      "grad_norm": 0.2085290104150772,
+      "learning_rate": 1.7569852651026048e-06,
+      "loss": 0.3322,
+      "step": 2778
+    },
+    {
+      "epoch": 2.2562922868741544,
+      "grad_norm": 0.1869528889656067,
+      "learning_rate": 1.7533888353109029e-06,
+      "loss": 0.3381,
+      "step": 2779
+    },
+    {
+      "epoch": 2.257104194857916,
+      "grad_norm": 0.20236369967460632,
+      "learning_rate": 1.7497953073314433e-06,
+      "loss": 0.3584,
+      "step": 2780
+    },
+    {
+      "epoch": 2.257916102841678,
+      "grad_norm": 0.20568716526031494,
+      "learning_rate": 1.7462046843761078e-06,
+      "loss": 0.3161,
+      "step": 2781
+    },
+    {
+      "epoch": 2.2587280108254397,
+      "grad_norm": 0.19296897947788239,
+      "learning_rate": 1.7426169696541884e-06,
+      "loss": 0.3245,
+      "step": 2782
+    },
+    {
+      "epoch": 2.259539918809202,
+      "grad_norm": 0.19353382289409637,
+      "learning_rate": 1.739032166372372e-06,
+      "loss": 0.337,
+      "step": 2783
+    },
+    {
+      "epoch": 2.2603518267929634,
+      "grad_norm": 0.1905898004770279,
+      "learning_rate": 1.7354502777347487e-06,
+      "loss": 0.3016,
+      "step": 2784
+    },
+    {
+      "epoch": 2.2611637347767255,
+      "grad_norm": 0.19753225147724152,
+      "learning_rate": 1.731871306942799e-06,
+      "loss": 0.3441,
+      "step": 2785
+    },
+    {
+      "epoch": 2.261975642760487,
+      "grad_norm": 0.195291668176651,
+      "learning_rate": 1.7282952571953987e-06,
+      "loss": 0.332,
+      "step": 2786
+    },
+    {
+      "epoch": 2.2627875507442488,
+      "grad_norm": 0.18161047995090485,
+      "learning_rate": 1.724722131688809e-06,
+      "loss": 0.317,
+      "step": 2787
+    },
+    {
+      "epoch": 2.263599458728011,
+      "grad_norm": 0.20284727215766907,
+      "learning_rate": 1.721151933616682e-06,
+      "loss": 0.3542,
+      "step": 2788
+    },
+    {
+      "epoch": 2.264411366711773,
+      "grad_norm": 0.19864869117736816,
+      "learning_rate": 1.7175846661700518e-06,
+      "loss": 0.3331,
+      "step": 2789
+    },
+    {
+      "epoch": 2.2652232746955345,
+      "grad_norm": 0.18836060166358948,
+      "learning_rate": 1.7140203325373294e-06,
+      "loss": 0.3184,
+      "step": 2790
+    },
+    {
+      "epoch": 2.266035182679296,
+      "grad_norm": 0.21382378041744232,
+      "learning_rate": 1.7104589359043095e-06,
+      "loss": 0.3715,
+      "step": 2791
+    },
+    {
+      "epoch": 2.2668470906630582,
+      "grad_norm": 0.1958325207233429,
+      "learning_rate": 1.706900479454156e-06,
+      "loss": 0.3117,
+      "step": 2792
+    },
+    {
+      "epoch": 2.26765899864682,
+      "grad_norm": 0.1912974864244461,
+      "learning_rate": 1.7033449663674102e-06,
+      "loss": 0.2918,
+      "step": 2793
+    },
+    {
+      "epoch": 2.268470906630582,
+      "grad_norm": 0.22348739206790924,
+      "learning_rate": 1.6997923998219774e-06,
+      "loss": 0.355,
+      "step": 2794
+    },
+    {
+      "epoch": 2.2692828146143436,
+      "grad_norm": 0.2093338966369629,
+      "learning_rate": 1.6962427829931348e-06,
+      "loss": 0.3568,
+      "step": 2795
+    },
+    {
+      "epoch": 2.2700947225981056,
+      "grad_norm": 0.18856249749660492,
+      "learning_rate": 1.6926961190535152e-06,
+      "loss": 0.2963,
+      "step": 2796
+    },
+    {
+      "epoch": 2.2709066305818673,
+      "grad_norm": 0.1943671852350235,
+      "learning_rate": 1.6891524111731246e-06,
+      "loss": 0.3502,
+      "step": 2797
+    },
+    {
+      "epoch": 2.2717185385656293,
+      "grad_norm": 0.1829935908317566,
+      "learning_rate": 1.6856116625193135e-06,
+      "loss": 0.2924,
+      "step": 2798
+    },
+    {
+      "epoch": 2.272530446549391,
+      "grad_norm": 0.20973965525627136,
+      "learning_rate": 1.6820738762567978e-06,
+      "loss": 0.3769,
+      "step": 2799
+    },
+    {
+      "epoch": 2.273342354533153,
+      "grad_norm": 0.21303915977478027,
+      "learning_rate": 1.678539055547637e-06,
+      "loss": 0.2947,
+      "step": 2800
+    },
+    {
+      "epoch": 2.2741542625169147,
+      "grad_norm": 0.2122032642364502,
+      "learning_rate": 1.6750072035512482e-06,
+      "loss": 0.3236,
+      "step": 2801
+    },
+    {
+      "epoch": 2.2749661705006767,
+      "grad_norm": 0.20635607838630676,
+      "learning_rate": 1.6714783234243875e-06,
+      "loss": 0.3208,
+      "step": 2802
+    },
+    {
+      "epoch": 2.2757780784844384,
+      "grad_norm": 0.19804933667182922,
+      "learning_rate": 1.6679524183211616e-06,
+      "loss": 0.3403,
+      "step": 2803
+    },
+    {
+      "epoch": 2.2765899864682004,
+      "grad_norm": 0.1975712776184082,
+      "learning_rate": 1.6644294913930115e-06,
+      "loss": 0.3345,
+      "step": 2804
+    },
+    {
+      "epoch": 2.277401894451962,
+      "grad_norm": 0.19292321801185608,
+      "learning_rate": 1.6609095457887237e-06,
+      "loss": 0.3541,
+      "step": 2805
+    },
+    {
+      "epoch": 2.278213802435724,
+      "grad_norm": 0.18958859145641327,
+      "learning_rate": 1.657392584654412e-06,
+      "loss": 0.3197,
+      "step": 2806
+    },
+    {
+      "epoch": 2.2790257104194858,
+      "grad_norm": 0.2157866209745407,
+      "learning_rate": 1.6538786111335303e-06,
+      "loss": 0.3195,
+      "step": 2807
+    },
+    {
+      "epoch": 2.2798376184032474,
+      "grad_norm": 0.19900549948215485,
+      "learning_rate": 1.6503676283668553e-06,
+      "loss": 0.2994,
+      "step": 2808
+    },
+    {
+      "epoch": 2.2806495263870095,
+      "grad_norm": 0.2425675392150879,
+      "learning_rate": 1.6468596394924974e-06,
+      "loss": 0.3294,
+      "step": 2809
+    },
+    {
+      "epoch": 2.2814614343707715,
+      "grad_norm": 0.20827148854732513,
+      "learning_rate": 1.6433546476458833e-06,
+      "loss": 0.3293,
+      "step": 2810
+    },
+    {
+      "epoch": 2.282273342354533,
+      "grad_norm": 0.19143660366535187,
+      "learning_rate": 1.63985265595977e-06,
+      "loss": 0.34,
+      "step": 2811
+    },
+    {
+      "epoch": 2.283085250338295,
+      "grad_norm": 0.24166226387023926,
+      "learning_rate": 1.6363536675642233e-06,
+      "loss": 0.3513,
+      "step": 2812
+    },
+    {
+      "epoch": 2.283897158322057,
+      "grad_norm": 0.200519397854805,
+      "learning_rate": 1.632857685586632e-06,
+      "loss": 0.2981,
+      "step": 2813
+    },
+    {
+      "epoch": 2.2847090663058185,
+      "grad_norm": 0.2051820009946823,
+      "learning_rate": 1.629364713151697e-06,
+      "loss": 0.3173,
+      "step": 2814
+    },
+    {
+      "epoch": 2.2855209742895806,
+      "grad_norm": 0.21330997347831726,
+      "learning_rate": 1.625874753381424e-06,
+      "loss": 0.3518,
+      "step": 2815
+    },
+    {
+      "epoch": 2.286332882273342,
+      "grad_norm": 0.19235005974769592,
+      "learning_rate": 1.6223878093951329e-06,
+      "loss": 0.311,
+      "step": 2816
+    },
+    {
+      "epoch": 2.2871447902571043,
+      "grad_norm": 0.19947923719882965,
+      "learning_rate": 1.6189038843094419e-06,
+      "loss": 0.3717,
+      "step": 2817
+    },
+    {
+      "epoch": 2.287956698240866,
+      "grad_norm": 0.19066692888736725,
+      "learning_rate": 1.615422981238276e-06,
+      "loss": 0.3355,
+      "step": 2818
+    },
+    {
+      "epoch": 2.288768606224628,
+      "grad_norm": 0.18103188276290894,
+      "learning_rate": 1.6119451032928547e-06,
+      "loss": 0.2994,
+      "step": 2819
+    },
+    {
+      "epoch": 2.2895805142083896,
+      "grad_norm": 0.18511269986629486,
+      "learning_rate": 1.6084702535816987e-06,
+      "loss": 0.3534,
+      "step": 2820
+    },
+    {
+      "epoch": 2.2903924221921517,
+      "grad_norm": 0.19247010350227356,
+      "learning_rate": 1.6049984352106163e-06,
+      "loss": 0.3343,
+      "step": 2821
+    },
+    {
+      "epoch": 2.2912043301759133,
+      "grad_norm": 0.1847759485244751,
+      "learning_rate": 1.6015296512827122e-06,
+      "loss": 0.3248,
+      "step": 2822
+    },
+    {
+      "epoch": 2.2920162381596754,
+      "grad_norm": 0.19152933359146118,
+      "learning_rate": 1.5980639048983726e-06,
+      "loss": 0.2989,
+      "step": 2823
+    },
+    {
+      "epoch": 2.292828146143437,
+      "grad_norm": 0.2042330503463745,
+      "learning_rate": 1.5946011991552757e-06,
+      "loss": 0.3261,
+      "step": 2824
+    },
+    {
+      "epoch": 2.293640054127199,
+      "grad_norm": 0.20546279847621918,
+      "learning_rate": 1.5911415371483757e-06,
+      "loss": 0.3284,
+      "step": 2825
+    },
+    {
+      "epoch": 2.2944519621109607,
+      "grad_norm": 0.1969582587480545,
+      "learning_rate": 1.587684921969912e-06,
+      "loss": 0.338,
+      "step": 2826
+    },
+    {
+      "epoch": 2.2952638700947228,
+      "grad_norm": 0.18959590792655945,
+      "learning_rate": 1.5842313567093942e-06,
+      "loss": 0.319,
+      "step": 2827
+    },
+    {
+      "epoch": 2.2960757780784844,
+      "grad_norm": 0.1920270323753357,
+      "learning_rate": 1.5807808444536132e-06,
+      "loss": 0.304,
+      "step": 2828
+    },
+    {
+      "epoch": 2.2968876860622465,
+      "grad_norm": 0.20454080402851105,
+      "learning_rate": 1.5773333882866242e-06,
+      "loss": 0.3247,
+      "step": 2829
+    },
+    {
+      "epoch": 2.297699594046008,
+      "grad_norm": 0.22462552785873413,
+      "learning_rate": 1.5738889912897575e-06,
+      "loss": 0.3489,
+      "step": 2830
+    },
+    {
+      "epoch": 2.29851150202977,
+      "grad_norm": 0.20337285101413727,
+      "learning_rate": 1.5704476565416027e-06,
+      "loss": 0.3629,
+      "step": 2831
+    },
+    {
+      "epoch": 2.299323410013532,
+      "grad_norm": 0.18424689769744873,
+      "learning_rate": 1.5670093871180165e-06,
+      "loss": 0.2794,
+      "step": 2832
+    },
+    {
+      "epoch": 2.3001353179972934,
+      "grad_norm": 0.18091656267642975,
+      "learning_rate": 1.5635741860921166e-06,
+      "loss": 0.3203,
+      "step": 2833
+    },
+    {
+      "epoch": 2.3009472259810555,
+      "grad_norm": 0.20084039866924286,
+      "learning_rate": 1.5601420565342734e-06,
+      "loss": 0.3218,
+      "step": 2834
+    },
+    {
+      "epoch": 2.301759133964817,
+      "grad_norm": 0.20008179545402527,
+      "learning_rate": 1.5567130015121169e-06,
+      "loss": 0.3252,
+      "step": 2835
+    },
+    {
+      "epoch": 2.302571041948579,
+      "grad_norm": 0.19702613353729248,
+      "learning_rate": 1.5532870240905272e-06,
+      "loss": 0.3685,
+      "step": 2836
+    },
+    {
+      "epoch": 2.303382949932341,
+      "grad_norm": 0.19093456864356995,
+      "learning_rate": 1.5498641273316322e-06,
+      "loss": 0.29,
+      "step": 2837
+    },
+    {
+      "epoch": 2.304194857916103,
+      "grad_norm": 0.19893111288547516,
+      "learning_rate": 1.5464443142948083e-06,
+      "loss": 0.3257,
+      "step": 2838
+    },
+    {
+      "epoch": 2.3050067658998645,
+      "grad_norm": 0.1792554259300232,
+      "learning_rate": 1.5430275880366774e-06,
+      "loss": 0.3218,
+      "step": 2839
+    },
+    {
+      "epoch": 2.3058186738836266,
+      "grad_norm": 0.1939542591571808,
+      "learning_rate": 1.5396139516110959e-06,
+      "loss": 0.3224,
+      "step": 2840
+    },
+    {
+      "epoch": 2.3066305818673882,
+      "grad_norm": 0.2188585102558136,
+      "learning_rate": 1.5362034080691673e-06,
+      "loss": 0.3157,
+      "step": 2841
+    },
+    {
+      "epoch": 2.3074424898511503,
+      "grad_norm": 0.20758715271949768,
+      "learning_rate": 1.5327959604592218e-06,
+      "loss": 0.3292,
+      "step": 2842
+    },
+    {
+      "epoch": 2.308254397834912,
+      "grad_norm": 0.20759567618370056,
+      "learning_rate": 1.5293916118268304e-06,
+      "loss": 0.3254,
+      "step": 2843
+    },
+    {
+      "epoch": 2.309066305818674,
+      "grad_norm": 0.1941070407629013,
+      "learning_rate": 1.5259903652147878e-06,
+      "loss": 0.3159,
+      "step": 2844
+    },
+    {
+      "epoch": 2.3098782138024356,
+      "grad_norm": 0.1911478489637375,
+      "learning_rate": 1.5225922236631225e-06,
+      "loss": 0.2925,
+      "step": 2845
+    },
+    {
+      "epoch": 2.3106901217861977,
+      "grad_norm": 0.2098912000656128,
+      "learning_rate": 1.5191971902090797e-06,
+      "loss": 0.3542,
+      "step": 2846
+    },
+    {
+      "epoch": 2.3115020297699593,
+      "grad_norm": 0.23497377336025238,
+      "learning_rate": 1.515805267887135e-06,
+      "loss": 0.3594,
+      "step": 2847
+    },
+    {
+      "epoch": 2.3123139377537214,
+      "grad_norm": 0.17733074724674225,
+      "learning_rate": 1.5124164597289765e-06,
+      "loss": 0.2798,
+      "step": 2848
+    },
+    {
+      "epoch": 2.313125845737483,
+      "grad_norm": 0.19751563668251038,
+      "learning_rate": 1.5090307687635141e-06,
+      "loss": 0.3544,
+      "step": 2849
+    },
+    {
+      "epoch": 2.313937753721245,
+      "grad_norm": 0.19295023381710052,
+      "learning_rate": 1.505648198016867e-06,
+      "loss": 0.3475,
+      "step": 2850
+    },
+    {
+      "epoch": 2.3147496617050067,
+      "grad_norm": 0.1934005469083786,
+      "learning_rate": 1.5022687505123702e-06,
+      "loss": 0.3009,
+      "step": 2851
+    },
+    {
+      "epoch": 2.315561569688769,
+      "grad_norm": 0.1989288181066513,
+      "learning_rate": 1.4988924292705615e-06,
+      "loss": 0.33,
+      "step": 2852
+    },
+    {
+      "epoch": 2.3163734776725304,
+      "grad_norm": 0.20251977443695068,
+      "learning_rate": 1.4955192373091915e-06,
+      "loss": 0.3377,
+      "step": 2853
+    },
+    {
+      "epoch": 2.317185385656292,
+      "grad_norm": 0.19406268000602722,
+      "learning_rate": 1.4921491776432062e-06,
+      "loss": 0.3333,
+      "step": 2854
+    },
+    {
+      "epoch": 2.317997293640054,
+      "grad_norm": 0.20036713778972626,
+      "learning_rate": 1.488782253284759e-06,
+      "loss": 0.3013,
+      "step": 2855
+    },
+    {
+      "epoch": 2.318809201623816,
+      "grad_norm": 0.20798751711845398,
+      "learning_rate": 1.4854184672431948e-06,
+      "loss": 0.3577,
+      "step": 2856
+    },
+    {
+      "epoch": 2.319621109607578,
+      "grad_norm": 0.19803696870803833,
+      "learning_rate": 1.4820578225250582e-06,
+      "loss": 0.329,
+      "step": 2857
+    },
+    {
+      "epoch": 2.3204330175913395,
+      "grad_norm": 0.19647663831710815,
+      "learning_rate": 1.4787003221340851e-06,
+      "loss": 0.3157,
+      "step": 2858
+    },
+    {
+      "epoch": 2.3212449255751015,
+      "grad_norm": 0.19605553150177002,
+      "learning_rate": 1.475345969071198e-06,
+      "loss": 0.35,
+      "step": 2859
+    },
+    {
+      "epoch": 2.322056833558863,
+      "grad_norm": 0.20682701468467712,
+      "learning_rate": 1.4719947663345112e-06,
+      "loss": 0.3276,
+      "step": 2860
+    },
+    {
+      "epoch": 2.3228687415426252,
+      "grad_norm": 0.19280420243740082,
+      "learning_rate": 1.468646716919318e-06,
+      "loss": 0.3212,
+      "step": 2861
+    },
+    {
+      "epoch": 2.323680649526387,
+      "grad_norm": 0.20546646416187286,
+      "learning_rate": 1.4653018238180982e-06,
+      "loss": 0.3618,
+      "step": 2862
+    },
+    {
+      "epoch": 2.324492557510149,
+      "grad_norm": 0.18528850376605988,
+      "learning_rate": 1.461960090020505e-06,
+      "loss": 0.3192,
+      "step": 2863
+    },
+    {
+      "epoch": 2.3253044654939106,
+      "grad_norm": 0.19321291148662567,
+      "learning_rate": 1.458621518513375e-06,
+      "loss": 0.3186,
+      "step": 2864
+    },
+    {
+      "epoch": 2.3261163734776726,
+      "grad_norm": 0.2004014402627945,
+      "learning_rate": 1.455286112280711e-06,
+      "loss": 0.3086,
+      "step": 2865
+    },
+    {
+      "epoch": 2.3269282814614343,
+      "grad_norm": 0.19191843271255493,
+      "learning_rate": 1.4519538743036927e-06,
+      "loss": 0.3265,
+      "step": 2866
+    },
+    {
+      "epoch": 2.3277401894451963,
+      "grad_norm": 0.20411525666713715,
+      "learning_rate": 1.4486248075606624e-06,
+      "loss": 0.341,
+      "step": 2867
+    },
+    {
+      "epoch": 2.328552097428958,
+      "grad_norm": 0.19159169495105743,
+      "learning_rate": 1.445298915027134e-06,
+      "loss": 0.2995,
+      "step": 2868
+    },
+    {
+      "epoch": 2.32936400541272,
+      "grad_norm": 0.19621224701404572,
+      "learning_rate": 1.441976199675778e-06,
+      "loss": 0.3365,
+      "step": 2869
+    },
+    {
+      "epoch": 2.3301759133964817,
+      "grad_norm": 0.20010292530059814,
+      "learning_rate": 1.4386566644764321e-06,
+      "loss": 0.3333,
+      "step": 2870
+    },
+    {
+      "epoch": 2.3309878213802437,
+      "grad_norm": 0.19238407909870148,
+      "learning_rate": 1.435340312396084e-06,
+      "loss": 0.3042,
+      "step": 2871
+    },
+    {
+      "epoch": 2.3317997293640054,
+      "grad_norm": 0.20835940539836884,
+      "learning_rate": 1.4320271463988822e-06,
+      "loss": 0.3482,
+      "step": 2872
+    },
+    {
+      "epoch": 2.3326116373477674,
+      "grad_norm": 0.19879306852817535,
+      "learning_rate": 1.428717169446125e-06,
+      "loss": 0.3037,
+      "step": 2873
+    },
+    {
+      "epoch": 2.333423545331529,
+      "grad_norm": 0.2168373465538025,
+      "learning_rate": 1.425410384496263e-06,
+      "loss": 0.3598,
+      "step": 2874
+    },
+    {
+      "epoch": 2.3342354533152907,
+      "grad_norm": 0.1968943178653717,
+      "learning_rate": 1.4221067945048877e-06,
+      "loss": 0.3244,
+      "step": 2875
+    },
+    {
+      "epoch": 2.3350473612990528,
+      "grad_norm": 0.17006225883960724,
+      "learning_rate": 1.4188064024247428e-06,
+      "loss": 0.279,
+      "step": 2876
+    },
+    {
+      "epoch": 2.335859269282815,
+      "grad_norm": 0.19064393639564514,
+      "learning_rate": 1.415509211205706e-06,
+      "loss": 0.3884,
+      "step": 2877
+    },
+    {
+      "epoch": 2.3366711772665765,
+      "grad_norm": 0.1822357326745987,
+      "learning_rate": 1.4122152237948018e-06,
+      "loss": 0.3041,
+      "step": 2878
+    },
+    {
+      "epoch": 2.337483085250338,
+      "grad_norm": 0.20772208273410797,
+      "learning_rate": 1.4089244431361826e-06,
+      "loss": 0.3413,
+      "step": 2879
+    },
+    {
+      "epoch": 2.3382949932341,
+      "grad_norm": 0.2106371372938156,
+      "learning_rate": 1.405636872171144e-06,
+      "loss": 0.3498,
+      "step": 2880
+    },
+    {
+      "epoch": 2.339106901217862,
+      "grad_norm": 0.19371894001960754,
+      "learning_rate": 1.4023525138381039e-06,
+      "loss": 0.3216,
+      "step": 2881
+    },
+    {
+      "epoch": 2.339918809201624,
+      "grad_norm": 0.17456111311912537,
+      "learning_rate": 1.3990713710726145e-06,
+      "loss": 0.3044,
+      "step": 2882
+    },
+    {
+      "epoch": 2.3407307171853855,
+      "grad_norm": 0.2012059986591339,
+      "learning_rate": 1.395793446807354e-06,
+      "loss": 0.3053,
+      "step": 2883
+    },
+    {
+      "epoch": 2.3415426251691476,
+      "grad_norm": 0.20504657924175262,
+      "learning_rate": 1.3925187439721193e-06,
+      "loss": 0.3428,
+      "step": 2884
+    },
+    {
+      "epoch": 2.342354533152909,
+      "grad_norm": 0.18713970482349396,
+      "learning_rate": 1.3892472654938338e-06,
+      "loss": 0.3249,
+      "step": 2885
+    },
+    {
+      "epoch": 2.3431664411366713,
+      "grad_norm": 0.2063145786523819,
+      "learning_rate": 1.385979014296533e-06,
+      "loss": 0.3397,
+      "step": 2886
+    },
+    {
+      "epoch": 2.343978349120433,
+      "grad_norm": 0.19832271337509155,
+      "learning_rate": 1.3827139933013739e-06,
+      "loss": 0.3523,
+      "step": 2887
+    },
+    {
+      "epoch": 2.344790257104195,
+      "grad_norm": 0.19000336527824402,
+      "learning_rate": 1.3794522054266207e-06,
+      "loss": 0.3081,
+      "step": 2888
+    },
+    {
+      "epoch": 2.3456021650879566,
+      "grad_norm": 0.20842434465885162,
+      "learning_rate": 1.376193653587653e-06,
+      "loss": 0.3669,
+      "step": 2889
+    },
+    {
+      "epoch": 2.3464140730717187,
+      "grad_norm": 0.1995701938867569,
+      "learning_rate": 1.3729383406969533e-06,
+      "loss": 0.3301,
+      "step": 2890
+    },
+    {
+      "epoch": 2.3472259810554803,
+      "grad_norm": 0.20981673896312714,
+      "learning_rate": 1.3696862696641133e-06,
+      "loss": 0.3311,
+      "step": 2891
+    },
+    {
+      "epoch": 2.3480378890392424,
+      "grad_norm": 0.2053629457950592,
+      "learning_rate": 1.366437443395824e-06,
+      "loss": 0.3159,
+      "step": 2892
+    },
+    {
+      "epoch": 2.348849797023004,
+      "grad_norm": 0.18204864859580994,
+      "learning_rate": 1.3631918647958786e-06,
+      "loss": 0.3075,
+      "step": 2893
+    },
+    {
+      "epoch": 2.349661705006766,
+      "grad_norm": 0.20075322687625885,
+      "learning_rate": 1.3599495367651645e-06,
+      "loss": 0.3036,
+      "step": 2894
+    },
+    {
+      "epoch": 2.3504736129905277,
+      "grad_norm": 0.20387734472751617,
+      "learning_rate": 1.3567104622016685e-06,
+      "loss": 0.3272,
+      "step": 2895
+    },
+    {
+      "epoch": 2.3512855209742893,
+      "grad_norm": 0.20250089466571808,
+      "learning_rate": 1.3534746440004637e-06,
+      "loss": 0.3156,
+      "step": 2896
+    },
+    {
+      "epoch": 2.3520974289580514,
+      "grad_norm": 0.17583473026752472,
+      "learning_rate": 1.3502420850537185e-06,
+      "loss": 0.3245,
+      "step": 2897
+    },
+    {
+      "epoch": 2.3529093369418135,
+      "grad_norm": 0.19450092315673828,
+      "learning_rate": 1.3470127882506824e-06,
+      "loss": 0.3522,
+      "step": 2898
+    },
+    {
+      "epoch": 2.353721244925575,
+      "grad_norm": 0.2139914631843567,
+      "learning_rate": 1.3437867564776952e-06,
+      "loss": 0.361,
+      "step": 2899
+    },
+    {
+      "epoch": 2.3545331529093367,
+      "grad_norm": 0.19017352163791656,
+      "learning_rate": 1.3405639926181718e-06,
+      "loss": 0.3126,
+      "step": 2900
+    },
+    {
+      "epoch": 2.355345060893099,
+      "grad_norm": 0.1771569848060608,
+      "learning_rate": 1.337344499552612e-06,
+      "loss": 0.283,
+      "step": 2901
+    },
+    {
+      "epoch": 2.3561569688768604,
+      "grad_norm": 0.20805060863494873,
+      "learning_rate": 1.3341282801585914e-06,
+      "loss": 0.367,
+      "step": 2902
+    },
+    {
+      "epoch": 2.3569688768606225,
+      "grad_norm": 0.1791091114282608,
+      "learning_rate": 1.3309153373107547e-06,
+      "loss": 0.3232,
+      "step": 2903
+    },
+    {
+      "epoch": 2.357780784844384,
+      "grad_norm": 0.18573522567749023,
+      "learning_rate": 1.3277056738808242e-06,
+      "loss": 0.319,
+      "step": 2904
+    },
+    {
+      "epoch": 2.358592692828146,
+      "grad_norm": 0.2060178965330124,
+      "learning_rate": 1.3244992927375865e-06,
+      "loss": 0.3469,
+      "step": 2905
+    },
+    {
+      "epoch": 2.359404600811908,
+      "grad_norm": 0.18865546584129333,
+      "learning_rate": 1.3212961967468985e-06,
+      "loss": 0.3205,
+      "step": 2906
+    },
+    {
+      "epoch": 2.36021650879567,
+      "grad_norm": 0.18897266685962677,
+      "learning_rate": 1.3180963887716758e-06,
+      "loss": 0.3371,
+      "step": 2907
+    },
+    {
+      "epoch": 2.3610284167794315,
+      "grad_norm": 0.19292959570884705,
+      "learning_rate": 1.3148998716719014e-06,
+      "loss": 0.3493,
+      "step": 2908
+    },
+    {
+      "epoch": 2.3618403247631936,
+      "grad_norm": 0.19258268177509308,
+      "learning_rate": 1.3117066483046098e-06,
+      "loss": 0.317,
+      "step": 2909
+    },
+    {
+      "epoch": 2.3626522327469552,
+      "grad_norm": 0.1931302845478058,
+      "learning_rate": 1.3085167215238982e-06,
+      "loss": 0.3207,
+      "step": 2910
+    },
+    {
+      "epoch": 2.3634641407307173,
+      "grad_norm": 0.19533517956733704,
+      "learning_rate": 1.3053300941809132e-06,
+      "loss": 0.3498,
+      "step": 2911
+    },
+    {
+      "epoch": 2.364276048714479,
+      "grad_norm": 0.1890619546175003,
+      "learning_rate": 1.302146769123856e-06,
+      "loss": 0.3169,
+      "step": 2912
+    },
+    {
+      "epoch": 2.365087956698241,
+      "grad_norm": 0.18475040793418884,
+      "learning_rate": 1.2989667491979702e-06,
+      "loss": 0.3279,
+      "step": 2913
+    },
+    {
+      "epoch": 2.3658998646820026,
+      "grad_norm": 0.17871330678462982,
+      "learning_rate": 1.2957900372455529e-06,
+      "loss": 0.3267,
+      "step": 2914
+    },
+    {
+      "epoch": 2.3667117726657647,
+      "grad_norm": 0.21414300799369812,
+      "learning_rate": 1.2926166361059366e-06,
+      "loss": 0.3338,
+      "step": 2915
+    },
+    {
+      "epoch": 2.3675236806495263,
+      "grad_norm": 0.18526801466941833,
+      "learning_rate": 1.2894465486155028e-06,
+      "loss": 0.3078,
+      "step": 2916
+    },
+    {
+      "epoch": 2.3683355886332884,
+      "grad_norm": 0.18869346380233765,
+      "learning_rate": 1.2862797776076624e-06,
+      "loss": 0.3329,
+      "step": 2917
+    },
+    {
+      "epoch": 2.36914749661705,
+      "grad_norm": 0.1963004618883133,
+      "learning_rate": 1.2831163259128714e-06,
+      "loss": 0.3069,
+      "step": 2918
+    },
+    {
+      "epoch": 2.369959404600812,
+      "grad_norm": 0.21695619821548462,
+      "learning_rate": 1.2799561963586104e-06,
+      "loss": 0.3367,
+      "step": 2919
+    },
+    {
+      "epoch": 2.3707713125845737,
+      "grad_norm": 0.18267785012722015,
+      "learning_rate": 1.2767993917693984e-06,
+      "loss": 0.3074,
+      "step": 2920
+    },
+    {
+      "epoch": 2.3715832205683354,
+      "grad_norm": 0.178884819149971,
+      "learning_rate": 1.2736459149667747e-06,
+      "loss": 0.3116,
+      "step": 2921
+    },
+    {
+      "epoch": 2.3723951285520974,
+      "grad_norm": 0.19302940368652344,
+      "learning_rate": 1.2704957687693114e-06,
+      "loss": 0.3399,
+      "step": 2922
+    },
+    {
+      "epoch": 2.3732070365358595,
+      "grad_norm": 0.19638453423976898,
+      "learning_rate": 1.2673489559925989e-06,
+      "loss": 0.3184,
+      "step": 2923
+    },
+    {
+      "epoch": 2.374018944519621,
+      "grad_norm": 0.19511476159095764,
+      "learning_rate": 1.2642054794492514e-06,
+      "loss": 0.3524,
+      "step": 2924
+    },
+    {
+      "epoch": 2.3748308525033828,
+      "grad_norm": 0.1733398288488388,
+      "learning_rate": 1.2610653419488978e-06,
+      "loss": 0.3123,
+      "step": 2925
+    },
+    {
+      "epoch": 2.375642760487145,
+      "grad_norm": 0.1919705718755722,
+      "learning_rate": 1.2579285462981855e-06,
+      "loss": 0.3447,
+      "step": 2926
+    },
+    {
+      "epoch": 2.3764546684709065,
+      "grad_norm": 0.20587553083896637,
+      "learning_rate": 1.2547950953007754e-06,
+      "loss": 0.3537,
+      "step": 2927
+    },
+    {
+      "epoch": 2.3772665764546685,
+      "grad_norm": 0.1848933845758438,
+      "learning_rate": 1.2516649917573342e-06,
+      "loss": 0.3253,
+      "step": 2928
+    },
+    {
+      "epoch": 2.37807848443843,
+      "grad_norm": 0.18180131912231445,
+      "learning_rate": 1.2485382384655437e-06,
+      "loss": 0.2871,
+      "step": 2929
+    },
+    {
+      "epoch": 2.3788903924221922,
+      "grad_norm": 0.17156387865543365,
+      "learning_rate": 1.2454148382200837e-06,
+      "loss": 0.3184,
+      "step": 2930
+    },
+    {
+      "epoch": 2.379702300405954,
+      "grad_norm": 0.18868547677993774,
+      "learning_rate": 1.2422947938126445e-06,
+      "loss": 0.3,
+      "step": 2931
+    },
+    {
+      "epoch": 2.380514208389716,
+      "grad_norm": 0.19234851002693176,
+      "learning_rate": 1.2391781080319098e-06,
+      "loss": 0.3425,
+      "step": 2932
+    },
+    {
+      "epoch": 2.3813261163734776,
+      "grad_norm": 0.2022353708744049,
+      "learning_rate": 1.236064783663568e-06,
+      "loss": 0.3518,
+      "step": 2933
+    },
+    {
+      "epoch": 2.3821380243572396,
+      "grad_norm": 0.19255325198173523,
+      "learning_rate": 1.2329548234902976e-06,
+      "loss": 0.3284,
+      "step": 2934
+    },
+    {
+      "epoch": 2.3829499323410013,
+      "grad_norm": 0.1860717236995697,
+      "learning_rate": 1.2298482302917752e-06,
+      "loss": 0.3245,
+      "step": 2935
+    },
+    {
+      "epoch": 2.3837618403247633,
+      "grad_norm": 0.19691437482833862,
+      "learning_rate": 1.2267450068446617e-06,
+      "loss": 0.3315,
+      "step": 2936
+    },
+    {
+      "epoch": 2.384573748308525,
+      "grad_norm": 0.1917988806962967,
+      "learning_rate": 1.2236451559226137e-06,
+      "loss": 0.3168,
+      "step": 2937
+    },
+    {
+      "epoch": 2.385385656292287,
+      "grad_norm": 0.21340516209602356,
+      "learning_rate": 1.2205486802962656e-06,
+      "loss": 0.3505,
+      "step": 2938
+    },
+    {
+      "epoch": 2.3861975642760487,
+      "grad_norm": 0.18313954770565033,
+      "learning_rate": 1.217455582733243e-06,
+      "loss": 0.2964,
+      "step": 2939
+    },
+    {
+      "epoch": 2.3870094722598107,
+      "grad_norm": 0.19814564287662506,
+      "learning_rate": 1.2143658659981444e-06,
+      "loss": 0.3049,
+      "step": 2940
+    },
+    {
+      "epoch": 2.3878213802435724,
+      "grad_norm": 0.1965659111738205,
+      "learning_rate": 1.2112795328525529e-06,
+      "loss": 0.3556,
+      "step": 2941
+    },
+    {
+      "epoch": 2.388633288227334,
+      "grad_norm": 0.19271732866764069,
+      "learning_rate": 1.2081965860550227e-06,
+      "loss": 0.3527,
+      "step": 2942
+    },
+    {
+      "epoch": 2.389445196211096,
+      "grad_norm": 0.18494363129138947,
+      "learning_rate": 1.2051170283610858e-06,
+      "loss": 0.3007,
+      "step": 2943
+    },
+    {
+      "epoch": 2.390257104194858,
+      "grad_norm": 0.1734316200017929,
+      "learning_rate": 1.2020408625232388e-06,
+      "loss": 0.308,
+      "step": 2944
+    },
+    {
+      "epoch": 2.39106901217862,
+      "grad_norm": 0.2032386064529419,
+      "learning_rate": 1.198968091290953e-06,
+      "loss": 0.3545,
+      "step": 2945
+    },
+    {
+      "epoch": 2.3918809201623814,
+      "grad_norm": 0.19505706429481506,
+      "learning_rate": 1.195898717410664e-06,
+      "loss": 0.314,
+      "step": 2946
+    },
+    {
+      "epoch": 2.3926928281461435,
+      "grad_norm": 0.21441584825515747,
+      "learning_rate": 1.1928327436257669e-06,
+      "loss": 0.3648,
+      "step": 2947
+    },
+    {
+      "epoch": 2.393504736129905,
+      "grad_norm": 0.193422332406044,
+      "learning_rate": 1.189770172676623e-06,
+      "loss": 0.3286,
+      "step": 2948
+    },
+    {
+      "epoch": 2.394316644113667,
+      "grad_norm": 0.19635632634162903,
+      "learning_rate": 1.1867110073005483e-06,
+      "loss": 0.3129,
+      "step": 2949
+    },
+    {
+      "epoch": 2.395128552097429,
+      "grad_norm": 0.19537708163261414,
+      "learning_rate": 1.1836552502318171e-06,
+      "loss": 0.3411,
+      "step": 2950
+    },
+    {
+      "epoch": 2.395940460081191,
+      "grad_norm": 0.19608810544013977,
+      "learning_rate": 1.1806029042016581e-06,
+      "loss": 0.3407,
+      "step": 2951
+    },
+    {
+      "epoch": 2.3967523680649525,
+      "grad_norm": 0.1956206113100052,
+      "learning_rate": 1.1775539719382473e-06,
+      "loss": 0.3227,
+      "step": 2952
+    },
+    {
+      "epoch": 2.3975642760487146,
+      "grad_norm": 0.18886259198188782,
+      "learning_rate": 1.1745084561667136e-06,
+      "loss": 0.3065,
+      "step": 2953
+    },
+    {
+      "epoch": 2.398376184032476,
+      "grad_norm": 0.21628929674625397,
+      "learning_rate": 1.1714663596091319e-06,
+      "loss": 0.3669,
+      "step": 2954
+    },
+    {
+      "epoch": 2.3991880920162383,
+      "grad_norm": 0.21439270675182343,
+      "learning_rate": 1.1684276849845171e-06,
+      "loss": 0.3118,
+      "step": 2955
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.20910055935382843,
+      "learning_rate": 1.1653924350088319e-06,
+      "loss": 0.3421,
+      "step": 2956
+    },
+    {
+      "epoch": 2.400811907983762,
+      "grad_norm": 0.19569739699363708,
+      "learning_rate": 1.1623606123949705e-06,
+      "loss": 0.3346,
+      "step": 2957
+    },
+    {
+      "epoch": 2.4016238159675236,
+      "grad_norm": 0.18719682097434998,
+      "learning_rate": 1.1593322198527707e-06,
+      "loss": 0.3276,
+      "step": 2958
+    },
+    {
+      "epoch": 2.4024357239512857,
+      "grad_norm": 0.18585561215877533,
+      "learning_rate": 1.1563072600889992e-06,
+      "loss": 0.3132,
+      "step": 2959
+    },
+    {
+      "epoch": 2.4032476319350473,
+      "grad_norm": 0.19666001200675964,
+      "learning_rate": 1.1532857358073597e-06,
+      "loss": 0.3373,
+      "step": 2960
+    },
+    {
+      "epoch": 2.4040595399188094,
+      "grad_norm": 0.1873420923948288,
+      "learning_rate": 1.1502676497084786e-06,
+      "loss": 0.2839,
+      "step": 2961
+    },
+    {
+      "epoch": 2.404871447902571,
+      "grad_norm": 0.2008202075958252,
+      "learning_rate": 1.147253004489917e-06,
+      "loss": 0.3673,
+      "step": 2962
+    },
+    {
+      "epoch": 2.4056833558863326,
+      "grad_norm": 0.19980937242507935,
+      "learning_rate": 1.1442418028461537e-06,
+      "loss": 0.3227,
+      "step": 2963
+    },
+    {
+      "epoch": 2.4064952638700947,
+      "grad_norm": 0.20811808109283447,
+      "learning_rate": 1.1412340474685946e-06,
+      "loss": 0.3549,
+      "step": 2964
+    },
+    {
+      "epoch": 2.407307171853857,
+      "grad_norm": 0.19051776826381683,
+      "learning_rate": 1.1382297410455612e-06,
+      "loss": 0.3325,
+      "step": 2965
+    },
+    {
+      "epoch": 2.4081190798376184,
+      "grad_norm": 0.1915096491575241,
+      "learning_rate": 1.1352288862622968e-06,
+      "loss": 0.3174,
+      "step": 2966
+    },
+    {
+      "epoch": 2.40893098782138,
+      "grad_norm": 0.1905997395515442,
+      "learning_rate": 1.1322314858009547e-06,
+      "loss": 0.3297,
+      "step": 2967
+    },
+    {
+      "epoch": 2.409742895805142,
+      "grad_norm": 0.22135958075523376,
+      "learning_rate": 1.129237542340606e-06,
+      "loss": 0.3732,
+      "step": 2968
+    },
+    {
+      "epoch": 2.4105548037889037,
+      "grad_norm": 0.18565407395362854,
+      "learning_rate": 1.126247058557226e-06,
+      "loss": 0.3219,
+      "step": 2969
+    },
+    {
+      "epoch": 2.411366711772666,
+      "grad_norm": 0.1978808492422104,
+      "learning_rate": 1.1232600371237035e-06,
+      "loss": 0.3185,
+      "step": 2970
+    },
+    {
+      "epoch": 2.4121786197564274,
+      "grad_norm": 0.1982131451368332,
+      "learning_rate": 1.1202764807098305e-06,
+      "loss": 0.3358,
+      "step": 2971
+    },
+    {
+      "epoch": 2.4129905277401895,
+      "grad_norm": 0.19569136202335358,
+      "learning_rate": 1.1172963919822988e-06,
+      "loss": 0.3767,
+      "step": 2972
+    },
+    {
+      "epoch": 2.413802435723951,
+      "grad_norm": 0.1900157332420349,
+      "learning_rate": 1.1143197736047073e-06,
+      "loss": 0.3277,
+      "step": 2973
+    },
+    {
+      "epoch": 2.414614343707713,
+      "grad_norm": 0.18981783092021942,
+      "learning_rate": 1.1113466282375457e-06,
+      "loss": 0.3171,
+      "step": 2974
+    },
+    {
+      "epoch": 2.415426251691475,
+      "grad_norm": 0.20192484557628632,
+      "learning_rate": 1.108376958538207e-06,
+      "loss": 0.3705,
+      "step": 2975
+    },
+    {
+      "epoch": 2.416238159675237,
+      "grad_norm": 0.17099638283252716,
+      "learning_rate": 1.1054107671609697e-06,
+      "loss": 0.3171,
+      "step": 2976
+    },
+    {
+      "epoch": 2.4170500676589985,
+      "grad_norm": 0.18442241847515106,
+      "learning_rate": 1.1024480567570123e-06,
+      "loss": 0.3271,
+      "step": 2977
+    },
+    {
+      "epoch": 2.4178619756427606,
+      "grad_norm": 0.1920735388994217,
+      "learning_rate": 1.0994888299743929e-06,
+      "loss": 0.363,
+      "step": 2978
+    },
+    {
+      "epoch": 2.4186738836265222,
+      "grad_norm": 0.18039803206920624,
+      "learning_rate": 1.0965330894580638e-06,
+      "loss": 0.3069,
+      "step": 2979
+    },
+    {
+      "epoch": 2.4194857916102843,
+      "grad_norm": 0.18314382433891296,
+      "learning_rate": 1.0935808378498547e-06,
+      "loss": 0.3072,
+      "step": 2980
+    },
+    {
+      "epoch": 2.420297699594046,
+      "grad_norm": 0.1894654780626297,
+      "learning_rate": 1.0906320777884833e-06,
+      "loss": 0.3486,
+      "step": 2981
+    },
+    {
+      "epoch": 2.421109607577808,
+      "grad_norm": 0.1932295262813568,
+      "learning_rate": 1.0876868119095407e-06,
+      "loss": 0.3563,
+      "step": 2982
+    },
+    {
+      "epoch": 2.4219215155615696,
+      "grad_norm": 0.1948682814836502,
+      "learning_rate": 1.0847450428454997e-06,
+      "loss": 0.3595,
+      "step": 2983
+    },
+    {
+      "epoch": 2.4227334235453317,
+      "grad_norm": 0.18310394883155823,
+      "learning_rate": 1.0818067732257032e-06,
+      "loss": 0.2842,
+      "step": 2984
+    },
+    {
+      "epoch": 2.4235453315290933,
+      "grad_norm": 0.18354585766792297,
+      "learning_rate": 1.0788720056763712e-06,
+      "loss": 0.3199,
+      "step": 2985
+    },
+    {
+      "epoch": 2.4243572395128554,
+      "grad_norm": 0.19106504321098328,
+      "learning_rate": 1.075940742820588e-06,
+      "loss": 0.3357,
+      "step": 2986
+    },
+    {
+      "epoch": 2.425169147496617,
+      "grad_norm": 0.18747876584529877,
+      "learning_rate": 1.0730129872783102e-06,
+      "loss": 0.33,
+      "step": 2987
+    },
+    {
+      "epoch": 2.4259810554803787,
+      "grad_norm": 0.17471006512641907,
+      "learning_rate": 1.0700887416663575e-06,
+      "loss": 0.3337,
+      "step": 2988
+    },
+    {
+      "epoch": 2.4267929634641408,
+      "grad_norm": 0.19256801903247833,
+      "learning_rate": 1.0671680085984137e-06,
+      "loss": 0.321,
+      "step": 2989
+    },
+    {
+      "epoch": 2.4276048714479024,
+      "grad_norm": 0.2284218817949295,
+      "learning_rate": 1.0642507906850202e-06,
+      "loss": 0.3452,
+      "step": 2990
+    },
+    {
+      "epoch": 2.4284167794316645,
+      "grad_norm": 0.20350132882595062,
+      "learning_rate": 1.0613370905335801e-06,
+      "loss": 0.3217,
+      "step": 2991
+    },
+    {
+      "epoch": 2.429228687415426,
+      "grad_norm": 0.20776550471782684,
+      "learning_rate": 1.058426910748349e-06,
+      "loss": 0.3394,
+      "step": 2992
+    },
+    {
+      "epoch": 2.430040595399188,
+      "grad_norm": 0.17248664796352386,
+      "learning_rate": 1.0555202539304398e-06,
+      "loss": 0.3069,
+      "step": 2993
+    },
+    {
+      "epoch": 2.43085250338295,
+      "grad_norm": 0.20249295234680176,
+      "learning_rate": 1.052617122677812e-06,
+      "loss": 0.3528,
+      "step": 2994
+    },
+    {
+      "epoch": 2.431664411366712,
+      "grad_norm": 0.2047611027956009,
+      "learning_rate": 1.0497175195852777e-06,
+      "loss": 0.3214,
+      "step": 2995
+    },
+    {
+      "epoch": 2.4324763193504735,
+      "grad_norm": 0.20809674263000488,
+      "learning_rate": 1.0468214472444955e-06,
+      "loss": 0.3717,
+      "step": 2996
+    },
+    {
+      "epoch": 2.4332882273342356,
+      "grad_norm": 0.18549399077892303,
+      "learning_rate": 1.0439289082439647e-06,
+      "loss": 0.2788,
+      "step": 2997
+    },
+    {
+      "epoch": 2.434100135317997,
+      "grad_norm": 0.19999492168426514,
+      "learning_rate": 1.0410399051690306e-06,
+      "loss": 0.3173,
+      "step": 2998
+    },
+    {
+      "epoch": 2.4349120433017593,
+      "grad_norm": 0.200644388794899,
+      "learning_rate": 1.0381544406018751e-06,
+      "loss": 0.3395,
+      "step": 2999
+    },
+    {
+      "epoch": 2.435723951285521,
+      "grad_norm": 0.19019752740859985,
+      "learning_rate": 1.0352725171215194e-06,
+      "loss": 0.3534,
+      "step": 3000
+    },
+    {
+      "epoch": 2.436535859269283,
+      "grad_norm": 0.20732276141643524,
+      "learning_rate": 1.0323941373038177e-06,
+      "loss": 0.3805,
+      "step": 3001
+    },
+    {
+      "epoch": 2.4373477672530446,
+      "grad_norm": 0.1586148589849472,
+      "learning_rate": 1.0295193037214606e-06,
+      "loss": 0.2641,
+      "step": 3002
+    },
+    {
+      "epoch": 2.4381596752368067,
+      "grad_norm": 0.18795549869537354,
+      "learning_rate": 1.0266480189439631e-06,
+      "loss": 0.3447,
+      "step": 3003
+    },
+    {
+      "epoch": 2.4389715832205683,
+      "grad_norm": 0.19631880521774292,
+      "learning_rate": 1.0237802855376745e-06,
+      "loss": 0.3487,
+      "step": 3004
+    },
+    {
+      "epoch": 2.4397834912043304,
+      "grad_norm": 0.20051532983779907,
+      "learning_rate": 1.0209161060657651e-06,
+      "loss": 0.3448,
+      "step": 3005
+    },
+    {
+      "epoch": 2.440595399188092,
+      "grad_norm": 0.18827258050441742,
+      "learning_rate": 1.0180554830882333e-06,
+      "loss": 0.3001,
+      "step": 3006
+    },
+    {
+      "epoch": 2.441407307171854,
+      "grad_norm": 0.21202042698860168,
+      "learning_rate": 1.015198419161893e-06,
+      "loss": 0.3781,
+      "step": 3007
+    },
+    {
+      "epoch": 2.4422192151556157,
+      "grad_norm": 0.19589298963546753,
+      "learning_rate": 1.0123449168403826e-06,
+      "loss": 0.3183,
+      "step": 3008
+    },
+    {
+      "epoch": 2.4430311231393773,
+      "grad_norm": 0.18542811274528503,
+      "learning_rate": 1.0094949786741531e-06,
+      "loss": 0.3397,
+      "step": 3009
+    },
+    {
+      "epoch": 2.4438430311231394,
+      "grad_norm": 0.1781880259513855,
+      "learning_rate": 1.0066486072104736e-06,
+      "loss": 0.3066,
+      "step": 3010
+    },
+    {
+      "epoch": 2.4446549391069015,
+      "grad_norm": 0.21066296100616455,
+      "learning_rate": 1.0038058049934203e-06,
+      "loss": 0.3411,
+      "step": 3011
+    },
+    {
+      "epoch": 2.445466847090663,
+      "grad_norm": 0.1871166229248047,
+      "learning_rate": 1.0009665745638852e-06,
+      "loss": 0.3129,
+      "step": 3012
+    },
+    {
+      "epoch": 2.4462787550744247,
+      "grad_norm": 0.19006915390491486,
+      "learning_rate": 9.981309184595617e-07,
+      "loss": 0.3341,
+      "step": 3013
+    },
+    {
+      "epoch": 2.447090663058187,
+      "grad_norm": 0.19085726141929626,
+      "learning_rate": 9.952988392149532e-07,
+      "loss": 0.3266,
+      "step": 3014
+    },
+    {
+      "epoch": 2.4479025710419484,
+      "grad_norm": 0.17809265851974487,
+      "learning_rate": 9.924703393613655e-07,
+      "loss": 0.3021,
+      "step": 3015
+    },
+    {
+      "epoch": 2.4487144790257105,
+      "grad_norm": 0.1888776272535324,
+      "learning_rate": 9.896454214269014e-07,
+      "loss": 0.3423,
+      "step": 3016
+    },
+    {
+      "epoch": 2.449526387009472,
+      "grad_norm": 0.17341932654380798,
+      "learning_rate": 9.868240879364676e-07,
+      "loss": 0.3018,
+      "step": 3017
+    },
+    {
+      "epoch": 2.450338294993234,
+      "grad_norm": 0.19802053272724152,
+      "learning_rate": 9.840063414117618e-07,
+      "loss": 0.3718,
+      "step": 3018
+    },
+    {
+      "epoch": 2.451150202976996,
+      "grad_norm": 0.18607288599014282,
+      "learning_rate": 9.811921843712813e-07,
+      "loss": 0.3188,
+      "step": 3019
+    },
+    {
+      "epoch": 2.451962110960758,
+      "grad_norm": 0.18426305055618286,
+      "learning_rate": 9.78381619330308e-07,
+      "loss": 0.335,
+      "step": 3020
+    },
+    {
+      "epoch": 2.4527740189445195,
+      "grad_norm": 0.1852353811264038,
+      "learning_rate": 9.75574648800921e-07,
+      "loss": 0.337,
+      "step": 3021
+    },
+    {
+      "epoch": 2.4535859269282816,
+      "grad_norm": 0.20497792959213257,
+      "learning_rate": 9.727712752919804e-07,
+      "loss": 0.339,
+      "step": 3022
+    },
+    {
+      "epoch": 2.454397834912043,
+      "grad_norm": 0.19332203269004822,
+      "learning_rate": 9.699715013091354e-07,
+      "loss": 0.3304,
+      "step": 3023
+    },
+    {
+      "epoch": 2.4552097428958053,
+      "grad_norm": 0.1976161003112793,
+      "learning_rate": 9.671753293548153e-07,
+      "loss": 0.3189,
+      "step": 3024
+    },
+    {
+      "epoch": 2.456021650879567,
+      "grad_norm": 0.19223108887672424,
+      "learning_rate": 9.643827619282309e-07,
+      "loss": 0.3385,
+      "step": 3025
+    },
+    {
+      "epoch": 2.456833558863329,
+      "grad_norm": 0.19086366891860962,
+      "learning_rate": 9.615938015253723e-07,
+      "loss": 0.3602,
+      "step": 3026
+    },
+    {
+      "epoch": 2.4576454668470906,
+      "grad_norm": 0.1918683499097824,
+      "learning_rate": 9.588084506390055e-07,
+      "loss": 0.3421,
+      "step": 3027
+    },
+    {
+      "epoch": 2.4584573748308527,
+      "grad_norm": 0.18355713784694672,
+      "learning_rate": 9.56026711758667e-07,
+      "loss": 0.3233,
+      "step": 3028
+    },
+    {
+      "epoch": 2.4592692828146143,
+      "grad_norm": 0.18370778858661652,
+      "learning_rate": 9.532485873706704e-07,
+      "loss": 0.3575,
+      "step": 3029
+    },
+    {
+      "epoch": 2.460081190798376,
+      "grad_norm": 0.19700896739959717,
+      "learning_rate": 9.504740799580931e-07,
+      "loss": 0.3047,
+      "step": 3030
+    },
+    {
+      "epoch": 2.460893098782138,
+      "grad_norm": 0.18893669545650482,
+      "learning_rate": 9.477031920007845e-07,
+      "loss": 0.3588,
+      "step": 3031
+    },
+    {
+      "epoch": 2.4617050067659,
+      "grad_norm": 0.18968765437602997,
+      "learning_rate": 9.449359259753543e-07,
+      "loss": 0.3489,
+      "step": 3032
+    },
+    {
+      "epoch": 2.4625169147496617,
+      "grad_norm": 0.19850139319896698,
+      "learning_rate": 9.421722843551801e-07,
+      "loss": 0.3226,
+      "step": 3033
+    },
+    {
+      "epoch": 2.4633288227334234,
+      "grad_norm": 0.20679272711277008,
+      "learning_rate": 9.394122696103947e-07,
+      "loss": 0.3033,
+      "step": 3034
+    },
+    {
+      "epoch": 2.4641407307171854,
+      "grad_norm": 0.19632133841514587,
+      "learning_rate": 9.366558842078943e-07,
+      "loss": 0.3334,
+      "step": 3035
+    },
+    {
+      "epoch": 2.464952638700947,
+      "grad_norm": 0.18622373044490814,
+      "learning_rate": 9.339031306113256e-07,
+      "loss": 0.332,
+      "step": 3036
+    },
+    {
+      "epoch": 2.465764546684709,
+      "grad_norm": 0.19979478418827057,
+      "learning_rate": 9.311540112810957e-07,
+      "loss": 0.3613,
+      "step": 3037
+    },
+    {
+      "epoch": 2.4665764546684708,
+      "grad_norm": 0.17775622010231018,
+      "learning_rate": 9.284085286743571e-07,
+      "loss": 0.2968,
+      "step": 3038
+    },
+    {
+      "epoch": 2.467388362652233,
+      "grad_norm": 0.19438128173351288,
+      "learning_rate": 9.256666852450158e-07,
+      "loss": 0.3679,
+      "step": 3039
+    },
+    {
+      "epoch": 2.4682002706359945,
+      "grad_norm": 0.19148623943328857,
+      "learning_rate": 9.229284834437252e-07,
+      "loss": 0.3294,
+      "step": 3040
+    },
+    {
+      "epoch": 2.4690121786197565,
+      "grad_norm": 0.19394387304782867,
+      "learning_rate": 9.20193925717881e-07,
+      "loss": 0.3237,
+      "step": 3041
+    },
+    {
+      "epoch": 2.469824086603518,
+      "grad_norm": 0.19533659517765045,
+      "learning_rate": 9.174630145116248e-07,
+      "loss": 0.3272,
+      "step": 3042
+    },
+    {
+      "epoch": 2.4706359945872802,
+      "grad_norm": 0.1746395230293274,
+      "learning_rate": 9.147357522658351e-07,
+      "loss": 0.323,
+      "step": 3043
+    },
+    {
+      "epoch": 2.471447902571042,
+      "grad_norm": 0.17495569586753845,
+      "learning_rate": 9.120121414181354e-07,
+      "loss": 0.3311,
+      "step": 3044
+    },
+    {
+      "epoch": 2.472259810554804,
+      "grad_norm": 0.1893194168806076,
+      "learning_rate": 9.092921844028779e-07,
+      "loss": 0.3393,
+      "step": 3045
+    },
+    {
+      "epoch": 2.4730717185385656,
+      "grad_norm": 0.19884805381298065,
+      "learning_rate": 9.065758836511556e-07,
+      "loss": 0.3604,
+      "step": 3046
+    },
+    {
+      "epoch": 2.4738836265223276,
+      "grad_norm": 0.18353500962257385,
+      "learning_rate": 9.038632415907883e-07,
+      "loss": 0.3267,
+      "step": 3047
+    },
+    {
+      "epoch": 2.4746955345060893,
+      "grad_norm": 0.1812758445739746,
+      "learning_rate": 9.011542606463309e-07,
+      "loss": 0.3507,
+      "step": 3048
+    },
+    {
+      "epoch": 2.4755074424898513,
+      "grad_norm": 0.177732914686203,
+      "learning_rate": 8.984489432390597e-07,
+      "loss": 0.3426,
+      "step": 3049
+    },
+    {
+      "epoch": 2.476319350473613,
+      "grad_norm": 0.2040518969297409,
+      "learning_rate": 8.957472917869836e-07,
+      "loss": 0.3681,
+      "step": 3050
+    },
+    {
+      "epoch": 2.4771312584573746,
+      "grad_norm": 0.18057014048099518,
+      "learning_rate": 8.930493087048286e-07,
+      "loss": 0.3136,
+      "step": 3051
+    },
+    {
+      "epoch": 2.4779431664411367,
+      "grad_norm": 0.17835043370723724,
+      "learning_rate": 8.90354996404047e-07,
+      "loss": 0.3129,
+      "step": 3052
+    },
+    {
+      "epoch": 2.4787550744248987,
+      "grad_norm": 0.18347781896591187,
+      "learning_rate": 8.876643572928056e-07,
+      "loss": 0.317,
+      "step": 3053
+    },
+    {
+      "epoch": 2.4795669824086604,
+      "grad_norm": 0.1835402101278305,
+      "learning_rate": 8.84977393775992e-07,
+      "loss": 0.3196,
+      "step": 3054
+    },
+    {
+      "epoch": 2.480378890392422,
+      "grad_norm": 0.21044647693634033,
+      "learning_rate": 8.822941082552055e-07,
+      "loss": 0.3542,
+      "step": 3055
+    },
+    {
+      "epoch": 2.481190798376184,
+      "grad_norm": 0.18420448899269104,
+      "learning_rate": 8.796145031287612e-07,
+      "loss": 0.303,
+      "step": 3056
+    },
+    {
+      "epoch": 2.4820027063599457,
+      "grad_norm": 0.19496576488018036,
+      "learning_rate": 8.769385807916808e-07,
+      "loss": 0.342,
+      "step": 3057
+    },
+    {
+      "epoch": 2.4828146143437078,
+      "grad_norm": 0.19800378382205963,
+      "learning_rate": 8.742663436356969e-07,
+      "loss": 0.3151,
+      "step": 3058
+    },
+    {
+      "epoch": 2.4836265223274694,
+      "grad_norm": 0.20701348781585693,
+      "learning_rate": 8.715977940492504e-07,
+      "loss": 0.3324,
+      "step": 3059
+    },
+    {
+      "epoch": 2.4844384303112315,
+      "grad_norm": 0.18038778007030487,
+      "learning_rate": 8.689329344174797e-07,
+      "loss": 0.3219,
+      "step": 3060
+    },
+    {
+      "epoch": 2.485250338294993,
+      "grad_norm": 0.17740096151828766,
+      "learning_rate": 8.662717671222326e-07,
+      "loss": 0.3416,
+      "step": 3061
+    },
+    {
+      "epoch": 2.486062246278755,
+      "grad_norm": 0.1861264854669571,
+      "learning_rate": 8.6361429454205e-07,
+      "loss": 0.3311,
+      "step": 3062
+    },
+    {
+      "epoch": 2.486874154262517,
+      "grad_norm": 0.18973012268543243,
+      "learning_rate": 8.609605190521752e-07,
+      "loss": 0.3605,
+      "step": 3063
+    },
+    {
+      "epoch": 2.487686062246279,
+      "grad_norm": 0.18656058609485626,
+      "learning_rate": 8.583104430245476e-07,
+      "loss": 0.3279,
+      "step": 3064
+    },
+    {
+      "epoch": 2.4884979702300405,
+      "grad_norm": 0.19464436173439026,
+      "learning_rate": 8.55664068827794e-07,
+      "loss": 0.3169,
+      "step": 3065
+    },
+    {
+      "epoch": 2.4893098782138026,
+      "grad_norm": 0.1877729445695877,
+      "learning_rate": 8.53021398827239e-07,
+      "loss": 0.3229,
+      "step": 3066
+    },
+    {
+      "epoch": 2.490121786197564,
+      "grad_norm": 0.19486500322818756,
+      "learning_rate": 8.503824353848944e-07,
+      "loss": 0.3427,
+      "step": 3067
+    },
+    {
+      "epoch": 2.4909336941813263,
+      "grad_norm": 0.18277627229690552,
+      "learning_rate": 8.477471808594556e-07,
+      "loss": 0.2802,
+      "step": 3068
+    },
+    {
+      "epoch": 2.491745602165088,
+      "grad_norm": 0.19243121147155762,
+      "learning_rate": 8.451156376063096e-07,
+      "loss": 0.3471,
+      "step": 3069
+    },
+    {
+      "epoch": 2.49255751014885,
+      "grad_norm": 0.19316624104976654,
+      "learning_rate": 8.424878079775195e-07,
+      "loss": 0.3465,
+      "step": 3070
+    },
+    {
+      "epoch": 2.4933694181326116,
+      "grad_norm": 0.20623767375946045,
+      "learning_rate": 8.398636943218353e-07,
+      "loss": 0.3266,
+      "step": 3071
+    },
+    {
+      "epoch": 2.4941813261163737,
+      "grad_norm": 0.18773795664310455,
+      "learning_rate": 8.37243298984679e-07,
+      "loss": 0.3207,
+      "step": 3072
+    },
+    {
+      "epoch": 2.4949932341001353,
+      "grad_norm": 0.1857750415802002,
+      "learning_rate": 8.346266243081563e-07,
+      "loss": 0.3024,
+      "step": 3073
+    },
+    {
+      "epoch": 2.4958051420838974,
+      "grad_norm": 0.21834681928157806,
+      "learning_rate": 8.320136726310413e-07,
+      "loss": 0.3563,
+      "step": 3074
+    },
+    {
+      "epoch": 2.496617050067659,
+      "grad_norm": 0.18611454963684082,
+      "learning_rate": 8.294044462887852e-07,
+      "loss": 0.3186,
+      "step": 3075
+    },
+    {
+      "epoch": 2.4974289580514206,
+      "grad_norm": 0.18615387380123138,
+      "learning_rate": 8.267989476135052e-07,
+      "loss": 0.3338,
+      "step": 3076
+    },
+    {
+      "epoch": 2.4982408660351827,
+      "grad_norm": 0.1886281818151474,
+      "learning_rate": 8.24197178933992e-07,
+      "loss": 0.3208,
+      "step": 3077
+    },
+    {
+      "epoch": 2.4990527740189448,
+      "grad_norm": 0.1949118673801422,
+      "learning_rate": 8.215991425756959e-07,
+      "loss": 0.3592,
+      "step": 3078
+    },
+    {
+      "epoch": 2.4998646820027064,
+      "grad_norm": 0.2695879638195038,
+      "learning_rate": 8.190048408607376e-07,
+      "loss": 0.3141,
+      "step": 3079
+    },
+    {
+      "epoch": 2.500676589986468,
+      "grad_norm": 0.20285677909851074,
+      "learning_rate": 8.164142761078941e-07,
+      "loss": 0.401,
+      "step": 3080
+    },
+    {
+      "epoch": 2.50148849797023,
+      "grad_norm": 0.1762535572052002,
+      "learning_rate": 8.138274506326088e-07,
+      "loss": 0.3102,
+      "step": 3081
+    },
+    {
+      "epoch": 2.5023004059539917,
+      "grad_norm": 0.1744300276041031,
+      "learning_rate": 8.112443667469754e-07,
+      "loss": 0.2994,
+      "step": 3082
+    },
+    {
+      "epoch": 2.503112313937754,
+      "grad_norm": 0.1919785887002945,
+      "learning_rate": 8.08665026759749e-07,
+      "loss": 0.3508,
+      "step": 3083
+    },
+    {
+      "epoch": 2.5039242219215154,
+      "grad_norm": 0.18033501505851746,
+      "learning_rate": 8.060894329763386e-07,
+      "loss": 0.3173,
+      "step": 3084
+    },
+    {
+      "epoch": 2.5047361299052775,
+      "grad_norm": 0.18745407462120056,
+      "learning_rate": 8.035175876987994e-07,
+      "loss": 0.3785,
+      "step": 3085
+    },
+    {
+      "epoch": 2.505548037889039,
+      "grad_norm": 0.19211943447589874,
+      "learning_rate": 8.009494932258427e-07,
+      "loss": 0.3031,
+      "step": 3086
+    },
+    {
+      "epoch": 2.506359945872801,
+      "grad_norm": 0.18943564593791962,
+      "learning_rate": 7.983851518528224e-07,
+      "loss": 0.3294,
+      "step": 3087
+    },
+    {
+      "epoch": 2.507171853856563,
+      "grad_norm": 0.18988443911075592,
+      "learning_rate": 7.958245658717411e-07,
+      "loss": 0.3335,
+      "step": 3088
+    },
+    {
+      "epoch": 2.507983761840325,
+      "grad_norm": 0.19199185073375702,
+      "learning_rate": 7.932677375712423e-07,
+      "loss": 0.3326,
+      "step": 3089
+    },
+    {
+      "epoch": 2.5087956698240865,
+      "grad_norm": 0.18789389729499817,
+      "learning_rate": 7.907146692366135e-07,
+      "loss": 0.3454,
+      "step": 3090
+    },
+    {
+      "epoch": 2.5096075778078486,
+      "grad_norm": 0.19735769927501678,
+      "learning_rate": 7.881653631497793e-07,
+      "loss": 0.3506,
+      "step": 3091
+    },
+    {
+      "epoch": 2.5104194857916102,
+      "grad_norm": 0.17913809418678284,
+      "learning_rate": 7.856198215893035e-07,
+      "loss": 0.2907,
+      "step": 3092
+    },
+    {
+      "epoch": 2.511231393775372,
+      "grad_norm": 0.19348320364952087,
+      "learning_rate": 7.830780468303828e-07,
+      "loss": 0.3172,
+      "step": 3093
+    },
+    {
+      "epoch": 2.512043301759134,
+      "grad_norm": 0.1913117915391922,
+      "learning_rate": 7.805400411448505e-07,
+      "loss": 0.323,
+      "step": 3094
+    },
+    {
+      "epoch": 2.512855209742896,
+      "grad_norm": 0.18038423359394073,
+      "learning_rate": 7.78005806801167e-07,
+      "loss": 0.3113,
+      "step": 3095
+    },
+    {
+      "epoch": 2.5136671177266576,
+      "grad_norm": 0.18813300132751465,
+      "learning_rate": 7.754753460644266e-07,
+      "loss": 0.3504,
+      "step": 3096
+    },
+    {
+      "epoch": 2.5144790257104193,
+      "grad_norm": 0.19093145430088043,
+      "learning_rate": 7.729486611963449e-07,
+      "loss": 0.3308,
+      "step": 3097
+    },
+    {
+      "epoch": 2.5152909336941813,
+      "grad_norm": 0.18804404139518738,
+      "learning_rate": 7.704257544552696e-07,
+      "loss": 0.3173,
+      "step": 3098
+    },
+    {
+      "epoch": 2.5161028416779434,
+      "grad_norm": 0.18301160633563995,
+      "learning_rate": 7.67906628096165e-07,
+      "loss": 0.3291,
+      "step": 3099
+    },
+    {
+      "epoch": 2.516914749661705,
+      "grad_norm": 0.1713382452726364,
+      "learning_rate": 7.653912843706219e-07,
+      "loss": 0.3023,
+      "step": 3100
+    },
+    {
+      "epoch": 2.5177266576454667,
+      "grad_norm": 0.18733292818069458,
+      "learning_rate": 7.628797255268439e-07,
+      "loss": 0.3327,
+      "step": 3101
+    },
+    {
+      "epoch": 2.5185385656292287,
+      "grad_norm": 0.18120600283145905,
+      "learning_rate": 7.603719538096604e-07,
+      "loss": 0.3257,
+      "step": 3102
+    },
+    {
+      "epoch": 2.519350473612991,
+      "grad_norm": 0.18089093267917633,
+      "learning_rate": 7.57867971460507e-07,
+      "loss": 0.3301,
+      "step": 3103
+    },
+    {
+      "epoch": 2.5201623815967524,
+      "grad_norm": 0.19535592198371887,
+      "learning_rate": 7.553677807174392e-07,
+      "loss": 0.3503,
+      "step": 3104
+    },
+    {
+      "epoch": 2.520974289580514,
+      "grad_norm": 0.18661540746688843,
+      "learning_rate": 7.52871383815118e-07,
+      "loss": 0.3195,
+      "step": 3105
+    },
+    {
+      "epoch": 2.521786197564276,
+      "grad_norm": 0.20230573415756226,
+      "learning_rate": 7.503787829848191e-07,
+      "loss": 0.3286,
+      "step": 3106
+    },
+    {
+      "epoch": 2.5225981055480378,
+      "grad_norm": 0.1875879168510437,
+      "learning_rate": 7.478899804544193e-07,
+      "loss": 0.3333,
+      "step": 3107
+    },
+    {
+      "epoch": 2.5234100135318,
+      "grad_norm": 0.178261399269104,
+      "learning_rate": 7.454049784484063e-07,
+      "loss": 0.3453,
+      "step": 3108
+    },
+    {
+      "epoch": 2.5242219215155615,
+      "grad_norm": 0.1892920583486557,
+      "learning_rate": 7.429237791878652e-07,
+      "loss": 0.3379,
+      "step": 3109
+    },
+    {
+      "epoch": 2.5250338294993235,
+      "grad_norm": 0.21874766051769257,
+      "learning_rate": 7.40446384890487e-07,
+      "loss": 0.3077,
+      "step": 3110
+    },
+    {
+      "epoch": 2.525845737483085,
+      "grad_norm": 0.19929294288158417,
+      "learning_rate": 7.379727977705614e-07,
+      "loss": 0.3427,
+      "step": 3111
+    },
+    {
+      "epoch": 2.5266576454668472,
+      "grad_norm": 0.1907694786787033,
+      "learning_rate": 7.355030200389706e-07,
+      "loss": 0.3107,
+      "step": 3112
+    },
+    {
+      "epoch": 2.527469553450609,
+      "grad_norm": 0.1870553195476532,
+      "learning_rate": 7.330370539031978e-07,
+      "loss": 0.3238,
+      "step": 3113
+    },
+    {
+      "epoch": 2.528281461434371,
+      "grad_norm": 0.1752476692199707,
+      "learning_rate": 7.305749015673153e-07,
+      "loss": 0.3469,
+      "step": 3114
+    },
+    {
+      "epoch": 2.5290933694181326,
+      "grad_norm": 0.1864292472600937,
+      "learning_rate": 7.281165652319893e-07,
+      "loss": 0.3138,
+      "step": 3115
+    },
+    {
+      "epoch": 2.5299052774018946,
+      "grad_norm": 0.19614975154399872,
+      "learning_rate": 7.25662047094473e-07,
+      "loss": 0.3429,
+      "step": 3116
+    },
+    {
+      "epoch": 2.5307171853856563,
+      "grad_norm": 0.18571297824382782,
+      "learning_rate": 7.232113493486099e-07,
+      "loss": 0.2989,
+      "step": 3117
+    },
+    {
+      "epoch": 2.531529093369418,
+      "grad_norm": 0.19236604869365692,
+      "learning_rate": 7.207644741848241e-07,
+      "loss": 0.3231,
+      "step": 3118
+    },
+    {
+      "epoch": 2.53234100135318,
+      "grad_norm": 0.20378991961479187,
+      "learning_rate": 7.183214237901287e-07,
+      "loss": 0.3062,
+      "step": 3119
+    },
+    {
+      "epoch": 2.533152909336942,
+      "grad_norm": 0.2041609138250351,
+      "learning_rate": 7.158822003481136e-07,
+      "loss": 0.3369,
+      "step": 3120
+    },
+    {
+      "epoch": 2.5339648173207037,
+      "grad_norm": 0.19786538183689117,
+      "learning_rate": 7.134468060389521e-07,
+      "loss": 0.3713,
+      "step": 3121
+    },
+    {
+      "epoch": 2.5347767253044653,
+      "grad_norm": 0.17813830077648163,
+      "learning_rate": 7.110152430393902e-07,
+      "loss": 0.3088,
+      "step": 3122
+    },
+    {
+      "epoch": 2.5355886332882274,
+      "grad_norm": 0.17688030004501343,
+      "learning_rate": 7.085875135227555e-07,
+      "loss": 0.3167,
+      "step": 3123
+    },
+    {
+      "epoch": 2.5364005412719894,
+      "grad_norm": 0.19104081392288208,
+      "learning_rate": 7.061636196589433e-07,
+      "loss": 0.325,
+      "step": 3124
+    },
+    {
+      "epoch": 2.537212449255751,
+      "grad_norm": 0.18981610238552094,
+      "learning_rate": 7.037435636144252e-07,
+      "loss": 0.3209,
+      "step": 3125
+    },
+    {
+      "epoch": 2.5380243572395127,
+      "grad_norm": 0.18333403766155243,
+      "learning_rate": 7.013273475522392e-07,
+      "loss": 0.3382,
+      "step": 3126
+    },
+    {
+      "epoch": 2.5388362652232748,
+      "grad_norm": 0.1912931203842163,
+      "learning_rate": 6.989149736319939e-07,
+      "loss": 0.3325,
+      "step": 3127
+    },
+    {
+      "epoch": 2.5396481732070364,
+      "grad_norm": 0.19140057265758514,
+      "learning_rate": 6.965064440098629e-07,
+      "loss": 0.3282,
+      "step": 3128
+    },
+    {
+      "epoch": 2.5404600811907985,
+      "grad_norm": 0.18376098573207855,
+      "learning_rate": 6.941017608385813e-07,
+      "loss": 0.3192,
+      "step": 3129
+    },
+    {
+      "epoch": 2.54127198917456,
+      "grad_norm": 0.18468251824378967,
+      "learning_rate": 6.917009262674512e-07,
+      "loss": 0.3074,
+      "step": 3130
+    },
+    {
+      "epoch": 2.542083897158322,
+      "grad_norm": 0.2143646627664566,
+      "learning_rate": 6.893039424423298e-07,
+      "loss": 0.3206,
+      "step": 3131
+    },
+    {
+      "epoch": 2.542895805142084,
+      "grad_norm": 0.1823718398809433,
+      "learning_rate": 6.869108115056361e-07,
+      "loss": 0.3247,
+      "step": 3132
+    },
+    {
+      "epoch": 2.543707713125846,
+      "grad_norm": 0.231570303440094,
+      "learning_rate": 6.845215355963425e-07,
+      "loss": 0.3637,
+      "step": 3133
+    },
+    {
+      "epoch": 2.5445196211096075,
+      "grad_norm": 0.19110940396785736,
+      "learning_rate": 6.821361168499796e-07,
+      "loss": 0.324,
+      "step": 3134
+    },
+    {
+      "epoch": 2.5453315290933696,
+      "grad_norm": 0.19192618131637573,
+      "learning_rate": 6.797545573986253e-07,
+      "loss": 0.3431,
+      "step": 3135
+    },
+    {
+      "epoch": 2.546143437077131,
+      "grad_norm": 0.1862684041261673,
+      "learning_rate": 6.77376859370914e-07,
+      "loss": 0.335,
+      "step": 3136
+    },
+    {
+      "epoch": 2.5469553450608933,
+      "grad_norm": 0.17893125116825104,
+      "learning_rate": 6.750030248920231e-07,
+      "loss": 0.3356,
+      "step": 3137
+    },
+    {
+      "epoch": 2.547767253044655,
+      "grad_norm": 0.21611236035823822,
+      "learning_rate": 6.726330560836813e-07,
+      "loss": 0.3347,
+      "step": 3138
+    },
+    {
+      "epoch": 2.5485791610284165,
+      "grad_norm": 0.1950601190328598,
+      "learning_rate": 6.702669550641589e-07,
+      "loss": 0.3483,
+      "step": 3139
+    },
+    {
+      "epoch": 2.5493910690121786,
+      "grad_norm": 0.18337251245975494,
+      "learning_rate": 6.679047239482711e-07,
+      "loss": 0.3179,
+      "step": 3140
+    },
+    {
+      "epoch": 2.5502029769959407,
+      "grad_norm": 0.17243637144565582,
+      "learning_rate": 6.655463648473736e-07,
+      "loss": 0.293,
+      "step": 3141
+    },
+    {
+      "epoch": 2.5510148849797023,
+      "grad_norm": 0.5626041889190674,
+      "learning_rate": 6.631918798693632e-07,
+      "loss": 0.3679,
+      "step": 3142
+    },
+    {
+      "epoch": 2.551826792963464,
+      "grad_norm": 0.20070084929466248,
+      "learning_rate": 6.60841271118669e-07,
+      "loss": 0.3274,
+      "step": 3143
+    },
+    {
+      "epoch": 2.552638700947226,
+      "grad_norm": 0.18911679089069366,
+      "learning_rate": 6.584945406962617e-07,
+      "loss": 0.3135,
+      "step": 3144
+    },
+    {
+      "epoch": 2.553450608930988,
+      "grad_norm": 0.19112491607666016,
+      "learning_rate": 6.561516906996395e-07,
+      "loss": 0.3297,
+      "step": 3145
+    },
+    {
+      "epoch": 2.5542625169147497,
+      "grad_norm": 0.21130810678005219,
+      "learning_rate": 6.53812723222838e-07,
+      "loss": 0.3202,
+      "step": 3146
+    },
+    {
+      "epoch": 2.5550744248985113,
+      "grad_norm": 0.19666330516338348,
+      "learning_rate": 6.514776403564183e-07,
+      "loss": 0.3224,
+      "step": 3147
+    },
+    {
+      "epoch": 2.5558863328822734,
+      "grad_norm": 0.18539170920848846,
+      "learning_rate": 6.491464441874728e-07,
+      "loss": 0.3311,
+      "step": 3148
+    },
+    {
+      "epoch": 2.556698240866035,
+      "grad_norm": 0.2082895040512085,
+      "learning_rate": 6.468191367996163e-07,
+      "loss": 0.3666,
+      "step": 3149
+    },
+    {
+      "epoch": 2.557510148849797,
+      "grad_norm": 0.17891336977481842,
+      "learning_rate": 6.444957202729918e-07,
+      "loss": 0.2843,
+      "step": 3150
+    },
+    {
+      "epoch": 2.5583220568335587,
+      "grad_norm": 0.20524373650550842,
+      "learning_rate": 6.421761966842604e-07,
+      "loss": 0.344,
+      "step": 3151
+    },
+    {
+      "epoch": 2.559133964817321,
+      "grad_norm": 0.19362302124500275,
+      "learning_rate": 6.398605681066083e-07,
+      "loss": 0.3492,
+      "step": 3152
+    },
+    {
+      "epoch": 2.5599458728010824,
+      "grad_norm": 0.1935529261827469,
+      "learning_rate": 6.37548836609736e-07,
+      "loss": 0.3424,
+      "step": 3153
+    },
+    {
+      "epoch": 2.5607577807848445,
+      "grad_norm": 0.19712930917739868,
+      "learning_rate": 6.352410042598639e-07,
+      "loss": 0.3608,
+      "step": 3154
+    },
+    {
+      "epoch": 2.561569688768606,
+      "grad_norm": 0.18271997570991516,
+      "learning_rate": 6.329370731197276e-07,
+      "loss": 0.327,
+      "step": 3155
+    },
+    {
+      "epoch": 2.562381596752368,
+      "grad_norm": 0.19526724517345428,
+      "learning_rate": 6.306370452485716e-07,
+      "loss": 0.3368,
+      "step": 3156
+    },
+    {
+      "epoch": 2.56319350473613,
+      "grad_norm": 0.19420628249645233,
+      "learning_rate": 6.283409227021575e-07,
+      "loss": 0.2857,
+      "step": 3157
+    },
+    {
+      "epoch": 2.564005412719892,
+      "grad_norm": 0.19973738491535187,
+      "learning_rate": 6.260487075327504e-07,
+      "loss": 0.3414,
+      "step": 3158
+    },
+    {
+      "epoch": 2.5648173207036535,
+      "grad_norm": 0.18292838335037231,
+      "learning_rate": 6.237604017891291e-07,
+      "loss": 0.3151,
+      "step": 3159
+    },
+    {
+      "epoch": 2.565629228687415,
+      "grad_norm": 0.19613362848758698,
+      "learning_rate": 6.214760075165721e-07,
+      "loss": 0.3556,
+      "step": 3160
+    },
+    {
+      "epoch": 2.5664411366711772,
+      "grad_norm": 0.19968324899673462,
+      "learning_rate": 6.191955267568678e-07,
+      "loss": 0.3318,
+      "step": 3161
+    },
+    {
+      "epoch": 2.5672530446549393,
+      "grad_norm": 0.18328046798706055,
+      "learning_rate": 6.169189615483007e-07,
+      "loss": 0.3389,
+      "step": 3162
+    },
+    {
+      "epoch": 2.568064952638701,
+      "grad_norm": 0.19672614336013794,
+      "learning_rate": 6.146463139256609e-07,
+      "loss": 0.344,
+      "step": 3163
+    },
+    {
+      "epoch": 2.5688768606224626,
+      "grad_norm": 0.2144729644060135,
+      "learning_rate": 6.12377585920233e-07,
+      "loss": 0.3271,
+      "step": 3164
+    },
+    {
+      "epoch": 2.5696887686062246,
+      "grad_norm": 0.18910811841487885,
+      "learning_rate": 6.101127795598016e-07,
+      "loss": 0.3155,
+      "step": 3165
+    },
+    {
+      "epoch": 2.5705006765899867,
+      "grad_norm": 0.18413443863391876,
+      "learning_rate": 6.078518968686426e-07,
+      "loss": 0.3276,
+      "step": 3166
+    },
+    {
+      "epoch": 2.5713125845737483,
+      "grad_norm": 0.1885359287261963,
+      "learning_rate": 6.055949398675293e-07,
+      "loss": 0.3144,
+      "step": 3167
+    },
+    {
+      "epoch": 2.57212449255751,
+      "grad_norm": 0.20480728149414062,
+      "learning_rate": 6.033419105737204e-07,
+      "loss": 0.3167,
+      "step": 3168
+    },
+    {
+      "epoch": 2.572936400541272,
+      "grad_norm": 0.19453857839107513,
+      "learning_rate": 6.010928110009711e-07,
+      "loss": 0.3171,
+      "step": 3169
+    },
+    {
+      "epoch": 2.5737483085250337,
+      "grad_norm": 0.18475186824798584,
+      "learning_rate": 5.98847643159517e-07,
+      "loss": 0.3281,
+      "step": 3170
+    },
+    {
+      "epoch": 2.5745602165087957,
+      "grad_norm": 0.1922556459903717,
+      "learning_rate": 5.966064090560847e-07,
+      "loss": 0.3811,
+      "step": 3171
+    },
+    {
+      "epoch": 2.5753721244925574,
+      "grad_norm": 0.19418615102767944,
+      "learning_rate": 5.943691106938843e-07,
+      "loss": 0.3259,
+      "step": 3172
+    },
+    {
+      "epoch": 2.5761840324763194,
+      "grad_norm": 0.1885213702917099,
+      "learning_rate": 5.921357500726044e-07,
+      "loss": 0.3453,
+      "step": 3173
+    },
+    {
+      "epoch": 2.576995940460081,
+      "grad_norm": 0.19406382739543915,
+      "learning_rate": 5.899063291884193e-07,
+      "loss": 0.2981,
+      "step": 3174
+    },
+    {
+      "epoch": 2.577807848443843,
+      "grad_norm": 0.19989027082920074,
+      "learning_rate": 5.87680850033977e-07,
+      "loss": 0.3126,
+      "step": 3175
+    },
+    {
+      "epoch": 2.5786197564276048,
+      "grad_norm": 0.20401804149150848,
+      "learning_rate": 5.854593145984067e-07,
+      "loss": 0.3338,
+      "step": 3176
+    },
+    {
+      "epoch": 2.579431664411367,
+      "grad_norm": 0.18977415561676025,
+      "learning_rate": 5.832417248673089e-07,
+      "loss": 0.3394,
+      "step": 3177
+    },
+    {
+      "epoch": 2.5802435723951285,
+      "grad_norm": 0.1779099553823471,
+      "learning_rate": 5.810280828227599e-07,
+      "loss": 0.3085,
+      "step": 3178
+    },
+    {
+      "epoch": 2.5810554803788905,
+      "grad_norm": 0.1864883005619049,
+      "learning_rate": 5.788183904433076e-07,
+      "loss": 0.3674,
+      "step": 3179
+    },
+    {
+      "epoch": 2.581867388362652,
+      "grad_norm": 0.17859680950641632,
+      "learning_rate": 5.766126497039698e-07,
+      "loss": 0.2967,
+      "step": 3180
+    },
+    {
+      "epoch": 2.582679296346414,
+      "grad_norm": 0.17314811050891876,
+      "learning_rate": 5.744108625762291e-07,
+      "loss": 0.2854,
+      "step": 3181
+    },
+    {
+      "epoch": 2.583491204330176,
+      "grad_norm": 0.20772767066955566,
+      "learning_rate": 5.722130310280394e-07,
+      "loss": 0.4072,
+      "step": 3182
+    },
+    {
+      "epoch": 2.584303112313938,
+      "grad_norm": 0.18678788840770721,
+      "learning_rate": 5.70019157023814e-07,
+      "loss": 0.3507,
+      "step": 3183
+    },
+    {
+      "epoch": 2.5851150202976996,
+      "grad_norm": 0.1905314028263092,
+      "learning_rate": 5.678292425244347e-07,
+      "loss": 0.3129,
+      "step": 3184
+    },
+    {
+      "epoch": 2.585926928281461,
+      "grad_norm": 0.1910918951034546,
+      "learning_rate": 5.656432894872376e-07,
+      "loss": 0.3302,
+      "step": 3185
+    },
+    {
+      "epoch": 2.5867388362652233,
+      "grad_norm": 0.1832975447177887,
+      "learning_rate": 5.634612998660249e-07,
+      "loss": 0.3119,
+      "step": 3186
+    },
+    {
+      "epoch": 2.5875507442489853,
+      "grad_norm": 0.18062268197536469,
+      "learning_rate": 5.612832756110498e-07,
+      "loss": 0.3309,
+      "step": 3187
+    },
+    {
+      "epoch": 2.588362652232747,
+      "grad_norm": 0.1919523924589157,
+      "learning_rate": 5.591092186690278e-07,
+      "loss": 0.3463,
+      "step": 3188
+    },
+    {
+      "epoch": 2.5891745602165086,
+      "grad_norm": 0.17138129472732544,
+      "learning_rate": 5.569391309831218e-07,
+      "loss": 0.3171,
+      "step": 3189
+    },
+    {
+      "epoch": 2.5899864682002707,
+      "grad_norm": 0.18817903101444244,
+      "learning_rate": 5.547730144929531e-07,
+      "loss": 0.347,
+      "step": 3190
+    },
+    {
+      "epoch": 2.5907983761840327,
+      "grad_norm": 0.1885135918855667,
+      "learning_rate": 5.526108711345884e-07,
+      "loss": 0.3274,
+      "step": 3191
+    },
+    {
+      "epoch": 2.5916102841677944,
+      "grad_norm": 0.1917082816362381,
+      "learning_rate": 5.504527028405482e-07,
+      "loss": 0.3124,
+      "step": 3192
+    },
+    {
+      "epoch": 2.592422192151556,
+      "grad_norm": 0.19591057300567627,
+      "learning_rate": 5.482985115397943e-07,
+      "loss": 0.3503,
+      "step": 3193
+    },
+    {
+      "epoch": 2.593234100135318,
+      "grad_norm": 0.18408437073230743,
+      "learning_rate": 5.461482991577399e-07,
+      "loss": 0.3245,
+      "step": 3194
+    },
+    {
+      "epoch": 2.5940460081190797,
+      "grad_norm": 0.1710149198770523,
+      "learning_rate": 5.440020676162367e-07,
+      "loss": 0.3046,
+      "step": 3195
+    },
+    {
+      "epoch": 2.5948579161028418,
+      "grad_norm": 0.18940046429634094,
+      "learning_rate": 5.418598188335816e-07,
+      "loss": 0.3384,
+      "step": 3196
+    },
+    {
+      "epoch": 2.5956698240866034,
+      "grad_norm": 0.18744125962257385,
+      "learning_rate": 5.397215547245116e-07,
+      "loss": 0.3403,
+      "step": 3197
+    },
+    {
+      "epoch": 2.5964817320703655,
+      "grad_norm": 0.18122704327106476,
+      "learning_rate": 5.375872772002e-07,
+      "loss": 0.2977,
+      "step": 3198
+    },
+    {
+      "epoch": 2.597293640054127,
+      "grad_norm": 0.19043493270874023,
+      "learning_rate": 5.354569881682592e-07,
+      "loss": 0.3544,
+      "step": 3199
+    },
+    {
+      "epoch": 2.598105548037889,
+      "grad_norm": 0.1788402497768402,
+      "learning_rate": 5.33330689532735e-07,
+      "loss": 0.3324,
+      "step": 3200
+    },
+    {
+      "epoch": 2.598917456021651,
+      "grad_norm": 0.17642781138420105,
+      "learning_rate": 5.312083831941084e-07,
+      "loss": 0.3242,
+      "step": 3201
+    },
+    {
+      "epoch": 2.599729364005413,
+      "grad_norm": 0.18630655109882355,
+      "learning_rate": 5.290900710492891e-07,
+      "loss": 0.3354,
+      "step": 3202
+    },
+    {
+      "epoch": 2.6005412719891745,
+      "grad_norm": 0.19044356048107147,
+      "learning_rate": 5.269757549916216e-07,
+      "loss": 0.331,
+      "step": 3203
+    },
+    {
+      "epoch": 2.6013531799729366,
+      "grad_norm": 0.1953808069229126,
+      "learning_rate": 5.24865436910873e-07,
+      "loss": 0.3766,
+      "step": 3204
+    },
+    {
+      "epoch": 2.602165087956698,
+      "grad_norm": 0.1799723356962204,
+      "learning_rate": 5.227591186932418e-07,
+      "loss": 0.2829,
+      "step": 3205
+    },
+    {
+      "epoch": 2.60297699594046,
+      "grad_norm": 0.19711996614933014,
+      "learning_rate": 5.206568022213482e-07,
+      "loss": 0.345,
+      "step": 3206
+    },
+    {
+      "epoch": 2.603788903924222,
+      "grad_norm": 0.17252270877361298,
+      "learning_rate": 5.185584893742379e-07,
+      "loss": 0.3323,
+      "step": 3207
+    },
+    {
+      "epoch": 2.604600811907984,
+      "grad_norm": 0.18615341186523438,
+      "learning_rate": 5.164641820273764e-07,
+      "loss": 0.3307,
+      "step": 3208
+    },
+    {
+      "epoch": 2.6054127198917456,
+      "grad_norm": 0.1939464956521988,
+      "learning_rate": 5.143738820526517e-07,
+      "loss": 0.368,
+      "step": 3209
+    },
+    {
+      "epoch": 2.6062246278755072,
+      "grad_norm": 0.19792777299880981,
+      "learning_rate": 5.122875913183656e-07,
+      "loss": 0.3075,
+      "step": 3210
+    },
+    {
+      "epoch": 2.6070365358592693,
+      "grad_norm": 0.18236370384693146,
+      "learning_rate": 5.102053116892414e-07,
+      "loss": 0.318,
+      "step": 3211
+    },
+    {
+      "epoch": 2.6078484438430314,
+      "grad_norm": 0.17754267156124115,
+      "learning_rate": 5.081270450264125e-07,
+      "loss": 0.3053,
+      "step": 3212
+    },
+    {
+      "epoch": 2.608660351826793,
+      "grad_norm": 0.20079311728477478,
+      "learning_rate": 5.060527931874309e-07,
+      "loss": 0.3494,
+      "step": 3213
+    },
+    {
+      "epoch": 2.6094722598105546,
+      "grad_norm": 0.17388774454593658,
+      "learning_rate": 5.039825580262553e-07,
+      "loss": 0.3175,
+      "step": 3214
+    },
+    {
+      "epoch": 2.6102841677943167,
+      "grad_norm": 0.18794164061546326,
+      "learning_rate": 5.019163413932565e-07,
+      "loss": 0.3296,
+      "step": 3215
+    },
+    {
+      "epoch": 2.6110960757780783,
+      "grad_norm": 0.30149030685424805,
+      "learning_rate": 4.998541451352134e-07,
+      "loss": 0.332,
+      "step": 3216
+    },
+    {
+      "epoch": 2.6119079837618404,
+      "grad_norm": 0.1879594326019287,
+      "learning_rate": 4.977959710953128e-07,
+      "loss": 0.3236,
+      "step": 3217
+    },
+    {
+      "epoch": 2.612719891745602,
+      "grad_norm": 0.1791881024837494,
+      "learning_rate": 4.957418211131432e-07,
+      "loss": 0.3227,
+      "step": 3218
+    },
+    {
+      "epoch": 2.613531799729364,
+      "grad_norm": 0.2038021832704544,
+      "learning_rate": 4.936916970246997e-07,
+      "loss": 0.3534,
+      "step": 3219
+    },
+    {
+      "epoch": 2.6143437077131257,
+      "grad_norm": 0.20205388963222504,
+      "learning_rate": 4.91645600662376e-07,
+      "loss": 0.3416,
+      "step": 3220
+    },
+    {
+      "epoch": 2.615155615696888,
+      "grad_norm": 0.19113677740097046,
+      "learning_rate": 4.896035338549693e-07,
+      "loss": 0.3108,
+      "step": 3221
+    },
+    {
+      "epoch": 2.6159675236806494,
+      "grad_norm": 0.18627013266086578,
+      "learning_rate": 4.875654984276718e-07,
+      "loss": 0.2935,
+      "step": 3222
+    },
+    {
+      "epoch": 2.6167794316644115,
+      "grad_norm": 0.18686647713184357,
+      "learning_rate": 4.855314962020746e-07,
+      "loss": 0.3297,
+      "step": 3223
+    },
+    {
+      "epoch": 2.617591339648173,
+      "grad_norm": 0.18558327853679657,
+      "learning_rate": 4.835015289961642e-07,
+      "loss": 0.3726,
+      "step": 3224
+    },
+    {
+      "epoch": 2.618403247631935,
+      "grad_norm": 0.1926698386669159,
+      "learning_rate": 4.81475598624318e-07,
+      "loss": 0.3432,
+      "step": 3225
+    },
+    {
+      "epoch": 2.619215155615697,
+      "grad_norm": 0.19128067791461945,
+      "learning_rate": 4.794537068973093e-07,
+      "loss": 0.338,
+      "step": 3226
+    },
+    {
+      "epoch": 2.6200270635994585,
+      "grad_norm": 0.1853039562702179,
+      "learning_rate": 4.774358556222974e-07,
+      "loss": 0.3434,
+      "step": 3227
+    },
+    {
+      "epoch": 2.6208389715832205,
+      "grad_norm": 0.19414806365966797,
+      "learning_rate": 4.7542204660283363e-07,
+      "loss": 0.328,
+      "step": 3228
+    },
+    {
+      "epoch": 2.6216508795669826,
+      "grad_norm": 0.18154501914978027,
+      "learning_rate": 4.7341228163885424e-07,
+      "loss": 0.3263,
+      "step": 3229
+    },
+    {
+      "epoch": 2.6224627875507442,
+      "grad_norm": 0.16941721737384796,
+      "learning_rate": 4.7140656252668226e-07,
+      "loss": 0.3184,
+      "step": 3230
+    },
+    {
+      "epoch": 2.623274695534506,
+      "grad_norm": 0.18542702496051788,
+      "learning_rate": 4.6940489105902323e-07,
+      "loss": 0.3266,
+      "step": 3231
+    },
+    {
+      "epoch": 2.624086603518268,
+      "grad_norm": 0.18007929623126984,
+      "learning_rate": 4.67407269024967e-07,
+      "loss": 0.3211,
+      "step": 3232
+    },
+    {
+      "epoch": 2.62489851150203,
+      "grad_norm": 0.18457120656967163,
+      "learning_rate": 4.654136982099805e-07,
+      "loss": 0.3193,
+      "step": 3233
+    },
+    {
+      "epoch": 2.6257104194857916,
+      "grad_norm": 0.1823306828737259,
+      "learning_rate": 4.634241803959144e-07,
+      "loss": 0.324,
+      "step": 3234
+    },
+    {
+      "epoch": 2.6265223274695533,
+      "grad_norm": 0.18947546184062958,
+      "learning_rate": 4.6143871736099246e-07,
+      "loss": 0.3223,
+      "step": 3235
+    },
+    {
+      "epoch": 2.6273342354533153,
+      "grad_norm": 0.22879110276699066,
+      "learning_rate": 4.5945731087981783e-07,
+      "loss": 0.368,
+      "step": 3236
+    },
+    {
+      "epoch": 2.628146143437077,
+      "grad_norm": 0.1855212301015854,
+      "learning_rate": 4.574799627233639e-07,
+      "loss": 0.3115,
+      "step": 3237
+    },
+    {
+      "epoch": 2.628958051420839,
+      "grad_norm": 0.169219970703125,
+      "learning_rate": 4.5550667465898204e-07,
+      "loss": 0.3114,
+      "step": 3238
+    },
+    {
+      "epoch": 2.6297699594046007,
+      "grad_norm": 0.1842631995677948,
+      "learning_rate": 4.535374484503885e-07,
+      "loss": 0.3338,
+      "step": 3239
+    },
+    {
+      "epoch": 2.6305818673883627,
+      "grad_norm": 0.19539353251457214,
+      "learning_rate": 4.515722858576743e-07,
+      "loss": 0.3191,
+      "step": 3240
+    },
+    {
+      "epoch": 2.6313937753721244,
+      "grad_norm": 0.18358981609344482,
+      "learning_rate": 4.4961118863729725e-07,
+      "loss": 0.3382,
+      "step": 3241
+    },
+    {
+      "epoch": 2.6322056833558864,
+      "grad_norm": 0.19726337492465973,
+      "learning_rate": 4.476541585420785e-07,
+      "loss": 0.3393,
+      "step": 3242
+    },
+    {
+      "epoch": 2.633017591339648,
+      "grad_norm": 0.18975725769996643,
+      "learning_rate": 4.457011973212083e-07,
+      "loss": 0.2944,
+      "step": 3243
+    },
+    {
+      "epoch": 2.63382949932341,
+      "grad_norm": 0.2006247192621231,
+      "learning_rate": 4.437523067202365e-07,
+      "loss": 0.3376,
+      "step": 3244
+    },
+    {
+      "epoch": 2.634641407307172,
+      "grad_norm": 0.19280685484409332,
+      "learning_rate": 4.4180748848107745e-07,
+      "loss": 0.3467,
+      "step": 3245
+    },
+    {
+      "epoch": 2.635453315290934,
+      "grad_norm": 0.19434425234794617,
+      "learning_rate": 4.398667443420029e-07,
+      "loss": 0.3422,
+      "step": 3246
+    },
+    {
+      "epoch": 2.6362652232746955,
+      "grad_norm": 0.17332780361175537,
+      "learning_rate": 4.379300760376465e-07,
+      "loss": 0.2957,
+      "step": 3247
+    },
+    {
+      "epoch": 2.637077131258457,
+      "grad_norm": 0.18899551033973694,
+      "learning_rate": 4.359974852989951e-07,
+      "loss": 0.3638,
+      "step": 3248
+    },
+    {
+      "epoch": 2.637889039242219,
+      "grad_norm": 0.17284271121025085,
+      "learning_rate": 4.3406897385339544e-07,
+      "loss": 0.3115,
+      "step": 3249
+    },
+    {
+      "epoch": 2.6387009472259813,
+      "grad_norm": 0.18669451773166656,
+      "learning_rate": 4.3214454342454357e-07,
+      "loss": 0.3183,
+      "step": 3250
+    },
+    {
+      "epoch": 2.639512855209743,
+      "grad_norm": 0.1893383413553238,
+      "learning_rate": 4.302241957324915e-07,
+      "loss": 0.3284,
+      "step": 3251
+    },
+    {
+      "epoch": 2.6403247631935045,
+      "grad_norm": 0.17861998081207275,
+      "learning_rate": 4.283079324936401e-07,
+      "loss": 0.311,
+      "step": 3252
+    },
+    {
+      "epoch": 2.6411366711772666,
+      "grad_norm": 0.19436894357204437,
+      "learning_rate": 4.263957554207415e-07,
+      "loss": 0.3237,
+      "step": 3253
+    },
+    {
+      "epoch": 2.6419485791610287,
+      "grad_norm": 0.17673245072364807,
+      "learning_rate": 4.244876662228914e-07,
+      "loss": 0.321,
+      "step": 3254
+    },
+    {
+      "epoch": 2.6427604871447903,
+      "grad_norm": 0.17087873816490173,
+      "learning_rate": 4.2258366660553885e-07,
+      "loss": 0.3298,
+      "step": 3255
+    },
+    {
+      "epoch": 2.643572395128552,
+      "grad_norm": 0.1881648600101471,
+      "learning_rate": 4.206837582704704e-07,
+      "loss": 0.334,
+      "step": 3256
+    },
+    {
+      "epoch": 2.644384303112314,
+      "grad_norm": 0.18569831550121307,
+      "learning_rate": 4.187879429158204e-07,
+      "loss": 0.3512,
+      "step": 3257
+    },
+    {
+      "epoch": 2.645196211096076,
+      "grad_norm": 0.19067305326461792,
+      "learning_rate": 4.1689622223606283e-07,
+      "loss": 0.3245,
+      "step": 3258
+    },
+    {
+      "epoch": 2.6460081190798377,
+      "grad_norm": 0.17806801199913025,
+      "learning_rate": 4.1500859792201363e-07,
+      "loss": 0.3022,
+      "step": 3259
+    },
+    {
+      "epoch": 2.6468200270635993,
+      "grad_norm": 0.19906294345855713,
+      "learning_rate": 4.1312507166082425e-07,
+      "loss": 0.3642,
+      "step": 3260
+    },
+    {
+      "epoch": 2.6476319350473614,
+      "grad_norm": 0.17814120650291443,
+      "learning_rate": 4.1124564513598755e-07,
+      "loss": 0.2922,
+      "step": 3261
+    },
+    {
+      "epoch": 2.648443843031123,
+      "grad_norm": 0.2004735767841339,
+      "learning_rate": 4.09370320027328e-07,
+      "loss": 0.3434,
+      "step": 3262
+    },
+    {
+      "epoch": 2.649255751014885,
+      "grad_norm": 0.20643550157546997,
+      "learning_rate": 4.0749909801100775e-07,
+      "loss": 0.3191,
+      "step": 3263
+    },
+    {
+      "epoch": 2.6500676589986467,
+      "grad_norm": 0.19981984794139862,
+      "learning_rate": 4.0563198075951824e-07,
+      "loss": 0.3336,
+      "step": 3264
+    },
+    {
+      "epoch": 2.650879566982409,
+      "grad_norm": 0.17844019830226898,
+      "learning_rate": 4.037689699416858e-07,
+      "loss": 0.3396,
+      "step": 3265
+    },
+    {
+      "epoch": 2.6516914749661704,
+      "grad_norm": 0.17575010657310486,
+      "learning_rate": 4.019100672226617e-07,
+      "loss": 0.3183,
+      "step": 3266
+    },
+    {
+      "epoch": 2.6525033829499325,
+      "grad_norm": 0.1943996101617813,
+      "learning_rate": 4.000552742639302e-07,
+      "loss": 0.3354,
+      "step": 3267
+    },
+    {
+      "epoch": 2.653315290933694,
+      "grad_norm": 0.19651736319065094,
+      "learning_rate": 3.982045927233008e-07,
+      "loss": 0.3589,
+      "step": 3268
+    },
+    {
+      "epoch": 2.654127198917456,
+      "grad_norm": 0.1850006878376007,
+      "learning_rate": 3.963580242549053e-07,
+      "loss": 0.3117,
+      "step": 3269
+    },
+    {
+      "epoch": 2.654939106901218,
+      "grad_norm": 0.1965353786945343,
+      "learning_rate": 3.9451557050920406e-07,
+      "loss": 0.3282,
+      "step": 3270
+    },
+    {
+      "epoch": 2.65575101488498,
+      "grad_norm": 0.18477128446102142,
+      "learning_rate": 3.9267723313297535e-07,
+      "loss": 0.331,
+      "step": 3271
+    },
+    {
+      "epoch": 2.6565629228687415,
+      "grad_norm": 0.20113030076026917,
+      "learning_rate": 3.9084301376932143e-07,
+      "loss": 0.3708,
+      "step": 3272
+    },
+    {
+      "epoch": 2.657374830852503,
+      "grad_norm": 0.21849849820137024,
+      "learning_rate": 3.8901291405766097e-07,
+      "loss": 0.3349,
+      "step": 3273
+    },
+    {
+      "epoch": 2.658186738836265,
+      "grad_norm": 0.19243885576725006,
+      "learning_rate": 3.8718693563373376e-07,
+      "loss": 0.3288,
+      "step": 3274
+    },
+    {
+      "epoch": 2.6589986468200273,
+      "grad_norm": 0.1897915005683899,
+      "learning_rate": 3.853650801295933e-07,
+      "loss": 0.3424,
+      "step": 3275
+    },
+    {
+      "epoch": 2.659810554803789,
+      "grad_norm": 0.17580671608448029,
+      "learning_rate": 3.8354734917361024e-07,
+      "loss": 0.3226,
+      "step": 3276
+    },
+    {
+      "epoch": 2.6606224627875505,
+      "grad_norm": 0.17831555008888245,
+      "learning_rate": 3.817337443904662e-07,
+      "loss": 0.32,
+      "step": 3277
+    },
+    {
+      "epoch": 2.6614343707713126,
+      "grad_norm": 0.22693440318107605,
+      "learning_rate": 3.799242674011572e-07,
+      "loss": 0.2977,
+      "step": 3278
+    },
+    {
+      "epoch": 2.6622462787550747,
+      "grad_norm": 0.19768564403057098,
+      "learning_rate": 3.7811891982298843e-07,
+      "loss": 0.3166,
+      "step": 3279
+    },
+    {
+      "epoch": 2.6630581867388363,
+      "grad_norm": 0.19116057455539703,
+      "learning_rate": 3.7631770326957475e-07,
+      "loss": 0.2985,
+      "step": 3280
+    },
+    {
+      "epoch": 2.663870094722598,
+      "grad_norm": 0.1851990520954132,
+      "learning_rate": 3.7452061935083826e-07,
+      "loss": 0.3297,
+      "step": 3281
+    },
+    {
+      "epoch": 2.66468200270636,
+      "grad_norm": 0.17983731627464294,
+      "learning_rate": 3.7272766967300864e-07,
+      "loss": 0.3205,
+      "step": 3282
+    },
+    {
+      "epoch": 2.6654939106901216,
+      "grad_norm": 0.18241935968399048,
+      "learning_rate": 3.709388558386179e-07,
+      "loss": 0.3337,
+      "step": 3283
+    },
+    {
+      "epoch": 2.6663058186738837,
+      "grad_norm": 0.1868843138217926,
+      "learning_rate": 3.691541794465042e-07,
+      "loss": 0.3594,
+      "step": 3284
+    },
+    {
+      "epoch": 2.6671177266576453,
+      "grad_norm": 0.20077484846115112,
+      "learning_rate": 3.673736420918067e-07,
+      "loss": 0.2974,
+      "step": 3285
+    },
+    {
+      "epoch": 2.6679296346414074,
+      "grad_norm": 0.18288151919841766,
+      "learning_rate": 3.65597245365964e-07,
+      "loss": 0.3062,
+      "step": 3286
+    },
+    {
+      "epoch": 2.668741542625169,
+      "grad_norm": 0.1806783378124237,
+      "learning_rate": 3.638249908567154e-07,
+      "loss": 0.3269,
+      "step": 3287
+    },
+    {
+      "epoch": 2.669553450608931,
+      "grad_norm": 0.17670366168022156,
+      "learning_rate": 3.6205688014809615e-07,
+      "loss": 0.3303,
+      "step": 3288
+    },
+    {
+      "epoch": 2.6703653585926928,
+      "grad_norm": 0.1930493861436844,
+      "learning_rate": 3.602929148204398e-07,
+      "loss": 0.3601,
+      "step": 3289
+    },
+    {
+      "epoch": 2.671177266576455,
+      "grad_norm": 0.17786367237567902,
+      "learning_rate": 3.5853309645037306e-07,
+      "loss": 0.311,
+      "step": 3290
+    },
+    {
+      "epoch": 2.6719891745602165,
+      "grad_norm": 0.1777045875787735,
+      "learning_rate": 3.567774266108176e-07,
+      "loss": 0.3237,
+      "step": 3291
+    },
+    {
+      "epoch": 2.6728010825439785,
+      "grad_norm": 0.17502866685390472,
+      "learning_rate": 3.550259068709844e-07,
+      "loss": 0.301,
+      "step": 3292
+    },
+    {
+      "epoch": 2.67361299052774,
+      "grad_norm": 0.21089668571949005,
+      "learning_rate": 3.5327853879637994e-07,
+      "loss": 0.3367,
+      "step": 3293
+    },
+    {
+      "epoch": 2.674424898511502,
+      "grad_norm": 0.18312060832977295,
+      "learning_rate": 3.5153532394879507e-07,
+      "loss": 0.345,
+      "step": 3294
+    },
+    {
+      "epoch": 2.675236806495264,
+      "grad_norm": 0.18578796088695526,
+      "learning_rate": 3.4979626388631174e-07,
+      "loss": 0.3386,
+      "step": 3295
+    },
+    {
+      "epoch": 2.676048714479026,
+      "grad_norm": 0.19639916718006134,
+      "learning_rate": 3.4806136016329537e-07,
+      "loss": 0.3533,
+      "step": 3296
+    },
+    {
+      "epoch": 2.6768606224627876,
+      "grad_norm": 0.18068018555641174,
+      "learning_rate": 3.4633061433040014e-07,
+      "loss": 0.296,
+      "step": 3297
+    },
+    {
+      "epoch": 2.677672530446549,
+      "grad_norm": 0.1743270605802536,
+      "learning_rate": 3.4460402793456083e-07,
+      "loss": 0.3245,
+      "step": 3298
+    },
+    {
+      "epoch": 2.6784844384303113,
+      "grad_norm": 0.18615074455738068,
+      "learning_rate": 3.4288160251899625e-07,
+      "loss": 0.3472,
+      "step": 3299
+    },
+    {
+      "epoch": 2.6792963464140733,
+      "grad_norm": 0.19390513002872467,
+      "learning_rate": 3.4116333962320445e-07,
+      "loss": 0.3349,
+      "step": 3300
+    },
+    {
+      "epoch": 2.680108254397835,
+      "grad_norm": 0.18291126191616058,
+      "learning_rate": 3.394492407829658e-07,
+      "loss": 0.331,
+      "step": 3301
+    },
+    {
+      "epoch": 2.6809201623815966,
+      "grad_norm": 0.19587011635303497,
+      "learning_rate": 3.3773930753033535e-07,
+      "loss": 0.3408,
+      "step": 3302
+    },
+    {
+      "epoch": 2.6817320703653587,
+      "grad_norm": 0.19236768782138824,
+      "learning_rate": 3.360335413936489e-07,
+      "loss": 0.3394,
+      "step": 3303
+    },
+    {
+      "epoch": 2.6825439783491203,
+      "grad_norm": 0.17384736239910126,
+      "learning_rate": 3.343319438975134e-07,
+      "loss": 0.2776,
+      "step": 3304
+    },
+    {
+      "epoch": 2.6833558863328824,
+      "grad_norm": 0.19803327322006226,
+      "learning_rate": 3.3263451656281364e-07,
+      "loss": 0.3614,
+      "step": 3305
+    },
+    {
+      "epoch": 2.684167794316644,
+      "grad_norm": 0.17739756405353546,
+      "learning_rate": 3.3094126090670477e-07,
+      "loss": 0.283,
+      "step": 3306
+    },
+    {
+      "epoch": 2.684979702300406,
+      "grad_norm": 0.1925831437110901,
+      "learning_rate": 3.2925217844261503e-07,
+      "loss": 0.3381,
+      "step": 3307
+    },
+    {
+      "epoch": 2.6857916102841677,
+      "grad_norm": 0.18470527231693268,
+      "learning_rate": 3.275672706802402e-07,
+      "loss": 0.3255,
+      "step": 3308
+    },
+    {
+      "epoch": 2.6866035182679298,
+      "grad_norm": 0.1773693859577179,
+      "learning_rate": 3.258865391255478e-07,
+      "loss": 0.299,
+      "step": 3309
+    },
+    {
+      "epoch": 2.6874154262516914,
+      "grad_norm": 0.18930688500404358,
+      "learning_rate": 3.242099852807695e-07,
+      "loss": 0.3764,
+      "step": 3310
+    },
+    {
+      "epoch": 2.6882273342354535,
+      "grad_norm": 0.20800574123859406,
+      "learning_rate": 3.225376106444056e-07,
+      "loss": 0.3092,
+      "step": 3311
+    },
+    {
+      "epoch": 2.689039242219215,
+      "grad_norm": 0.19169723987579346,
+      "learning_rate": 3.2086941671122084e-07,
+      "loss": 0.3825,
+      "step": 3312
+    },
+    {
+      "epoch": 2.689851150202977,
+      "grad_norm": 0.16201266646385193,
+      "learning_rate": 3.1920540497224095e-07,
+      "loss": 0.2845,
+      "step": 3313
+    },
+    {
+      "epoch": 2.690663058186739,
+      "grad_norm": 0.17472629249095917,
+      "learning_rate": 3.1754557691475617e-07,
+      "loss": 0.3433,
+      "step": 3314
+    },
+    {
+      "epoch": 2.6914749661705004,
+      "grad_norm": 0.1810588389635086,
+      "learning_rate": 3.158899340223154e-07,
+      "loss": 0.3449,
+      "step": 3315
+    },
+    {
+      "epoch": 2.6922868741542625,
+      "grad_norm": 0.17703375220298767,
+      "learning_rate": 3.142384777747287e-07,
+      "loss": 0.3443,
+      "step": 3316
+    },
+    {
+      "epoch": 2.6930987821380246,
+      "grad_norm": 0.18263068795204163,
+      "learning_rate": 3.125912096480621e-07,
+      "loss": 0.3284,
+      "step": 3317
+    },
+    {
+      "epoch": 2.693910690121786,
+      "grad_norm": 0.19459748268127441,
+      "learning_rate": 3.1094813111464004e-07,
+      "loss": 0.3258,
+      "step": 3318
+    },
+    {
+      "epoch": 2.694722598105548,
+      "grad_norm": 0.19226200878620148,
+      "learning_rate": 3.0930924364304117e-07,
+      "loss": 0.3314,
+      "step": 3319
+    },
+    {
+      "epoch": 2.69553450608931,
+      "grad_norm": 0.1767144352197647,
+      "learning_rate": 3.076745486981003e-07,
+      "loss": 0.3199,
+      "step": 3320
+    },
+    {
+      "epoch": 2.696346414073072,
+      "grad_norm": 0.1810034215450287,
+      "learning_rate": 3.060440477409005e-07,
+      "loss": 0.3345,
+      "step": 3321
+    },
+    {
+      "epoch": 2.6971583220568336,
+      "grad_norm": 0.1674317568540573,
+      "learning_rate": 3.0441774222878194e-07,
+      "loss": 0.3461,
+      "step": 3322
+    },
+    {
+      "epoch": 2.697970230040595,
+      "grad_norm": 0.17940710484981537,
+      "learning_rate": 3.027956336153298e-07,
+      "loss": 0.3138,
+      "step": 3323
+    },
+    {
+      "epoch": 2.6987821380243573,
+      "grad_norm": 0.18338648974895477,
+      "learning_rate": 3.011777233503821e-07,
+      "loss": 0.334,
+      "step": 3324
+    },
+    {
+      "epoch": 2.699594046008119,
+      "grad_norm": 0.2249225378036499,
+      "learning_rate": 2.995640128800209e-07,
+      "loss": 0.362,
+      "step": 3325
+    },
+    {
+      "epoch": 2.700405953991881,
+      "grad_norm": 0.17677171528339386,
+      "learning_rate": 2.9795450364657865e-07,
+      "loss": 0.3034,
+      "step": 3326
+    },
+    {
+      "epoch": 2.7012178619756426,
+      "grad_norm": 0.18953631818294525,
+      "learning_rate": 2.963491970886273e-07,
+      "loss": 0.341,
+      "step": 3327
+    },
+    {
+      "epoch": 2.7020297699594047,
+      "grad_norm": 0.17596660554409027,
+      "learning_rate": 2.947480946409875e-07,
+      "loss": 0.3512,
+      "step": 3328
+    },
+    {
+      "epoch": 2.7028416779431663,
+      "grad_norm": 0.17635232210159302,
+      "learning_rate": 2.9315119773472067e-07,
+      "loss": 0.3036,
+      "step": 3329
+    },
+    {
+      "epoch": 2.7036535859269284,
+      "grad_norm": 0.1864037960767746,
+      "learning_rate": 2.9155850779712723e-07,
+      "loss": 0.3501,
+      "step": 3330
+    },
+    {
+      "epoch": 2.70446549391069,
+      "grad_norm": 0.17953728139400482,
+      "learning_rate": 2.899700262517502e-07,
+      "loss": 0.3025,
+      "step": 3331
+    },
+    {
+      "epoch": 2.705277401894452,
+      "grad_norm": 0.1900356262922287,
+      "learning_rate": 2.8838575451837046e-07,
+      "loss": 0.3275,
+      "step": 3332
+    },
+    {
+      "epoch": 2.7060893098782137,
+      "grad_norm": 0.18842825293540955,
+      "learning_rate": 2.868056940130048e-07,
+      "loss": 0.3588,
+      "step": 3333
+    },
+    {
+      "epoch": 2.706901217861976,
+      "grad_norm": 0.1944694221019745,
+      "learning_rate": 2.852298461479075e-07,
+      "loss": 0.333,
+      "step": 3334
+    },
+    {
+      "epoch": 2.7077131258457374,
+      "grad_norm": 0.1897508054971695,
+      "learning_rate": 2.8365821233156677e-07,
+      "loss": 0.3388,
+      "step": 3335
+    },
+    {
+      "epoch": 2.708525033829499,
+      "grad_norm": 0.17850112915039062,
+      "learning_rate": 2.8209079396870454e-07,
+      "loss": 0.3257,
+      "step": 3336
+    },
+    {
+      "epoch": 2.709336941813261,
+      "grad_norm": 0.19416135549545288,
+      "learning_rate": 2.8052759246027626e-07,
+      "loss": 0.3309,
+      "step": 3337
+    },
+    {
+      "epoch": 2.710148849797023,
+      "grad_norm": 0.16008704900741577,
+      "learning_rate": 2.7896860920346605e-07,
+      "loss": 0.3003,
+      "step": 3338
+    },
+    {
+      "epoch": 2.710960757780785,
+      "grad_norm": 0.1698494255542755,
+      "learning_rate": 2.774138455916897e-07,
+      "loss": 0.2993,
+      "step": 3339
+    },
+    {
+      "epoch": 2.7117726657645465,
+      "grad_norm": 0.21187497675418854,
+      "learning_rate": 2.758633030145891e-07,
+      "loss": 0.3522,
+      "step": 3340
+    },
+    {
+      "epoch": 2.7125845737483085,
+      "grad_norm": 0.16605067253112793,
+      "learning_rate": 2.743169828580372e-07,
+      "loss": 0.2881,
+      "step": 3341
+    },
+    {
+      "epoch": 2.7133964817320706,
+      "grad_norm": 0.18292470276355743,
+      "learning_rate": 2.727748865041291e-07,
+      "loss": 0.3514,
+      "step": 3342
+    },
+    {
+      "epoch": 2.7142083897158322,
+      "grad_norm": 0.193888857960701,
+      "learning_rate": 2.712370153311872e-07,
+      "loss": 0.2947,
+      "step": 3343
+    },
+    {
+      "epoch": 2.715020297699594,
+      "grad_norm": 0.18411600589752197,
+      "learning_rate": 2.6970337071375584e-07,
+      "loss": 0.3171,
+      "step": 3344
+    },
+    {
+      "epoch": 2.715832205683356,
+      "grad_norm": 0.20722849667072296,
+      "learning_rate": 2.681739540226036e-07,
+      "loss": 0.3392,
+      "step": 3345
+    },
+    {
+      "epoch": 2.716644113667118,
+      "grad_norm": 0.18540751934051514,
+      "learning_rate": 2.6664876662471697e-07,
+      "loss": 0.3503,
+      "step": 3346
+    },
+    {
+      "epoch": 2.7174560216508796,
+      "grad_norm": 0.1810639202594757,
+      "learning_rate": 2.6512780988330654e-07,
+      "loss": 0.3114,
+      "step": 3347
+    },
+    {
+      "epoch": 2.7182679296346413,
+      "grad_norm": 0.19516801834106445,
+      "learning_rate": 2.6361108515779777e-07,
+      "loss": 0.3071,
+      "step": 3348
+    },
+    {
+      "epoch": 2.7190798376184033,
+      "grad_norm": 0.1803337186574936,
+      "learning_rate": 2.6209859380383673e-07,
+      "loss": 0.2832,
+      "step": 3349
+    },
+    {
+      "epoch": 2.719891745602165,
+      "grad_norm": 0.18409375846385956,
+      "learning_rate": 2.6059033717328154e-07,
+      "loss": 0.3114,
+      "step": 3350
+    },
+    {
+      "epoch": 2.720703653585927,
+      "grad_norm": 0.17622490227222443,
+      "learning_rate": 2.590863166142105e-07,
+      "loss": 0.3397,
+      "step": 3351
+    },
+    {
+      "epoch": 2.7215155615696887,
+      "grad_norm": 0.1979091316461563,
+      "learning_rate": 2.5758653347091103e-07,
+      "loss": 0.3126,
+      "step": 3352
+    },
+    {
+      "epoch": 2.7223274695534507,
+      "grad_norm": 0.18453000485897064,
+      "learning_rate": 2.560909890838864e-07,
+      "loss": 0.3208,
+      "step": 3353
+    },
+    {
+      "epoch": 2.7231393775372124,
+      "grad_norm": 0.19602401554584503,
+      "learning_rate": 2.5459968478985074e-07,
+      "loss": 0.3418,
+      "step": 3354
+    },
+    {
+      "epoch": 2.7239512855209744,
+      "grad_norm": 0.16978849470615387,
+      "learning_rate": 2.531126219217256e-07,
+      "loss": 0.3095,
+      "step": 3355
+    },
+    {
+      "epoch": 2.724763193504736,
+      "grad_norm": 0.1866815835237503,
+      "learning_rate": 2.5162980180864605e-07,
+      "loss": 0.3407,
+      "step": 3356
+    },
+    {
+      "epoch": 2.725575101488498,
+      "grad_norm": 0.18332692980766296,
+      "learning_rate": 2.5015122577595084e-07,
+      "loss": 0.3386,
+      "step": 3357
+    },
+    {
+      "epoch": 2.7263870094722598,
+      "grad_norm": 0.17671167850494385,
+      "learning_rate": 2.48676895145189e-07,
+      "loss": 0.3384,
+      "step": 3358
+    },
+    {
+      "epoch": 2.727198917456022,
+      "grad_norm": 0.2008109837770462,
+      "learning_rate": 2.472068112341114e-07,
+      "loss": 0.3159,
+      "step": 3359
+    },
+    {
+      "epoch": 2.7280108254397835,
+      "grad_norm": 0.1930491179227829,
+      "learning_rate": 2.4574097535667696e-07,
+      "loss": 0.323,
+      "step": 3360
+    },
+    {
+      "epoch": 2.728822733423545,
+      "grad_norm": 0.19536791741847992,
+      "learning_rate": 2.4427938882304383e-07,
+      "loss": 0.3625,
+      "step": 3361
+    },
+    {
+      "epoch": 2.729634641407307,
+      "grad_norm": 0.19736993312835693,
+      "learning_rate": 2.428220529395753e-07,
+      "loss": 0.3547,
+      "step": 3362
+    },
+    {
+      "epoch": 2.7304465493910692,
+      "grad_norm": 0.19344642758369446,
+      "learning_rate": 2.413689690088339e-07,
+      "loss": 0.3108,
+      "step": 3363
+    },
+    {
+      "epoch": 2.731258457374831,
+      "grad_norm": 0.1765623241662979,
+      "learning_rate": 2.399201383295824e-07,
+      "loss": 0.3,
+      "step": 3364
+    },
+    {
+      "epoch": 2.7320703653585925,
+      "grad_norm": 0.18518368899822235,
+      "learning_rate": 2.3847556219678105e-07,
+      "loss": 0.3445,
+      "step": 3365
+    },
+    {
+      "epoch": 2.7328822733423546,
+      "grad_norm": 0.1820385456085205,
+      "learning_rate": 2.370352419015892e-07,
+      "loss": 0.3499,
+      "step": 3366
+    },
+    {
+      "epoch": 2.7336941813261166,
+      "grad_norm": 0.1803084760904312,
+      "learning_rate": 2.3559917873135996e-07,
+      "loss": 0.3556,
+      "step": 3367
+    },
+    {
+      "epoch": 2.7345060893098783,
+      "grad_norm": 0.17809373140335083,
+      "learning_rate": 2.3416737396964373e-07,
+      "loss": 0.2991,
+      "step": 3368
+    },
+    {
+      "epoch": 2.73531799729364,
+      "grad_norm": 0.17525096237659454,
+      "learning_rate": 2.3273982889618186e-07,
+      "loss": 0.3225,
+      "step": 3369
+    },
+    {
+      "epoch": 2.736129905277402,
+      "grad_norm": 0.16653640568256378,
+      "learning_rate": 2.3131654478691258e-07,
+      "loss": 0.3325,
+      "step": 3370
+    },
+    {
+      "epoch": 2.7369418132611636,
+      "grad_norm": 0.1823665052652359,
+      "learning_rate": 2.2989752291396218e-07,
+      "loss": 0.3163,
+      "step": 3371
+    },
+    {
+      "epoch": 2.7377537212449257,
+      "grad_norm": 0.19122840464115143,
+      "learning_rate": 2.2848276454564943e-07,
+      "loss": 0.3492,
+      "step": 3372
+    },
+    {
+      "epoch": 2.7385656292286873,
+      "grad_norm": 0.17212402820587158,
+      "learning_rate": 2.270722709464801e-07,
+      "loss": 0.2844,
+      "step": 3373
+    },
+    {
+      "epoch": 2.7393775372124494,
+      "grad_norm": 0.19412873685359955,
+      "learning_rate": 2.2566604337715127e-07,
+      "loss": 0.3348,
+      "step": 3374
+    },
+    {
+      "epoch": 2.740189445196211,
+      "grad_norm": 0.18164421617984772,
+      "learning_rate": 2.242640830945436e-07,
+      "loss": 0.2953,
+      "step": 3375
+    },
+    {
+      "epoch": 2.741001353179973,
+      "grad_norm": 0.1983167827129364,
+      "learning_rate": 2.228663913517265e-07,
+      "loss": 0.3474,
+      "step": 3376
+    },
+    {
+      "epoch": 2.7418132611637347,
+      "grad_norm": 0.17161297798156738,
+      "learning_rate": 2.2147296939795226e-07,
+      "loss": 0.3141,
+      "step": 3377
+    },
+    {
+      "epoch": 2.7426251691474968,
+      "grad_norm": 0.17643731832504272,
+      "learning_rate": 2.2008381847865911e-07,
+      "loss": 0.3309,
+      "step": 3378
+    },
+    {
+      "epoch": 2.7434370771312584,
+      "grad_norm": 0.18373622000217438,
+      "learning_rate": 2.1869893983546442e-07,
+      "loss": 0.3092,
+      "step": 3379
+    },
+    {
+      "epoch": 2.7442489851150205,
+      "grad_norm": 0.2007768452167511,
+      "learning_rate": 2.173183347061697e-07,
+      "loss": 0.3506,
+      "step": 3380
+    },
+    {
+      "epoch": 2.745060893098782,
+      "grad_norm": 0.20613454282283783,
+      "learning_rate": 2.1594200432475675e-07,
+      "loss": 0.3761,
+      "step": 3381
+    },
+    {
+      "epoch": 2.7458728010825437,
+      "grad_norm": 0.18432773649692535,
+      "learning_rate": 2.145699499213849e-07,
+      "loss": 0.303,
+      "step": 3382
+    },
+    {
+      "epoch": 2.746684709066306,
+      "grad_norm": 0.17016445100307465,
+      "learning_rate": 2.132021727223943e-07,
+      "loss": 0.3135,
+      "step": 3383
+    },
+    {
+      "epoch": 2.747496617050068,
+      "grad_norm": 0.17576153576374054,
+      "learning_rate": 2.1183867395029812e-07,
+      "loss": 0.3242,
+      "step": 3384
+    },
+    {
+      "epoch": 2.7483085250338295,
+      "grad_norm": 0.20130284130573273,
+      "learning_rate": 2.1047945482378985e-07,
+      "loss": 0.3728,
+      "step": 3385
+    },
+    {
+      "epoch": 2.749120433017591,
+      "grad_norm": 0.1804194450378418,
+      "learning_rate": 2.091245165577349e-07,
+      "loss": 0.3116,
+      "step": 3386
+    },
+    {
+      "epoch": 2.749932341001353,
+      "grad_norm": 0.18645258247852325,
+      "learning_rate": 2.0777386036317404e-07,
+      "loss": 0.3143,
+      "step": 3387
+    },
+    {
+      "epoch": 2.7507442489851153,
+      "grad_norm": 0.17165416479110718,
+      "learning_rate": 2.0642748744731878e-07,
+      "loss": 0.3117,
+      "step": 3388
+    },
+    {
+      "epoch": 2.751556156968877,
+      "grad_norm": 0.17007847130298615,
+      "learning_rate": 2.050853990135554e-07,
+      "loss": 0.3269,
+      "step": 3389
+    },
+    {
+      "epoch": 2.7523680649526385,
+      "grad_norm": 0.194077268242836,
+      "learning_rate": 2.0374759626143714e-07,
+      "loss": 0.3447,
+      "step": 3390
+    },
+    {
+      "epoch": 2.7531799729364006,
+      "grad_norm": 0.19125232100486755,
+      "learning_rate": 2.0241408038668976e-07,
+      "loss": 0.3366,
+      "step": 3391
+    },
+    {
+      "epoch": 2.7539918809201622,
+      "grad_norm": 0.18619458377361298,
+      "learning_rate": 2.0108485258120536e-07,
+      "loss": 0.3005,
+      "step": 3392
+    },
+    {
+      "epoch": 2.7548037889039243,
+      "grad_norm": 0.19056960940361023,
+      "learning_rate": 1.9975991403304529e-07,
+      "loss": 0.3716,
+      "step": 3393
+    },
+    {
+      "epoch": 2.755615696887686,
+      "grad_norm": 0.18696384131908417,
+      "learning_rate": 1.9843926592643392e-07,
+      "loss": 0.3357,
+      "step": 3394
+    },
+    {
+      "epoch": 2.756427604871448,
+      "grad_norm": 0.18273375928401947,
+      "learning_rate": 1.9712290944176537e-07,
+      "loss": 0.3347,
+      "step": 3395
+    },
+    {
+      "epoch": 2.7572395128552096,
+      "grad_norm": 0.1701890379190445,
+      "learning_rate": 1.9581084575559352e-07,
+      "loss": 0.2912,
+      "step": 3396
+    },
+    {
+      "epoch": 2.7580514208389717,
+      "grad_norm": 0.19522613286972046,
+      "learning_rate": 1.94503076040638e-07,
+      "loss": 0.3481,
+      "step": 3397
+    },
+    {
+      "epoch": 2.7588633288227333,
+      "grad_norm": 0.1846839189529419,
+      "learning_rate": 1.9319960146578055e-07,
+      "loss": 0.3325,
+      "step": 3398
+    },
+    {
+      "epoch": 2.7596752368064954,
+      "grad_norm": 0.183439701795578,
+      "learning_rate": 1.919004231960625e-07,
+      "loss": 0.2976,
+      "step": 3399
+    },
+    {
+      "epoch": 2.760487144790257,
+      "grad_norm": 0.1829131841659546,
+      "learning_rate": 1.9060554239268613e-07,
+      "loss": 0.3472,
+      "step": 3400
+    },
+    {
+      "epoch": 2.761299052774019,
+      "grad_norm": 0.18440894782543182,
+      "learning_rate": 1.8931496021301178e-07,
+      "loss": 0.3183,
+      "step": 3401
+    },
+    {
+      "epoch": 2.7621109607577807,
+      "grad_norm": 0.18943548202514648,
+      "learning_rate": 1.8802867781055943e-07,
+      "loss": 0.3548,
+      "step": 3402
+    },
+    {
+      "epoch": 2.7629228687415424,
+      "grad_norm": 0.16972792148590088,
+      "learning_rate": 1.8674669633500342e-07,
+      "loss": 0.3101,
+      "step": 3403
+    },
+    {
+      "epoch": 2.7637347767253044,
+      "grad_norm": 0.18760138750076294,
+      "learning_rate": 1.8546901693217655e-07,
+      "loss": 0.3251,
+      "step": 3404
+    },
+    {
+      "epoch": 2.7645466847090665,
+      "grad_norm": 0.18381749093532562,
+      "learning_rate": 1.8419564074406427e-07,
+      "loss": 0.3243,
+      "step": 3405
+    },
+    {
+      "epoch": 2.765358592692828,
+      "grad_norm": 0.1822693943977356,
+      "learning_rate": 1.8292656890880722e-07,
+      "loss": 0.2927,
+      "step": 3406
+    },
+    {
+      "epoch": 2.7661705006765898,
+      "grad_norm": 0.16798989474773407,
+      "learning_rate": 1.8166180256069753e-07,
+      "loss": 0.3506,
+      "step": 3407
+    },
+    {
+      "epoch": 2.766982408660352,
+      "grad_norm": 0.17436732351779938,
+      "learning_rate": 1.8040134283018207e-07,
+      "loss": 0.3635,
+      "step": 3408
+    },
+    {
+      "epoch": 2.767794316644114,
+      "grad_norm": 0.1733822524547577,
+      "learning_rate": 1.7914519084385407e-07,
+      "loss": 0.3371,
+      "step": 3409
+    },
+    {
+      "epoch": 2.7686062246278755,
+      "grad_norm": 0.1660536229610443,
+      "learning_rate": 1.7789334772445987e-07,
+      "loss": 0.282,
+      "step": 3410
+    },
+    {
+      "epoch": 2.769418132611637,
+      "grad_norm": 0.1862877607345581,
+      "learning_rate": 1.7664581459089335e-07,
+      "loss": 0.3716,
+      "step": 3411
+    },
+    {
+      "epoch": 2.7702300405953992,
+      "grad_norm": 0.18035022914409637,
+      "learning_rate": 1.754025925581959e-07,
+      "loss": 0.3149,
+      "step": 3412
+    },
+    {
+      "epoch": 2.771041948579161,
+      "grad_norm": 0.1925985962152481,
+      "learning_rate": 1.7416368273755646e-07,
+      "loss": 0.329,
+      "step": 3413
+    },
+    {
+      "epoch": 2.771853856562923,
+      "grad_norm": 0.1805017590522766,
+      "learning_rate": 1.7292908623630866e-07,
+      "loss": 0.3736,
+      "step": 3414
+    },
+    {
+      "epoch": 2.7726657645466846,
+      "grad_norm": 0.19073544442653656,
+      "learning_rate": 1.7169880415793206e-07,
+      "loss": 0.3225,
+      "step": 3415
+    },
+    {
+      "epoch": 2.7734776725304466,
+      "grad_norm": 0.18778644502162933,
+      "learning_rate": 1.7047283760204925e-07,
+      "loss": 0.2763,
+      "step": 3416
+    },
+    {
+      "epoch": 2.7742895805142083,
+      "grad_norm": 0.1920059621334076,
+      "learning_rate": 1.6925118766442538e-07,
+      "loss": 0.3772,
+      "step": 3417
+    },
+    {
+      "epoch": 2.7751014884979703,
+      "grad_norm": 0.1962202489376068,
+      "learning_rate": 1.6803385543696872e-07,
+      "loss": 0.3401,
+      "step": 3418
+    },
+    {
+      "epoch": 2.775913396481732,
+      "grad_norm": 0.17882809042930603,
+      "learning_rate": 1.6682084200772663e-07,
+      "loss": 0.3167,
+      "step": 3419
+    },
+    {
+      "epoch": 2.776725304465494,
+      "grad_norm": 0.17723269760608673,
+      "learning_rate": 1.65612148460888e-07,
+      "loss": 0.3339,
+      "step": 3420
+    },
+    {
+      "epoch": 2.7775372124492557,
+      "grad_norm": 0.17593781650066376,
+      "learning_rate": 1.6440777587677924e-07,
+      "loss": 0.2935,
+      "step": 3421
+    },
+    {
+      "epoch": 2.7783491204330177,
+      "grad_norm": 0.1868167221546173,
+      "learning_rate": 1.6320772533186536e-07,
+      "loss": 0.3573,
+      "step": 3422
+    },
+    {
+      "epoch": 2.7791610284167794,
+      "grad_norm": 0.185824453830719,
+      "learning_rate": 1.6201199789874833e-07,
+      "loss": 0.3534,
+      "step": 3423
+    },
+    {
+      "epoch": 2.779972936400541,
+      "grad_norm": 0.1980564147233963,
+      "learning_rate": 1.6082059464616662e-07,
+      "loss": 0.3302,
+      "step": 3424
+    },
+    {
+      "epoch": 2.780784844384303,
+      "grad_norm": 0.19022195041179657,
+      "learning_rate": 1.5963351663899284e-07,
+      "loss": 0.3351,
+      "step": 3425
+    },
+    {
+      "epoch": 2.781596752368065,
+      "grad_norm": 0.19187839329242706,
+      "learning_rate": 1.5845076493823331e-07,
+      "loss": 0.3393,
+      "step": 3426
+    },
+    {
+      "epoch": 2.7824086603518268,
+      "grad_norm": 0.19183887541294098,
+      "learning_rate": 1.572723406010296e-07,
+      "loss": 0.3202,
+      "step": 3427
+    },
+    {
+      "epoch": 2.7832205683355884,
+      "grad_norm": 0.1861354261636734,
+      "learning_rate": 1.5609824468065306e-07,
+      "loss": 0.3397,
+      "step": 3428
+    },
+    {
+      "epoch": 2.7840324763193505,
+      "grad_norm": 0.17855960130691528,
+      "learning_rate": 1.549284782265087e-07,
+      "loss": 0.3163,
+      "step": 3429
+    },
+    {
+      "epoch": 2.7848443843031125,
+      "grad_norm": 0.19907405972480774,
+      "learning_rate": 1.537630422841291e-07,
+      "loss": 0.3194,
+      "step": 3430
+    },
+    {
+      "epoch": 2.785656292286874,
+      "grad_norm": 0.1716957837343216,
+      "learning_rate": 1.526019378951793e-07,
+      "loss": 0.3003,
+      "step": 3431
+    },
+    {
+      "epoch": 2.786468200270636,
+      "grad_norm": 0.1809748411178589,
+      "learning_rate": 1.5144516609744975e-07,
+      "loss": 0.3466,
+      "step": 3432
+    },
+    {
+      "epoch": 2.787280108254398,
+      "grad_norm": 0.17897577583789825,
+      "learning_rate": 1.5029272792486116e-07,
+      "loss": 0.2989,
+      "step": 3433
+    },
+    {
+      "epoch": 2.78809201623816,
+      "grad_norm": 0.17736178636550903,
+      "learning_rate": 1.4914462440745847e-07,
+      "loss": 0.3594,
+      "step": 3434
+    },
+    {
+      "epoch": 2.7889039242219216,
+      "grad_norm": 0.16936039924621582,
+      "learning_rate": 1.480008565714147e-07,
+      "loss": 0.3325,
+      "step": 3435
+    },
+    {
+      "epoch": 2.789715832205683,
+      "grad_norm": 0.17033784091472626,
+      "learning_rate": 1.4686142543902548e-07,
+      "loss": 0.3078,
+      "step": 3436
+    },
+    {
+      "epoch": 2.7905277401894453,
+      "grad_norm": 0.1838013380765915,
+      "learning_rate": 1.4572633202871222e-07,
+      "loss": 0.3565,
+      "step": 3437
+    },
+    {
+      "epoch": 2.791339648173207,
+      "grad_norm": 0.18368671834468842,
+      "learning_rate": 1.4459557735501671e-07,
+      "loss": 0.3322,
+      "step": 3438
+    },
+    {
+      "epoch": 2.792151556156969,
+      "grad_norm": 0.19118914008140564,
+      "learning_rate": 1.4346916242860664e-07,
+      "loss": 0.3595,
+      "step": 3439
+    },
+    {
+      "epoch": 2.7929634641407306,
+      "grad_norm": 0.1766211837530136,
+      "learning_rate": 1.4234708825626664e-07,
+      "loss": 0.3203,
+      "step": 3440
+    },
+    {
+      "epoch": 2.7937753721244927,
+      "grad_norm": 0.1728927046060562,
+      "learning_rate": 1.4122935584090447e-07,
+      "loss": 0.3195,
+      "step": 3441
+    },
+    {
+      "epoch": 2.7945872801082543,
+      "grad_norm": 0.17690646648406982,
+      "learning_rate": 1.4011596618154655e-07,
+      "loss": 0.2948,
+      "step": 3442
+    },
+    {
+      "epoch": 2.7953991880920164,
+      "grad_norm": 0.17195844650268555,
+      "learning_rate": 1.3900692027333628e-07,
+      "loss": 0.3167,
+      "step": 3443
+    },
+    {
+      "epoch": 2.796211096075778,
+      "grad_norm": 0.17757652699947357,
+      "learning_rate": 1.37902219107538e-07,
+      "loss": 0.3486,
+      "step": 3444
+    },
+    {
+      "epoch": 2.79702300405954,
+      "grad_norm": 0.18571889400482178,
+      "learning_rate": 1.3680186367152848e-07,
+      "loss": 0.3476,
+      "step": 3445
+    },
+    {
+      "epoch": 2.7978349120433017,
+      "grad_norm": 0.18292756378650665,
+      "learning_rate": 1.3570585494880328e-07,
+      "loss": 0.275,
+      "step": 3446
+    },
+    {
+      "epoch": 2.7986468200270638,
+      "grad_norm": 0.19311662018299103,
+      "learning_rate": 1.3461419391897268e-07,
+      "loss": 0.3584,
+      "step": 3447
+    },
+    {
+      "epoch": 2.7994587280108254,
+      "grad_norm": 0.17698229849338531,
+      "learning_rate": 1.3352688155775896e-07,
+      "loss": 0.3056,
+      "step": 3448
+    },
+    {
+      "epoch": 2.800270635994587,
+      "grad_norm": 0.19521954655647278,
+      "learning_rate": 1.3244391883700025e-07,
+      "loss": 0.3454,
+      "step": 3449
+    },
+    {
+      "epoch": 2.801082543978349,
+      "grad_norm": 0.18056681752204895,
+      "learning_rate": 1.3136530672464564e-07,
+      "loss": 0.3317,
+      "step": 3450
+    },
+    {
+      "epoch": 2.801894451962111,
+      "grad_norm": 0.16839607059955597,
+      "learning_rate": 1.3029104618475397e-07,
+      "loss": 0.3303,
+      "step": 3451
+    },
+    {
+      "epoch": 2.802706359945873,
+      "grad_norm": 0.18150226771831512,
+      "learning_rate": 1.292211381774988e-07,
+      "loss": 0.3246,
+      "step": 3452
+    },
+    {
+      "epoch": 2.8035182679296344,
+      "grad_norm": 0.19041599333286285,
+      "learning_rate": 1.2815558365915916e-07,
+      "loss": 0.3395,
+      "step": 3453
+    },
+    {
+      "epoch": 2.8043301759133965,
+      "grad_norm": 0.172689750790596,
+      "learning_rate": 1.2709438358212644e-07,
+      "loss": 0.2953,
+      "step": 3454
+    },
+    {
+      "epoch": 2.8051420838971586,
+      "grad_norm": 0.18549509346485138,
+      "learning_rate": 1.2603753889489645e-07,
+      "loss": 0.3481,
+      "step": 3455
+    },
+    {
+      "epoch": 2.80595399188092,
+      "grad_norm": 0.1799917072057724,
+      "learning_rate": 1.2498505054207578e-07,
+      "loss": 0.3296,
+      "step": 3456
+    },
+    {
+      "epoch": 2.806765899864682,
+      "grad_norm": 0.18276475369930267,
+      "learning_rate": 1.2393691946437414e-07,
+      "loss": 0.3296,
+      "step": 3457
+    },
+    {
+      "epoch": 2.807577807848444,
+      "grad_norm": 0.1761421412229538,
+      "learning_rate": 1.228931465986094e-07,
+      "loss": 0.2923,
+      "step": 3458
+    },
+    {
+      "epoch": 2.8083897158322055,
+      "grad_norm": 0.1932811588048935,
+      "learning_rate": 1.2185373287770254e-07,
+      "loss": 0.3618,
+      "step": 3459
+    },
+    {
+      "epoch": 2.8092016238159676,
+      "grad_norm": 0.19021354615688324,
+      "learning_rate": 1.2081867923067925e-07,
+      "loss": 0.3285,
+      "step": 3460
+    },
+    {
+      "epoch": 2.8100135317997292,
+      "grad_norm": 0.18134503066539764,
+      "learning_rate": 1.1978798658266678e-07,
+      "loss": 0.3313,
+      "step": 3461
+    },
+    {
+      "epoch": 2.8108254397834913,
+      "grad_norm": 0.1679713875055313,
+      "learning_rate": 1.1876165585489596e-07,
+      "loss": 0.2833,
+      "step": 3462
+    },
+    {
+      "epoch": 2.811637347767253,
+      "grad_norm": 0.18133126199245453,
+      "learning_rate": 1.1773968796469748e-07,
+      "loss": 0.322,
+      "step": 3463
+    },
+    {
+      "epoch": 2.812449255751015,
+      "grad_norm": 0.18458953499794006,
+      "learning_rate": 1.1672208382550509e-07,
+      "loss": 0.3365,
+      "step": 3464
+    },
+    {
+      "epoch": 2.8132611637347766,
+      "grad_norm": 0.18291443586349487,
+      "learning_rate": 1.1570884434684959e-07,
+      "loss": 0.366,
+      "step": 3465
+    },
+    {
+      "epoch": 2.8140730717185387,
+      "grad_norm": 0.17690277099609375,
+      "learning_rate": 1.1469997043436154e-07,
+      "loss": 0.3415,
+      "step": 3466
+    },
+    {
+      "epoch": 2.8148849797023003,
+      "grad_norm": 0.1846897006034851,
+      "learning_rate": 1.1369546298977018e-07,
+      "loss": 0.31,
+      "step": 3467
+    },
+    {
+      "epoch": 2.8156968876860624,
+      "grad_norm": 0.1842087209224701,
+      "learning_rate": 1.1269532291090124e-07,
+      "loss": 0.3081,
+      "step": 3468
+    },
+    {
+      "epoch": 2.816508795669824,
+      "grad_norm": 0.18614192306995392,
+      "learning_rate": 1.1169955109167741e-07,
+      "loss": 0.3347,
+      "step": 3469
+    },
+    {
+      "epoch": 2.8173207036535857,
+      "grad_norm": 0.1972523331642151,
+      "learning_rate": 1.1070814842211675e-07,
+      "loss": 0.312,
+      "step": 3470
+    },
+    {
+      "epoch": 2.8181326116373477,
+      "grad_norm": 0.20379500091075897,
+      "learning_rate": 1.0972111578833266e-07,
+      "loss": 0.3531,
+      "step": 3471
+    },
+    {
+      "epoch": 2.81894451962111,
+      "grad_norm": 0.1844189465045929,
+      "learning_rate": 1.0873845407253114e-07,
+      "loss": 0.3085,
+      "step": 3472
+    },
+    {
+      "epoch": 2.8197564276048714,
+      "grad_norm": 0.18140117824077606,
+      "learning_rate": 1.0776016415301405e-07,
+      "loss": 0.289,
+      "step": 3473
+    },
+    {
+      "epoch": 2.820568335588633,
+      "grad_norm": 0.19003745913505554,
+      "learning_rate": 1.067862469041725e-07,
+      "loss": 0.3501,
+      "step": 3474
+    },
+    {
+      "epoch": 2.821380243572395,
+      "grad_norm": 0.1866818368434906,
+      "learning_rate": 1.0581670319649295e-07,
+      "loss": 0.3203,
+      "step": 3475
+    },
+    {
+      "epoch": 2.822192151556157,
+      "grad_norm": 0.18259680271148682,
+      "learning_rate": 1.0485153389654945e-07,
+      "loss": 0.3441,
+      "step": 3476
+    },
+    {
+      "epoch": 2.823004059539919,
+      "grad_norm": 0.16885769367218018,
+      "learning_rate": 1.038907398670086e-07,
+      "loss": 0.3259,
+      "step": 3477
+    },
+    {
+      "epoch": 2.8238159675236805,
+      "grad_norm": 0.1799483746290207,
+      "learning_rate": 1.0293432196662512e-07,
+      "loss": 0.2923,
+      "step": 3478
+    },
+    {
+      "epoch": 2.8246278755074425,
+      "grad_norm": 0.19754785299301147,
+      "learning_rate": 1.01982281050243e-07,
+      "loss": 0.3875,
+      "step": 3479
+    },
+    {
+      "epoch": 2.825439783491204,
+      "grad_norm": 0.1778147965669632,
+      "learning_rate": 1.0103461796879265e-07,
+      "loss": 0.3225,
+      "step": 3480
+    },
+    {
+      "epoch": 2.8262516914749662,
+      "grad_norm": 0.18646013736724854,
+      "learning_rate": 1.000913335692949e-07,
+      "loss": 0.3361,
+      "step": 3481
+    },
+    {
+      "epoch": 2.827063599458728,
+      "grad_norm": 0.1718693971633911,
+      "learning_rate": 9.915242869485309e-08,
+      "loss": 0.3234,
+      "step": 3482
+    },
+    {
+      "epoch": 2.82787550744249,
+      "grad_norm": 0.1789943277835846,
+      "learning_rate": 9.821790418465815e-08,
+      "loss": 0.3217,
+      "step": 3483
+    },
+    {
+      "epoch": 2.8286874154262516,
+      "grad_norm": 0.1991688460111618,
+      "learning_rate": 9.728776087398583e-08,
+      "loss": 0.3353,
+      "step": 3484
+    },
+    {
+      "epoch": 2.8294993234100136,
+      "grad_norm": 0.17854006588459015,
+      "learning_rate": 9.636199959419557e-08,
+      "loss": 0.3274,
+      "step": 3485
+    },
+    {
+      "epoch": 2.8303112313937753,
+      "grad_norm": 0.17426542937755585,
+      "learning_rate": 9.544062117273045e-08,
+      "loss": 0.3144,
+      "step": 3486
+    },
+    {
+      "epoch": 2.8311231393775373,
+      "grad_norm": 0.19132675230503082,
+      "learning_rate": 9.452362643311619e-08,
+      "loss": 0.3419,
+      "step": 3487
+    },
+    {
+      "epoch": 2.831935047361299,
+      "grad_norm": 0.1644316166639328,
+      "learning_rate": 9.361101619495994e-08,
+      "loss": 0.3131,
+      "step": 3488
+    },
+    {
+      "epoch": 2.832746955345061,
+      "grad_norm": 0.1818498969078064,
+      "learning_rate": 9.270279127394977e-08,
+      "loss": 0.3311,
+      "step": 3489
+    },
+    {
+      "epoch": 2.8335588633288227,
+      "grad_norm": 0.20542341470718384,
+      "learning_rate": 9.179895248185522e-08,
+      "loss": 0.3111,
+      "step": 3490
+    },
+    {
+      "epoch": 2.8343707713125843,
+      "grad_norm": 0.18407122790813446,
+      "learning_rate": 9.089950062652508e-08,
+      "loss": 0.3726,
+      "step": 3491
+    },
+    {
+      "epoch": 2.8351826792963464,
+      "grad_norm": 0.1940666288137436,
+      "learning_rate": 9.000443651188628e-08,
+      "loss": 0.3041,
+      "step": 3492
+    },
+    {
+      "epoch": 2.8359945872801084,
+      "grad_norm": 0.19594602286815643,
+      "learning_rate": 8.911376093794555e-08,
+      "loss": 0.3312,
+      "step": 3493
+    },
+    {
+      "epoch": 2.83680649526387,
+      "grad_norm": 0.19601556658744812,
+      "learning_rate": 8.822747470078552e-08,
+      "loss": 0.3215,
+      "step": 3494
+    },
+    {
+      "epoch": 2.8376184032476317,
+      "grad_norm": 0.17945493757724762,
+      "learning_rate": 8.734557859256698e-08,
+      "loss": 0.334,
+      "step": 3495
+    },
+    {
+      "epoch": 2.8384303112313938,
+      "grad_norm": 0.19256268441677094,
+      "learning_rate": 8.646807340152607e-08,
+      "loss": 0.3491,
+      "step": 3496
+    },
+    {
+      "epoch": 2.839242219215156,
+      "grad_norm": 0.1757120043039322,
+      "learning_rate": 8.55949599119743e-08,
+      "loss": 0.3111,
+      "step": 3497
+    },
+    {
+      "epoch": 2.8400541271989175,
+      "grad_norm": 0.1711449921131134,
+      "learning_rate": 8.47262389042991e-08,
+      "loss": 0.3234,
+      "step": 3498
+    },
+    {
+      "epoch": 2.840866035182679,
+      "grad_norm": 0.1831902712583542,
+      "learning_rate": 8.386191115495934e-08,
+      "loss": 0.3346,
+      "step": 3499
+    },
+    {
+      "epoch": 2.841677943166441,
+      "grad_norm": 0.16271920502185822,
+      "learning_rate": 8.300197743649041e-08,
+      "loss": 0.2799,
+      "step": 3500
+    },
+    {
+      "epoch": 2.8424898511502033,
+      "grad_norm": 0.18514792621135712,
+      "learning_rate": 8.214643851749748e-08,
+      "loss": 0.3351,
+      "step": 3501
+    },
+    {
+      "epoch": 2.843301759133965,
+      "grad_norm": 0.19730618596076965,
+      "learning_rate": 8.129529516266e-08,
+      "loss": 0.3365,
+      "step": 3502
+    },
+    {
+      "epoch": 2.8441136671177265,
+      "grad_norm": 0.1617821753025055,
+      "learning_rate": 8.044854813272662e-08,
+      "loss": 0.2666,
+      "step": 3503
+    },
+    {
+      "epoch": 2.8449255751014886,
+      "grad_norm": 0.2031940519809723,
+      "learning_rate": 7.960619818451865e-08,
+      "loss": 0.3894,
+      "step": 3504
+    },
+    {
+      "epoch": 2.84573748308525,
+      "grad_norm": 0.18865470588207245,
+      "learning_rate": 7.876824607092437e-08,
+      "loss": 0.3089,
+      "step": 3505
+    },
+    {
+      "epoch": 2.8465493910690123,
+      "grad_norm": 0.18493427336215973,
+      "learning_rate": 7.793469254090524e-08,
+      "loss": 0.3214,
+      "step": 3506
+    },
+    {
+      "epoch": 2.847361299052774,
+      "grad_norm": 0.17331187427043915,
+      "learning_rate": 7.710553833948753e-08,
+      "loss": 0.3425,
+      "step": 3507
+    },
+    {
+      "epoch": 2.848173207036536,
+      "grad_norm": 0.1908278912305832,
+      "learning_rate": 7.628078420776786e-08,
+      "loss": 0.3589,
+      "step": 3508
+    },
+    {
+      "epoch": 2.8489851150202976,
+      "grad_norm": 0.19553807377815247,
+      "learning_rate": 7.546043088290878e-08,
+      "loss": 0.309,
+      "step": 3509
+    },
+    {
+      "epoch": 2.8497970230040597,
+      "grad_norm": 0.1721663922071457,
+      "learning_rate": 7.464447909814043e-08,
+      "loss": 0.3065,
+      "step": 3510
+    },
+    {
+      "epoch": 2.8506089309878213,
+      "grad_norm": 0.17896386981010437,
+      "learning_rate": 7.383292958275778e-08,
+      "loss": 0.3329,
+      "step": 3511
+    },
+    {
+      "epoch": 2.8514208389715834,
+      "grad_norm": 0.2096192091703415,
+      "learning_rate": 7.302578306212116e-08,
+      "loss": 0.363,
+      "step": 3512
+    },
+    {
+      "epoch": 2.852232746955345,
+      "grad_norm": 0.17473061382770538,
+      "learning_rate": 7.222304025765736e-08,
+      "loss": 0.2984,
+      "step": 3513
+    },
+    {
+      "epoch": 2.853044654939107,
+      "grad_norm": 0.18695694208145142,
+      "learning_rate": 7.142470188685413e-08,
+      "loss": 0.3294,
+      "step": 3514
+    },
+    {
+      "epoch": 2.8538565629228687,
+      "grad_norm": 0.16983966529369354,
+      "learning_rate": 7.063076866326568e-08,
+      "loss": 0.3205,
+      "step": 3515
+    },
+    {
+      "epoch": 2.8546684709066303,
+      "grad_norm": 0.19969230890274048,
+      "learning_rate": 6.984124129650605e-08,
+      "loss": 0.3501,
+      "step": 3516
+    },
+    {
+      "epoch": 2.8554803788903924,
+      "grad_norm": 0.1917487382888794,
+      "learning_rate": 6.905612049225352e-08,
+      "loss": 0.3232,
+      "step": 3517
+    },
+    {
+      "epoch": 2.8562922868741545,
+      "grad_norm": 0.17636620998382568,
+      "learning_rate": 6.827540695224622e-08,
+      "loss": 0.3292,
+      "step": 3518
+    },
+    {
+      "epoch": 2.857104194857916,
+      "grad_norm": 0.18421196937561035,
+      "learning_rate": 6.749910137428484e-08,
+      "loss": 0.315,
+      "step": 3519
+    },
+    {
+      "epoch": 2.8579161028416777,
+      "grad_norm": 0.18648824095726013,
+      "learning_rate": 6.672720445222769e-08,
+      "loss": 0.32,
+      "step": 3520
+    },
+    {
+      "epoch": 2.85872801082544,
+      "grad_norm": 0.186990424990654,
+      "learning_rate": 6.595971687599567e-08,
+      "loss": 0.3518,
+      "step": 3521
+    },
+    {
+      "epoch": 2.859539918809202,
+      "grad_norm": 0.18758608400821686,
+      "learning_rate": 6.519663933156562e-08,
+      "loss": 0.3505,
+      "step": 3522
+    },
+    {
+      "epoch": 2.8603518267929635,
+      "grad_norm": 0.18786361813545227,
+      "learning_rate": 6.443797250097528e-08,
+      "loss": 0.3346,
+      "step": 3523
+    },
+    {
+      "epoch": 2.861163734776725,
+      "grad_norm": 0.18394331634044647,
+      "learning_rate": 6.368371706231891e-08,
+      "loss": 0.3338,
+      "step": 3524
+    },
+    {
+      "epoch": 2.861975642760487,
+      "grad_norm": 0.1868656426668167,
+      "learning_rate": 6.29338736897478e-08,
+      "loss": 0.3156,
+      "step": 3525
+    },
+    {
+      "epoch": 2.862787550744249,
+      "grad_norm": 0.19027630984783173,
+      "learning_rate": 6.218844305346916e-08,
+      "loss": 0.3327,
+      "step": 3526
+    },
+    {
+      "epoch": 2.863599458728011,
+      "grad_norm": 0.17146709561347961,
+      "learning_rate": 6.144742581974839e-08,
+      "loss": 0.308,
+      "step": 3527
+    },
+    {
+      "epoch": 2.8644113667117725,
+      "grad_norm": 0.17006371915340424,
+      "learning_rate": 6.071082265090344e-08,
+      "loss": 0.3243,
+      "step": 3528
+    },
+    {
+      "epoch": 2.8652232746955346,
+      "grad_norm": 0.19989944994449615,
+      "learning_rate": 5.99786342053088e-08,
+      "loss": 0.3275,
+      "step": 3529
+    },
+    {
+      "epoch": 2.8660351826792962,
+      "grad_norm": 0.1919189691543579,
+      "learning_rate": 5.925086113739209e-08,
+      "loss": 0.3556,
+      "step": 3530
+    },
+    {
+      "epoch": 2.8668470906630583,
+      "grad_norm": 0.19643646478652954,
+      "learning_rate": 5.852750409763575e-08,
+      "loss": 0.2735,
+      "step": 3531
+    },
+    {
+      "epoch": 2.86765899864682,
+      "grad_norm": 0.18624305725097656,
+      "learning_rate": 5.7808563732573155e-08,
+      "loss": 0.3637,
+      "step": 3532
+    },
+    {
+      "epoch": 2.868470906630582,
+      "grad_norm": 0.16751444339752197,
+      "learning_rate": 5.709404068479196e-08,
+      "loss": 0.3061,
+      "step": 3533
+    },
+    {
+      "epoch": 2.8692828146143436,
+      "grad_norm": 0.176624596118927,
+      "learning_rate": 5.638393559293076e-08,
+      "loss": 0.3139,
+      "step": 3534
+    },
+    {
+      "epoch": 2.8700947225981057,
+      "grad_norm": 0.1830405741930008,
+      "learning_rate": 5.5678249091679624e-08,
+      "loss": 0.3199,
+      "step": 3535
+    },
+    {
+      "epoch": 2.8709066305818673,
+      "grad_norm": 0.18323978781700134,
+      "learning_rate": 5.4976981811779575e-08,
+      "loss": 0.3335,
+      "step": 3536
+    },
+    {
+      "epoch": 2.871718538565629,
+      "grad_norm": 0.16508954763412476,
+      "learning_rate": 5.4280134380020336e-08,
+      "loss": 0.3139,
+      "step": 3537
+    },
+    {
+      "epoch": 2.872530446549391,
+      "grad_norm": 0.1806226372718811,
+      "learning_rate": 5.358770741924368e-08,
+      "loss": 0.3325,
+      "step": 3538
+    },
+    {
+      "epoch": 2.873342354533153,
+      "grad_norm": 0.17620451748371124,
+      "learning_rate": 5.28997015483379e-08,
+      "loss": 0.3311,
+      "step": 3539
+    },
+    {
+      "epoch": 2.8741542625169147,
+      "grad_norm": 0.1819583624601364,
+      "learning_rate": 5.2216117382241084e-08,
+      "loss": 0.3708,
+      "step": 3540
+    },
+    {
+      "epoch": 2.8749661705006764,
+      "grad_norm": 0.17088262736797333,
+      "learning_rate": 5.1536955531938937e-08,
+      "loss": 0.2857,
+      "step": 3541
+    },
+    {
+      "epoch": 2.8757780784844384,
+      "grad_norm": 0.16675534844398499,
+      "learning_rate": 5.086221660446422e-08,
+      "loss": 0.3079,
+      "step": 3542
+    },
+    {
+      "epoch": 2.8765899864682005,
+      "grad_norm": 0.17822113633155823,
+      "learning_rate": 5.0191901202897295e-08,
+      "loss": 0.3143,
+      "step": 3543
+    },
+    {
+      "epoch": 2.877401894451962,
+      "grad_norm": 0.17477335035800934,
+      "learning_rate": 4.952600992636392e-08,
+      "loss": 0.3055,
+      "step": 3544
+    },
+    {
+      "epoch": 2.878213802435724,
+      "grad_norm": 0.1897391974925995,
+      "learning_rate": 4.8864543370035764e-08,
+      "loss": 0.344,
+      "step": 3545
+    },
+    {
+      "epoch": 2.879025710419486,
+      "grad_norm": 0.1744416058063507,
+      "learning_rate": 4.820750212513048e-08,
+      "loss": 0.2882,
+      "step": 3546
+    },
+    {
+      "epoch": 2.8798376184032475,
+      "grad_norm": 0.19691471755504608,
+      "learning_rate": 4.755488677890829e-08,
+      "loss": 0.3186,
+      "step": 3547
+    },
+    {
+      "epoch": 2.8806495263870096,
+      "grad_norm": 0.17443303763866425,
+      "learning_rate": 4.6906697914677055e-08,
+      "loss": 0.2823,
+      "step": 3548
+    },
+    {
+      "epoch": 2.881461434370771,
+      "grad_norm": 0.18256105482578278,
+      "learning_rate": 4.626293611178445e-08,
+      "loss": 0.3435,
+      "step": 3549
+    },
+    {
+      "epoch": 2.8822733423545333,
+      "grad_norm": 0.17572712898254395,
+      "learning_rate": 4.5623601945624096e-08,
+      "loss": 0.3238,
+      "step": 3550
+    },
+    {
+      "epoch": 2.883085250338295,
+      "grad_norm": 0.17860786616802216,
+      "learning_rate": 4.498869598763056e-08,
+      "loss": 0.3254,
+      "step": 3551
+    },
+    {
+      "epoch": 2.883897158322057,
+      "grad_norm": 0.19168518483638763,
+      "learning_rate": 4.435821880528157e-08,
+      "loss": 0.3648,
+      "step": 3552
+    },
+    {
+      "epoch": 2.8847090663058186,
+      "grad_norm": 0.19023928046226501,
+      "learning_rate": 4.373217096209526e-08,
+      "loss": 0.3221,
+      "step": 3553
+    },
+    {
+      "epoch": 2.8855209742895807,
+      "grad_norm": 0.18133315443992615,
+      "learning_rate": 4.3110553017631786e-08,
+      "loss": 0.2765,
+      "step": 3554
+    },
+    {
+      "epoch": 2.8863328822733423,
+      "grad_norm": 0.2157335877418518,
+      "learning_rate": 4.249336552749172e-08,
+      "loss": 0.3895,
+      "step": 3555
+    },
+    {
+      "epoch": 2.8871447902571044,
+      "grad_norm": 0.19117408990859985,
+      "learning_rate": 4.188060904331548e-08,
+      "loss": 0.3273,
+      "step": 3556
+    },
+    {
+      "epoch": 2.887956698240866,
+      "grad_norm": 0.17874912917613983,
+      "learning_rate": 4.127228411278383e-08,
+      "loss": 0.3359,
+      "step": 3557
+    },
+    {
+      "epoch": 2.8887686062246276,
+      "grad_norm": 0.1810157299041748,
+      "learning_rate": 4.0668391279614085e-08,
+      "loss": 0.3292,
+      "step": 3558
+    },
+    {
+      "epoch": 2.8895805142083897,
+      "grad_norm": 0.18089009821414948,
+      "learning_rate": 4.0068931083566134e-08,
+      "loss": 0.328,
+      "step": 3559
+    },
+    {
+      "epoch": 2.8903924221921518,
+      "grad_norm": 0.1822344958782196,
+      "learning_rate": 3.947390406043472e-08,
+      "loss": 0.3154,
+      "step": 3560
+    },
+    {
+      "epoch": 2.8912043301759134,
+      "grad_norm": 0.17798471450805664,
+      "learning_rate": 3.8883310742053316e-08,
+      "loss": 0.3244,
+      "step": 3561
+    },
+    {
+      "epoch": 2.892016238159675,
+      "grad_norm": 0.1792553961277008,
+      "learning_rate": 3.8297151656293e-08,
+      "loss": 0.3149,
+      "step": 3562
+    },
+    {
+      "epoch": 2.892828146143437,
+      "grad_norm": 0.16871392726898193,
+      "learning_rate": 3.771542732706135e-08,
+      "loss": 0.292,
+      "step": 3563
+    },
+    {
+      "epoch": 2.893640054127199,
+      "grad_norm": 0.180922269821167,
+      "learning_rate": 3.713813827430135e-08,
+      "loss": 0.3773,
+      "step": 3564
+    },
+    {
+      "epoch": 2.894451962110961,
+      "grad_norm": 0.1872747540473938,
+      "learning_rate": 3.656528501399359e-08,
+      "loss": 0.3075,
+      "step": 3565
+    },
+    {
+      "epoch": 2.8952638700947224,
+      "grad_norm": 0.1845092922449112,
+      "learning_rate": 3.599686805815128e-08,
+      "loss": 0.3281,
+      "step": 3566
+    },
+    {
+      "epoch": 2.8960757780784845,
+      "grad_norm": 0.1991731822490692,
+      "learning_rate": 3.543288791482469e-08,
+      "loss": 0.3619,
+      "step": 3567
+    },
+    {
+      "epoch": 2.896887686062246,
+      "grad_norm": 0.1940995156764984,
+      "learning_rate": 3.4873345088097255e-08,
+      "loss": 0.2843,
+      "step": 3568
+    },
+    {
+      "epoch": 2.897699594046008,
+      "grad_norm": 0.1890859752893448,
+      "learning_rate": 3.4318240078087816e-08,
+      "loss": 0.3288,
+      "step": 3569
+    },
+    {
+      "epoch": 2.89851150202977,
+      "grad_norm": 0.17690710723400116,
+      "learning_rate": 3.3767573380946717e-08,
+      "loss": 0.3525,
+      "step": 3570
+    },
+    {
+      "epoch": 2.899323410013532,
+      "grad_norm": 0.18428310751914978,
+      "learning_rate": 3.3221345488859134e-08,
+      "loss": 0.333,
+      "step": 3571
+    },
+    {
+      "epoch": 2.9001353179972935,
+      "grad_norm": 0.19804523885250092,
+      "learning_rate": 3.267955689004176e-08,
+      "loss": 0.3274,
+      "step": 3572
+    },
+    {
+      "epoch": 2.9009472259810556,
+      "grad_norm": 0.17699231207370758,
+      "learning_rate": 3.214220806874335e-08,
+      "loss": 0.3535,
+      "step": 3573
+    },
+    {
+      "epoch": 2.901759133964817,
+      "grad_norm": 0.18369637429714203,
+      "learning_rate": 3.1609299505245275e-08,
+      "loss": 0.3435,
+      "step": 3574
+    },
+    {
+      "epoch": 2.9025710419485793,
+      "grad_norm": 0.19758519530296326,
+      "learning_rate": 3.1080831675859845e-08,
+      "loss": 0.3663,
+      "step": 3575
+    },
+    {
+      "epoch": 2.903382949932341,
+      "grad_norm": 0.19301550090312958,
+      "learning_rate": 3.055680505292979e-08,
+      "loss": 0.3226,
+      "step": 3576
+    },
+    {
+      "epoch": 2.904194857916103,
+      "grad_norm": 0.18160277605056763,
+      "learning_rate": 3.0037220104829344e-08,
+      "loss": 0.3043,
+      "step": 3577
+    },
+    {
+      "epoch": 2.9050067658998646,
+      "grad_norm": 0.18017329275608063,
+      "learning_rate": 2.952207729596146e-08,
+      "loss": 0.3327,
+      "step": 3578
+    },
+    {
+      "epoch": 2.9058186738836262,
+      "grad_norm": 0.1879228800535202,
+      "learning_rate": 2.9011377086759496e-08,
+      "loss": 0.3332,
+      "step": 3579
+    },
+    {
+      "epoch": 2.9066305818673883,
+      "grad_norm": 0.17808818817138672,
+      "learning_rate": 2.85051199336861e-08,
+      "loss": 0.3067,
+      "step": 3580
+    },
+    {
+      "epoch": 2.9074424898511504,
+      "grad_norm": 0.18098266422748566,
+      "learning_rate": 2.800330628923209e-08,
+      "loss": 0.3349,
+      "step": 3581
+    },
+    {
+      "epoch": 2.908254397834912,
+      "grad_norm": 0.16781920194625854,
+      "learning_rate": 2.7505936601918137e-08,
+      "loss": 0.2974,
+      "step": 3582
+    },
+    {
+      "epoch": 2.9090663058186736,
+      "grad_norm": 0.18088236451148987,
+      "learning_rate": 2.701301131629086e-08,
+      "loss": 0.3485,
+      "step": 3583
+    },
+    {
+      "epoch": 2.9098782138024357,
+      "grad_norm": 0.19744667410850525,
+      "learning_rate": 2.652453087292617e-08,
+      "loss": 0.3291,
+      "step": 3584
+    },
+    {
+      "epoch": 2.910690121786198,
+      "grad_norm": 0.19625650346279144,
+      "learning_rate": 2.6040495708425928e-08,
+      "loss": 0.3439,
+      "step": 3585
+    },
+    {
+      "epoch": 2.9115020297699594,
+      "grad_norm": 0.4273628890514374,
+      "learning_rate": 2.5560906255420737e-08,
+      "loss": 0.3362,
+      "step": 3586
+    },
+    {
+      "epoch": 2.912313937753721,
+      "grad_norm": 0.18096792697906494,
+      "learning_rate": 2.5085762942564927e-08,
+      "loss": 0.3204,
+      "step": 3587
+    },
+    {
+      "epoch": 2.913125845737483,
+      "grad_norm": 0.17710824310779572,
+      "learning_rate": 2.4615066194541015e-08,
+      "loss": 0.3464,
+      "step": 3588
+    },
+    {
+      "epoch": 2.913937753721245,
+      "grad_norm": 0.17082080245018005,
+      "learning_rate": 2.41488164320558e-08,
+      "loss": 0.3144,
+      "step": 3589
+    },
+    {
+      "epoch": 2.914749661705007,
+      "grad_norm": 0.20494933426380157,
+      "learning_rate": 2.368701407184315e-08,
+      "loss": 0.369,
+      "step": 3590
+    },
+    {
+      "epoch": 2.9155615696887685,
+      "grad_norm": 0.17302276194095612,
+      "learning_rate": 2.322965952666012e-08,
+      "loss": 0.3241,
+      "step": 3591
+    },
+    {
+      "epoch": 2.9163734776725305,
+      "grad_norm": 0.17284782230854034,
+      "learning_rate": 2.277675320528916e-08,
+      "loss": 0.307,
+      "step": 3592
+    },
+    {
+      "epoch": 2.917185385656292,
+      "grad_norm": 0.1808546930551529,
+      "learning_rate": 2.2328295512535902e-08,
+      "loss": 0.3375,
+      "step": 3593
+    },
+    {
+      "epoch": 2.9179972936400542,
+      "grad_norm": 0.18827113509178162,
+      "learning_rate": 2.188428684923194e-08,
+      "loss": 0.3443,
+      "step": 3594
+    },
+    {
+      "epoch": 2.918809201623816,
+      "grad_norm": 0.17342112958431244,
+      "learning_rate": 2.144472761222982e-08,
+      "loss": 0.3097,
+      "step": 3595
+    },
+    {
+      "epoch": 2.919621109607578,
+      "grad_norm": 0.1894642859697342,
+      "learning_rate": 2.10096181944075e-08,
+      "loss": 0.3275,
+      "step": 3596
+    },
+    {
+      "epoch": 2.9204330175913396,
+      "grad_norm": 0.1770143210887909,
+      "learning_rate": 2.0578958984663888e-08,
+      "loss": 0.3175,
+      "step": 3597
+    },
+    {
+      "epoch": 2.9212449255751016,
+      "grad_norm": 0.1824568212032318,
+      "learning_rate": 2.015275036792108e-08,
+      "loss": 0.3501,
+      "step": 3598
+    },
+    {
+      "epoch": 2.9220568335588633,
+      "grad_norm": 0.17532792687416077,
+      "learning_rate": 1.973099272512269e-08,
+      "loss": 0.2896,
+      "step": 3599
+    },
+    {
+      "epoch": 2.9228687415426253,
+      "grad_norm": 0.18696366250514984,
+      "learning_rate": 1.9313686433236057e-08,
+      "loss": 0.343,
+      "step": 3600
+    },
+    {
+      "epoch": 2.923680649526387,
+      "grad_norm": 0.17934131622314453,
+      "learning_rate": 1.8900831865246717e-08,
+      "loss": 0.3207,
+      "step": 3601
+    },
+    {
+      "epoch": 2.924492557510149,
+      "grad_norm": 0.19648589193820953,
+      "learning_rate": 1.849242939016449e-08,
+      "loss": 0.3624,
+      "step": 3602
+    },
+    {
+      "epoch": 2.9253044654939107,
+      "grad_norm": 0.17450867593288422,
+      "learning_rate": 1.8088479373016833e-08,
+      "loss": 0.2694,
+      "step": 3603
+    },
+    {
+      "epoch": 2.9261163734776723,
+      "grad_norm": 0.2019394487142563,
+      "learning_rate": 1.768898217485493e-08,
+      "loss": 0.3714,
+      "step": 3604
+    },
+    {
+      "epoch": 2.9269282814614344,
+      "grad_norm": 0.17694585025310516,
+      "learning_rate": 1.7293938152746493e-08,
+      "loss": 0.2974,
+      "step": 3605
+    },
+    {
+      "epoch": 2.9277401894451964,
+      "grad_norm": 0.1911110132932663,
+      "learning_rate": 1.6903347659781856e-08,
+      "loss": 0.3202,
+      "step": 3606
+    },
+    {
+      "epoch": 2.928552097428958,
+      "grad_norm": 0.19160738587379456,
+      "learning_rate": 1.6517211045069537e-08,
+      "loss": 0.3157,
+      "step": 3607
+    },
+    {
+      "epoch": 2.9293640054127197,
+      "grad_norm": 0.20632876455783844,
+      "learning_rate": 1.6135528653737353e-08,
+      "loss": 0.3547,
+      "step": 3608
+    },
+    {
+      "epoch": 2.9301759133964818,
+      "grad_norm": 0.18852488696575165,
+      "learning_rate": 1.575830082693186e-08,
+      "loss": 0.3078,
+      "step": 3609
+    },
+    {
+      "epoch": 2.930987821380244,
+      "grad_norm": 0.17725138366222382,
+      "learning_rate": 1.5385527901818907e-08,
+      "loss": 0.3209,
+      "step": 3610
+    },
+    {
+      "epoch": 2.9317997293640055,
+      "grad_norm": 0.17734643816947937,
+      "learning_rate": 1.5017210211580868e-08,
+      "loss": 0.3586,
+      "step": 3611
+    },
+    {
+      "epoch": 2.932611637347767,
+      "grad_norm": 0.1784542351961136,
+      "learning_rate": 1.4653348085419961e-08,
+      "loss": 0.3206,
+      "step": 3612
+    },
+    {
+      "epoch": 2.933423545331529,
+      "grad_norm": 0.18330296874046326,
+      "learning_rate": 1.4293941848554926e-08,
+      "loss": 0.3211,
+      "step": 3613
+    },
+    {
+      "epoch": 2.934235453315291,
+      "grad_norm": 0.17470023036003113,
+      "learning_rate": 1.3938991822222137e-08,
+      "loss": 0.3316,
+      "step": 3614
+    },
+    {
+      "epoch": 2.935047361299053,
+      "grad_norm": 0.1952422708272934,
+      "learning_rate": 1.3588498323674482e-08,
+      "loss": 0.3658,
+      "step": 3615
+    },
+    {
+      "epoch": 2.9358592692828145,
+      "grad_norm": 0.17796678841114044,
+      "learning_rate": 1.3242461666182483e-08,
+      "loss": 0.3163,
+      "step": 3616
+    },
+    {
+      "epoch": 2.9366711772665766,
+      "grad_norm": 0.1717871129512787,
+      "learning_rate": 1.2900882159033178e-08,
+      "loss": 0.3113,
+      "step": 3617
+    },
+    {
+      "epoch": 2.937483085250338,
+      "grad_norm": 0.18539494276046753,
+      "learning_rate": 1.2563760107528466e-08,
+      "loss": 0.3672,
+      "step": 3618
+    },
+    {
+      "epoch": 2.9382949932341003,
+      "grad_norm": 0.18299327790737152,
+      "learning_rate": 1.223109581298787e-08,
+      "loss": 0.3237,
+      "step": 3619
+    },
+    {
+      "epoch": 2.939106901217862,
+      "grad_norm": 0.17467635869979858,
+      "learning_rate": 1.1902889572745213e-08,
+      "loss": 0.2968,
+      "step": 3620
+    },
+    {
+      "epoch": 2.939918809201624,
+      "grad_norm": 0.18281607329845428,
+      "learning_rate": 1.1579141680150841e-08,
+      "loss": 0.342,
+      "step": 3621
+    },
+    {
+      "epoch": 2.9407307171853856,
+      "grad_norm": 0.19280225038528442,
+      "learning_rate": 1.125985242456995e-08,
+      "loss": 0.3478,
+      "step": 3622
+    },
+    {
+      "epoch": 2.9415426251691477,
+      "grad_norm": 0.17374856770038605,
+      "learning_rate": 1.094502209138204e-08,
+      "loss": 0.3012,
+      "step": 3623
+    },
+    {
+      "epoch": 2.9423545331529093,
+      "grad_norm": 0.17550191283226013,
+      "learning_rate": 1.0634650961981462e-08,
+      "loss": 0.3761,
+      "step": 3624
+    },
+    {
+      "epoch": 2.943166441136671,
+      "grad_norm": 0.16769611835479736,
+      "learning_rate": 1.0328739313776869e-08,
+      "loss": 0.295,
+      "step": 3625
+    },
+    {
+      "epoch": 2.943978349120433,
+      "grad_norm": 0.18039877712726593,
+      "learning_rate": 1.0027287420192322e-08,
+      "loss": 0.3126,
+      "step": 3626
+    },
+    {
+      "epoch": 2.944790257104195,
+      "grad_norm": 0.19193851947784424,
+      "learning_rate": 9.730295550663959e-09,
+      "loss": 0.3492,
+      "step": 3627
+    },
+    {
+      "epoch": 2.9456021650879567,
+      "grad_norm": 0.1801174134016037,
+      "learning_rate": 9.437763970642223e-09,
+      "loss": 0.3039,
+      "step": 3628
+    },
+    {
+      "epoch": 2.9464140730717183,
+      "grad_norm": 0.19072727859020233,
+      "learning_rate": 9.149692941590738e-09,
+      "loss": 0.3148,
+      "step": 3629
+    },
+    {
+      "epoch": 2.9472259810554804,
+      "grad_norm": 0.19023066759109497,
+      "learning_rate": 8.866082720987989e-09,
+      "loss": 0.3404,
+      "step": 3630
+    },
+    {
+      "epoch": 2.9480378890392425,
+      "grad_norm": 0.19011977314949036,
+      "learning_rate": 8.58693356232232e-09,
+      "loss": 0.3554,
+      "step": 3631
+    },
+    {
+      "epoch": 2.948849797023004,
+      "grad_norm": 0.18086890876293182,
+      "learning_rate": 8.312245715097478e-09,
+      "loss": 0.3308,
+      "step": 3632
+    },
+    {
+      "epoch": 2.9496617050067657,
+      "grad_norm": 0.18266567587852478,
+      "learning_rate": 8.042019424828739e-09,
+      "loss": 0.3268,
+      "step": 3633
+    },
+    {
+      "epoch": 2.950473612990528,
+      "grad_norm": 0.18250097334384918,
+      "learning_rate": 7.776254933043458e-09,
+      "loss": 0.3426,
+      "step": 3634
+    },
+    {
+      "epoch": 2.9512855209742894,
+      "grad_norm": 0.18050798773765564,
+      "learning_rate": 7.514952477281067e-09,
+      "loss": 0.3246,
+      "step": 3635
+    },
+    {
+      "epoch": 2.9520974289580515,
+      "grad_norm": 0.17372667789459229,
+      "learning_rate": 7.258112291093078e-09,
+      "loss": 0.3144,
+      "step": 3636
+    },
+    {
+      "epoch": 2.952909336941813,
+      "grad_norm": 0.19550307095050812,
+      "learning_rate": 7.005734604041969e-09,
+      "loss": 0.3465,
+      "step": 3637
+    },
+    {
+      "epoch": 2.953721244925575,
+      "grad_norm": 0.20049066841602325,
+      "learning_rate": 6.757819641703966e-09,
+      "loss": 0.3268,
+      "step": 3638
+    },
+    {
+      "epoch": 2.954533152909337,
+      "grad_norm": 0.1733388453722,
+      "learning_rate": 6.514367625662932e-09,
+      "loss": 0.3394,
+      "step": 3639
+    },
+    {
+      "epoch": 2.955345060893099,
+      "grad_norm": 0.18839865922927856,
+      "learning_rate": 6.275378773516472e-09,
+      "loss": 0.3207,
+      "step": 3640
+    },
+    {
+      "epoch": 2.9561569688768605,
+      "grad_norm": 0.17455987632274628,
+      "learning_rate": 6.040853298872051e-09,
+      "loss": 0.3144,
+      "step": 3641
+    },
+    {
+      "epoch": 2.9569688768606226,
+      "grad_norm": 0.18060794472694397,
+      "learning_rate": 5.8107914113475494e-09,
+      "loss": 0.3123,
+      "step": 3642
+    },
+    {
+      "epoch": 2.9577807848443842,
+      "grad_norm": 0.19623500108718872,
+      "learning_rate": 5.585193316572368e-09,
+      "loss": 0.3462,
+      "step": 3643
+    },
+    {
+      "epoch": 2.9585926928281463,
+      "grad_norm": 0.18604233860969543,
+      "learning_rate": 5.364059216185213e-09,
+      "loss": 0.3359,
+      "step": 3644
+    },
+    {
+      "epoch": 2.959404600811908,
+      "grad_norm": 0.1711633801460266,
+      "learning_rate": 5.147389307834649e-09,
+      "loss": 0.3159,
+      "step": 3645
+    },
+    {
+      "epoch": 2.9602165087956696,
+      "grad_norm": 0.18190641701221466,
+      "learning_rate": 4.935183785180209e-09,
+      "loss": 0.3477,
+      "step": 3646
+    },
+    {
+      "epoch": 2.9610284167794316,
+      "grad_norm": 0.18316449224948883,
+      "learning_rate": 4.727442837890173e-09,
+      "loss": 0.3613,
+      "step": 3647
+    },
+    {
+      "epoch": 2.9618403247631937,
+      "grad_norm": 0.17387297749519348,
+      "learning_rate": 4.5241666516432355e-09,
+      "loss": 0.2957,
+      "step": 3648
+    },
+    {
+      "epoch": 2.9626522327469553,
+      "grad_norm": 0.17778031527996063,
+      "learning_rate": 4.325355408127396e-09,
+      "loss": 0.327,
+      "step": 3649
+    },
+    {
+      "epoch": 2.963464140730717,
+      "grad_norm": 0.18560712039470673,
+      "learning_rate": 4.131009285038845e-09,
+      "loss": 0.3315,
+      "step": 3650
+    },
+    {
+      "epoch": 2.964276048714479,
+      "grad_norm": 0.1886829137802124,
+      "learning_rate": 3.941128456083631e-09,
+      "loss": 0.2867,
+      "step": 3651
+    },
+    {
+      "epoch": 2.965087956698241,
+      "grad_norm": 0.1909620463848114,
+      "learning_rate": 3.755713090977664e-09,
+      "loss": 0.3275,
+      "step": 3652
+    },
+    {
+      "epoch": 2.9658998646820027,
+      "grad_norm": 0.19645372033119202,
+      "learning_rate": 3.5747633554433782e-09,
+      "loss": 0.3515,
+      "step": 3653
+    },
+    {
+      "epoch": 2.9667117726657644,
+      "grad_norm": 0.2362673133611679,
+      "learning_rate": 3.3982794112141783e-09,
+      "loss": 0.3022,
+      "step": 3654
+    },
+    {
+      "epoch": 2.9675236806495264,
+      "grad_norm": 0.17852936685085297,
+      "learning_rate": 3.226261416030552e-09,
+      "loss": 0.3266,
+      "step": 3655
+    },
+    {
+      "epoch": 2.9683355886332885,
+      "grad_norm": 0.18304862082004547,
+      "learning_rate": 3.0587095236417342e-09,
+      "loss": 0.3394,
+      "step": 3656
+    },
+    {
+      "epoch": 2.96914749661705,
+      "grad_norm": 0.180111363530159,
+      "learning_rate": 2.8956238838051544e-09,
+      "loss": 0.3396,
+      "step": 3657
+    },
+    {
+      "epoch": 2.9699594046008118,
+      "grad_norm": 0.17617453634738922,
+      "learning_rate": 2.737004642286434e-09,
+      "loss": 0.3428,
+      "step": 3658
+    },
+    {
+      "epoch": 2.970771312584574,
+      "grad_norm": 0.1832631528377533,
+      "learning_rate": 2.5828519408593877e-09,
+      "loss": 0.3416,
+      "step": 3659
+    },
+    {
+      "epoch": 2.9715832205683355,
+      "grad_norm": 0.18564113974571228,
+      "learning_rate": 2.4331659173038038e-09,
+      "loss": 0.3307,
+      "step": 3660
+    },
+    {
+      "epoch": 2.9723951285520975,
+      "grad_norm": 0.16720937192440033,
+      "learning_rate": 2.2879467054104378e-09,
+      "loss": 0.3025,
+      "step": 3661
+    },
+    {
+      "epoch": 2.973207036535859,
+      "grad_norm": 0.1826496571302414,
+      "learning_rate": 2.1471944349749086e-09,
+      "loss": 0.3583,
+      "step": 3662
+    },
+    {
+      "epoch": 2.9740189445196212,
+      "grad_norm": 0.1708817481994629,
+      "learning_rate": 2.0109092318010283e-09,
+      "loss": 0.3188,
+      "step": 3663
+    },
+    {
+      "epoch": 2.974830852503383,
+      "grad_norm": 0.17912131547927856,
+      "learning_rate": 1.8790912177008015e-09,
+      "loss": 0.3237,
+      "step": 3664
+    },
+    {
+      "epoch": 2.975642760487145,
+      "grad_norm": 0.1775650680065155,
+      "learning_rate": 1.7517405104922057e-09,
+      "loss": 0.3346,
+      "step": 3665
+    },
+    {
+      "epoch": 2.9764546684709066,
+      "grad_norm": 0.17787154018878937,
+      "learning_rate": 1.6288572240014123e-09,
+      "loss": 0.3431,
+      "step": 3666
+    },
+    {
+      "epoch": 2.9772665764546686,
+      "grad_norm": 0.17713086307048798,
+      "learning_rate": 1.510441468061119e-09,
+      "loss": 0.3014,
+      "step": 3667
+    },
+    {
+      "epoch": 2.9780784844384303,
+      "grad_norm": 0.17799222469329834,
+      "learning_rate": 1.3964933485105526e-09,
+      "loss": 0.3263,
+      "step": 3668
+    },
+    {
+      "epoch": 2.9788903924221923,
+      "grad_norm": 0.1770942062139511,
+      "learning_rate": 1.287012967196577e-09,
+      "loss": 0.3189,
+      "step": 3669
+    },
+    {
+      "epoch": 2.979702300405954,
+      "grad_norm": 0.17684818804264069,
+      "learning_rate": 1.1820004219725845e-09,
+      "loss": 0.3242,
+      "step": 3670
+    },
+    {
+      "epoch": 2.9805142083897156,
+      "grad_norm": 0.1918574869632721,
+      "learning_rate": 1.0814558066979397e-09,
+      "loss": 0.3293,
+      "step": 3671
+    },
+    {
+      "epoch": 2.9813261163734777,
+      "grad_norm": 0.18301308155059814,
+      "learning_rate": 9.853792112402005e-10,
+      "loss": 0.3255,
+      "step": 3672
+    },
+    {
+      "epoch": 2.9821380243572397,
+      "grad_norm": 0.1659892052412033,
+      "learning_rate": 8.937707214712321e-10,
+      "loss": 0.311,
+      "step": 3673
+    },
+    {
+      "epoch": 2.9829499323410014,
+      "grad_norm": 0.1725640445947647,
+      "learning_rate": 8.066304192710927e-10,
+      "loss": 0.3263,
+      "step": 3674
+    },
+    {
+      "epoch": 2.983761840324763,
+      "grad_norm": 0.17038391530513763,
+      "learning_rate": 7.239583825252583e-10,
+      "loss": 0.3102,
+      "step": 3675
+    },
+    {
+      "epoch": 2.984573748308525,
+      "grad_norm": 0.1800076961517334,
+      "learning_rate": 6.457546851262875e-10,
+      "loss": 0.3271,
+      "step": 3676
+    },
+    {
+      "epoch": 2.985385656292287,
+      "grad_norm": 0.18058328330516815,
+      "learning_rate": 5.720193969716015e-10,
+      "loss": 0.3181,
+      "step": 3677
+    },
+    {
+      "epoch": 2.9861975642760488,
+      "grad_norm": 0.17571358382701874,
+      "learning_rate": 5.027525839662594e-10,
+      "loss": 0.3156,
+      "step": 3678
+    },
+    {
+      "epoch": 2.9870094722598104,
+      "grad_norm": 0.1721496880054474,
+      "learning_rate": 4.3795430802073825e-10,
+      "loss": 0.3202,
+      "step": 3679
+    },
+    {
+      "epoch": 2.9878213802435725,
+      "grad_norm": 0.18645229935646057,
+      "learning_rate": 3.7762462705093204e-10,
+      "loss": 0.3368,
+      "step": 3680
+    },
+    {
+      "epoch": 2.988633288227334,
+      "grad_norm": 0.1774570196866989,
+      "learning_rate": 3.2176359498037325e-10,
+      "loss": 0.3202,
+      "step": 3681
+    },
+    {
+      "epoch": 2.989445196211096,
+      "grad_norm": 0.18650133907794952,
+      "learning_rate": 2.703712617363463e-10,
+      "loss": 0.3246,
+      "step": 3682
+    },
+    {
+      "epoch": 2.990257104194858,
+      "grad_norm": 0.20218618214130402,
+      "learning_rate": 2.234476732537738e-10,
+      "loss": 0.3032,
+      "step": 3683
+    },
+    {
+      "epoch": 2.99106901217862,
+      "grad_norm": 0.2038549780845642,
+      "learning_rate": 1.8099287147355095e-10,
+      "loss": 0.3631,
+      "step": 3684
+    },
+    {
+      "epoch": 2.9918809201623815,
+      "grad_norm": 0.1977202147245407,
+      "learning_rate": 1.4300689434032512e-10,
+      "loss": 0.3217,
+      "step": 3685
+    },
+    {
+      "epoch": 2.9926928281461436,
+      "grad_norm": 0.1887689232826233,
+      "learning_rate": 1.0948977580638176e-10,
+      "loss": 0.3043,
+      "step": 3686
+    },
+    {
+      "epoch": 2.993504736129905,
+      "grad_norm": 0.18132632970809937,
+      "learning_rate": 8.044154582942388e-11,
+      "loss": 0.3342,
+      "step": 3687
+    },
+    {
+      "epoch": 2.9943166441136673,
+      "grad_norm": 0.19394326210021973,
+      "learning_rate": 5.586223037257199e-11,
+      "loss": 0.3499,
+      "step": 3688
+    },
+    {
+      "epoch": 2.995128552097429,
+      "grad_norm": 0.18086886405944824,
+      "learning_rate": 3.575185140436421e-11,
+      "loss": 0.3123,
+      "step": 3689
+    },
+    {
+      "epoch": 2.995940460081191,
+      "grad_norm": 0.20476791262626648,
+      "learning_rate": 2.0110426900421532e-11,
+      "loss": 0.3498,
+      "step": 3690
+    },
+    {
+      "epoch": 2.9967523680649526,
+      "grad_norm": 0.18325304985046387,
+      "learning_rate": 8.937970840117161e-12,
+      "loss": 0.3346,
+      "step": 3691
+    },
+    {
+      "epoch": 2.9975642760487142,
+      "grad_norm": 0.18689973652362823,
+      "learning_rate": 2.2344932093520955e-12,
+      "loss": 0.3275,
+      "step": 3692
+    },
+    {
+      "epoch": 2.9983761840324763,
+      "grad_norm": 0.1948087066411972,
+      "learning_rate": 0.0,
+      "loss": 0.3689,
+      "step": 3693
+    },
+    {
+      "epoch": 2.9983761840324763,
+      "step": 3693,
+      "total_flos": 1.5512490720886784e+16,
+      "train_loss": 0.3700987050083314,
+      "train_runtime": 454895.6696,
+      "train_samples_per_second": 0.78,
+      "train_steps_per_second": 0.008
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3693,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.5512490720886784e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}