diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,23748 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9993349589891376,
+  "eval_steps": 564,
+  "global_step": 3381,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008867213478164486,
+      "grad_norm": 105.35820770263672,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 3.2849,
+      "step": 1
+    },
+    {
+      "epoch": 0.0008867213478164486,
+      "eval_loss": 3.391964912414551,
+      "eval_runtime": 71.6314,
+      "eval_samples_per_second": 2.68,
+      "eval_steps_per_second": 0.67,
+      "step": 1
+    },
+    {
+      "epoch": 0.0017734426956328973,
+      "grad_norm": 97.53443908691406,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 3.1563,
+      "step": 2
+    },
+    {
+      "epoch": 0.002660164043449346,
+      "grad_norm": 101.90808868408203,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 3.2752,
+      "step": 3
+    },
+    {
+      "epoch": 0.0035468853912657946,
+      "grad_norm": 99.9984359741211,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 3.2379,
+      "step": 4
+    },
+    {
+      "epoch": 0.004433606739082243,
+      "grad_norm": 93.44331359863281,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 3.0293,
+      "step": 5
+    },
+    {
+      "epoch": 0.005320328086898692,
+      "grad_norm": 98.4569091796875,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 3.2555,
+      "step": 6
+    },
+    {
+      "epoch": 0.006207049434715141,
+      "grad_norm": 96.81407165527344,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 3.2231,
+      "step": 7
+    },
+    {
+      "epoch": 0.007093770782531589,
+      "grad_norm": 81.63873291015625,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 2.8762,
+      "step": 8
+    },
+    {
+      "epoch": 0.007980492130348038,
+      "grad_norm": 94.48062896728516,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 3.1835,
+      "step": 9
+    },
+    {
+      "epoch": 0.008867213478164486,
+      "grad_norm": 82.22791290283203,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 2.9473,
+      "step": 10
+    },
+    {
+      "epoch": 0.009753934825980935,
+      "grad_norm": 66.42625427246094,
+      "learning_rate": 1.1e-06,
+      "loss": 2.819,
+      "step": 11
+    },
+    {
+      "epoch": 0.010640656173797385,
+      "grad_norm": 63.793827056884766,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 2.8275,
+      "step": 12
+    },
+    {
+      "epoch": 0.011527377521613832,
+      "grad_norm": 55.935733795166016,
+      "learning_rate": 1.3e-06,
+      "loss": 2.6193,
+      "step": 13
+    },
+    {
+      "epoch": 0.012414098869430281,
+      "grad_norm": 37.10558319091797,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 2.2145,
+      "step": 14
+    },
+    {
+      "epoch": 0.01330082021724673,
+      "grad_norm": 31.75094985961914,
+      "learning_rate": 1.5e-06,
+      "loss": 2.0544,
+      "step": 15
+    },
+    {
+      "epoch": 0.014187541565063178,
+      "grad_norm": 33.343467712402344,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 2.0847,
+      "step": 16
+    },
+    {
+      "epoch": 0.015074262912879628,
+      "grad_norm": 30.80988311767578,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 1.9482,
+      "step": 17
+    },
+    {
+      "epoch": 0.015960984260696077,
+      "grad_norm": 29.666048049926758,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 1.8699,
+      "step": 18
+    },
+    {
+      "epoch": 0.016847705608512526,
+      "grad_norm": 29.294734954833984,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 1.4459,
+      "step": 19
+    },
+    {
+      "epoch": 0.017734426956328972,
+      "grad_norm": 31.5433292388916,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.2445,
+      "step": 20
+    },
+    {
+      "epoch": 0.01862114830414542,
+      "grad_norm": 33.008087158203125,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 1.2146,
+      "step": 21
+    },
+    {
+      "epoch": 0.01950786965196187,
+      "grad_norm": 30.054840087890625,
+      "learning_rate": 2.2e-06,
+      "loss": 1.0768,
+      "step": 22
+    },
+    {
+      "epoch": 0.02039459099977832,
+      "grad_norm": 34.325687408447266,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 1.0071,
+      "step": 23
+    },
+    {
+      "epoch": 0.02128131234759477,
+      "grad_norm": 37.760826110839844,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.7746,
+      "step": 24
+    },
+    {
+      "epoch": 0.02216803369541122,
+      "grad_norm": 32.44926834106445,
+      "learning_rate": 2.5e-06,
+      "loss": 0.583,
+      "step": 25
+    },
+    {
+      "epoch": 0.023054755043227664,
+      "grad_norm": 43.93800735473633,
+      "learning_rate": 2.6e-06,
+      "loss": 0.311,
+      "step": 26
+    },
+    {
+      "epoch": 0.023941476391044113,
+      "grad_norm": 34.61675262451172,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 0.2481,
+      "step": 27
+    },
+    {
+      "epoch": 0.024828197738860563,
+      "grad_norm": 16.67363166809082,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.1261,
+      "step": 28
+    },
+    {
+      "epoch": 0.025714919086677012,
+      "grad_norm": 16.36240005493164,
+      "learning_rate": 2.9e-06,
+      "loss": 0.1044,
+      "step": 29
+    },
+    {
+      "epoch": 0.02660164043449346,
+      "grad_norm": 13.118659019470215,
+      "learning_rate": 3e-06,
+      "loss": 0.1006,
+      "step": 30
+    },
+    {
+      "epoch": 0.02748836178230991,
+      "grad_norm": 15.39649486541748,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 0.1069,
+      "step": 31
+    },
+    {
+      "epoch": 0.028375083130126356,
+      "grad_norm": 14.789741516113281,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.113,
+      "step": 32
+    },
+    {
+      "epoch": 0.029261804477942806,
+      "grad_norm": 8.225541114807129,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.0753,
+      "step": 33
+    },
+    {
+      "epoch": 0.030148525825759255,
+      "grad_norm": 5.805243968963623,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.0788,
+      "step": 34
+    },
+    {
+      "epoch": 0.031035247173575704,
+      "grad_norm": 9.23648738861084,
+      "learning_rate": 3.5e-06,
+      "loss": 0.0662,
+      "step": 35
+    },
+    {
+      "epoch": 0.031921968521392154,
+      "grad_norm": 3.07475209236145,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.0452,
+      "step": 36
+    },
+    {
+      "epoch": 0.0328086898692086,
+      "grad_norm": 18.152185440063477,
+      "learning_rate": 3.7e-06,
+      "loss": 0.0722,
+      "step": 37
+    },
+    {
+      "epoch": 0.03369541121702505,
+      "grad_norm": 8.058207511901855,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.0592,
+      "step": 38
+    },
+    {
+      "epoch": 0.0345821325648415,
+      "grad_norm": 6.9130353927612305,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.0616,
+      "step": 39
+    },
+    {
+      "epoch": 0.035468853912657944,
+      "grad_norm": 6.1149516105651855,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.0545,
+      "step": 40
+    },
+    {
+      "epoch": 0.03635557526047439,
+      "grad_norm": 2.7555227279663086,
+      "learning_rate": 4.1e-06,
+      "loss": 0.05,
+      "step": 41
+    },
+    {
+      "epoch": 0.03724229660829084,
+      "grad_norm": 7.199471950531006,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.0536,
+      "step": 42
+    },
+    {
+      "epoch": 0.03812901795610729,
+      "grad_norm": 5.784890174865723,
+      "learning_rate": 4.3e-06,
+      "loss": 0.0534,
+      "step": 43
+    },
+    {
+      "epoch": 0.03901573930392374,
+      "grad_norm": 2.663280487060547,
+      "learning_rate": 4.4e-06,
+      "loss": 0.0459,
+      "step": 44
+    },
+    {
+      "epoch": 0.03990246065174019,
+      "grad_norm": 2.6677751541137695,
+      "learning_rate": 4.5e-06,
+      "loss": 0.0427,
+      "step": 45
+    },
+    {
+      "epoch": 0.04078918199955664,
+      "grad_norm": 6.349153518676758,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.0711,
+      "step": 46
+    },
+    {
+      "epoch": 0.04167590334737309,
+      "grad_norm": 2.082395076751709,
+      "learning_rate": 4.7e-06,
+      "loss": 0.0546,
+      "step": 47
+    },
+    {
+      "epoch": 0.04256262469518954,
+      "grad_norm": 3.0053043365478516,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.0461,
+      "step": 48
+    },
+    {
+      "epoch": 0.04344934604300599,
+      "grad_norm": 4.920153617858887,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.0537,
+      "step": 49
+    },
+    {
+      "epoch": 0.04433606739082244,
+      "grad_norm": 1.2779656648635864,
+      "learning_rate": 5e-06,
+      "loss": 0.0419,
+      "step": 50
+    },
+    {
+      "epoch": 0.045222788738638886,
+      "grad_norm": 2.678823471069336,
+      "learning_rate": 5.1e-06,
+      "loss": 0.0385,
+      "step": 51
+    },
+    {
+      "epoch": 0.04610951008645533,
+      "grad_norm": 2.920421600341797,
+      "learning_rate": 5.2e-06,
+      "loss": 0.0518,
+      "step": 52
+    },
+    {
+      "epoch": 0.04699623143427178,
+      "grad_norm": 2.3334977626800537,
+      "learning_rate": 5.300000000000001e-06,
+      "loss": 0.0397,
+      "step": 53
+    },
+    {
+      "epoch": 0.04788295278208823,
+      "grad_norm": 2.648272752761841,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 0.0374,
+      "step": 54
+    },
+    {
+      "epoch": 0.048769674129904676,
+      "grad_norm": 6.664809703826904,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 0.052,
+      "step": 55
+    },
+    {
+      "epoch": 0.049656395477721126,
+      "grad_norm": 2.4327173233032227,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.0272,
+      "step": 56
+    },
+    {
+      "epoch": 0.050543116825537575,
+      "grad_norm": 1.9386110305786133,
+      "learning_rate": 5.7e-06,
+      "loss": 0.0285,
+      "step": 57
+    },
+    {
+      "epoch": 0.051429838173354024,
+      "grad_norm": 2.0530142784118652,
+      "learning_rate": 5.8e-06,
+      "loss": 0.0353,
+      "step": 58
+    },
+    {
+      "epoch": 0.05231655952117047,
+      "grad_norm": 1.730095624923706,
+      "learning_rate": 5.9e-06,
+      "loss": 0.0393,
+      "step": 59
+    },
+    {
+      "epoch": 0.05320328086898692,
+      "grad_norm": 1.3259291648864746,
+      "learning_rate": 6e-06,
+      "loss": 0.0245,
+      "step": 60
+    },
+    {
+      "epoch": 0.05409000221680337,
+      "grad_norm": 2.2212586402893066,
+      "learning_rate": 6.1e-06,
+      "loss": 0.038,
+      "step": 61
+    },
+    {
+      "epoch": 0.05497672356461982,
+      "grad_norm": 2.2513837814331055,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 0.0351,
+      "step": 62
+    },
+    {
+      "epoch": 0.055863444912436264,
+      "grad_norm": 1.917279601097107,
+      "learning_rate": 6.300000000000001e-06,
+      "loss": 0.0277,
+      "step": 63
+    },
+    {
+      "epoch": 0.05675016626025271,
+      "grad_norm": 4.1914544105529785,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.0375,
+      "step": 64
+    },
+    {
+      "epoch": 0.05763688760806916,
+      "grad_norm": 2.5768065452575684,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 0.0366,
+      "step": 65
+    },
+    {
+      "epoch": 0.05852360895588561,
+      "grad_norm": 2.6516637802124023,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 0.0259,
+      "step": 66
+    },
+    {
+      "epoch": 0.05941033030370206,
+      "grad_norm": 6.144593238830566,
+      "learning_rate": 6.700000000000001e-06,
+      "loss": 0.0416,
+      "step": 67
+    },
+    {
+      "epoch": 0.06029705165151851,
+      "grad_norm": 2.34024977684021,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.0236,
+      "step": 68
+    },
+    {
+      "epoch": 0.06118377299933496,
+      "grad_norm": 1.2529782056808472,
+      "learning_rate": 6.9e-06,
+      "loss": 0.0205,
+      "step": 69
+    },
+    {
+      "epoch": 0.06207049434715141,
+      "grad_norm": 2.8280766010284424,
+      "learning_rate": 7e-06,
+      "loss": 0.024,
+      "step": 70
+    },
+    {
+      "epoch": 0.06295721569496786,
+      "grad_norm": 2.5235207080841064,
+      "learning_rate": 7.100000000000001e-06,
+      "loss": 0.046,
+      "step": 71
+    },
+    {
+      "epoch": 0.06384393704278431,
+      "grad_norm": 2.7281038761138916,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 0.036,
+      "step": 72
+    },
+    {
+      "epoch": 0.06473065839060076,
+      "grad_norm": 1.456838846206665,
+      "learning_rate": 7.3e-06,
+      "loss": 0.0224,
+      "step": 73
+    },
+    {
+      "epoch": 0.0656173797384172,
+      "grad_norm": 2.595977306365967,
+      "learning_rate": 7.4e-06,
+      "loss": 0.0377,
+      "step": 74
+    },
+    {
+      "epoch": 0.06650410108623366,
+      "grad_norm": 1.190896987915039,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.0244,
+      "step": 75
+    },
+    {
+      "epoch": 0.0673908224340501,
+      "grad_norm": 3.9588253498077393,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 0.0471,
+      "step": 76
+    },
+    {
+      "epoch": 0.06827754378186655,
+      "grad_norm": 1.1399868726730347,
+      "learning_rate": 7.7e-06,
+      "loss": 0.0167,
+      "step": 77
+    },
+    {
+      "epoch": 0.069164265129683,
+      "grad_norm": 2.5425806045532227,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 0.0395,
+      "step": 78
+    },
+    {
+      "epoch": 0.07005098647749945,
+      "grad_norm": 1.4164979457855225,
+      "learning_rate": 7.9e-06,
+      "loss": 0.015,
+      "step": 79
+    },
+    {
+      "epoch": 0.07093770782531589,
+      "grad_norm": 2.7378995418548584,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.0232,
+      "step": 80
+    },
+    {
+      "epoch": 0.07182442917313234,
+      "grad_norm": 1.0994404554367065,
+      "learning_rate": 8.1e-06,
+      "loss": 0.014,
+      "step": 81
+    },
+    {
+      "epoch": 0.07271115052094879,
+      "grad_norm": 1.4709571599960327,
+      "learning_rate": 8.2e-06,
+      "loss": 0.0347,
+      "step": 82
+    },
+    {
+      "epoch": 0.07359787186876524,
+      "grad_norm": 1.0280840396881104,
+      "learning_rate": 8.3e-06,
+      "loss": 0.0181,
+      "step": 83
+    },
+    {
+      "epoch": 0.07448459321658168,
+      "grad_norm": 4.092020034790039,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.0395,
+      "step": 84
+    },
+    {
+      "epoch": 0.07537131456439813,
+      "grad_norm": 3.000915050506592,
+      "learning_rate": 8.5e-06,
+      "loss": 0.0356,
+      "step": 85
+    },
+    {
+      "epoch": 0.07625803591221458,
+      "grad_norm": 1.4135894775390625,
+      "learning_rate": 8.6e-06,
+      "loss": 0.0283,
+      "step": 86
+    },
+    {
+      "epoch": 0.07714475726003103,
+      "grad_norm": 2.2624311447143555,
+      "learning_rate": 8.700000000000001e-06,
+      "loss": 0.0329,
+      "step": 87
+    },
+    {
+      "epoch": 0.07803147860784748,
+      "grad_norm": 3.6346747875213623,
+      "learning_rate": 8.8e-06,
+      "loss": 0.0394,
+      "step": 88
+    },
+    {
+      "epoch": 0.07891819995566393,
+      "grad_norm": 6.760399341583252,
+      "learning_rate": 8.900000000000001e-06,
+      "loss": 0.0298,
+      "step": 89
+    },
+    {
+      "epoch": 0.07980492130348038,
+      "grad_norm": 2.7166779041290283,
+      "learning_rate": 9e-06,
+      "loss": 0.0364,
+      "step": 90
+    },
+    {
+      "epoch": 0.08069164265129683,
+      "grad_norm": 1.9583271741867065,
+      "learning_rate": 9.100000000000001e-06,
+      "loss": 0.0257,
+      "step": 91
+    },
+    {
+      "epoch": 0.08157836399911328,
+      "grad_norm": 0.892090916633606,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 0.0242,
+      "step": 92
+    },
+    {
+      "epoch": 0.08246508534692973,
+      "grad_norm": 2.415807008743286,
+      "learning_rate": 9.3e-06,
+      "loss": 0.022,
+      "step": 93
+    },
+    {
+      "epoch": 0.08335180669474618,
+      "grad_norm": 1.4595792293548584,
+      "learning_rate": 9.4e-06,
+      "loss": 0.0227,
+      "step": 94
+    },
+    {
+      "epoch": 0.08423852804256263,
+      "grad_norm": 2.336437702178955,
+      "learning_rate": 9.5e-06,
+      "loss": 0.0395,
+      "step": 95
+    },
+    {
+      "epoch": 0.08512524939037908,
+      "grad_norm": 3.5256662368774414,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.0396,
+      "step": 96
+    },
+    {
+      "epoch": 0.08601197073819553,
+      "grad_norm": 2.1209492683410645,
+      "learning_rate": 9.7e-06,
+      "loss": 0.0289,
+      "step": 97
+    },
+    {
+      "epoch": 0.08689869208601197,
+      "grad_norm": 1.9491901397705078,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.0299,
+      "step": 98
+    },
+    {
+      "epoch": 0.08778541343382842,
+      "grad_norm": 1.8206737041473389,
+      "learning_rate": 9.9e-06,
+      "loss": 0.0285,
+      "step": 99
+    },
+    {
+      "epoch": 0.08867213478164487,
+      "grad_norm": 3.291557788848877,
+      "learning_rate": 1e-05,
+      "loss": 0.0391,
+      "step": 100
+    },
+    {
+      "epoch": 0.08955885612946132,
+      "grad_norm": 2.652033805847168,
+      "learning_rate": 9.999997707933255e-06,
+      "loss": 0.0321,
+      "step": 101
+    },
+    {
+      "epoch": 0.09044557747727777,
+      "grad_norm": 0.7913987040519714,
+      "learning_rate": 9.999990831735122e-06,
+      "loss": 0.0211,
+      "step": 102
+    },
+    {
+      "epoch": 0.09133229882509421,
+      "grad_norm": 2.8626115322113037,
+      "learning_rate": 9.999979371411906e-06,
+      "loss": 0.0317,
+      "step": 103
+    },
+    {
+      "epoch": 0.09221902017291066,
+      "grad_norm": 1.934716820716858,
+      "learning_rate": 9.999963326974111e-06,
+      "loss": 0.0227,
+      "step": 104
+    },
+    {
+      "epoch": 0.0931057415207271,
+      "grad_norm": 1.4647551774978638,
+      "learning_rate": 9.999942698436452e-06,
+      "loss": 0.0253,
+      "step": 105
+    },
+    {
+      "epoch": 0.09399246286854356,
+      "grad_norm": 1.2313497066497803,
+      "learning_rate": 9.999917485817836e-06,
+      "loss": 0.0291,
+      "step": 106
+    },
+    {
+      "epoch": 0.09487918421636,
+      "grad_norm": 3.5174624919891357,
+      "learning_rate": 9.999887689141383e-06,
+      "loss": 0.0495,
+      "step": 107
+    },
+    {
+      "epoch": 0.09576590556417645,
+      "grad_norm": 3.038734197616577,
+      "learning_rate": 9.99985330843441e-06,
+      "loss": 0.0317,
+      "step": 108
+    },
+    {
+      "epoch": 0.0966526269119929,
+      "grad_norm": 1.4445629119873047,
+      "learning_rate": 9.999814343728439e-06,
+      "loss": 0.0235,
+      "step": 109
+    },
+    {
+      "epoch": 0.09753934825980935,
+      "grad_norm": 1.7566113471984863,
+      "learning_rate": 9.999770795059189e-06,
+      "loss": 0.0268,
+      "step": 110
+    },
+    {
+      "epoch": 0.0984260696076258,
+      "grad_norm": 2.2457563877105713,
+      "learning_rate": 9.999722662466594e-06,
+      "loss": 0.0287,
+      "step": 111
+    },
+    {
+      "epoch": 0.09931279095544225,
+      "grad_norm": 2.0211808681488037,
+      "learning_rate": 9.999669945994778e-06,
+      "loss": 0.0241,
+      "step": 112
+    },
+    {
+      "epoch": 0.1001995123032587,
+      "grad_norm": 1.2760939598083496,
+      "learning_rate": 9.999612645692077e-06,
+      "loss": 0.0212,
+      "step": 113
+    },
+    {
+      "epoch": 0.10108623365107515,
+      "grad_norm": 0.9905439019203186,
+      "learning_rate": 9.999550761611021e-06,
+      "loss": 0.0195,
+      "step": 114
+    },
+    {
+      "epoch": 0.1019729549988916,
+      "grad_norm": 1.5984809398651123,
+      "learning_rate": 9.999484293808351e-06,
+      "loss": 0.0175,
+      "step": 115
+    },
+    {
+      "epoch": 0.10285967634670805,
+      "grad_norm": 0.9621034860610962,
+      "learning_rate": 9.999413242345001e-06,
+      "loss": 0.0132,
+      "step": 116
+    },
+    {
+      "epoch": 0.1037463976945245,
+      "grad_norm": 1.1000981330871582,
+      "learning_rate": 9.99933760728612e-06,
+      "loss": 0.0158,
+      "step": 117
+    },
+    {
+      "epoch": 0.10463311904234095,
+      "grad_norm": 0.9676356911659241,
+      "learning_rate": 9.999257388701046e-06,
+      "loss": 0.0227,
+      "step": 118
+    },
+    {
+      "epoch": 0.1055198403901574,
+      "grad_norm": 2.129136323928833,
+      "learning_rate": 9.999172586663331e-06,
+      "loss": 0.028,
+      "step": 119
+    },
+    {
+      "epoch": 0.10640656173797385,
+      "grad_norm": 2.1270437240600586,
+      "learning_rate": 9.99908320125072e-06,
+      "loss": 0.0169,
+      "step": 120
+    },
+    {
+      "epoch": 0.1072932830857903,
+      "grad_norm": 2.0533697605133057,
+      "learning_rate": 9.998989232545164e-06,
+      "loss": 0.024,
+      "step": 121
+    },
+    {
+      "epoch": 0.10818000443360674,
+      "grad_norm": 2.021212577819824,
+      "learning_rate": 9.998890680632818e-06,
+      "loss": 0.0236,
+      "step": 122
+    },
+    {
+      "epoch": 0.1090667257814232,
+      "grad_norm": 3.157949686050415,
+      "learning_rate": 9.998787545604034e-06,
+      "loss": 0.0125,
+      "step": 123
+    },
+    {
+      "epoch": 0.10995344712923964,
+      "grad_norm": 1.5896356105804443,
+      "learning_rate": 9.998679827553372e-06,
+      "loss": 0.0222,
+      "step": 124
+    },
+    {
+      "epoch": 0.11084016847705609,
+      "grad_norm": 1.598581075668335,
+      "learning_rate": 9.99856752657959e-06,
+      "loss": 0.0271,
+      "step": 125
+    },
+    {
+      "epoch": 0.11172688982487253,
+      "grad_norm": 0.9781515598297119,
+      "learning_rate": 9.998450642785649e-06,
+      "loss": 0.0122,
+      "step": 126
+    },
+    {
+      "epoch": 0.11261361117268898,
+      "grad_norm": 1.5786858797073364,
+      "learning_rate": 9.99832917627871e-06,
+      "loss": 0.0219,
+      "step": 127
+    },
+    {
+      "epoch": 0.11350033252050543,
+      "grad_norm": 1.4831783771514893,
+      "learning_rate": 9.998203127170137e-06,
+      "loss": 0.0215,
+      "step": 128
+    },
+    {
+      "epoch": 0.11438705386832188,
+      "grad_norm": 2.904634952545166,
+      "learning_rate": 9.998072495575493e-06,
+      "loss": 0.0269,
+      "step": 129
+    },
+    {
+      "epoch": 0.11527377521613832,
+      "grad_norm": 1.4405618906021118,
+      "learning_rate": 9.997937281614549e-06,
+      "loss": 0.0281,
+      "step": 130
+    },
+    {
+      "epoch": 0.11616049656395477,
+      "grad_norm": 1.347447156906128,
+      "learning_rate": 9.99779748541127e-06,
+      "loss": 0.0225,
+      "step": 131
+    },
+    {
+      "epoch": 0.11704721791177122,
+      "grad_norm": 0.973578691482544,
+      "learning_rate": 9.997653107093825e-06,
+      "loss": 0.0232,
+      "step": 132
+    },
+    {
+      "epoch": 0.11793393925958767,
+      "grad_norm": 1.3047363758087158,
+      "learning_rate": 9.997504146794587e-06,
+      "loss": 0.0279,
+      "step": 133
+    },
+    {
+      "epoch": 0.11882066060740412,
+      "grad_norm": 1.9343928098678589,
+      "learning_rate": 9.997350604650123e-06,
+      "loss": 0.0334,
+      "step": 134
+    },
+    {
+      "epoch": 0.11970738195522057,
+      "grad_norm": 0.9066912531852722,
+      "learning_rate": 9.997192480801203e-06,
+      "loss": 0.0143,
+      "step": 135
+    },
+    {
+      "epoch": 0.12059410330303702,
+      "grad_norm": 1.209761381149292,
+      "learning_rate": 9.997029775392805e-06,
+      "loss": 0.0216,
+      "step": 136
+    },
+    {
+      "epoch": 0.12148082465085347,
+      "grad_norm": 2.670605182647705,
+      "learning_rate": 9.996862488574098e-06,
+      "loss": 0.0264,
+      "step": 137
+    },
+    {
+      "epoch": 0.12236754599866992,
+      "grad_norm": 1.5146721601486206,
+      "learning_rate": 9.996690620498454e-06,
+      "loss": 0.0262,
+      "step": 138
+    },
+    {
+      "epoch": 0.12325426734648637,
+      "grad_norm": 2.121723175048828,
+      "learning_rate": 9.996514171323449e-06,
+      "loss": 0.0345,
+      "step": 139
+    },
+    {
+      "epoch": 0.12414098869430282,
+      "grad_norm": 0.9533706307411194,
+      "learning_rate": 9.996333141210855e-06,
+      "loss": 0.0192,
+      "step": 140
+    },
+    {
+      "epoch": 0.12502771004211927,
+      "grad_norm": 0.7458922863006592,
+      "learning_rate": 9.996147530326645e-06,
+      "loss": 0.0172,
+      "step": 141
+    },
+    {
+      "epoch": 0.12591443138993572,
+      "grad_norm": 1.2241960763931274,
+      "learning_rate": 9.995957338840993e-06,
+      "loss": 0.0152,
+      "step": 142
+    },
+    {
+      "epoch": 0.12680115273775217,
+      "grad_norm": 1.9640858173370361,
+      "learning_rate": 9.99576256692827e-06,
+      "loss": 0.0206,
+      "step": 143
+    },
+    {
+      "epoch": 0.12768787408556861,
+      "grad_norm": 0.6544690728187561,
+      "learning_rate": 9.99556321476705e-06,
+      "loss": 0.0105,
+      "step": 144
+    },
+    {
+      "epoch": 0.12857459543338506,
+      "grad_norm": 0.9590530395507812,
+      "learning_rate": 9.995359282540105e-06,
+      "loss": 0.0212,
+      "step": 145
+    },
+    {
+      "epoch": 0.1294613167812015,
+      "grad_norm": 1.0109769105911255,
+      "learning_rate": 9.995150770434401e-06,
+      "loss": 0.0105,
+      "step": 146
+    },
+    {
+      "epoch": 0.13034803812901796,
+      "grad_norm": 1.1498576402664185,
+      "learning_rate": 9.994937678641113e-06,
+      "loss": 0.0255,
+      "step": 147
+    },
+    {
+      "epoch": 0.1312347594768344,
+      "grad_norm": 1.3174915313720703,
+      "learning_rate": 9.994720007355604e-06,
+      "loss": 0.0188,
+      "step": 148
+    },
+    {
+      "epoch": 0.13212148082465086,
+      "grad_norm": 2.1449038982391357,
+      "learning_rate": 9.994497756777445e-06,
+      "loss": 0.0254,
+      "step": 149
+    },
+    {
+      "epoch": 0.1330082021724673,
+      "grad_norm": 2.002474784851074,
+      "learning_rate": 9.994270927110402e-06,
+      "loss": 0.0178,
+      "step": 150
+    },
+    {
+      "epoch": 0.13389492352028376,
+      "grad_norm": 3.6749556064605713,
+      "learning_rate": 9.994039518562433e-06,
+      "loss": 0.0374,
+      "step": 151
+    },
+    {
+      "epoch": 0.1347816448681002,
+      "grad_norm": 0.846824049949646,
+      "learning_rate": 9.993803531345703e-06,
+      "loss": 0.0239,
+      "step": 152
+    },
+    {
+      "epoch": 0.13566836621591666,
+      "grad_norm": 1.4229305982589722,
+      "learning_rate": 9.993562965676572e-06,
+      "loss": 0.0242,
+      "step": 153
+    },
+    {
+      "epoch": 0.1365550875637331,
+      "grad_norm": 3.022684097290039,
+      "learning_rate": 9.993317821775597e-06,
+      "loss": 0.0362,
+      "step": 154
+    },
+    {
+      "epoch": 0.13744180891154956,
+      "grad_norm": 1.6719801425933838,
+      "learning_rate": 9.99306809986753e-06,
+      "loss": 0.0212,
+      "step": 155
+    },
+    {
+      "epoch": 0.138328530259366,
+      "grad_norm": 2.7858753204345703,
+      "learning_rate": 9.992813800181326e-06,
+      "loss": 0.0241,
+      "step": 156
+    },
+    {
+      "epoch": 0.13921525160718246,
+      "grad_norm": 0.45698562264442444,
+      "learning_rate": 9.992554922950133e-06,
+      "loss": 0.0133,
+      "step": 157
+    },
+    {
+      "epoch": 0.1401019729549989,
+      "grad_norm": 1.5275429487228394,
+      "learning_rate": 9.992291468411293e-06,
+      "loss": 0.0248,
+      "step": 158
+    },
+    {
+      "epoch": 0.14098869430281535,
+      "grad_norm": 2.2717480659484863,
+      "learning_rate": 9.992023436806353e-06,
+      "loss": 0.0272,
+      "step": 159
+    },
+    {
+      "epoch": 0.14187541565063178,
+      "grad_norm": 1.3149230480194092,
+      "learning_rate": 9.991750828381048e-06,
+      "loss": 0.0182,
+      "step": 160
+    },
+    {
+      "epoch": 0.14276213699844822,
+      "grad_norm": 1.2786978483200073,
+      "learning_rate": 9.991473643385317e-06,
+      "loss": 0.0187,
+      "step": 161
+    },
+    {
+      "epoch": 0.14364885834626467,
+      "grad_norm": 1.7859911918640137,
+      "learning_rate": 9.991191882073285e-06,
+      "loss": 0.0215,
+      "step": 162
+    },
+    {
+      "epoch": 0.14453557969408112,
+      "grad_norm": 1.611259937286377,
+      "learning_rate": 9.990905544703281e-06,
+      "loss": 0.0181,
+      "step": 163
+    },
+    {
+      "epoch": 0.14542230104189757,
+      "grad_norm": 0.6689061522483826,
+      "learning_rate": 9.990614631537827e-06,
+      "loss": 0.0124,
+      "step": 164
+    },
+    {
+      "epoch": 0.14630902238971402,
+      "grad_norm": 1.0409244298934937,
+      "learning_rate": 9.990319142843641e-06,
+      "loss": 0.0204,
+      "step": 165
+    },
+    {
+      "epoch": 0.14719574373753047,
+      "grad_norm": 0.7819463014602661,
+      "learning_rate": 9.990019078891633e-06,
+      "loss": 0.0223,
+      "step": 166
+    },
+    {
+      "epoch": 0.14808246508534692,
+      "grad_norm": 1.9662169218063354,
+      "learning_rate": 9.989714439956909e-06,
+      "loss": 0.0119,
+      "step": 167
+    },
+    {
+      "epoch": 0.14896918643316337,
+      "grad_norm": 1.353804111480713,
+      "learning_rate": 9.989405226318772e-06,
+      "loss": 0.0276,
+      "step": 168
+    },
+    {
+      "epoch": 0.14985590778097982,
+      "grad_norm": 0.8875593543052673,
+      "learning_rate": 9.989091438260718e-06,
+      "loss": 0.0224,
+      "step": 169
+    },
+    {
+      "epoch": 0.15074262912879627,
+      "grad_norm": 0.862114429473877,
+      "learning_rate": 9.988773076070433e-06,
+      "loss": 0.0159,
+      "step": 170
+    },
+    {
+      "epoch": 0.15162935047661272,
+      "grad_norm": 2.4255783557891846,
+      "learning_rate": 9.988450140039802e-06,
+      "loss": 0.0381,
+      "step": 171
+    },
+    {
+      "epoch": 0.15251607182442917,
+      "grad_norm": 0.9882796406745911,
+      "learning_rate": 9.988122630464902e-06,
+      "loss": 0.0182,
+      "step": 172
+    },
+    {
+      "epoch": 0.15340279317224562,
+      "grad_norm": 1.3533726930618286,
+      "learning_rate": 9.987790547646003e-06,
+      "loss": 0.0253,
+      "step": 173
+    },
+    {
+      "epoch": 0.15428951452006207,
+      "grad_norm": 0.563122034072876,
+      "learning_rate": 9.987453891887567e-06,
+      "loss": 0.0185,
+      "step": 174
+    },
+    {
+      "epoch": 0.15517623586787851,
+      "grad_norm": 0.98648601770401,
+      "learning_rate": 9.987112663498245e-06,
+      "loss": 0.0175,
+      "step": 175
+    },
+    {
+      "epoch": 0.15606295721569496,
+      "grad_norm": 3.2445409297943115,
+      "learning_rate": 9.986766862790888e-06,
+      "loss": 0.0206,
+      "step": 176
+    },
+    {
+      "epoch": 0.1569496785635114,
+      "grad_norm": 0.9300917387008667,
+      "learning_rate": 9.986416490082537e-06,
+      "loss": 0.0115,
+      "step": 177
+    },
+    {
+      "epoch": 0.15783639991132786,
+      "grad_norm": 0.5182569026947021,
+      "learning_rate": 9.98606154569442e-06,
+      "loss": 0.0107,
+      "step": 178
+    },
+    {
+      "epoch": 0.1587231212591443,
+      "grad_norm": 1.1416966915130615,
+      "learning_rate": 9.98570202995196e-06,
+      "loss": 0.0116,
+      "step": 179
+    },
+    {
+      "epoch": 0.15960984260696076,
+      "grad_norm": 0.6441078782081604,
+      "learning_rate": 9.98533794318477e-06,
+      "loss": 0.0147,
+      "step": 180
+    },
+    {
+      "epoch": 0.1604965639547772,
+      "grad_norm": 1.4156014919281006,
+      "learning_rate": 9.984969285726656e-06,
+      "loss": 0.0199,
+      "step": 181
+    },
+    {
+      "epoch": 0.16138328530259366,
+      "grad_norm": 2.0149669647216797,
+      "learning_rate": 9.984596057915613e-06,
+      "loss": 0.0323,
+      "step": 182
+    },
+    {
+      "epoch": 0.1622700066504101,
+      "grad_norm": 1.7955539226531982,
+      "learning_rate": 9.984218260093826e-06,
+      "loss": 0.0332,
+      "step": 183
+    },
+    {
+      "epoch": 0.16315672799822656,
+      "grad_norm": 1.028401494026184,
+      "learning_rate": 9.98383589260767e-06,
+      "loss": 0.0174,
+      "step": 184
+    },
+    {
+      "epoch": 0.164043449346043,
+      "grad_norm": 1.5321674346923828,
+      "learning_rate": 9.983448955807708e-06,
+      "loss": 0.028,
+      "step": 185
+    },
+    {
+      "epoch": 0.16493017069385946,
+      "grad_norm": 1.3364337682724,
+      "learning_rate": 9.983057450048697e-06,
+      "loss": 0.0178,
+      "step": 186
+    },
+    {
+      "epoch": 0.1658168920416759,
+      "grad_norm": 1.4727017879486084,
+      "learning_rate": 9.982661375689577e-06,
+      "loss": 0.0177,
+      "step": 187
+    },
+    {
+      "epoch": 0.16670361338949236,
+      "grad_norm": 0.46738162636756897,
+      "learning_rate": 9.982260733093482e-06,
+      "loss": 0.0085,
+      "step": 188
+    },
+    {
+      "epoch": 0.1675903347373088,
+      "grad_norm": 0.4024590849876404,
+      "learning_rate": 9.98185552262773e-06,
+      "loss": 0.0131,
+      "step": 189
+    },
+    {
+      "epoch": 0.16847705608512525,
+      "grad_norm": 0.5752727389335632,
+      "learning_rate": 9.98144574466383e-06,
+      "loss": 0.0111,
+      "step": 190
+    },
+    {
+      "epoch": 0.1693637774329417,
+      "grad_norm": 0.593545138835907,
+      "learning_rate": 9.981031399577479e-06,
+      "loss": 0.012,
+      "step": 191
+    },
+    {
+      "epoch": 0.17025049878075815,
+      "grad_norm": 0.9298077821731567,
+      "learning_rate": 9.980612487748556e-06,
+      "loss": 0.0143,
+      "step": 192
+    },
+    {
+      "epoch": 0.1711372201285746,
+      "grad_norm": 3.385441780090332,
+      "learning_rate": 9.980189009561131e-06,
+      "loss": 0.0192,
+      "step": 193
+    },
+    {
+      "epoch": 0.17202394147639105,
+      "grad_norm": 1.0534871816635132,
+      "learning_rate": 9.979760965403462e-06,
+      "loss": 0.018,
+      "step": 194
+    },
+    {
+      "epoch": 0.1729106628242075,
+      "grad_norm": 0.6916846036911011,
+      "learning_rate": 9.97932835566799e-06,
+      "loss": 0.0078,
+      "step": 195
+    },
+    {
+      "epoch": 0.17379738417202395,
+      "grad_norm": 0.8603850603103638,
+      "learning_rate": 9.978891180751346e-06,
+      "loss": 0.0101,
+      "step": 196
+    },
+    {
+      "epoch": 0.1746841055198404,
+      "grad_norm": 0.9636616110801697,
+      "learning_rate": 9.97844944105434e-06,
+      "loss": 0.0111,
+      "step": 197
+    },
+    {
+      "epoch": 0.17557082686765685,
+      "grad_norm": 1.4222532510757446,
+      "learning_rate": 9.978003136981972e-06,
+      "loss": 0.0158,
+      "step": 198
+    },
+    {
+      "epoch": 0.1764575482154733,
+      "grad_norm": 1.2078766822814941,
+      "learning_rate": 9.977552268943426e-06,
+      "loss": 0.0227,
+      "step": 199
+    },
+    {
+      "epoch": 0.17734426956328975,
+      "grad_norm": 0.7552307844161987,
+      "learning_rate": 9.97709683735207e-06,
+      "loss": 0.0125,
+      "step": 200
+    },
+    {
+      "epoch": 0.1782309909111062,
+      "grad_norm": 1.0908719301223755,
+      "learning_rate": 9.976636842625454e-06,
+      "loss": 0.0235,
+      "step": 201
+    },
+    {
+      "epoch": 0.17911771225892265,
+      "grad_norm": 1.2128921747207642,
+      "learning_rate": 9.976172285185315e-06,
+      "loss": 0.0128,
+      "step": 202
+    },
+    {
+      "epoch": 0.1800044336067391,
+      "grad_norm": 0.7293077707290649,
+      "learning_rate": 9.975703165457571e-06,
+      "loss": 0.0146,
+      "step": 203
+    },
+    {
+      "epoch": 0.18089115495455554,
+      "grad_norm": 0.995217502117157,
+      "learning_rate": 9.975229483872325e-06,
+      "loss": 0.0166,
+      "step": 204
+    },
+    {
+      "epoch": 0.181777876302372,
+      "grad_norm": 0.7026538252830505,
+      "learning_rate": 9.974751240863858e-06,
+      "loss": 0.02,
+      "step": 205
+    },
+    {
+      "epoch": 0.18266459765018841,
+      "grad_norm": 1.7263981103897095,
+      "learning_rate": 9.97426843687064e-06,
+      "loss": 0.0286,
+      "step": 206
+    },
+    {
+      "epoch": 0.18355131899800486,
+      "grad_norm": 1.4515742063522339,
+      "learning_rate": 9.973781072335315e-06,
+      "loss": 0.015,
+      "step": 207
+    },
+    {
+      "epoch": 0.1844380403458213,
+      "grad_norm": 0.743718683719635,
+      "learning_rate": 9.973289147704714e-06,
+      "loss": 0.0183,
+      "step": 208
+    },
+    {
+      "epoch": 0.18532476169363776,
+      "grad_norm": 0.5708590745925903,
+      "learning_rate": 9.972792663429847e-06,
+      "loss": 0.0107,
+      "step": 209
+    },
+    {
+      "epoch": 0.1862114830414542,
+      "grad_norm": 1.0680079460144043,
+      "learning_rate": 9.972291619965901e-06,
+      "loss": 0.0198,
+      "step": 210
+    },
+    {
+      "epoch": 0.18709820438927066,
+      "grad_norm": 1.7404032945632935,
+      "learning_rate": 9.971786017772249e-06,
+      "loss": 0.0187,
+      "step": 211
+    },
+    {
+      "epoch": 0.1879849257370871,
+      "grad_norm": 1.116582989692688,
+      "learning_rate": 9.971275857312438e-06,
+      "loss": 0.0124,
+      "step": 212
+    },
+    {
+      "epoch": 0.18887164708490356,
+      "grad_norm": 0.9854468107223511,
+      "learning_rate": 9.9707611390542e-06,
+      "loss": 0.0146,
+      "step": 213
+    },
+    {
+      "epoch": 0.18975836843272,
+      "grad_norm": 1.0946894884109497,
+      "learning_rate": 9.970241863469439e-06,
+      "loss": 0.0165,
+      "step": 214
+    },
+    {
+      "epoch": 0.19064508978053646,
+      "grad_norm": 0.6762576699256897,
+      "learning_rate": 9.969718031034243e-06,
+      "loss": 0.0083,
+      "step": 215
+    },
+    {
+      "epoch": 0.1915318111283529,
+      "grad_norm": 0.8591622114181519,
+      "learning_rate": 9.969189642228874e-06,
+      "loss": 0.02,
+      "step": 216
+    },
+    {
+      "epoch": 0.19241853247616936,
+      "grad_norm": 0.9623995423316956,
+      "learning_rate": 9.968656697537776e-06,
+      "loss": 0.0145,
+      "step": 217
+    },
+    {
+      "epoch": 0.1933052538239858,
+      "grad_norm": 1.2979151010513306,
+      "learning_rate": 9.968119197449564e-06,
+      "loss": 0.0143,
+      "step": 218
+    },
+    {
+      "epoch": 0.19419197517180226,
+      "grad_norm": 1.155766487121582,
+      "learning_rate": 9.967577142457031e-06,
+      "loss": 0.0167,
+      "step": 219
+    },
+    {
+      "epoch": 0.1950786965196187,
+      "grad_norm": 2.1985487937927246,
+      "learning_rate": 9.967030533057155e-06,
+      "loss": 0.0305,
+      "step": 220
+    },
+    {
+      "epoch": 0.19596541786743515,
+      "grad_norm": 0.6944717168807983,
+      "learning_rate": 9.966479369751072e-06,
+      "loss": 0.0157,
+      "step": 221
+    },
+    {
+      "epoch": 0.1968521392152516,
+      "grad_norm": 0.7124226093292236,
+      "learning_rate": 9.96592365304411e-06,
+      "loss": 0.0255,
+      "step": 222
+    },
+    {
+      "epoch": 0.19773886056306805,
+      "grad_norm": 1.8563802242279053,
+      "learning_rate": 9.965363383445762e-06,
+      "loss": 0.0189,
+      "step": 223
+    },
+    {
+      "epoch": 0.1986255819108845,
+      "grad_norm": 2.422700881958008,
+      "learning_rate": 9.9647985614697e-06,
+      "loss": 0.0249,
+      "step": 224
+    },
+    {
+      "epoch": 0.19951230325870095,
+      "grad_norm": 1.5919915437698364,
+      "learning_rate": 9.964229187633767e-06,
+      "loss": 0.0243,
+      "step": 225
+    },
+    {
+      "epoch": 0.2003990246065174,
+      "grad_norm": 0.6012934446334839,
+      "learning_rate": 9.963655262459978e-06,
+      "loss": 0.0228,
+      "step": 226
+    },
+    {
+      "epoch": 0.20128574595433385,
+      "grad_norm": 0.7975476384162903,
+      "learning_rate": 9.963076786474529e-06,
+      "loss": 0.0203,
+      "step": 227
+    },
+    {
+      "epoch": 0.2021724673021503,
+      "grad_norm": 0.9103793501853943,
+      "learning_rate": 9.962493760207775e-06,
+      "loss": 0.0199,
+      "step": 228
+    },
+    {
+      "epoch": 0.20305918864996675,
+      "grad_norm": 2.0518851280212402,
+      "learning_rate": 9.961906184194255e-06,
+      "loss": 0.0284,
+      "step": 229
+    },
+    {
+      "epoch": 0.2039459099977832,
+      "grad_norm": 0.6736752390861511,
+      "learning_rate": 9.961314058972672e-06,
+      "loss": 0.0115,
+      "step": 230
+    },
+    {
+      "epoch": 0.20483263134559965,
+      "grad_norm": 0.31807392835617065,
+      "learning_rate": 9.960717385085904e-06,
+      "loss": 0.0126,
+      "step": 231
+    },
+    {
+      "epoch": 0.2057193526934161,
+      "grad_norm": 1.0832537412643433,
+      "learning_rate": 9.960116163080995e-06,
+      "loss": 0.0179,
+      "step": 232
+    },
+    {
+      "epoch": 0.20660607404123255,
+      "grad_norm": 1.3503203392028809,
+      "learning_rate": 9.959510393509163e-06,
+      "loss": 0.0191,
+      "step": 233
+    },
+    {
+      "epoch": 0.207492795389049,
+      "grad_norm": 0.45156577229499817,
+      "learning_rate": 9.958900076925793e-06,
+      "loss": 0.0088,
+      "step": 234
+    },
+    {
+      "epoch": 0.20837951673686544,
+      "grad_norm": 0.5721056461334229,
+      "learning_rate": 9.958285213890442e-06,
+      "loss": 0.0159,
+      "step": 235
+    },
+    {
+      "epoch": 0.2092662380846819,
+      "grad_norm": 1.2899566888809204,
+      "learning_rate": 9.95766580496683e-06,
+      "loss": 0.0179,
+      "step": 236
+    },
+    {
+      "epoch": 0.21015295943249834,
+      "grad_norm": 0.5771740674972534,
+      "learning_rate": 9.957041850722848e-06,
+      "loss": 0.0077,
+      "step": 237
+    },
+    {
+      "epoch": 0.2110396807803148,
+      "grad_norm": 0.7620949745178223,
+      "learning_rate": 9.956413351730556e-06,
+      "loss": 0.0163,
+      "step": 238
+    },
+    {
+      "epoch": 0.21192640212813124,
+      "grad_norm": 1.0351495742797852,
+      "learning_rate": 9.955780308566174e-06,
+      "loss": 0.0178,
+      "step": 239
+    },
+    {
+      "epoch": 0.2128131234759477,
+      "grad_norm": 0.6831864714622498,
+      "learning_rate": 9.955142721810099e-06,
+      "loss": 0.0087,
+      "step": 240
+    },
+    {
+      "epoch": 0.21369984482376414,
+      "grad_norm": 0.6774961948394775,
+      "learning_rate": 9.954500592046883e-06,
+      "loss": 0.0123,
+      "step": 241
+    },
+    {
+      "epoch": 0.2145865661715806,
+      "grad_norm": 1.7236403226852417,
+      "learning_rate": 9.953853919865251e-06,
+      "loss": 0.0232,
+      "step": 242
+    },
+    {
+      "epoch": 0.21547328751939704,
+      "grad_norm": 1.017462134361267,
+      "learning_rate": 9.953202705858087e-06,
+      "loss": 0.0176,
+      "step": 243
+    },
+    {
+      "epoch": 0.2163600088672135,
+      "grad_norm": 0.37702977657318115,
+      "learning_rate": 9.95254695062244e-06,
+      "loss": 0.01,
+      "step": 244
+    },
+    {
+      "epoch": 0.21724673021502994,
+      "grad_norm": 1.5439709424972534,
+      "learning_rate": 9.951886654759528e-06,
+      "loss": 0.0193,
+      "step": 245
+    },
+    {
+      "epoch": 0.2181334515628464,
+      "grad_norm": 0.5491088628768921,
+      "learning_rate": 9.951221818874724e-06,
+      "loss": 0.0095,
+      "step": 246
+    },
+    {
+      "epoch": 0.21902017291066284,
+      "grad_norm": 0.6317766308784485,
+      "learning_rate": 9.950552443577571e-06,
+      "loss": 0.0164,
+      "step": 247
+    },
+    {
+      "epoch": 0.21990689425847929,
+      "grad_norm": 0.9085983037948608,
+      "learning_rate": 9.949878529481767e-06,
+      "loss": 0.0137,
+      "step": 248
+    },
+    {
+      "epoch": 0.22079361560629573,
+      "grad_norm": 0.583099365234375,
+      "learning_rate": 9.949200077205177e-06,
+      "loss": 0.012,
+      "step": 249
+    },
+    {
+      "epoch": 0.22168033695411218,
+      "grad_norm": 0.4570126533508301,
+      "learning_rate": 9.948517087369822e-06,
+      "loss": 0.008,
+      "step": 250
+    },
+    {
+      "epoch": 0.2225670583019286,
+      "grad_norm": 1.1167882680892944,
+      "learning_rate": 9.947829560601884e-06,
+      "loss": 0.0181,
+      "step": 251
+    },
+    {
+      "epoch": 0.22345377964974505,
+      "grad_norm": 0.6278082728385925,
+      "learning_rate": 9.947137497531711e-06,
+      "loss": 0.0154,
+      "step": 252
+    },
+    {
+      "epoch": 0.2243405009975615,
+      "grad_norm": 0.43254172801971436,
+      "learning_rate": 9.9464408987938e-06,
+      "loss": 0.0093,
+      "step": 253
+    },
+    {
+      "epoch": 0.22522722234537795,
+      "grad_norm": 0.45759662985801697,
+      "learning_rate": 9.945739765026814e-06,
+      "loss": 0.0105,
+      "step": 254
+    },
+    {
+      "epoch": 0.2261139436931944,
+      "grad_norm": 0.6438144445419312,
+      "learning_rate": 9.945034096873571e-06,
+      "loss": 0.0136,
+      "step": 255
+    },
+    {
+      "epoch": 0.22700066504101085,
+      "grad_norm": 1.2649673223495483,
+      "learning_rate": 9.944323894981045e-06,
+      "loss": 0.0223,
+      "step": 256
+    },
+    {
+      "epoch": 0.2278873863888273,
+      "grad_norm": 0.7128615975379944,
+      "learning_rate": 9.943609160000369e-06,
+      "loss": 0.0123,
+      "step": 257
+    },
+    {
+      "epoch": 0.22877410773664375,
+      "grad_norm": 1.0316693782806396,
+      "learning_rate": 9.94288989258683e-06,
+      "loss": 0.0163,
+      "step": 258
+    },
+    {
+      "epoch": 0.2296608290844602,
+      "grad_norm": 1.885494589805603,
+      "learning_rate": 9.942166093399875e-06,
+      "loss": 0.0174,
+      "step": 259
+    },
+    {
+      "epoch": 0.23054755043227665,
+      "grad_norm": 0.5127606391906738,
+      "learning_rate": 9.9414377631031e-06,
+      "loss": 0.0091,
+      "step": 260
+    },
+    {
+      "epoch": 0.2314342717800931,
+      "grad_norm": 0.6301166415214539,
+      "learning_rate": 9.940704902364254e-06,
+      "loss": 0.0128,
+      "step": 261
+    },
+    {
+      "epoch": 0.23232099312790955,
+      "grad_norm": 0.6273002624511719,
+      "learning_rate": 9.93996751185525e-06,
+      "loss": 0.0131,
+      "step": 262
+    },
+    {
+      "epoch": 0.233207714475726,
+      "grad_norm": 0.4122670888900757,
+      "learning_rate": 9.939225592252143e-06,
+      "loss": 0.006,
+      "step": 263
+    },
+    {
+      "epoch": 0.23409443582354245,
+      "grad_norm": 0.3675723075866699,
+      "learning_rate": 9.938479144235146e-06,
+      "loss": 0.0076,
+      "step": 264
+    },
+    {
+      "epoch": 0.2349811571713589,
+      "grad_norm": 0.5711640119552612,
+      "learning_rate": 9.937728168488622e-06,
+      "loss": 0.0123,
+      "step": 265
+    },
+    {
+      "epoch": 0.23586787851917534,
+      "grad_norm": 0.5491669774055481,
+      "learning_rate": 9.936972665701086e-06,
+      "loss": 0.0094,
+      "step": 266
+    },
+    {
+      "epoch": 0.2367545998669918,
+      "grad_norm": 0.25534459948539734,
+      "learning_rate": 9.936212636565205e-06,
+      "loss": 0.0032,
+      "step": 267
+    },
+    {
+      "epoch": 0.23764132121480824,
+      "grad_norm": 0.7696008682250977,
+      "learning_rate": 9.93544808177779e-06,
+      "loss": 0.0118,
+      "step": 268
+    },
+    {
+      "epoch": 0.2385280425626247,
+      "grad_norm": 0.8142785429954529,
+      "learning_rate": 9.934679002039809e-06,
+      "loss": 0.0202,
+      "step": 269
+    },
+    {
+      "epoch": 0.23941476391044114,
+      "grad_norm": 0.9304322004318237,
+      "learning_rate": 9.933905398056371e-06,
+      "loss": 0.0158,
+      "step": 270
+    },
+    {
+      "epoch": 0.2403014852582576,
+      "grad_norm": 0.6360547542572021,
+      "learning_rate": 9.93312727053674e-06,
+      "loss": 0.0091,
+      "step": 271
+    },
+    {
+      "epoch": 0.24118820660607404,
+      "grad_norm": 0.5449398756027222,
+      "learning_rate": 9.932344620194322e-06,
+      "loss": 0.0112,
+      "step": 272
+    },
+    {
+      "epoch": 0.2420749279538905,
+      "grad_norm": 0.4042613208293915,
+      "learning_rate": 9.931557447746675e-06,
+      "loss": 0.0056,
+      "step": 273
+    },
+    {
+      "epoch": 0.24296164930170694,
+      "grad_norm": 0.8487951755523682,
+      "learning_rate": 9.930765753915497e-06,
+      "loss": 0.012,
+      "step": 274
+    },
+    {
+      "epoch": 0.2438483706495234,
+      "grad_norm": 0.47606247663497925,
+      "learning_rate": 9.929969539426634e-06,
+      "loss": 0.0075,
+      "step": 275
+    },
+    {
+      "epoch": 0.24473509199733984,
+      "grad_norm": 0.7890058159828186,
+      "learning_rate": 9.929168805010078e-06,
+      "loss": 0.0142,
+      "step": 276
+    },
+    {
+      "epoch": 0.2456218133451563,
+      "grad_norm": 0.6880226135253906,
+      "learning_rate": 9.928363551399961e-06,
+      "loss": 0.0123,
+      "step": 277
+    },
+    {
+      "epoch": 0.24650853469297274,
+      "grad_norm": 0.7886592745780945,
+      "learning_rate": 9.927553779334565e-06,
+      "loss": 0.0122,
+      "step": 278
+    },
+    {
+      "epoch": 0.24739525604078919,
+      "grad_norm": 0.44544312357902527,
+      "learning_rate": 9.926739489556308e-06,
+      "loss": 0.006,
+      "step": 279
+    },
+    {
+      "epoch": 0.24828197738860563,
+      "grad_norm": 0.6304795742034912,
+      "learning_rate": 9.925920682811752e-06,
+      "loss": 0.0145,
+      "step": 280
+    },
+    {
+      "epoch": 0.24916869873642208,
+      "grad_norm": 0.47169551253318787,
+      "learning_rate": 9.925097359851603e-06,
+      "loss": 0.0063,
+      "step": 281
+    },
+    {
+      "epoch": 0.25005542008423853,
+      "grad_norm": 0.43053027987480164,
+      "learning_rate": 9.924269521430705e-06,
+      "loss": 0.005,
+      "step": 282
+    },
+    {
+      "epoch": 0.25094214143205495,
+      "grad_norm": 0.7776206731796265,
+      "learning_rate": 9.923437168308042e-06,
+      "loss": 0.0152,
+      "step": 283
+    },
+    {
+      "epoch": 0.25182886277987143,
+      "grad_norm": 1.4472633600234985,
+      "learning_rate": 9.922600301246736e-06,
+      "loss": 0.0307,
+      "step": 284
+    },
+    {
+      "epoch": 0.25271558412768785,
+      "grad_norm": 0.6039207577705383,
+      "learning_rate": 9.921758921014052e-06,
+      "loss": 0.0144,
+      "step": 285
+    },
+    {
+      "epoch": 0.25360230547550433,
+      "grad_norm": 0.8766511082649231,
+      "learning_rate": 9.92091302838139e-06,
+      "loss": 0.0138,
+      "step": 286
+    },
+    {
+      "epoch": 0.25448902682332075,
+      "grad_norm": 1.0067875385284424,
+      "learning_rate": 9.920062624124282e-06,
+      "loss": 0.0193,
+      "step": 287
+    },
+    {
+      "epoch": 0.25537574817113723,
+      "grad_norm": 0.9771963953971863,
+      "learning_rate": 9.919207709022407e-06,
+      "loss": 0.0289,
+      "step": 288
+    },
+    {
+      "epoch": 0.25626246951895365,
+      "grad_norm": 0.8762062191963196,
+      "learning_rate": 9.91834828385957e-06,
+      "loss": 0.0139,
+      "step": 289
+    },
+    {
+      "epoch": 0.2571491908667701,
+      "grad_norm": 0.9518645405769348,
+      "learning_rate": 9.917484349423719e-06,
+      "loss": 0.0079,
+      "step": 290
+    },
+    {
+      "epoch": 0.25803591221458655,
+      "grad_norm": 0.36570677161216736,
+      "learning_rate": 9.916615906506927e-06,
+      "loss": 0.0119,
+      "step": 291
+    },
+    {
+      "epoch": 0.258922633562403,
+      "grad_norm": 0.5024058818817139,
+      "learning_rate": 9.91574295590541e-06,
+      "loss": 0.0105,
+      "step": 292
+    },
+    {
+      "epoch": 0.25980935491021945,
+      "grad_norm": 1.2302806377410889,
+      "learning_rate": 9.91486549841951e-06,
+      "loss": 0.0122,
+      "step": 293
+    },
+    {
+      "epoch": 0.2606960762580359,
+      "grad_norm": 0.4594174921512604,
+      "learning_rate": 9.913983534853703e-06,
+      "loss": 0.0101,
+      "step": 294
+    },
+    {
+      "epoch": 0.26158279760585235,
+      "grad_norm": 1.1169053316116333,
+      "learning_rate": 9.913097066016601e-06,
+      "loss": 0.0161,
+      "step": 295
+    },
+    {
+      "epoch": 0.2624695189536688,
+      "grad_norm": 0.719704806804657,
+      "learning_rate": 9.912206092720939e-06,
+      "loss": 0.008,
+      "step": 296
+    },
+    {
+      "epoch": 0.26335624030148524,
+      "grad_norm": 0.5716864466667175,
+      "learning_rate": 9.911310615783583e-06,
+      "loss": 0.022,
+      "step": 297
+    },
+    {
+      "epoch": 0.2642429616493017,
+      "grad_norm": 0.6044641137123108,
+      "learning_rate": 9.910410636025536e-06,
+      "loss": 0.0152,
+      "step": 298
+    },
+    {
+      "epoch": 0.26512968299711814,
+      "grad_norm": 0.4611899256706238,
+      "learning_rate": 9.909506154271917e-06,
+      "loss": 0.0105,
+      "step": 299
+    },
+    {
+      "epoch": 0.2660164043449346,
+      "grad_norm": 1.0533437728881836,
+      "learning_rate": 9.908597171351984e-06,
+      "loss": 0.0098,
+      "step": 300
+    },
+    {
+      "epoch": 0.26690312569275104,
+      "grad_norm": 0.5442478060722351,
+      "learning_rate": 9.907683688099114e-06,
+      "loss": 0.0069,
+      "step": 301
+    },
+    {
+      "epoch": 0.2677898470405675,
+      "grad_norm": 0.6688672304153442,
+      "learning_rate": 9.906765705350814e-06,
+      "loss": 0.0143,
+      "step": 302
+    },
+    {
+      "epoch": 0.26867656838838394,
+      "grad_norm": 0.7143788933753967,
+      "learning_rate": 9.905843223948715e-06,
+      "loss": 0.0138,
+      "step": 303
+    },
+    {
+      "epoch": 0.2695632897362004,
+      "grad_norm": 1.6420787572860718,
+      "learning_rate": 9.904916244738572e-06,
+      "loss": 0.0206,
+      "step": 304
+    },
+    {
+      "epoch": 0.27045001108401684,
+      "grad_norm": 0.9799435138702393,
+      "learning_rate": 9.903984768570264e-06,
+      "loss": 0.0118,
+      "step": 305
+    },
+    {
+      "epoch": 0.2713367324318333,
+      "grad_norm": 1.430992841720581,
+      "learning_rate": 9.903048796297794e-06,
+      "loss": 0.0176,
+      "step": 306
+    },
+    {
+      "epoch": 0.27222345377964974,
+      "grad_norm": 6.0022125244140625,
+      "learning_rate": 9.902108328779287e-06,
+      "loss": 0.0162,
+      "step": 307
+    },
+    {
+      "epoch": 0.2731101751274662,
+      "grad_norm": 0.5858060717582703,
+      "learning_rate": 9.90116336687699e-06,
+      "loss": 0.0112,
+      "step": 308
+    },
+    {
+      "epoch": 0.27399689647528264,
+      "grad_norm": 0.7073821425437927,
+      "learning_rate": 9.900213911457263e-06,
+      "loss": 0.0156,
+      "step": 309
+    },
+    {
+      "epoch": 0.2748836178230991,
+      "grad_norm": 0.3150938153266907,
+      "learning_rate": 9.8992599633906e-06,
+      "loss": 0.0085,
+      "step": 310
+    },
+    {
+      "epoch": 0.27577033917091553,
+      "grad_norm": 0.5376273989677429,
+      "learning_rate": 9.898301523551601e-06,
+      "loss": 0.0115,
+      "step": 311
+    },
+    {
+      "epoch": 0.276657060518732,
+      "grad_norm": 0.32037392258644104,
+      "learning_rate": 9.89733859281899e-06,
+      "loss": 0.0046,
+      "step": 312
+    },
+    {
+      "epoch": 0.27754378186654843,
+      "grad_norm": 0.9390332698822021,
+      "learning_rate": 9.89637117207561e-06,
+      "loss": 0.0109,
+      "step": 313
+    },
+    {
+      "epoch": 0.2784305032143649,
+      "grad_norm": 0.5948623418807983,
+      "learning_rate": 9.895399262208414e-06,
+      "loss": 0.0051,
+      "step": 314
+    },
+    {
+      "epoch": 0.27931722456218133,
+      "grad_norm": 0.5544528365135193,
+      "learning_rate": 9.894422864108479e-06,
+      "loss": 0.0132,
+      "step": 315
+    },
+    {
+      "epoch": 0.2802039459099978,
+      "grad_norm": 0.9701108336448669,
+      "learning_rate": 9.893441978670993e-06,
+      "loss": 0.0088,
+      "step": 316
+    },
+    {
+      "epoch": 0.28109066725781423,
+      "grad_norm": 0.6148018836975098,
+      "learning_rate": 9.892456606795254e-06,
+      "loss": 0.0048,
+      "step": 317
+    },
+    {
+      "epoch": 0.2819773886056307,
+      "grad_norm": 0.563202440738678,
+      "learning_rate": 9.891466749384679e-06,
+      "loss": 0.0115,
+      "step": 318
+    },
+    {
+      "epoch": 0.28286410995344713,
+      "grad_norm": 0.6775363683700562,
+      "learning_rate": 9.890472407346796e-06,
+      "loss": 0.007,
+      "step": 319
+    },
+    {
+      "epoch": 0.28375083130126355,
+      "grad_norm": 0.820683479309082,
+      "learning_rate": 9.889473581593245e-06,
+      "loss": 0.0162,
+      "step": 320
+    },
+    {
+      "epoch": 0.28463755264908003,
+      "grad_norm": 1.3535810708999634,
+      "learning_rate": 9.888470273039776e-06,
+      "loss": 0.016,
+      "step": 321
+    },
+    {
+      "epoch": 0.28552427399689645,
+      "grad_norm": 0.42245036363601685,
+      "learning_rate": 9.88746248260625e-06,
+      "loss": 0.0038,
+      "step": 322
+    },
+    {
+      "epoch": 0.2864109953447129,
+      "grad_norm": 0.38233160972595215,
+      "learning_rate": 9.886450211216631e-06,
+      "loss": 0.002,
+      "step": 323
+    },
+    {
+      "epoch": 0.28729771669252935,
+      "grad_norm": 1.296417236328125,
+      "learning_rate": 9.885433459799003e-06,
+      "loss": 0.0111,
+      "step": 324
+    },
+    {
+      "epoch": 0.2881844380403458,
+      "grad_norm": 1.1593570709228516,
+      "learning_rate": 9.884412229285547e-06,
+      "loss": 0.0101,
+      "step": 325
+    },
+    {
+      "epoch": 0.28907115938816225,
+      "grad_norm": 0.44105803966522217,
+      "learning_rate": 9.883386520612556e-06,
+      "loss": 0.0113,
+      "step": 326
+    },
+    {
+      "epoch": 0.2899578807359787,
+      "grad_norm": 1.3377820253372192,
+      "learning_rate": 9.882356334720426e-06,
+      "loss": 0.0083,
+      "step": 327
+    },
+    {
+      "epoch": 0.29084460208379515,
+      "grad_norm": 1.594089150428772,
+      "learning_rate": 9.881321672553661e-06,
+      "loss": 0.0242,
+      "step": 328
+    },
+    {
+      "epoch": 0.2917313234316116,
+      "grad_norm": 0.6307036876678467,
+      "learning_rate": 9.880282535060863e-06,
+      "loss": 0.015,
+      "step": 329
+    },
+    {
+      "epoch": 0.29261804477942804,
+      "grad_norm": 1.0397772789001465,
+      "learning_rate": 9.879238923194746e-06,
+      "loss": 0.0209,
+      "step": 330
+    },
+    {
+      "epoch": 0.2935047661272445,
+      "grad_norm": 1.5129303932189941,
+      "learning_rate": 9.878190837912118e-06,
+      "loss": 0.0241,
+      "step": 331
+    },
+    {
+      "epoch": 0.29439148747506094,
+      "grad_norm": 1.1227725744247437,
+      "learning_rate": 9.87713828017389e-06,
+      "loss": 0.0224,
+      "step": 332
+    },
+    {
+      "epoch": 0.2952782088228774,
+      "grad_norm": 1.045912504196167,
+      "learning_rate": 9.87608125094508e-06,
+      "loss": 0.0106,
+      "step": 333
+    },
+    {
+      "epoch": 0.29616493017069384,
+      "grad_norm": 0.5505847930908203,
+      "learning_rate": 9.875019751194796e-06,
+      "loss": 0.0078,
+      "step": 334
+    },
+    {
+      "epoch": 0.2970516515185103,
+      "grad_norm": 1.4759756326675415,
+      "learning_rate": 9.873953781896251e-06,
+      "loss": 0.0229,
+      "step": 335
+    },
+    {
+      "epoch": 0.29793837286632674,
+      "grad_norm": 1.1660606861114502,
+      "learning_rate": 9.872883344026755e-06,
+      "loss": 0.0147,
+      "step": 336
+    },
+    {
+      "epoch": 0.2988250942141432,
+      "grad_norm": 0.9158754944801331,
+      "learning_rate": 9.871808438567713e-06,
+      "loss": 0.01,
+      "step": 337
+    },
+    {
+      "epoch": 0.29971181556195964,
+      "grad_norm": 0.3273065388202667,
+      "learning_rate": 9.870729066504629e-06,
+      "loss": 0.005,
+      "step": 338
+    },
+    {
+      "epoch": 0.3005985369097761,
+      "grad_norm": 0.857749342918396,
+      "learning_rate": 9.869645228827097e-06,
+      "loss": 0.0219,
+      "step": 339
+    },
+    {
+      "epoch": 0.30148525825759254,
+      "grad_norm": 1.448362112045288,
+      "learning_rate": 9.868556926528809e-06,
+      "loss": 0.0227,
+      "step": 340
+    },
+    {
+      "epoch": 0.302371979605409,
+      "grad_norm": 1.0792773962020874,
+      "learning_rate": 9.867464160607552e-06,
+      "loss": 0.0189,
+      "step": 341
+    },
+    {
+      "epoch": 0.30325870095322544,
+      "grad_norm": 0.3688802421092987,
+      "learning_rate": 9.866366932065199e-06,
+      "loss": 0.0032,
+      "step": 342
+    },
+    {
+      "epoch": 0.3041454223010419,
+      "grad_norm": 0.6941332221031189,
+      "learning_rate": 9.865265241907722e-06,
+      "loss": 0.0066,
+      "step": 343
+    },
+    {
+      "epoch": 0.30503214364885833,
+      "grad_norm": 0.35826510190963745,
+      "learning_rate": 9.86415909114518e-06,
+      "loss": 0.0093,
+      "step": 344
+    },
+    {
+      "epoch": 0.3059188649966748,
+      "grad_norm": 0.45968568325042725,
+      "learning_rate": 9.863048480791718e-06,
+      "loss": 0.005,
+      "step": 345
+    },
+    {
+      "epoch": 0.30680558634449123,
+      "grad_norm": 0.7145997881889343,
+      "learning_rate": 9.861933411865576e-06,
+      "loss": 0.0139,
+      "step": 346
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 1.11654794216156,
+      "learning_rate": 9.86081388538908e-06,
+      "loss": 0.0064,
+      "step": 347
+    },
+    {
+      "epoch": 0.30857902904012413,
+      "grad_norm": 0.5863871574401855,
+      "learning_rate": 9.859689902388639e-06,
+      "loss": 0.0119,
+      "step": 348
+    },
+    {
+      "epoch": 0.3094657503879406,
+      "grad_norm": 0.5604485869407654,
+      "learning_rate": 9.858561463894752e-06,
+      "loss": 0.0046,
+      "step": 349
+    },
+    {
+      "epoch": 0.31035247173575703,
+      "grad_norm": 0.575162410736084,
+      "learning_rate": 9.857428570942e-06,
+      "loss": 0.0082,
+      "step": 350
+    },
+    {
+      "epoch": 0.3112391930835735,
+      "grad_norm": 0.6239818930625916,
+      "learning_rate": 9.856291224569052e-06,
+      "loss": 0.0104,
+      "step": 351
+    },
+    {
+      "epoch": 0.31212591443138993,
+      "grad_norm": 0.768072247505188,
+      "learning_rate": 9.855149425818657e-06,
+      "loss": 0.0113,
+      "step": 352
+    },
+    {
+      "epoch": 0.3130126357792064,
+      "grad_norm": 0.9411647915840149,
+      "learning_rate": 9.854003175737645e-06,
+      "loss": 0.0178,
+      "step": 353
+    },
+    {
+      "epoch": 0.3138993571270228,
+      "grad_norm": 0.3603442907333374,
+      "learning_rate": 9.85285247537693e-06,
+      "loss": 0.0067,
+      "step": 354
+    },
+    {
+      "epoch": 0.3147860784748393,
+      "grad_norm": 2.249364137649536,
+      "learning_rate": 9.851697325791505e-06,
+      "loss": 0.0368,
+      "step": 355
+    },
+    {
+      "epoch": 0.3156727998226557,
+      "grad_norm": 10.260486602783203,
+      "learning_rate": 9.850537728040442e-06,
+      "loss": 0.0724,
+      "step": 356
+    },
+    {
+      "epoch": 0.3165595211704722,
+      "grad_norm": 0.7737874984741211,
+      "learning_rate": 9.849373683186887e-06,
+      "loss": 0.0153,
+      "step": 357
+    },
+    {
+      "epoch": 0.3174462425182886,
+      "grad_norm": 0.9794998168945312,
+      "learning_rate": 9.848205192298075e-06,
+      "loss": 0.01,
+      "step": 358
+    },
+    {
+      "epoch": 0.3183329638661051,
+      "grad_norm": 0.3936000168323517,
+      "learning_rate": 9.847032256445303e-06,
+      "loss": 0.0078,
+      "step": 359
+    },
+    {
+      "epoch": 0.3192196852139215,
+      "grad_norm": 1.2881196737289429,
+      "learning_rate": 9.845854876703955e-06,
+      "loss": 0.0279,
+      "step": 360
+    },
+    {
+      "epoch": 0.320106406561738,
+      "grad_norm": 0.8685862421989441,
+      "learning_rate": 9.84467305415348e-06,
+      "loss": 0.0194,
+      "step": 361
+    },
+    {
+      "epoch": 0.3209931279095544,
+      "grad_norm": 0.48143163323402405,
+      "learning_rate": 9.843486789877405e-06,
+      "loss": 0.0098,
+      "step": 362
+    },
+    {
+      "epoch": 0.3218798492573709,
+      "grad_norm": 0.6279829144477844,
+      "learning_rate": 9.842296084963332e-06,
+      "loss": 0.0139,
+      "step": 363
+    },
+    {
+      "epoch": 0.3227665706051873,
+      "grad_norm": 0.955162763595581,
+      "learning_rate": 9.84110094050293e-06,
+      "loss": 0.0157,
+      "step": 364
+    },
+    {
+      "epoch": 0.32365329195300374,
+      "grad_norm": 0.7307074666023254,
+      "learning_rate": 9.839901357591935e-06,
+      "loss": 0.0095,
+      "step": 365
+    },
+    {
+      "epoch": 0.3245400133008202,
+      "grad_norm": 2.4147238731384277,
+      "learning_rate": 9.838697337330163e-06,
+      "loss": 0.0207,
+      "step": 366
+    },
+    {
+      "epoch": 0.32542673464863664,
+      "grad_norm": 0.3981778025627136,
+      "learning_rate": 9.837488880821487e-06,
+      "loss": 0.0051,
+      "step": 367
+    },
+    {
+      "epoch": 0.3263134559964531,
+      "grad_norm": 0.6395242810249329,
+      "learning_rate": 9.836275989173855e-06,
+      "loss": 0.0132,
+      "step": 368
+    },
+    {
+      "epoch": 0.32720017734426954,
+      "grad_norm": 0.3053138256072998,
+      "learning_rate": 9.835058663499277e-06,
+      "loss": 0.0073,
+      "step": 369
+    },
+    {
+      "epoch": 0.328086898692086,
+      "grad_norm": 1.085937261581421,
+      "learning_rate": 9.833836904913831e-06,
+      "loss": 0.011,
+      "step": 370
+    },
+    {
+      "epoch": 0.32897362003990244,
+      "grad_norm": 0.8119531273841858,
+      "learning_rate": 9.832610714537655e-06,
+      "loss": 0.01,
+      "step": 371
+    },
+    {
+      "epoch": 0.3298603413877189,
+      "grad_norm": 0.6228383183479309,
+      "learning_rate": 9.831380093494957e-06,
+      "loss": 0.0047,
+      "step": 372
+    },
+    {
+      "epoch": 0.33074706273553534,
+      "grad_norm": 0.8070245385169983,
+      "learning_rate": 9.830145042914003e-06,
+      "loss": 0.0167,
+      "step": 373
+    },
+    {
+      "epoch": 0.3316337840833518,
+      "grad_norm": 0.6485422849655151,
+      "learning_rate": 9.828905563927117e-06,
+      "loss": 0.015,
+      "step": 374
+    },
+    {
+      "epoch": 0.33252050543116823,
+      "grad_norm": 1.153606653213501,
+      "learning_rate": 9.82766165767069e-06,
+      "loss": 0.0068,
+      "step": 375
+    },
+    {
+      "epoch": 0.3334072267789847,
+      "grad_norm": 1.0364385843276978,
+      "learning_rate": 9.826413325285162e-06,
+      "loss": 0.0159,
+      "step": 376
+    },
+    {
+      "epoch": 0.33429394812680113,
+      "grad_norm": 0.6059282422065735,
+      "learning_rate": 9.825160567915045e-06,
+      "loss": 0.0057,
+      "step": 377
+    },
+    {
+      "epoch": 0.3351806694746176,
+      "grad_norm": 1.1247456073760986,
+      "learning_rate": 9.823903386708897e-06,
+      "loss": 0.0119,
+      "step": 378
+    },
+    {
+      "epoch": 0.33606739082243403,
+      "grad_norm": 1.2303680181503296,
+      "learning_rate": 9.822641782819337e-06,
+      "loss": 0.0223,
+      "step": 379
+    },
+    {
+      "epoch": 0.3369541121702505,
+      "grad_norm": 0.7362028956413269,
+      "learning_rate": 9.821375757403035e-06,
+      "loss": 0.0143,
+      "step": 380
+    },
+    {
+      "epoch": 0.33784083351806693,
+      "grad_norm": 0.8117135763168335,
+      "learning_rate": 9.820105311620717e-06,
+      "loss": 0.0153,
+      "step": 381
+    },
+    {
+      "epoch": 0.3387275548658834,
+      "grad_norm": 0.6031385064125061,
+      "learning_rate": 9.818830446637165e-06,
+      "loss": 0.0177,
+      "step": 382
+    },
+    {
+      "epoch": 0.33961427621369983,
+      "grad_norm": 0.5231031179428101,
+      "learning_rate": 9.817551163621205e-06,
+      "loss": 0.0117,
+      "step": 383
+    },
+    {
+      "epoch": 0.3405009975615163,
+      "grad_norm": 0.6594051122665405,
+      "learning_rate": 9.81626746374572e-06,
+      "loss": 0.0141,
+      "step": 384
+    },
+    {
+      "epoch": 0.3413877189093327,
+      "grad_norm": 0.4554224908351898,
+      "learning_rate": 9.814979348187639e-06,
+      "loss": 0.0128,
+      "step": 385
+    },
+    {
+      "epoch": 0.3422744402571492,
+      "grad_norm": 0.5283925533294678,
+      "learning_rate": 9.813686818127943e-06,
+      "loss": 0.0115,
+      "step": 386
+    },
+    {
+      "epoch": 0.3431611616049656,
+      "grad_norm": 0.5425133109092712,
+      "learning_rate": 9.812389874751656e-06,
+      "loss": 0.0115,
+      "step": 387
+    },
+    {
+      "epoch": 0.3440478829527821,
+      "grad_norm": 0.5007458925247192,
+      "learning_rate": 9.81108851924785e-06,
+      "loss": 0.0115,
+      "step": 388
+    },
+    {
+      "epoch": 0.3449346043005985,
+      "grad_norm": 0.5451713800430298,
+      "learning_rate": 9.809782752809644e-06,
+      "loss": 0.0093,
+      "step": 389
+    },
+    {
+      "epoch": 0.345821325648415,
+      "grad_norm": 0.6177766919136047,
+      "learning_rate": 9.808472576634199e-06,
+      "loss": 0.0087,
+      "step": 390
+    },
+    {
+      "epoch": 0.3467080469962314,
+      "grad_norm": 0.5689627528190613,
+      "learning_rate": 9.80715799192272e-06,
+      "loss": 0.0125,
+      "step": 391
+    },
+    {
+      "epoch": 0.3475947683440479,
+      "grad_norm": 0.36475062370300293,
+      "learning_rate": 9.805838999880453e-06,
+      "loss": 0.0071,
+      "step": 392
+    },
+    {
+      "epoch": 0.3484814896918643,
+      "grad_norm": 0.7817016839981079,
+      "learning_rate": 9.804515601716685e-06,
+      "loss": 0.0151,
+      "step": 393
+    },
+    {
+      "epoch": 0.3493682110396808,
+      "grad_norm": 0.42941737174987793,
+      "learning_rate": 9.80318779864474e-06,
+      "loss": 0.0038,
+      "step": 394
+    },
+    {
+      "epoch": 0.3502549323874972,
+      "grad_norm": 0.5236615538597107,
+      "learning_rate": 9.801855591881989e-06,
+      "loss": 0.0089,
+      "step": 395
+    },
+    {
+      "epoch": 0.3511416537353137,
+      "grad_norm": 0.4579971134662628,
+      "learning_rate": 9.80051898264983e-06,
+      "loss": 0.0094,
+      "step": 396
+    },
+    {
+      "epoch": 0.3520283750831301,
+      "grad_norm": 0.9356654286384583,
+      "learning_rate": 9.799177972173706e-06,
+      "loss": 0.0188,
+      "step": 397
+    },
+    {
+      "epoch": 0.3529150964309466,
+      "grad_norm": 0.664996325969696,
+      "learning_rate": 9.797832561683087e-06,
+      "loss": 0.016,
+      "step": 398
+    },
+    {
+      "epoch": 0.353801817778763,
+      "grad_norm": 0.8114444613456726,
+      "learning_rate": 9.796482752411482e-06,
+      "loss": 0.0101,
+      "step": 399
+    },
+    {
+      "epoch": 0.3546885391265795,
+      "grad_norm": 0.5470486283302307,
+      "learning_rate": 9.795128545596436e-06,
+      "loss": 0.0082,
+      "step": 400
+    },
+    {
+      "epoch": 0.3555752604743959,
+      "grad_norm": 0.716001570224762,
+      "learning_rate": 9.793769942479518e-06,
+      "loss": 0.0154,
+      "step": 401
+    },
+    {
+      "epoch": 0.3564619818222124,
+      "grad_norm": 0.8142730593681335,
+      "learning_rate": 9.792406944306334e-06,
+      "loss": 0.0098,
+      "step": 402
+    },
+    {
+      "epoch": 0.3573487031700288,
+      "grad_norm": 0.8549275398254395,
+      "learning_rate": 9.791039552326515e-06,
+      "loss": 0.0065,
+      "step": 403
+    },
+    {
+      "epoch": 0.3582354245178453,
+      "grad_norm": 0.5257267951965332,
+      "learning_rate": 9.789667767793725e-06,
+      "loss": 0.015,
+      "step": 404
+    },
+    {
+      "epoch": 0.3591221458656617,
+      "grad_norm": 0.48406821489334106,
+      "learning_rate": 9.788291591965651e-06,
+      "loss": 0.0041,
+      "step": 405
+    },
+    {
+      "epoch": 0.3600088672134782,
+      "grad_norm": 1.1776056289672852,
+      "learning_rate": 9.786911026104007e-06,
+      "loss": 0.0137,
+      "step": 406
+    },
+    {
+      "epoch": 0.3608955885612946,
+      "grad_norm": 0.8189646005630493,
+      "learning_rate": 9.785526071474533e-06,
+      "loss": 0.0187,
+      "step": 407
+    },
+    {
+      "epoch": 0.3617823099091111,
+      "grad_norm": 1.4497982263565063,
+      "learning_rate": 9.784136729346994e-06,
+      "loss": 0.0203,
+      "step": 408
+    },
+    {
+      "epoch": 0.3626690312569275,
+      "grad_norm": 0.32031527161598206,
+      "learning_rate": 9.782743000995175e-06,
+      "loss": 0.0052,
+      "step": 409
+    },
+    {
+      "epoch": 0.363555752604744,
+      "grad_norm": 0.3057740330696106,
+      "learning_rate": 9.781344887696884e-06,
+      "loss": 0.0074,
+      "step": 410
+    },
+    {
+      "epoch": 0.3644424739525604,
+      "grad_norm": 0.6097119450569153,
+      "learning_rate": 9.779942390733948e-06,
+      "loss": 0.0124,
+      "step": 411
+    },
+    {
+      "epoch": 0.36532919530037683,
+      "grad_norm": 0.25449925661087036,
+      "learning_rate": 9.77853551139221e-06,
+      "loss": 0.0053,
+      "step": 412
+    },
+    {
+      "epoch": 0.3662159166481933,
+      "grad_norm": 0.43352335691452026,
+      "learning_rate": 9.777124250961543e-06,
+      "loss": 0.0094,
+      "step": 413
+    },
+    {
+      "epoch": 0.36710263799600973,
+      "grad_norm": 0.5348715782165527,
+      "learning_rate": 9.775708610735821e-06,
+      "loss": 0.0081,
+      "step": 414
+    },
+    {
+      "epoch": 0.3679893593438262,
+      "grad_norm": 0.3730040192604065,
+      "learning_rate": 9.774288592012944e-06,
+      "loss": 0.0052,
+      "step": 415
+    },
+    {
+      "epoch": 0.3688760806916426,
+      "grad_norm": 0.6741971373558044,
+      "learning_rate": 9.77286419609482e-06,
+      "loss": 0.0112,
+      "step": 416
+    },
+    {
+      "epoch": 0.3697628020394591,
+      "grad_norm": 0.6674507260322571,
+      "learning_rate": 9.771435424287377e-06,
+      "loss": 0.0128,
+      "step": 417
+    },
+    {
+      "epoch": 0.3706495233872755,
+      "grad_norm": 0.7770352363586426,
+      "learning_rate": 9.77000227790055e-06,
+      "loss": 0.0139,
+      "step": 418
+    },
+    {
+      "epoch": 0.371536244735092,
+      "grad_norm": 0.6249715089797974,
+      "learning_rate": 9.768564758248285e-06,
+      "loss": 0.0141,
+      "step": 419
+    },
+    {
+      "epoch": 0.3724229660829084,
+      "grad_norm": 0.42244604229927063,
+      "learning_rate": 9.76712286664854e-06,
+      "loss": 0.0047,
+      "step": 420
+    },
+    {
+      "epoch": 0.3733096874307249,
+      "grad_norm": 0.6065423488616943,
+      "learning_rate": 9.765676604423277e-06,
+      "loss": 0.0128,
+      "step": 421
+    },
+    {
+      "epoch": 0.3741964087785413,
+      "grad_norm": 0.9267554879188538,
+      "learning_rate": 9.76422597289847e-06,
+      "loss": 0.0172,
+      "step": 422
+    },
+    {
+      "epoch": 0.3750831301263578,
+      "grad_norm": 0.6917939782142639,
+      "learning_rate": 9.762770973404094e-06,
+      "loss": 0.0138,
+      "step": 423
+    },
+    {
+      "epoch": 0.3759698514741742,
+      "grad_norm": 0.4068993330001831,
+      "learning_rate": 9.761311607274135e-06,
+      "loss": 0.0036,
+      "step": 424
+    },
+    {
+      "epoch": 0.3768565728219907,
+      "grad_norm": 1.279029369354248,
+      "learning_rate": 9.759847875846578e-06,
+      "loss": 0.0099,
+      "step": 425
+    },
+    {
+      "epoch": 0.3777432941698071,
+      "grad_norm": 0.6679139137268066,
+      "learning_rate": 9.758379780463409e-06,
+      "loss": 0.0067,
+      "step": 426
+    },
+    {
+      "epoch": 0.3786300155176236,
+      "grad_norm": 1.1384234428405762,
+      "learning_rate": 9.756907322470619e-06,
+      "loss": 0.0142,
+      "step": 427
+    },
+    {
+      "epoch": 0.37951673686544,
+      "grad_norm": 0.9144220948219299,
+      "learning_rate": 9.755430503218197e-06,
+      "loss": 0.0088,
+      "step": 428
+    },
+    {
+      "epoch": 0.3804034582132565,
+      "grad_norm": 0.9335492849349976,
+      "learning_rate": 9.753949324060127e-06,
+      "loss": 0.0129,
+      "step": 429
+    },
+    {
+      "epoch": 0.3812901795610729,
+      "grad_norm": 0.5396351218223572,
+      "learning_rate": 9.7524637863544e-06,
+      "loss": 0.0096,
+      "step": 430
+    },
+    {
+      "epoch": 0.3821769009088894,
+      "grad_norm": 0.5314906239509583,
+      "learning_rate": 9.75097389146299e-06,
+      "loss": 0.0097,
+      "step": 431
+    },
+    {
+      "epoch": 0.3830636222567058,
+      "grad_norm": 0.6313880681991577,
+      "learning_rate": 9.749479640751876e-06,
+      "loss": 0.0095,
+      "step": 432
+    },
+    {
+      "epoch": 0.3839503436045223,
+      "grad_norm": 0.955323338508606,
+      "learning_rate": 9.747981035591025e-06,
+      "loss": 0.0121,
+      "step": 433
+    },
+    {
+      "epoch": 0.3848370649523387,
+      "grad_norm": 1.0209813117980957,
+      "learning_rate": 9.7464780773544e-06,
+      "loss": 0.0126,
+      "step": 434
+    },
+    {
+      "epoch": 0.3857237863001552,
+      "grad_norm": 0.4658578038215637,
+      "learning_rate": 9.744970767419952e-06,
+      "loss": 0.0082,
+      "step": 435
+    },
+    {
+      "epoch": 0.3866105076479716,
+      "grad_norm": 0.6823182702064514,
+      "learning_rate": 9.743459107169624e-06,
+      "loss": 0.013,
+      "step": 436
+    },
+    {
+      "epoch": 0.3874972289957881,
+      "grad_norm": 0.2446506917476654,
+      "learning_rate": 9.741943097989345e-06,
+      "loss": 0.0038,
+      "step": 437
+    },
+    {
+      "epoch": 0.3883839503436045,
+      "grad_norm": 0.42889147996902466,
+      "learning_rate": 9.740422741269035e-06,
+      "loss": 0.0099,
+      "step": 438
+    },
+    {
+      "epoch": 0.389270671691421,
+      "grad_norm": 0.30183982849121094,
+      "learning_rate": 9.738898038402597e-06,
+      "loss": 0.0064,
+      "step": 439
+    },
+    {
+      "epoch": 0.3901573930392374,
+      "grad_norm": 1.023352861404419,
+      "learning_rate": 9.737368990787917e-06,
+      "loss": 0.0071,
+      "step": 440
+    },
+    {
+      "epoch": 0.3910441143870539,
+      "grad_norm": 0.39613592624664307,
+      "learning_rate": 9.735835599826868e-06,
+      "loss": 0.0062,
+      "step": 441
+    },
+    {
+      "epoch": 0.3919308357348703,
+      "grad_norm": 1.0003458261489868,
+      "learning_rate": 9.734297866925305e-06,
+      "loss": 0.0076,
+      "step": 442
+    },
+    {
+      "epoch": 0.3928175570826868,
+      "grad_norm": 0.4582412838935852,
+      "learning_rate": 9.732755793493062e-06,
+      "loss": 0.008,
+      "step": 443
+    },
+    {
+      "epoch": 0.3937042784305032,
+      "grad_norm": 0.969808042049408,
+      "learning_rate": 9.731209380943951e-06,
+      "loss": 0.0152,
+      "step": 444
+    },
+    {
+      "epoch": 0.3945909997783197,
+      "grad_norm": 0.38053014874458313,
+      "learning_rate": 9.729658630695768e-06,
+      "loss": 0.0033,
+      "step": 445
+    },
+    {
+      "epoch": 0.3954777211261361,
+      "grad_norm": 0.8101393580436707,
+      "learning_rate": 9.72810354417028e-06,
+      "loss": 0.0111,
+      "step": 446
+    },
+    {
+      "epoch": 0.3963644424739526,
+      "grad_norm": 0.7292132377624512,
+      "learning_rate": 9.72654412279323e-06,
+      "loss": 0.0086,
+      "step": 447
+    },
+    {
+      "epoch": 0.397251163821769,
+      "grad_norm": 0.44195911288261414,
+      "learning_rate": 9.724980367994341e-06,
+      "loss": 0.0058,
+      "step": 448
+    },
+    {
+      "epoch": 0.3981378851695855,
+      "grad_norm": 0.2693403959274292,
+      "learning_rate": 9.723412281207304e-06,
+      "loss": 0.0019,
+      "step": 449
+    },
+    {
+      "epoch": 0.3990246065174019,
+      "grad_norm": 0.2798308730125427,
+      "learning_rate": 9.721839863869783e-06,
+      "loss": 0.0051,
+      "step": 450
+    },
+    {
+      "epoch": 0.3999113278652184,
+      "grad_norm": 0.8944824934005737,
+      "learning_rate": 9.720263117423408e-06,
+      "loss": 0.0111,
+      "step": 451
+    },
+    {
+      "epoch": 0.4007980492130348,
+      "grad_norm": 0.4507511258125305,
+      "learning_rate": 9.718682043313787e-06,
+      "loss": 0.0084,
+      "step": 452
+    },
+    {
+      "epoch": 0.4016847705608513,
+      "grad_norm": 0.4930645823478699,
+      "learning_rate": 9.71709664299049e-06,
+      "loss": 0.0149,
+      "step": 453
+    },
+    {
+      "epoch": 0.4025714919086677,
+      "grad_norm": 1.0915964841842651,
+      "learning_rate": 9.715506917907055e-06,
+      "loss": 0.0116,
+      "step": 454
+    },
+    {
+      "epoch": 0.4034582132564842,
+      "grad_norm": 0.45030245184898376,
+      "learning_rate": 9.713912869520982e-06,
+      "loss": 0.0067,
+      "step": 455
+    },
+    {
+      "epoch": 0.4043449346043006,
+      "grad_norm": 0.6651377081871033,
+      "learning_rate": 9.71231449929374e-06,
+      "loss": 0.0071,
+      "step": 456
+    },
+    {
+      "epoch": 0.405231655952117,
+      "grad_norm": 0.5960193276405334,
+      "learning_rate": 9.710711808690754e-06,
+      "loss": 0.0066,
+      "step": 457
+    },
+    {
+      "epoch": 0.4061183772999335,
+      "grad_norm": 0.290302038192749,
+      "learning_rate": 9.709104799181418e-06,
+      "loss": 0.0046,
+      "step": 458
+    },
+    {
+      "epoch": 0.4070050986477499,
+      "grad_norm": 0.7967752814292908,
+      "learning_rate": 9.707493472239075e-06,
+      "loss": 0.0062,
+      "step": 459
+    },
+    {
+      "epoch": 0.4078918199955664,
+      "grad_norm": 0.759286642074585,
+      "learning_rate": 9.70587782934104e-06,
+      "loss": 0.0184,
+      "step": 460
+    },
+    {
+      "epoch": 0.4087785413433828,
+      "grad_norm": 0.5127438306808472,
+      "learning_rate": 9.704257871968573e-06,
+      "loss": 0.0117,
+      "step": 461
+    },
+    {
+      "epoch": 0.4096652626911993,
+      "grad_norm": 0.4300234913825989,
+      "learning_rate": 9.702633601606893e-06,
+      "loss": 0.0052,
+      "step": 462
+    },
+    {
+      "epoch": 0.4105519840390157,
+      "grad_norm": 0.3789842128753662,
+      "learning_rate": 9.701005019745178e-06,
+      "loss": 0.0087,
+      "step": 463
+    },
+    {
+      "epoch": 0.4114387053868322,
+      "grad_norm": 0.6187935471534729,
+      "learning_rate": 9.699372127876555e-06,
+      "loss": 0.0104,
+      "step": 464
+    },
+    {
+      "epoch": 0.4123254267346486,
+      "grad_norm": 0.8972629904747009,
+      "learning_rate": 9.697734927498099e-06,
+      "loss": 0.0253,
+      "step": 465
+    },
+    {
+      "epoch": 0.4132121480824651,
+      "grad_norm": 0.24438129365444183,
+      "learning_rate": 9.696093420110843e-06,
+      "loss": 0.0052,
+      "step": 466
+    },
+    {
+      "epoch": 0.4140988694302815,
+      "grad_norm": 1.079038381576538,
+      "learning_rate": 9.694447607219763e-06,
+      "loss": 0.01,
+      "step": 467
+    },
+    {
+      "epoch": 0.414985590778098,
+      "grad_norm": 0.382589727640152,
+      "learning_rate": 9.692797490333785e-06,
+      "loss": 0.0049,
+      "step": 468
+    },
+    {
+      "epoch": 0.4158723121259144,
+      "grad_norm": 0.7200171947479248,
+      "learning_rate": 9.69114307096578e-06,
+      "loss": 0.0095,
+      "step": 469
+    },
+    {
+      "epoch": 0.4167590334737309,
+      "grad_norm": 0.6705244779586792,
+      "learning_rate": 9.689484350632564e-06,
+      "loss": 0.0085,
+      "step": 470
+    },
+    {
+      "epoch": 0.4176457548215473,
+      "grad_norm": 0.5431262254714966,
+      "learning_rate": 9.687821330854894e-06,
+      "loss": 0.0071,
+      "step": 471
+    },
+    {
+      "epoch": 0.4185324761693638,
+      "grad_norm": 0.5441691875457764,
+      "learning_rate": 9.686154013157474e-06,
+      "loss": 0.009,
+      "step": 472
+    },
+    {
+      "epoch": 0.4194191975171802,
+      "grad_norm": 1.1173381805419922,
+      "learning_rate": 9.684482399068944e-06,
+      "loss": 0.0156,
+      "step": 473
+    },
+    {
+      "epoch": 0.4203059188649967,
+      "grad_norm": 1.0084078311920166,
+      "learning_rate": 9.682806490121886e-06,
+      "loss": 0.0102,
+      "step": 474
+    },
+    {
+      "epoch": 0.4211926402128131,
+      "grad_norm": 0.3242226243019104,
+      "learning_rate": 9.681126287852814e-06,
+      "loss": 0.0048,
+      "step": 475
+    },
+    {
+      "epoch": 0.4220793615606296,
+      "grad_norm": 1.2636985778808594,
+      "learning_rate": 9.679441793802186e-06,
+      "loss": 0.0172,
+      "step": 476
+    },
+    {
+      "epoch": 0.422966082908446,
+      "grad_norm": 0.8478817343711853,
+      "learning_rate": 9.67775300951439e-06,
+      "loss": 0.0134,
+      "step": 477
+    },
+    {
+      "epoch": 0.4238528042562625,
+      "grad_norm": 0.8426191806793213,
+      "learning_rate": 9.67605993653775e-06,
+      "loss": 0.0132,
+      "step": 478
+    },
+    {
+      "epoch": 0.4247395256040789,
+      "grad_norm": 0.661734938621521,
+      "learning_rate": 9.674362576424516e-06,
+      "loss": 0.0096,
+      "step": 479
+    },
+    {
+      "epoch": 0.4256262469518954,
+      "grad_norm": 0.6070396900177002,
+      "learning_rate": 9.67266093073088e-06,
+      "loss": 0.0167,
+      "step": 480
+    },
+    {
+      "epoch": 0.4265129682997118,
+      "grad_norm": 0.9344643950462341,
+      "learning_rate": 9.670955001016949e-06,
+      "loss": 0.0115,
+      "step": 481
+    },
+    {
+      "epoch": 0.4273996896475283,
+      "grad_norm": 0.843940019607544,
+      "learning_rate": 9.669244788846769e-06,
+      "loss": 0.0079,
+      "step": 482
+    },
+    {
+      "epoch": 0.4282864109953447,
+      "grad_norm": 0.26441389322280884,
+      "learning_rate": 9.667530295788307e-06,
+      "loss": 0.0025,
+      "step": 483
+    },
+    {
+      "epoch": 0.4291731323431612,
+      "grad_norm": 0.34116652607917786,
+      "learning_rate": 9.665811523413457e-06,
+      "loss": 0.0047,
+      "step": 484
+    },
+    {
+      "epoch": 0.4300598536909776,
+      "grad_norm": 0.37784484028816223,
+      "learning_rate": 9.664088473298035e-06,
+      "loss": 0.0086,
+      "step": 485
+    },
+    {
+      "epoch": 0.4309465750387941,
+      "grad_norm": 0.44266486167907715,
+      "learning_rate": 9.66236114702178e-06,
+      "loss": 0.014,
+      "step": 486
+    },
+    {
+      "epoch": 0.4318332963866105,
+      "grad_norm": 0.689039409160614,
+      "learning_rate": 9.66062954616835e-06,
+      "loss": 0.005,
+      "step": 487
+    },
+    {
+      "epoch": 0.432720017734427,
+      "grad_norm": 1.1525181531906128,
+      "learning_rate": 9.65889367232532e-06,
+      "loss": 0.0115,
+      "step": 488
+    },
+    {
+      "epoch": 0.4336067390822434,
+      "grad_norm": 0.36996787786483765,
+      "learning_rate": 9.657153527084191e-06,
+      "loss": 0.0064,
+      "step": 489
+    },
+    {
+      "epoch": 0.4344934604300599,
+      "grad_norm": 0.2661356031894684,
+      "learning_rate": 9.655409112040372e-06,
+      "loss": 0.004,
+      "step": 490
+    },
+    {
+      "epoch": 0.4353801817778763,
+      "grad_norm": 0.7092905640602112,
+      "learning_rate": 9.653660428793188e-06,
+      "loss": 0.0134,
+      "step": 491
+    },
+    {
+      "epoch": 0.4362669031256928,
+      "grad_norm": 0.9240937829017639,
+      "learning_rate": 9.651907478945882e-06,
+      "loss": 0.0128,
+      "step": 492
+    },
+    {
+      "epoch": 0.4371536244735092,
+      "grad_norm": 0.1982317417860031,
+      "learning_rate": 9.650150264105599e-06,
+      "loss": 0.0021,
+      "step": 493
+    },
+    {
+      "epoch": 0.43804034582132567,
+      "grad_norm": 0.790959358215332,
+      "learning_rate": 9.648388785883407e-06,
+      "loss": 0.011,
+      "step": 494
+    },
+    {
+      "epoch": 0.4389270671691421,
+      "grad_norm": 0.6713638305664062,
+      "learning_rate": 9.646623045894274e-06,
+      "loss": 0.0118,
+      "step": 495
+    },
+    {
+      "epoch": 0.43981378851695857,
+      "grad_norm": 0.5016664862632751,
+      "learning_rate": 9.644853045757077e-06,
+      "loss": 0.0055,
+      "step": 496
+    },
+    {
+      "epoch": 0.440700509864775,
+      "grad_norm": 0.5184798836708069,
+      "learning_rate": 9.6430787870946e-06,
+      "loss": 0.0159,
+      "step": 497
+    },
+    {
+      "epoch": 0.44158723121259147,
+      "grad_norm": 0.4628726541996002,
+      "learning_rate": 9.641300271533528e-06,
+      "loss": 0.0029,
+      "step": 498
+    },
+    {
+      "epoch": 0.4424739525604079,
+      "grad_norm": 0.5552617311477661,
+      "learning_rate": 9.639517500704456e-06,
+      "loss": 0.0075,
+      "step": 499
+    },
+    {
+      "epoch": 0.44336067390822437,
+      "grad_norm": 0.5783096551895142,
+      "learning_rate": 9.637730476241874e-06,
+      "loss": 0.0074,
+      "step": 500
+    },
+    {
+      "epoch": 0.4442473952560408,
+      "grad_norm": 0.46593576669692993,
+      "learning_rate": 9.635939199784173e-06,
+      "loss": 0.0039,
+      "step": 501
+    },
+    {
+      "epoch": 0.4451341166038572,
+      "grad_norm": 0.29269683361053467,
+      "learning_rate": 9.634143672973644e-06,
+      "loss": 0.0023,
+      "step": 502
+    },
+    {
+      "epoch": 0.4460208379516737,
+      "grad_norm": 0.41590166091918945,
+      "learning_rate": 9.632343897456473e-06,
+      "loss": 0.0125,
+      "step": 503
+    },
+    {
+      "epoch": 0.4469075592994901,
+      "grad_norm": 0.27357831597328186,
+      "learning_rate": 9.630539874882742e-06,
+      "loss": 0.0054,
+      "step": 504
+    },
+    {
+      "epoch": 0.4477942806473066,
+      "grad_norm": 0.602550745010376,
+      "learning_rate": 9.62873160690643e-06,
+      "loss": 0.0069,
+      "step": 505
+    },
+    {
+      "epoch": 0.448681001995123,
+      "grad_norm": 0.5262980461120605,
+      "learning_rate": 9.626919095185403e-06,
+      "loss": 0.0041,
+      "step": 506
+    },
+    {
+      "epoch": 0.4495677233429395,
+      "grad_norm": 0.5450761914253235,
+      "learning_rate": 9.625102341381418e-06,
+      "loss": 0.0089,
+      "step": 507
+    },
+    {
+      "epoch": 0.4504544446907559,
+      "grad_norm": 0.6254885792732239,
+      "learning_rate": 9.623281347160129e-06,
+      "loss": 0.0071,
+      "step": 508
+    },
+    {
+      "epoch": 0.4513411660385724,
+      "grad_norm": 0.4506608545780182,
+      "learning_rate": 9.621456114191067e-06,
+      "loss": 0.0051,
+      "step": 509
+    },
+    {
+      "epoch": 0.4522278873863888,
+      "grad_norm": 0.663757860660553,
+      "learning_rate": 9.619626644147655e-06,
+      "loss": 0.0071,
+      "step": 510
+    },
+    {
+      "epoch": 0.4531146087342053,
+      "grad_norm": 0.5022022128105164,
+      "learning_rate": 9.617792938707204e-06,
+      "loss": 0.0098,
+      "step": 511
+    },
+    {
+      "epoch": 0.4540013300820217,
+      "grad_norm": 0.46550077199935913,
+      "learning_rate": 9.615954999550899e-06,
+      "loss": 0.0047,
+      "step": 512
+    },
+    {
+      "epoch": 0.4548880514298382,
+      "grad_norm": 0.8453051447868347,
+      "learning_rate": 9.614112828363814e-06,
+      "loss": 0.0121,
+      "step": 513
+    },
+    {
+      "epoch": 0.4557747727776546,
+      "grad_norm": 0.44652843475341797,
+      "learning_rate": 9.612266426834901e-06,
+      "loss": 0.004,
+      "step": 514
+    },
+    {
+      "epoch": 0.4566614941254711,
+      "grad_norm": 0.6860093474388123,
+      "learning_rate": 9.61041579665699e-06,
+      "loss": 0.0059,
+      "step": 515
+    },
+    {
+      "epoch": 0.4575482154732875,
+      "grad_norm": 0.7249019742012024,
+      "learning_rate": 9.60856093952679e-06,
+      "loss": 0.0057,
+      "step": 516
+    },
+    {
+      "epoch": 0.458434936821104,
+      "grad_norm": 0.5832751393318176,
+      "learning_rate": 9.60670185714488e-06,
+      "loss": 0.0079,
+      "step": 517
+    },
+    {
+      "epoch": 0.4593216581689204,
+      "grad_norm": 0.7885969281196594,
+      "learning_rate": 9.604838551215715e-06,
+      "loss": 0.0071,
+      "step": 518
+    },
+    {
+      "epoch": 0.4602083795167369,
+      "grad_norm": 0.6885710954666138,
+      "learning_rate": 9.602971023447632e-06,
+      "loss": 0.0135,
+      "step": 519
+    },
+    {
+      "epoch": 0.4610951008645533,
+      "grad_norm": 1.1160459518432617,
+      "learning_rate": 9.60109927555282e-06,
+      "loss": 0.0111,
+      "step": 520
+    },
+    {
+      "epoch": 0.4619818222123698,
+      "grad_norm": 1.2547369003295898,
+      "learning_rate": 9.599223309247356e-06,
+      "loss": 0.0192,
+      "step": 521
+    },
+    {
+      "epoch": 0.4628685435601862,
+      "grad_norm": 0.7981187701225281,
+      "learning_rate": 9.59734312625117e-06,
+      "loss": 0.0046,
+      "step": 522
+    },
+    {
+      "epoch": 0.4637552649080027,
+      "grad_norm": 0.8385836482048035,
+      "learning_rate": 9.595458728288068e-06,
+      "loss": 0.01,
+      "step": 523
+    },
+    {
+      "epoch": 0.4646419862558191,
+      "grad_norm": 0.44900402426719666,
+      "learning_rate": 9.593570117085714e-06,
+      "loss": 0.0127,
+      "step": 524
+    },
+    {
+      "epoch": 0.46552870760363557,
+      "grad_norm": 0.8929427266120911,
+      "learning_rate": 9.591677294375637e-06,
+      "loss": 0.0075,
+      "step": 525
+    },
+    {
+      "epoch": 0.466415428951452,
+      "grad_norm": 0.3609806299209595,
+      "learning_rate": 9.589780261893229e-06,
+      "loss": 0.0055,
+      "step": 526
+    },
+    {
+      "epoch": 0.46730215029926847,
+      "grad_norm": 1.0652682781219482,
+      "learning_rate": 9.587879021377739e-06,
+      "loss": 0.0136,
+      "step": 527
+    },
+    {
+      "epoch": 0.4681888716470849,
+      "grad_norm": 0.5709818601608276,
+      "learning_rate": 9.585973574572276e-06,
+      "loss": 0.0068,
+      "step": 528
+    },
+    {
+      "epoch": 0.46907559299490137,
+      "grad_norm": 0.6579173803329468,
+      "learning_rate": 9.584063923223804e-06,
+      "loss": 0.0057,
+      "step": 529
+    },
+    {
+      "epoch": 0.4699623143427178,
+      "grad_norm": 0.2939516603946686,
+      "learning_rate": 9.582150069083141e-06,
+      "loss": 0.0063,
+      "step": 530
+    },
+    {
+      "epoch": 0.47084903569053427,
+      "grad_norm": 0.4318031072616577,
+      "learning_rate": 9.58023201390496e-06,
+      "loss": 0.0048,
+      "step": 531
+    },
+    {
+      "epoch": 0.4717357570383507,
+      "grad_norm": 0.5636759996414185,
+      "learning_rate": 9.578309759447786e-06,
+      "loss": 0.0109,
+      "step": 532
+    },
+    {
+      "epoch": 0.47262247838616717,
+      "grad_norm": 0.8026735186576843,
+      "learning_rate": 9.576383307473995e-06,
+      "loss": 0.0105,
+      "step": 533
+    },
+    {
+      "epoch": 0.4735091997339836,
+      "grad_norm": 0.6912361979484558,
+      "learning_rate": 9.574452659749807e-06,
+      "loss": 0.0052,
+      "step": 534
+    },
+    {
+      "epoch": 0.47439592108180006,
+      "grad_norm": 0.6018496751785278,
+      "learning_rate": 9.572517818045291e-06,
+      "loss": 0.0186,
+      "step": 535
+    },
+    {
+      "epoch": 0.4752826424296165,
+      "grad_norm": 0.4667069911956787,
+      "learning_rate": 9.570578784134364e-06,
+      "loss": 0.0062,
+      "step": 536
+    },
+    {
+      "epoch": 0.47616936377743296,
+      "grad_norm": 0.8536850810050964,
+      "learning_rate": 9.568635559794782e-06,
+      "loss": 0.0088,
+      "step": 537
+    },
+    {
+      "epoch": 0.4770560851252494,
+      "grad_norm": 0.7802487015724182,
+      "learning_rate": 9.566688146808147e-06,
+      "loss": 0.0141,
+      "step": 538
+    },
+    {
+      "epoch": 0.47794280647306586,
+      "grad_norm": 0.5833753347396851,
+      "learning_rate": 9.564736546959896e-06,
+      "loss": 0.0085,
+      "step": 539
+    },
+    {
+      "epoch": 0.4788295278208823,
+      "grad_norm": 0.6617295742034912,
+      "learning_rate": 9.56278076203931e-06,
+      "loss": 0.0074,
+      "step": 540
+    },
+    {
+      "epoch": 0.47971624916869876,
+      "grad_norm": 1.122158408164978,
+      "learning_rate": 9.560820793839505e-06,
+      "loss": 0.011,
+      "step": 541
+    },
+    {
+      "epoch": 0.4806029705165152,
+      "grad_norm": 0.49062931537628174,
+      "learning_rate": 9.558856644157432e-06,
+      "loss": 0.0065,
+      "step": 542
+    },
+    {
+      "epoch": 0.48148969186433166,
+      "grad_norm": 0.7522017955780029,
+      "learning_rate": 9.556888314793875e-06,
+      "loss": 0.0126,
+      "step": 543
+    },
+    {
+      "epoch": 0.4823764132121481,
+      "grad_norm": 1.2089123725891113,
+      "learning_rate": 9.554915807553451e-06,
+      "loss": 0.007,
+      "step": 544
+    },
+    {
+      "epoch": 0.48326313455996456,
+      "grad_norm": 0.5262856483459473,
+      "learning_rate": 9.552939124244608e-06,
+      "loss": 0.0071,
+      "step": 545
+    },
+    {
+      "epoch": 0.484149855907781,
+      "grad_norm": 0.5715947151184082,
+      "learning_rate": 9.550958266679623e-06,
+      "loss": 0.0052,
+      "step": 546
+    },
+    {
+      "epoch": 0.4850365772555974,
+      "grad_norm": 0.6762679815292358,
+      "learning_rate": 9.548973236674598e-06,
+      "loss": 0.0089,
+      "step": 547
+    },
+    {
+      "epoch": 0.4859232986034139,
+      "grad_norm": 0.4889708161354065,
+      "learning_rate": 9.546984036049461e-06,
+      "loss": 0.0063,
+      "step": 548
+    },
+    {
+      "epoch": 0.4868100199512303,
+      "grad_norm": 0.41651228070259094,
+      "learning_rate": 9.544990666627964e-06,
+      "loss": 0.0129,
+      "step": 549
+    },
+    {
+      "epoch": 0.4876967412990468,
+      "grad_norm": 0.41512253880500793,
+      "learning_rate": 9.542993130237683e-06,
+      "loss": 0.0085,
+      "step": 550
+    },
+    {
+      "epoch": 0.4885834626468632,
+      "grad_norm": 0.5797951221466064,
+      "learning_rate": 9.54099142871001e-06,
+      "loss": 0.0089,
+      "step": 551
+    },
+    {
+      "epoch": 0.4894701839946797,
+      "grad_norm": 1.1996581554412842,
+      "learning_rate": 9.53898556388016e-06,
+      "loss": 0.0132,
+      "step": 552
+    },
+    {
+      "epoch": 0.4903569053424961,
+      "grad_norm": 0.9442786574363708,
+      "learning_rate": 9.536975537587167e-06,
+      "loss": 0.0093,
+      "step": 553
+    },
+    {
+      "epoch": 0.4912436266903126,
+      "grad_norm": 0.9179630279541016,
+      "learning_rate": 9.534961351673871e-06,
+      "loss": 0.0212,
+      "step": 554
+    },
+    {
+      "epoch": 0.492130348038129,
+      "grad_norm": 0.34069445729255676,
+      "learning_rate": 9.532943007986932e-06,
+      "loss": 0.0052,
+      "step": 555
+    },
+    {
+      "epoch": 0.4930170693859455,
+      "grad_norm": 1.1509453058242798,
+      "learning_rate": 9.530920508376824e-06,
+      "loss": 0.0125,
+      "step": 556
+    },
+    {
+      "epoch": 0.4939037907337619,
+      "grad_norm": 0.3179965019226074,
+      "learning_rate": 9.52889385469783e-06,
+      "loss": 0.0063,
+      "step": 557
+    },
+    {
+      "epoch": 0.49479051208157837,
+      "grad_norm": 0.35437366366386414,
+      "learning_rate": 9.526863048808035e-06,
+      "loss": 0.0051,
+      "step": 558
+    },
+    {
+      "epoch": 0.4956772334293948,
+      "grad_norm": 0.43903347849845886,
+      "learning_rate": 9.52482809256934e-06,
+      "loss": 0.0067,
+      "step": 559
+    },
+    {
+      "epoch": 0.49656395477721127,
+      "grad_norm": 0.5896758437156677,
+      "learning_rate": 9.522788987847447e-06,
+      "loss": 0.0101,
+      "step": 560
+    },
+    {
+      "epoch": 0.4974506761250277,
+      "grad_norm": 0.1301652044057846,
+      "learning_rate": 9.520745736511861e-06,
+      "loss": 0.0018,
+      "step": 561
+    },
+    {
+      "epoch": 0.49833739747284417,
+      "grad_norm": 1.0899876356124878,
+      "learning_rate": 9.518698340435886e-06,
+      "loss": 0.0127,
+      "step": 562
+    },
+    {
+      "epoch": 0.4992241188206606,
+      "grad_norm": 0.6057047247886658,
+      "learning_rate": 9.516646801496636e-06,
+      "loss": 0.0063,
+      "step": 563
+    },
+    {
+      "epoch": 0.5001108401684771,
+      "grad_norm": 0.6446989178657532,
+      "learning_rate": 9.514591121575012e-06,
+      "loss": 0.0059,
+      "step": 564
+    },
+    {
+      "epoch": 0.5001108401684771,
+      "eval_loss": 0.021343758329749107,
+      "eval_runtime": 59.8604,
+      "eval_samples_per_second": 3.207,
+      "eval_steps_per_second": 0.802,
+      "step": 564
+    },
+    {
+      "epoch": 0.5009975615162935,
+      "grad_norm": 1.243457555770874,
+      "learning_rate": 9.512531302555717e-06,
+      "loss": 0.015,
+      "step": 565
+    },
+    {
+      "epoch": 0.5018842828641099,
+      "grad_norm": 0.9451794028282166,
+      "learning_rate": 9.51046734632725e-06,
+      "loss": 0.013,
+      "step": 566
+    },
+    {
+      "epoch": 0.5027710042119264,
+      "grad_norm": 0.5004759430885315,
+      "learning_rate": 9.508399254781899e-06,
+      "loss": 0.0091,
+      "step": 567
+    },
+    {
+      "epoch": 0.5036577255597429,
+      "grad_norm": 0.5445528626441956,
+      "learning_rate": 9.506327029815746e-06,
+      "loss": 0.0099,
+      "step": 568
+    },
+    {
+      "epoch": 0.5045444469075593,
+      "grad_norm": 0.5810771584510803,
+      "learning_rate": 9.504250673328661e-06,
+      "loss": 0.0123,
+      "step": 569
+    },
+    {
+      "epoch": 0.5054311682553757,
+      "grad_norm": 0.7572499513626099,
+      "learning_rate": 9.502170187224307e-06,
+      "loss": 0.008,
+      "step": 570
+    },
+    {
+      "epoch": 0.5063178896031922,
+      "grad_norm": 0.4947316348552704,
+      "learning_rate": 9.500085573410126e-06,
+      "loss": 0.0038,
+      "step": 571
+    },
+    {
+      "epoch": 0.5072046109510087,
+      "grad_norm": 0.2767347991466522,
+      "learning_rate": 9.497996833797348e-06,
+      "loss": 0.0033,
+      "step": 572
+    },
+    {
+      "epoch": 0.5080913322988251,
+      "grad_norm": 0.7010060548782349,
+      "learning_rate": 9.495903970300986e-06,
+      "loss": 0.0101,
+      "step": 573
+    },
+    {
+      "epoch": 0.5089780536466415,
+      "grad_norm": 0.7284058332443237,
+      "learning_rate": 9.493806984839833e-06,
+      "loss": 0.0098,
+      "step": 574
+    },
+    {
+      "epoch": 0.509864774994458,
+      "grad_norm": 0.42266321182250977,
+      "learning_rate": 9.49170587933646e-06,
+      "loss": 0.0048,
+      "step": 575
+    },
+    {
+      "epoch": 0.5107514963422745,
+      "grad_norm": 0.40918147563934326,
+      "learning_rate": 9.489600655717217e-06,
+      "loss": 0.0071,
+      "step": 576
+    },
+    {
+      "epoch": 0.5116382176900909,
+      "grad_norm": 0.3350435793399811,
+      "learning_rate": 9.487491315912231e-06,
+      "loss": 0.0118,
+      "step": 577
+    },
+    {
+      "epoch": 0.5125249390379073,
+      "grad_norm": 0.4555787146091461,
+      "learning_rate": 9.485377861855398e-06,
+      "loss": 0.0057,
+      "step": 578
+    },
+    {
+      "epoch": 0.5134116603857238,
+      "grad_norm": 0.9364067912101746,
+      "learning_rate": 9.483260295484393e-06,
+      "loss": 0.0119,
+      "step": 579
+    },
+    {
+      "epoch": 0.5142983817335403,
+      "grad_norm": 1.019731879234314,
+      "learning_rate": 9.481138618740655e-06,
+      "loss": 0.0143,
+      "step": 580
+    },
+    {
+      "epoch": 0.5151851030813567,
+      "grad_norm": 0.4347856342792511,
+      "learning_rate": 9.479012833569394e-06,
+      "loss": 0.0046,
+      "step": 581
+    },
+    {
+      "epoch": 0.5160718244291731,
+      "grad_norm": 0.34786656498908997,
+      "learning_rate": 9.476882941919587e-06,
+      "loss": 0.0068,
+      "step": 582
+    },
+    {
+      "epoch": 0.5169585457769896,
+      "grad_norm": 0.9139060378074646,
+      "learning_rate": 9.474748945743974e-06,
+      "loss": 0.0094,
+      "step": 583
+    },
+    {
+      "epoch": 0.517845267124806,
+      "grad_norm": 0.23097054660320282,
+      "learning_rate": 9.472610846999061e-06,
+      "loss": 0.0021,
+      "step": 584
+    },
+    {
+      "epoch": 0.5187319884726225,
+      "grad_norm": 0.5011057257652283,
+      "learning_rate": 9.470468647645116e-06,
+      "loss": 0.0054,
+      "step": 585
+    },
+    {
+      "epoch": 0.5196187098204389,
+      "grad_norm": 0.5539005398750305,
+      "learning_rate": 9.468322349646162e-06,
+      "loss": 0.0113,
+      "step": 586
+    },
+    {
+      "epoch": 0.5205054311682554,
+      "grad_norm": 0.5280828475952148,
+      "learning_rate": 9.466171954969981e-06,
+      "loss": 0.0088,
+      "step": 587
+    },
+    {
+      "epoch": 0.5213921525160718,
+      "grad_norm": 0.7050261497497559,
+      "learning_rate": 9.464017465588115e-06,
+      "loss": 0.0099,
+      "step": 588
+    },
+    {
+      "epoch": 0.5222788738638883,
+      "grad_norm": 0.2560153007507324,
+      "learning_rate": 9.461858883475859e-06,
+      "loss": 0.0018,
+      "step": 589
+    },
+    {
+      "epoch": 0.5231655952117047,
+      "grad_norm": 0.5633629560470581,
+      "learning_rate": 9.459696210612253e-06,
+      "loss": 0.0105,
+      "step": 590
+    },
+    {
+      "epoch": 0.5240523165595212,
+      "grad_norm": 0.7544528841972351,
+      "learning_rate": 9.457529448980098e-06,
+      "loss": 0.0102,
+      "step": 591
+    },
+    {
+      "epoch": 0.5249390379073376,
+      "grad_norm": 0.36814677715301514,
+      "learning_rate": 9.455358600565936e-06,
+      "loss": 0.0059,
+      "step": 592
+    },
+    {
+      "epoch": 0.5258257592551541,
+      "grad_norm": 0.8715342283248901,
+      "learning_rate": 9.453183667360062e-06,
+      "loss": 0.0124,
+      "step": 593
+    },
+    {
+      "epoch": 0.5267124806029705,
+      "grad_norm": 0.36388614773750305,
+      "learning_rate": 9.451004651356511e-06,
+      "loss": 0.0033,
+      "step": 594
+    },
+    {
+      "epoch": 0.5275992019507869,
+      "grad_norm": 0.6503794193267822,
+      "learning_rate": 9.448821554553061e-06,
+      "loss": 0.007,
+      "step": 595
+    },
+    {
+      "epoch": 0.5284859232986034,
+      "grad_norm": 1.0000611543655396,
+      "learning_rate": 9.446634378951236e-06,
+      "loss": 0.0137,
+      "step": 596
+    },
+    {
+      "epoch": 0.5293726446464199,
+      "grad_norm": 0.4502328634262085,
+      "learning_rate": 9.444443126556297e-06,
+      "loss": 0.0058,
+      "step": 597
+    },
+    {
+      "epoch": 0.5302593659942363,
+      "grad_norm": 0.5919182300567627,
+      "learning_rate": 9.442247799377242e-06,
+      "loss": 0.0066,
+      "step": 598
+    },
+    {
+      "epoch": 0.5311460873420527,
+      "grad_norm": 0.4766238033771515,
+      "learning_rate": 9.440048399426805e-06,
+      "loss": 0.0048,
+      "step": 599
+    },
+    {
+      "epoch": 0.5320328086898692,
+      "grad_norm": 0.26894569396972656,
+      "learning_rate": 9.437844928721455e-06,
+      "loss": 0.007,
+      "step": 600
+    },
+    {
+      "epoch": 0.5329195300376857,
+      "grad_norm": 1.3799517154693604,
+      "learning_rate": 9.435637389281395e-06,
+      "loss": 0.0072,
+      "step": 601
+    },
+    {
+      "epoch": 0.5338062513855021,
+      "grad_norm": 0.3927518129348755,
+      "learning_rate": 9.433425783130554e-06,
+      "loss": 0.0024,
+      "step": 602
+    },
+    {
+      "epoch": 0.5346929727333185,
+      "grad_norm": 0.2583082616329193,
+      "learning_rate": 9.43121011229659e-06,
+      "loss": 0.0032,
+      "step": 603
+    },
+    {
+      "epoch": 0.535579694081135,
+      "grad_norm": 0.6682037115097046,
+      "learning_rate": 9.428990378810891e-06,
+      "loss": 0.0076,
+      "step": 604
+    },
+    {
+      "epoch": 0.5364664154289515,
+      "grad_norm": 1.4830431938171387,
+      "learning_rate": 9.42676658470857e-06,
+      "loss": 0.0072,
+      "step": 605
+    },
+    {
+      "epoch": 0.5373531367767679,
+      "grad_norm": 0.5184667706489563,
+      "learning_rate": 9.424538732028457e-06,
+      "loss": 0.0058,
+      "step": 606
+    },
+    {
+      "epoch": 0.5382398581245843,
+      "grad_norm": 0.9888753294944763,
+      "learning_rate": 9.422306822813108e-06,
+      "loss": 0.0104,
+      "step": 607
+    },
+    {
+      "epoch": 0.5391265794724008,
+      "grad_norm": 0.4412868916988373,
+      "learning_rate": 9.420070859108799e-06,
+      "loss": 0.0045,
+      "step": 608
+    },
+    {
+      "epoch": 0.5400133008202173,
+      "grad_norm": 0.5253145694732666,
+      "learning_rate": 9.417830842965519e-06,
+      "loss": 0.0064,
+      "step": 609
+    },
+    {
+      "epoch": 0.5409000221680337,
+      "grad_norm": 0.2791012227535248,
+      "learning_rate": 9.415586776436973e-06,
+      "loss": 0.0026,
+      "step": 610
+    },
+    {
+      "epoch": 0.5417867435158501,
+      "grad_norm": 1.9227204322814941,
+      "learning_rate": 9.413338661580587e-06,
+      "loss": 0.0086,
+      "step": 611
+    },
+    {
+      "epoch": 0.5426734648636666,
+      "grad_norm": 0.3563939332962036,
+      "learning_rate": 9.411086500457486e-06,
+      "loss": 0.0032,
+      "step": 612
+    },
+    {
+      "epoch": 0.543560186211483,
+      "grad_norm": 0.4303697645664215,
+      "learning_rate": 9.408830295132516e-06,
+      "loss": 0.0033,
+      "step": 613
+    },
+    {
+      "epoch": 0.5444469075592995,
+      "grad_norm": 0.502570629119873,
+      "learning_rate": 9.406570047674223e-06,
+      "loss": 0.0036,
+      "step": 614
+    },
+    {
+      "epoch": 0.5453336289071159,
+      "grad_norm": 0.7179825305938721,
+      "learning_rate": 9.404305760154866e-06,
+      "loss": 0.0112,
+      "step": 615
+    },
+    {
+      "epoch": 0.5462203502549324,
+      "grad_norm": 1.023728847503662,
+      "learning_rate": 9.402037434650402e-06,
+      "loss": 0.0149,
+      "step": 616
+    },
+    {
+      "epoch": 0.5471070716027489,
+      "grad_norm": 1.108109951019287,
+      "learning_rate": 9.399765073240491e-06,
+      "loss": 0.0105,
+      "step": 617
+    },
+    {
+      "epoch": 0.5479937929505653,
+      "grad_norm": 0.9911279082298279,
+      "learning_rate": 9.397488678008498e-06,
+      "loss": 0.0014,
+      "step": 618
+    },
+    {
+      "epoch": 0.5488805142983817,
+      "grad_norm": 0.9575443863868713,
+      "learning_rate": 9.395208251041482e-06,
+      "loss": 0.0132,
+      "step": 619
+    },
+    {
+      "epoch": 0.5497672356461982,
+      "grad_norm": 0.9429709315299988,
+      "learning_rate": 9.392923794430196e-06,
+      "loss": 0.0119,
+      "step": 620
+    },
+    {
+      "epoch": 0.5506539569940146,
+      "grad_norm": 0.7063897848129272,
+      "learning_rate": 9.390635310269094e-06,
+      "loss": 0.0094,
+      "step": 621
+    },
+    {
+      "epoch": 0.5515406783418311,
+      "grad_norm": 0.5705755352973938,
+      "learning_rate": 9.388342800656319e-06,
+      "loss": 0.009,
+      "step": 622
+    },
+    {
+      "epoch": 0.5524273996896475,
+      "grad_norm": 0.67104572057724,
+      "learning_rate": 9.386046267693705e-06,
+      "loss": 0.0069,
+      "step": 623
+    },
+    {
+      "epoch": 0.553314121037464,
+      "grad_norm": 0.44542428851127625,
+      "learning_rate": 9.383745713486774e-06,
+      "loss": 0.0095,
+      "step": 624
+    },
+    {
+      "epoch": 0.5542008423852804,
+      "grad_norm": 0.3004834055900574,
+      "learning_rate": 9.381441140144735e-06,
+      "loss": 0.004,
+      "step": 625
+    },
+    {
+      "epoch": 0.5550875637330969,
+      "grad_norm": 0.48007500171661377,
+      "learning_rate": 9.379132549780486e-06,
+      "loss": 0.008,
+      "step": 626
+    },
+    {
+      "epoch": 0.5559742850809133,
+      "grad_norm": 0.8524070382118225,
+      "learning_rate": 9.376819944510598e-06,
+      "loss": 0.0128,
+      "step": 627
+    },
+    {
+      "epoch": 0.5568610064287298,
+      "grad_norm": 1.298292875289917,
+      "learning_rate": 9.374503326455335e-06,
+      "loss": 0.0154,
+      "step": 628
+    },
+    {
+      "epoch": 0.5577477277765462,
+      "grad_norm": 0.2932794690132141,
+      "learning_rate": 9.37218269773863e-06,
+      "loss": 0.0027,
+      "step": 629
+    },
+    {
+      "epoch": 0.5586344491243627,
+      "grad_norm": 0.5051419138908386,
+      "learning_rate": 9.369858060488102e-06,
+      "loss": 0.0077,
+      "step": 630
+    },
+    {
+      "epoch": 0.5595211704721791,
+      "grad_norm": 0.48923778533935547,
+      "learning_rate": 9.367529416835038e-06,
+      "loss": 0.0108,
+      "step": 631
+    },
+    {
+      "epoch": 0.5604078918199956,
+      "grad_norm": 0.7815966606140137,
+      "learning_rate": 9.365196768914399e-06,
+      "loss": 0.0146,
+      "step": 632
+    },
+    {
+      "epoch": 0.561294613167812,
+      "grad_norm": 0.3096315562725067,
+      "learning_rate": 9.362860118864822e-06,
+      "loss": 0.005,
+      "step": 633
+    },
+    {
+      "epoch": 0.5621813345156285,
+      "grad_norm": 0.6307308077812195,
+      "learning_rate": 9.360519468828608e-06,
+      "loss": 0.0101,
+      "step": 634
+    },
+    {
+      "epoch": 0.5630680558634449,
+      "grad_norm": 0.5058624744415283,
+      "learning_rate": 9.358174820951727e-06,
+      "loss": 0.0071,
+      "step": 635
+    },
+    {
+      "epoch": 0.5639547772112614,
+      "grad_norm": 0.35703045129776,
+      "learning_rate": 9.355826177383818e-06,
+      "loss": 0.0054,
+      "step": 636
+    },
+    {
+      "epoch": 0.5648414985590778,
+      "grad_norm": 0.9188088178634644,
+      "learning_rate": 9.353473540278175e-06,
+      "loss": 0.0133,
+      "step": 637
+    },
+    {
+      "epoch": 0.5657282199068943,
+      "grad_norm": 0.7231817245483398,
+      "learning_rate": 9.351116911791765e-06,
+      "loss": 0.0061,
+      "step": 638
+    },
+    {
+      "epoch": 0.5666149412547107,
+      "grad_norm": 0.551129937171936,
+      "learning_rate": 9.348756294085202e-06,
+      "loss": 0.0045,
+      "step": 639
+    },
+    {
+      "epoch": 0.5675016626025271,
+      "grad_norm": 1.0952398777008057,
+      "learning_rate": 9.346391689322768e-06,
+      "loss": 0.0096,
+      "step": 640
+    },
+    {
+      "epoch": 0.5683883839503436,
+      "grad_norm": 2.5634279251098633,
+      "learning_rate": 9.344023099672392e-06,
+      "loss": 0.012,
+      "step": 641
+    },
+    {
+      "epoch": 0.5692751052981601,
+      "grad_norm": 0.5959585309028625,
+      "learning_rate": 9.34165052730566e-06,
+      "loss": 0.0038,
+      "step": 642
+    },
+    {
+      "epoch": 0.5701618266459765,
+      "grad_norm": 0.8079138398170471,
+      "learning_rate": 9.339273974397814e-06,
+      "loss": 0.0057,
+      "step": 643
+    },
+    {
+      "epoch": 0.5710485479937929,
+      "grad_norm": 1.0635994672775269,
+      "learning_rate": 9.336893443127739e-06,
+      "loss": 0.0041,
+      "step": 644
+    },
+    {
+      "epoch": 0.5719352693416094,
+      "grad_norm": 0.2653323709964752,
+      "learning_rate": 9.334508935677968e-06,
+      "loss": 0.0016,
+      "step": 645
+    },
+    {
+      "epoch": 0.5728219906894259,
+      "grad_norm": 0.32824158668518066,
+      "learning_rate": 9.332120454234682e-06,
+      "loss": 0.0032,
+      "step": 646
+    },
+    {
+      "epoch": 0.5737087120372423,
+      "grad_norm": 0.31941306591033936,
+      "learning_rate": 9.329728000987706e-06,
+      "loss": 0.0023,
+      "step": 647
+    },
+    {
+      "epoch": 0.5745954333850587,
+      "grad_norm": 0.3932683765888214,
+      "learning_rate": 9.327331578130503e-06,
+      "loss": 0.0021,
+      "step": 648
+    },
+    {
+      "epoch": 0.5754821547328752,
+      "grad_norm": 1.0215529203414917,
+      "learning_rate": 9.324931187860179e-06,
+      "loss": 0.01,
+      "step": 649
+    },
+    {
+      "epoch": 0.5763688760806917,
+      "grad_norm": 1.4922890663146973,
+      "learning_rate": 9.322526832377473e-06,
+      "loss": 0.0035,
+      "step": 650
+    },
+    {
+      "epoch": 0.5772555974285081,
+      "grad_norm": 0.16321395337581635,
+      "learning_rate": 9.32011851388677e-06,
+      "loss": 0.001,
+      "step": 651
+    },
+    {
+      "epoch": 0.5781423187763245,
+      "grad_norm": 0.8923147320747375,
+      "learning_rate": 9.31770623459607e-06,
+      "loss": 0.0041,
+      "step": 652
+    },
+    {
+      "epoch": 0.579029040124141,
+      "grad_norm": 2.212514877319336,
+      "learning_rate": 9.315289996717022e-06,
+      "loss": 0.0159,
+      "step": 653
+    },
+    {
+      "epoch": 0.5799157614719574,
+      "grad_norm": 0.5469567775726318,
+      "learning_rate": 9.312869802464896e-06,
+      "loss": 0.0026,
+      "step": 654
+    },
+    {
+      "epoch": 0.5808024828197739,
+      "grad_norm": 0.2924470901489258,
+      "learning_rate": 9.310445654058589e-06,
+      "loss": 0.0016,
+      "step": 655
+    },
+    {
+      "epoch": 0.5816892041675903,
+      "grad_norm": 0.450457364320755,
+      "learning_rate": 9.308017553720628e-06,
+      "loss": 0.005,
+      "step": 656
+    },
+    {
+      "epoch": 0.5825759255154068,
+      "grad_norm": 0.2175384908914566,
+      "learning_rate": 9.305585503677158e-06,
+      "loss": 0.0011,
+      "step": 657
+    },
+    {
+      "epoch": 0.5834626468632232,
+      "grad_norm": 1.5817656517028809,
+      "learning_rate": 9.303149506157948e-06,
+      "loss": 0.0081,
+      "step": 658
+    },
+    {
+      "epoch": 0.5843493682110397,
+      "grad_norm": 0.8667041063308716,
+      "learning_rate": 9.300709563396386e-06,
+      "loss": 0.0096,
+      "step": 659
+    },
+    {
+      "epoch": 0.5852360895588561,
+      "grad_norm": 0.665524423122406,
+      "learning_rate": 9.298265677629476e-06,
+      "loss": 0.0045,
+      "step": 660
+    },
+    {
+      "epoch": 0.5861228109066726,
+      "grad_norm": 0.8336466550827026,
+      "learning_rate": 9.295817851097836e-06,
+      "loss": 0.0079,
+      "step": 661
+    },
+    {
+      "epoch": 0.587009532254489,
+      "grad_norm": 0.8972154259681702,
+      "learning_rate": 9.293366086045703e-06,
+      "loss": 0.0069,
+      "step": 662
+    },
+    {
+      "epoch": 0.5878962536023055,
+      "grad_norm": 0.594464123249054,
+      "learning_rate": 9.290910384720918e-06,
+      "loss": 0.0127,
+      "step": 663
+    },
+    {
+      "epoch": 0.5887829749501219,
+      "grad_norm": 0.5330869555473328,
+      "learning_rate": 9.288450749374933e-06,
+      "loss": 0.0073,
+      "step": 664
+    },
+    {
+      "epoch": 0.5896696962979384,
+      "grad_norm": 0.673941433429718,
+      "learning_rate": 9.28598718226281e-06,
+      "loss": 0.0055,
+      "step": 665
+    },
+    {
+      "epoch": 0.5905564176457548,
+      "grad_norm": 0.4206015169620514,
+      "learning_rate": 9.28351968564321e-06,
+      "loss": 0.0045,
+      "step": 666
+    },
+    {
+      "epoch": 0.5914431389935713,
+      "grad_norm": 0.4695059657096863,
+      "learning_rate": 9.2810482617784e-06,
+      "loss": 0.007,
+      "step": 667
+    },
+    {
+      "epoch": 0.5923298603413877,
+      "grad_norm": 0.5791723728179932,
+      "learning_rate": 9.27857291293425e-06,
+      "loss": 0.0082,
+      "step": 668
+    },
+    {
+      "epoch": 0.5932165816892042,
+      "grad_norm": 0.4505319595336914,
+      "learning_rate": 9.276093641380224e-06,
+      "loss": 0.007,
+      "step": 669
+    },
+    {
+      "epoch": 0.5941033030370206,
+      "grad_norm": 0.38367578387260437,
+      "learning_rate": 9.273610449389384e-06,
+      "loss": 0.0061,
+      "step": 670
+    },
+    {
+      "epoch": 0.5949900243848371,
+      "grad_norm": 0.2096157670021057,
+      "learning_rate": 9.271123339238387e-06,
+      "loss": 0.0045,
+      "step": 671
+    },
+    {
+      "epoch": 0.5958767457326535,
+      "grad_norm": 0.2905729115009308,
+      "learning_rate": 9.268632313207484e-06,
+      "loss": 0.009,
+      "step": 672
+    },
+    {
+      "epoch": 0.59676346708047,
+      "grad_norm": 0.40436604619026184,
+      "learning_rate": 9.266137373580512e-06,
+      "loss": 0.0057,
+      "step": 673
+    },
+    {
+      "epoch": 0.5976501884282864,
+      "grad_norm": 0.6021425724029541,
+      "learning_rate": 9.263638522644898e-06,
+      "loss": 0.0095,
+      "step": 674
+    },
+    {
+      "epoch": 0.5985369097761029,
+      "grad_norm": 0.39547786116600037,
+      "learning_rate": 9.261135762691658e-06,
+      "loss": 0.0073,
+      "step": 675
+    },
+    {
+      "epoch": 0.5994236311239193,
+      "grad_norm": 0.4780794382095337,
+      "learning_rate": 9.258629096015385e-06,
+      "loss": 0.0098,
+      "step": 676
+    },
+    {
+      "epoch": 0.6003103524717358,
+      "grad_norm": 0.5975167751312256,
+      "learning_rate": 9.256118524914263e-06,
+      "loss": 0.0146,
+      "step": 677
+    },
+    {
+      "epoch": 0.6011970738195522,
+      "grad_norm": 0.6294581890106201,
+      "learning_rate": 9.253604051690047e-06,
+      "loss": 0.0063,
+      "step": 678
+    },
+    {
+      "epoch": 0.6020837951673687,
+      "grad_norm": 0.28941860795021057,
+      "learning_rate": 9.251085678648072e-06,
+      "loss": 0.0034,
+      "step": 679
+    },
+    {
+      "epoch": 0.6029705165151851,
+      "grad_norm": 0.46086496114730835,
+      "learning_rate": 9.248563408097252e-06,
+      "loss": 0.0045,
+      "step": 680
+    },
+    {
+      "epoch": 0.6038572378630016,
+      "grad_norm": 0.2607171833515167,
+      "learning_rate": 9.246037242350072e-06,
+      "loss": 0.0035,
+      "step": 681
+    },
+    {
+      "epoch": 0.604743959210818,
+      "grad_norm": 0.5795825123786926,
+      "learning_rate": 9.24350718372259e-06,
+      "loss": 0.0069,
+      "step": 682
+    },
+    {
+      "epoch": 0.6056306805586344,
+      "grad_norm": 0.5818445086479187,
+      "learning_rate": 9.240973234534426e-06,
+      "loss": 0.0096,
+      "step": 683
+    },
+    {
+      "epoch": 0.6065174019064509,
+      "grad_norm": 0.3651141822338104,
+      "learning_rate": 9.238435397108777e-06,
+      "loss": 0.0061,
+      "step": 684
+    },
+    {
+      "epoch": 0.6074041232542673,
+      "grad_norm": 0.5181947350502014,
+      "learning_rate": 9.235893673772397e-06,
+      "loss": 0.0123,
+      "step": 685
+    },
+    {
+      "epoch": 0.6082908446020838,
+      "grad_norm": 0.4905543923377991,
+      "learning_rate": 9.233348066855607e-06,
+      "loss": 0.0031,
+      "step": 686
+    },
+    {
+      "epoch": 0.6091775659499002,
+      "grad_norm": 1.2415827512741089,
+      "learning_rate": 9.23079857869229e-06,
+      "loss": 0.004,
+      "step": 687
+    },
+    {
+      "epoch": 0.6100642872977167,
+      "grad_norm": 0.5994253158569336,
+      "learning_rate": 9.228245211619881e-06,
+      "loss": 0.0076,
+      "step": 688
+    },
+    {
+      "epoch": 0.6109510086455331,
+      "grad_norm": 0.2408173829317093,
+      "learning_rate": 9.225687967979377e-06,
+      "loss": 0.0029,
+      "step": 689
+    },
+    {
+      "epoch": 0.6118377299933496,
+      "grad_norm": 0.7248222827911377,
+      "learning_rate": 9.223126850115327e-06,
+      "loss": 0.0074,
+      "step": 690
+    },
+    {
+      "epoch": 0.612724451341166,
+      "grad_norm": 1.5161305665969849,
+      "learning_rate": 9.220561860375831e-06,
+      "loss": 0.0124,
+      "step": 691
+    },
+    {
+      "epoch": 0.6136111726889825,
+      "grad_norm": 0.6087315678596497,
+      "learning_rate": 9.217993001112544e-06,
+      "loss": 0.0065,
+      "step": 692
+    },
+    {
+      "epoch": 0.6144978940367989,
+      "grad_norm": 0.4666323661804199,
+      "learning_rate": 9.215420274680658e-06,
+      "loss": 0.0035,
+      "step": 693
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.32539889216423035,
+      "learning_rate": 9.212843683438923e-06,
+      "loss": 0.0045,
+      "step": 694
+    },
+    {
+      "epoch": 0.6162713367324318,
+      "grad_norm": 0.44837233424186707,
+      "learning_rate": 9.210263229749626e-06,
+      "loss": 0.0073,
+      "step": 695
+    },
+    {
+      "epoch": 0.6171580580802483,
+      "grad_norm": 0.7816086411476135,
+      "learning_rate": 9.207678915978593e-06,
+      "loss": 0.007,
+      "step": 696
+    },
+    {
+      "epoch": 0.6180447794280647,
+      "grad_norm": 0.4992975890636444,
+      "learning_rate": 9.205090744495193e-06,
+      "loss": 0.0066,
+      "step": 697
+    },
+    {
+      "epoch": 0.6189315007758812,
+      "grad_norm": 0.5404097437858582,
+      "learning_rate": 9.202498717672332e-06,
+      "loss": 0.0058,
+      "step": 698
+    },
+    {
+      "epoch": 0.6198182221236976,
+      "grad_norm": 0.4112424850463867,
+      "learning_rate": 9.19990283788645e-06,
+      "loss": 0.0055,
+      "step": 699
+    },
+    {
+      "epoch": 0.6207049434715141,
+      "grad_norm": 0.5358926653862,
+      "learning_rate": 9.197303107517516e-06,
+      "loss": 0.0063,
+      "step": 700
+    },
+    {
+      "epoch": 0.6215916648193305,
+      "grad_norm": 0.4793187975883484,
+      "learning_rate": 9.194699528949034e-06,
+      "loss": 0.005,
+      "step": 701
+    },
+    {
+      "epoch": 0.622478386167147,
+      "grad_norm": 0.4965543746948242,
+      "learning_rate": 9.192092104568033e-06,
+      "loss": 0.0065,
+      "step": 702
+    },
+    {
+      "epoch": 0.6233651075149634,
+      "grad_norm": 0.31342917680740356,
+      "learning_rate": 9.189480836765071e-06,
+      "loss": 0.0048,
+      "step": 703
+    },
+    {
+      "epoch": 0.6242518288627799,
+      "grad_norm": 0.8319897651672363,
+      "learning_rate": 9.186865727934227e-06,
+      "loss": 0.0062,
+      "step": 704
+    },
+    {
+      "epoch": 0.6251385502105963,
+      "grad_norm": 0.6753369569778442,
+      "learning_rate": 9.184246780473105e-06,
+      "loss": 0.0092,
+      "step": 705
+    },
+    {
+      "epoch": 0.6260252715584128,
+      "grad_norm": 0.7414891719818115,
+      "learning_rate": 9.181623996782822e-06,
+      "loss": 0.0129,
+      "step": 706
+    },
+    {
+      "epoch": 0.6269119929062292,
+      "grad_norm": 0.48528799414634705,
+      "learning_rate": 9.178997379268018e-06,
+      "loss": 0.0049,
+      "step": 707
+    },
+    {
+      "epoch": 0.6277987142540457,
+      "grad_norm": 0.3304223418235779,
+      "learning_rate": 9.176366930336846e-06,
+      "loss": 0.0074,
+      "step": 708
+    },
+    {
+      "epoch": 0.6286854356018621,
+      "grad_norm": 0.19392916560173035,
+      "learning_rate": 9.173732652400972e-06,
+      "loss": 0.0031,
+      "step": 709
+    },
+    {
+      "epoch": 0.6295721569496786,
+      "grad_norm": 0.641016960144043,
+      "learning_rate": 9.171094547875574e-06,
+      "loss": 0.008,
+      "step": 710
+    },
+    {
+      "epoch": 0.630458878297495,
+      "grad_norm": 0.3769802451133728,
+      "learning_rate": 9.168452619179334e-06,
+      "loss": 0.0039,
+      "step": 711
+    },
+    {
+      "epoch": 0.6313455996453115,
+      "grad_norm": 0.6935088038444519,
+      "learning_rate": 9.165806868734444e-06,
+      "loss": 0.0057,
+      "step": 712
+    },
+    {
+      "epoch": 0.6322323209931279,
+      "grad_norm": 0.44623953104019165,
+      "learning_rate": 9.163157298966596e-06,
+      "loss": 0.0053,
+      "step": 713
+    },
+    {
+      "epoch": 0.6331190423409444,
+      "grad_norm": 0.2738344669342041,
+      "learning_rate": 9.16050391230499e-06,
+      "loss": 0.0019,
+      "step": 714
+    },
+    {
+      "epoch": 0.6340057636887608,
+      "grad_norm": 0.6001261472702026,
+      "learning_rate": 9.157846711182322e-06,
+      "loss": 0.0059,
+      "step": 715
+    },
+    {
+      "epoch": 0.6348924850365772,
+      "grad_norm": 0.6489748954772949,
+      "learning_rate": 9.15518569803478e-06,
+      "loss": 0.0094,
+      "step": 716
+    },
+    {
+      "epoch": 0.6357792063843937,
+      "grad_norm": 0.14771291613578796,
+      "learning_rate": 9.152520875302058e-06,
+      "loss": 0.0009,
+      "step": 717
+    },
+    {
+      "epoch": 0.6366659277322102,
+      "grad_norm": 1.016381025314331,
+      "learning_rate": 9.149852245427333e-06,
+      "loss": 0.0251,
+      "step": 718
+    },
+    {
+      "epoch": 0.6375526490800266,
+      "grad_norm": 0.5366118550300598,
+      "learning_rate": 9.147179810857277e-06,
+      "loss": 0.0039,
+      "step": 719
+    },
+    {
+      "epoch": 0.638439370427843,
+      "grad_norm": 0.6103517413139343,
+      "learning_rate": 9.144503574042047e-06,
+      "loss": 0.0071,
+      "step": 720
+    },
+    {
+      "epoch": 0.6393260917756595,
+      "grad_norm": 0.5155932903289795,
+      "learning_rate": 9.141823537435294e-06,
+      "loss": 0.0066,
+      "step": 721
+    },
+    {
+      "epoch": 0.640212813123476,
+      "grad_norm": 1.691421627998352,
+      "learning_rate": 9.139139703494141e-06,
+      "loss": 0.0066,
+      "step": 722
+    },
+    {
+      "epoch": 0.6410995344712924,
+      "grad_norm": 0.6509506702423096,
+      "learning_rate": 9.136452074679204e-06,
+      "loss": 0.0078,
+      "step": 723
+    },
+    {
+      "epoch": 0.6419862558191088,
+      "grad_norm": 0.38033562898635864,
+      "learning_rate": 9.133760653454568e-06,
+      "loss": 0.0058,
+      "step": 724
+    },
+    {
+      "epoch": 0.6428729771669253,
+      "grad_norm": 0.7521248459815979,
+      "learning_rate": 9.131065442287803e-06,
+      "loss": 0.0102,
+      "step": 725
+    },
+    {
+      "epoch": 0.6437596985147418,
+      "grad_norm": 0.6125398278236389,
+      "learning_rate": 9.12836644364995e-06,
+      "loss": 0.0059,
+      "step": 726
+    },
+    {
+      "epoch": 0.6446464198625582,
+      "grad_norm": 0.5620765686035156,
+      "learning_rate": 9.125663660015523e-06,
+      "loss": 0.0115,
+      "step": 727
+    },
+    {
+      "epoch": 0.6455331412103746,
+      "grad_norm": 0.38775548338890076,
+      "learning_rate": 9.122957093862504e-06,
+      "loss": 0.0042,
+      "step": 728
+    },
+    {
+      "epoch": 0.6464198625581911,
+      "grad_norm": 0.60417240858078,
+      "learning_rate": 9.120246747672347e-06,
+      "loss": 0.0091,
+      "step": 729
+    },
+    {
+      "epoch": 0.6473065839060075,
+      "grad_norm": 0.2996869385242462,
+      "learning_rate": 9.117532623929969e-06,
+      "loss": 0.0041,
+      "step": 730
+    },
+    {
+      "epoch": 0.648193305253824,
+      "grad_norm": 0.5002415180206299,
+      "learning_rate": 9.114814725123755e-06,
+      "loss": 0.0096,
+      "step": 731
+    },
+    {
+      "epoch": 0.6490800266016404,
+      "grad_norm": 0.34925490617752075,
+      "learning_rate": 9.112093053745541e-06,
+      "loss": 0.0086,
+      "step": 732
+    },
+    {
+      "epoch": 0.6499667479494569,
+      "grad_norm": 0.3351205885410309,
+      "learning_rate": 9.109367612290633e-06,
+      "loss": 0.0033,
+      "step": 733
+    },
+    {
+      "epoch": 0.6508534692972733,
+      "grad_norm": 0.7079493999481201,
+      "learning_rate": 9.106638403257786e-06,
+      "loss": 0.0115,
+      "step": 734
+    },
+    {
+      "epoch": 0.6517401906450898,
+      "grad_norm": 0.6996093392372131,
+      "learning_rate": 9.103905429149212e-06,
+      "loss": 0.0066,
+      "step": 735
+    },
+    {
+      "epoch": 0.6526269119929062,
+      "grad_norm": 0.8951324820518494,
+      "learning_rate": 9.101168692470574e-06,
+      "loss": 0.01,
+      "step": 736
+    },
+    {
+      "epoch": 0.6535136333407227,
+      "grad_norm": 0.5505021810531616,
+      "learning_rate": 9.098428195730987e-06,
+      "loss": 0.0062,
+      "step": 737
+    },
+    {
+      "epoch": 0.6544003546885391,
+      "grad_norm": 0.3233986496925354,
+      "learning_rate": 9.09568394144301e-06,
+      "loss": 0.0044,
+      "step": 738
+    },
+    {
+      "epoch": 0.6552870760363556,
+      "grad_norm": 0.3201133608818054,
+      "learning_rate": 9.092935932122648e-06,
+      "loss": 0.0045,
+      "step": 739
+    },
+    {
+      "epoch": 0.656173797384172,
+      "grad_norm": 0.7146524786949158,
+      "learning_rate": 9.090184170289351e-06,
+      "loss": 0.0066,
+      "step": 740
+    },
+    {
+      "epoch": 0.6570605187319885,
+      "grad_norm": 0.5312227606773376,
+      "learning_rate": 9.087428658466009e-06,
+      "loss": 0.0036,
+      "step": 741
+    },
+    {
+      "epoch": 0.6579472400798049,
+      "grad_norm": 0.8712502121925354,
+      "learning_rate": 9.084669399178945e-06,
+      "loss": 0.0074,
+      "step": 742
+    },
+    {
+      "epoch": 0.6588339614276214,
+      "grad_norm": 1.0163110494613647,
+      "learning_rate": 9.081906394957924e-06,
+      "loss": 0.0136,
+      "step": 743
+    },
+    {
+      "epoch": 0.6597206827754378,
+      "grad_norm": 0.27555593848228455,
+      "learning_rate": 9.079139648336142e-06,
+      "loss": 0.0013,
+      "step": 744
+    },
+    {
+      "epoch": 0.6606074041232542,
+      "grad_norm": 0.892341136932373,
+      "learning_rate": 9.076369161850227e-06,
+      "loss": 0.009,
+      "step": 745
+    },
+    {
+      "epoch": 0.6614941254710707,
+      "grad_norm": 0.9638225436210632,
+      "learning_rate": 9.073594938040231e-06,
+      "loss": 0.0083,
+      "step": 746
+    },
+    {
+      "epoch": 0.6623808468188872,
+      "grad_norm": 0.9928414821624756,
+      "learning_rate": 9.070816979449642e-06,
+      "loss": 0.0113,
+      "step": 747
+    },
+    {
+      "epoch": 0.6632675681667036,
+      "grad_norm": 0.5321499705314636,
+      "learning_rate": 9.068035288625363e-06,
+      "loss": 0.0094,
+      "step": 748
+    },
+    {
+      "epoch": 0.66415428951452,
+      "grad_norm": 0.3631543517112732,
+      "learning_rate": 9.065249868117723e-06,
+      "loss": 0.0042,
+      "step": 749
+    },
+    {
+      "epoch": 0.6650410108623365,
+      "grad_norm": 1.0786023139953613,
+      "learning_rate": 9.06246072048047e-06,
+      "loss": 0.007,
+      "step": 750
+    },
+    {
+      "epoch": 0.665927732210153,
+      "grad_norm": 0.5387954115867615,
+      "learning_rate": 9.05966784827077e-06,
+      "loss": 0.0068,
+      "step": 751
+    },
+    {
+      "epoch": 0.6668144535579694,
+      "grad_norm": 0.14312922954559326,
+      "learning_rate": 9.0568712540492e-06,
+      "loss": 0.0016,
+      "step": 752
+    },
+    {
+      "epoch": 0.6677011749057858,
+      "grad_norm": 0.19700045883655548,
+      "learning_rate": 9.054070940379756e-06,
+      "loss": 0.0022,
+      "step": 753
+    },
+    {
+      "epoch": 0.6685878962536023,
+      "grad_norm": 0.24065862596035004,
+      "learning_rate": 9.051266909829838e-06,
+      "loss": 0.0027,
+      "step": 754
+    },
+    {
+      "epoch": 0.6694746176014188,
+      "grad_norm": 0.7865183353424072,
+      "learning_rate": 9.048459164970255e-06,
+      "loss": 0.0125,
+      "step": 755
+    },
+    {
+      "epoch": 0.6703613389492352,
+      "grad_norm": 0.4594501554965973,
+      "learning_rate": 9.045647708375226e-06,
+      "loss": 0.0071,
+      "step": 756
+    },
+    {
+      "epoch": 0.6712480602970516,
+      "grad_norm": 0.41208380460739136,
+      "learning_rate": 9.042832542622369e-06,
+      "loss": 0.0044,
+      "step": 757
+    },
+    {
+      "epoch": 0.6721347816448681,
+      "grad_norm": 0.49289634823799133,
+      "learning_rate": 9.0400136702927e-06,
+      "loss": 0.0101,
+      "step": 758
+    },
+    {
+      "epoch": 0.6730215029926846,
+      "grad_norm": 0.605767011642456,
+      "learning_rate": 9.037191093970636e-06,
+      "loss": 0.0078,
+      "step": 759
+    },
+    {
+      "epoch": 0.673908224340501,
+      "grad_norm": 0.5509247779846191,
+      "learning_rate": 9.034364816243995e-06,
+      "loss": 0.0053,
+      "step": 760
+    },
+    {
+      "epoch": 0.6747949456883174,
+      "grad_norm": 0.36483317613601685,
+      "learning_rate": 9.03153483970398e-06,
+      "loss": 0.0053,
+      "step": 761
+    },
+    {
+      "epoch": 0.6756816670361339,
+      "grad_norm": 0.7200064063072205,
+      "learning_rate": 9.028701166945191e-06,
+      "loss": 0.0059,
+      "step": 762
+    },
+    {
+      "epoch": 0.6765683883839504,
+      "grad_norm": 0.3441831171512604,
+      "learning_rate": 9.025863800565614e-06,
+      "loss": 0.0062,
+      "step": 763
+    },
+    {
+      "epoch": 0.6774551097317668,
+      "grad_norm": 0.3961979150772095,
+      "learning_rate": 9.023022743166623e-06,
+      "loss": 0.0034,
+      "step": 764
+    },
+    {
+      "epoch": 0.6783418310795832,
+      "grad_norm": 0.16061465442180634,
+      "learning_rate": 9.020177997352973e-06,
+      "loss": 0.0013,
+      "step": 765
+    },
+    {
+      "epoch": 0.6792285524273997,
+      "grad_norm": 0.8293284773826599,
+      "learning_rate": 9.017329565732806e-06,
+      "loss": 0.0085,
+      "step": 766
+    },
+    {
+      "epoch": 0.6801152737752162,
+      "grad_norm": 0.32432472705841064,
+      "learning_rate": 9.014477450917637e-06,
+      "loss": 0.0029,
+      "step": 767
+    },
+    {
+      "epoch": 0.6810019951230326,
+      "grad_norm": 0.48310521245002747,
+      "learning_rate": 9.011621655522365e-06,
+      "loss": 0.0046,
+      "step": 768
+    },
+    {
+      "epoch": 0.681888716470849,
+      "grad_norm": 0.5245798230171204,
+      "learning_rate": 9.008762182165256e-06,
+      "loss": 0.0072,
+      "step": 769
+    },
+    {
+      "epoch": 0.6827754378186655,
+      "grad_norm": 0.42478105425834656,
+      "learning_rate": 9.005899033467952e-06,
+      "loss": 0.0043,
+      "step": 770
+    },
+    {
+      "epoch": 0.683662159166482,
+      "grad_norm": 0.23128758370876312,
+      "learning_rate": 9.003032212055467e-06,
+      "loss": 0.0027,
+      "step": 771
+    },
+    {
+      "epoch": 0.6845488805142984,
+      "grad_norm": 0.1778474599123001,
+      "learning_rate": 9.000161720556177e-06,
+      "loss": 0.0019,
+      "step": 772
+    },
+    {
+      "epoch": 0.6854356018621148,
+      "grad_norm": 0.6820386648178101,
+      "learning_rate": 8.997287561601825e-06,
+      "loss": 0.0076,
+      "step": 773
+    },
+    {
+      "epoch": 0.6863223232099313,
+      "grad_norm": 0.5324229001998901,
+      "learning_rate": 8.994409737827516e-06,
+      "loss": 0.0086,
+      "step": 774
+    },
+    {
+      "epoch": 0.6872090445577477,
+      "grad_norm": 0.9875842928886414,
+      "learning_rate": 8.991528251871717e-06,
+      "loss": 0.0101,
+      "step": 775
+    },
+    {
+      "epoch": 0.6880957659055642,
+      "grad_norm": 0.3629169464111328,
+      "learning_rate": 8.988643106376252e-06,
+      "loss": 0.0028,
+      "step": 776
+    },
+    {
+      "epoch": 0.6889824872533806,
+      "grad_norm": 0.5612802505493164,
+      "learning_rate": 8.985754303986298e-06,
+      "loss": 0.0053,
+      "step": 777
+    },
+    {
+      "epoch": 0.689869208601197,
+      "grad_norm": 0.327252060174942,
+      "learning_rate": 8.982861847350388e-06,
+      "loss": 0.0022,
+      "step": 778
+    },
+    {
+      "epoch": 0.6907559299490135,
+      "grad_norm": 0.6719294786453247,
+      "learning_rate": 8.979965739120402e-06,
+      "loss": 0.0089,
+      "step": 779
+    },
+    {
+      "epoch": 0.69164265129683,
+      "grad_norm": 0.5353887677192688,
+      "learning_rate": 8.977065981951567e-06,
+      "loss": 0.0132,
+      "step": 780
+    },
+    {
+      "epoch": 0.6925293726446464,
+      "grad_norm": 0.4548264443874359,
+      "learning_rate": 8.974162578502463e-06,
+      "loss": 0.0053,
+      "step": 781
+    },
+    {
+      "epoch": 0.6934160939924628,
+      "grad_norm": 0.5508539080619812,
+      "learning_rate": 8.971255531435004e-06,
+      "loss": 0.004,
+      "step": 782
+    },
+    {
+      "epoch": 0.6943028153402793,
+      "grad_norm": 0.24120058119297028,
+      "learning_rate": 8.96834484341445e-06,
+      "loss": 0.0029,
+      "step": 783
+    },
+    {
+      "epoch": 0.6951895366880958,
+      "grad_norm": 0.562086820602417,
+      "learning_rate": 8.965430517109397e-06,
+      "loss": 0.0065,
+      "step": 784
+    },
+    {
+      "epoch": 0.6960762580359122,
+      "grad_norm": 0.3117201328277588,
+      "learning_rate": 8.962512555191776e-06,
+      "loss": 0.0022,
+      "step": 785
+    },
+    {
+      "epoch": 0.6969629793837286,
+      "grad_norm": 0.4886745810508728,
+      "learning_rate": 8.959590960336854e-06,
+      "loss": 0.0136,
+      "step": 786
+    },
+    {
+      "epoch": 0.6978497007315451,
+      "grad_norm": 0.2502830922603607,
+      "learning_rate": 8.956665735223227e-06,
+      "loss": 0.0027,
+      "step": 787
+    },
+    {
+      "epoch": 0.6987364220793616,
+      "grad_norm": 0.8199619650840759,
+      "learning_rate": 8.953736882532819e-06,
+      "loss": 0.0068,
+      "step": 788
+    },
+    {
+      "epoch": 0.699623143427178,
+      "grad_norm": 0.48591238260269165,
+      "learning_rate": 8.95080440495088e-06,
+      "loss": 0.0043,
+      "step": 789
+    },
+    {
+      "epoch": 0.7005098647749944,
+      "grad_norm": 0.6423205137252808,
+      "learning_rate": 8.947868305165985e-06,
+      "loss": 0.0033,
+      "step": 790
+    },
+    {
+      "epoch": 0.7013965861228109,
+      "grad_norm": 0.122403085231781,
+      "learning_rate": 8.944928585870028e-06,
+      "loss": 0.0014,
+      "step": 791
+    },
+    {
+      "epoch": 0.7022833074706274,
+      "grad_norm": 0.41814419627189636,
+      "learning_rate": 8.941985249758221e-06,
+      "loss": 0.0066,
+      "step": 792
+    },
+    {
+      "epoch": 0.7031700288184438,
+      "grad_norm": 0.6854628324508667,
+      "learning_rate": 8.939038299529094e-06,
+      "loss": 0.0102,
+      "step": 793
+    },
+    {
+      "epoch": 0.7040567501662602,
+      "grad_norm": 0.3310871720314026,
+      "learning_rate": 8.936087737884492e-06,
+      "loss": 0.0058,
+      "step": 794
+    },
+    {
+      "epoch": 0.7049434715140767,
+      "grad_norm": 0.34315598011016846,
+      "learning_rate": 8.933133567529564e-06,
+      "loss": 0.0062,
+      "step": 795
+    },
+    {
+      "epoch": 0.7058301928618932,
+      "grad_norm": 0.17736274003982544,
+      "learning_rate": 8.930175791172776e-06,
+      "loss": 0.002,
+      "step": 796
+    },
+    {
+      "epoch": 0.7067169142097096,
+      "grad_norm": 0.8943119645118713,
+      "learning_rate": 8.927214411525895e-06,
+      "loss": 0.014,
+      "step": 797
+    },
+    {
+      "epoch": 0.707603635557526,
+      "grad_norm": 0.6538609862327576,
+      "learning_rate": 8.924249431303994e-06,
+      "loss": 0.0119,
+      "step": 798
+    },
+    {
+      "epoch": 0.7084903569053425,
+      "grad_norm": 1.4449471235275269,
+      "learning_rate": 8.921280853225443e-06,
+      "loss": 0.0204,
+      "step": 799
+    },
+    {
+      "epoch": 0.709377078253159,
+      "grad_norm": 0.2964303493499756,
+      "learning_rate": 8.918308680011915e-06,
+      "loss": 0.0065,
+      "step": 800
+    },
+    {
+      "epoch": 0.7102637996009754,
+      "grad_norm": 0.912601113319397,
+      "learning_rate": 8.915332914388382e-06,
+      "loss": 0.0083,
+      "step": 801
+    },
+    {
+      "epoch": 0.7111505209487918,
+      "grad_norm": 0.9547178149223328,
+      "learning_rate": 8.912353559083098e-06,
+      "loss": 0.0111,
+      "step": 802
+    },
+    {
+      "epoch": 0.7120372422966083,
+      "grad_norm": 0.3429473042488098,
+      "learning_rate": 8.909370616827621e-06,
+      "loss": 0.0057,
+      "step": 803
+    },
+    {
+      "epoch": 0.7129239636444248,
+      "grad_norm": 0.713907778263092,
+      "learning_rate": 8.90638409035679e-06,
+      "loss": 0.006,
+      "step": 804
+    },
+    {
+      "epoch": 0.7138106849922412,
+      "grad_norm": 0.40096068382263184,
+      "learning_rate": 8.903393982408734e-06,
+      "loss": 0.0098,
+      "step": 805
+    },
+    {
+      "epoch": 0.7146974063400576,
+      "grad_norm": 0.538386344909668,
+      "learning_rate": 8.90040029572486e-06,
+      "loss": 0.0051,
+      "step": 806
+    },
+    {
+      "epoch": 0.715584127687874,
+      "grad_norm": 1.1172994375228882,
+      "learning_rate": 8.897403033049866e-06,
+      "loss": 0.0154,
+      "step": 807
+    },
+    {
+      "epoch": 0.7164708490356906,
+      "grad_norm": 0.31477785110473633,
+      "learning_rate": 8.894402197131715e-06,
+      "loss": 0.0059,
+      "step": 808
+    },
+    {
+      "epoch": 0.717357570383507,
+      "grad_norm": 1.1198912858963013,
+      "learning_rate": 8.89139779072166e-06,
+      "loss": 0.0074,
+      "step": 809
+    },
+    {
+      "epoch": 0.7182442917313234,
+      "grad_norm": 0.4623129963874817,
+      "learning_rate": 8.888389816574214e-06,
+      "loss": 0.0067,
+      "step": 810
+    },
+    {
+      "epoch": 0.7191310130791398,
+      "grad_norm": 0.400272011756897,
+      "learning_rate": 8.885378277447176e-06,
+      "loss": 0.0125,
+      "step": 811
+    },
+    {
+      "epoch": 0.7200177344269564,
+      "grad_norm": 0.4467921257019043,
+      "learning_rate": 8.882363176101601e-06,
+      "loss": 0.0122,
+      "step": 812
+    },
+    {
+      "epoch": 0.7209044557747728,
+      "grad_norm": 0.26177653670310974,
+      "learning_rate": 8.879344515301814e-06,
+      "loss": 0.0032,
+      "step": 813
+    },
+    {
+      "epoch": 0.7217911771225892,
+      "grad_norm": 0.3401440978050232,
+      "learning_rate": 8.876322297815406e-06,
+      "loss": 0.0058,
+      "step": 814
+    },
+    {
+      "epoch": 0.7226778984704056,
+      "grad_norm": 0.5830419063568115,
+      "learning_rate": 8.873296526413224e-06,
+      "loss": 0.0069,
+      "step": 815
+    },
+    {
+      "epoch": 0.7235646198182222,
+      "grad_norm": 0.4246460795402527,
+      "learning_rate": 8.870267203869379e-06,
+      "loss": 0.009,
+      "step": 816
+    },
+    {
+      "epoch": 0.7244513411660386,
+      "grad_norm": 0.3305572271347046,
+      "learning_rate": 8.867234332961232e-06,
+      "loss": 0.0052,
+      "step": 817
+    },
+    {
+      "epoch": 0.725338062513855,
+      "grad_norm": 0.23400214314460754,
+      "learning_rate": 8.864197916469404e-06,
+      "loss": 0.004,
+      "step": 818
+    },
+    {
+      "epoch": 0.7262247838616714,
+      "grad_norm": 0.46885445713996887,
+      "learning_rate": 8.861157957177757e-06,
+      "loss": 0.0044,
+      "step": 819
+    },
+    {
+      "epoch": 0.727111505209488,
+      "grad_norm": 0.17335279285907745,
+      "learning_rate": 8.858114457873411e-06,
+      "loss": 0.0017,
+      "step": 820
+    },
+    {
+      "epoch": 0.7279982265573044,
+      "grad_norm": 0.42308181524276733,
+      "learning_rate": 8.855067421346727e-06,
+      "loss": 0.0066,
+      "step": 821
+    },
+    {
+      "epoch": 0.7288849479051208,
+      "grad_norm": 0.29205307364463806,
+      "learning_rate": 8.852016850391309e-06,
+      "loss": 0.0022,
+      "step": 822
+    },
+    {
+      "epoch": 0.7297716692529372,
+      "grad_norm": 0.7200466990470886,
+      "learning_rate": 8.848962747804e-06,
+      "loss": 0.0052,
+      "step": 823
+    },
+    {
+      "epoch": 0.7306583906007537,
+      "grad_norm": 0.43672290444374084,
+      "learning_rate": 8.845905116384885e-06,
+      "loss": 0.0044,
+      "step": 824
+    },
+    {
+      "epoch": 0.7315451119485702,
+      "grad_norm": 0.07030118256807327,
+      "learning_rate": 8.842843958937282e-06,
+      "loss": 0.0004,
+      "step": 825
+    },
+    {
+      "epoch": 0.7324318332963866,
+      "grad_norm": 0.30260026454925537,
+      "learning_rate": 8.839779278267742e-06,
+      "loss": 0.0014,
+      "step": 826
+    },
+    {
+      "epoch": 0.733318554644203,
+      "grad_norm": 0.479753702878952,
+      "learning_rate": 8.836711077186045e-06,
+      "loss": 0.0025,
+      "step": 827
+    },
+    {
+      "epoch": 0.7342052759920195,
+      "grad_norm": 0.42094236612319946,
+      "learning_rate": 8.833639358505202e-06,
+      "loss": 0.0035,
+      "step": 828
+    },
+    {
+      "epoch": 0.735091997339836,
+      "grad_norm": 0.7436134815216064,
+      "learning_rate": 8.830564125041443e-06,
+      "loss": 0.0041,
+      "step": 829
+    },
+    {
+      "epoch": 0.7359787186876524,
+      "grad_norm": 0.3896116018295288,
+      "learning_rate": 8.827485379614224e-06,
+      "loss": 0.0021,
+      "step": 830
+    },
+    {
+      "epoch": 0.7368654400354688,
+      "grad_norm": 0.4691396653652191,
+      "learning_rate": 8.824403125046225e-06,
+      "loss": 0.0075,
+      "step": 831
+    },
+    {
+      "epoch": 0.7377521613832853,
+      "grad_norm": 0.3711012303829193,
+      "learning_rate": 8.821317364163338e-06,
+      "loss": 0.0017,
+      "step": 832
+    },
+    {
+      "epoch": 0.7386388827311018,
+      "grad_norm": 0.4321559965610504,
+      "learning_rate": 8.81822809979467e-06,
+      "loss": 0.0047,
+      "step": 833
+    },
+    {
+      "epoch": 0.7395256040789182,
+      "grad_norm": 0.9089894890785217,
+      "learning_rate": 8.815135334772539e-06,
+      "loss": 0.0032,
+      "step": 834
+    },
+    {
+      "epoch": 0.7404123254267346,
+      "grad_norm": 0.034477680921554565,
+      "learning_rate": 8.81203907193248e-06,
+      "loss": 0.0002,
+      "step": 835
+    },
+    {
+      "epoch": 0.741299046774551,
+      "grad_norm": 0.4759688675403595,
+      "learning_rate": 8.808939314113225e-06,
+      "loss": 0.0115,
+      "step": 836
+    },
+    {
+      "epoch": 0.7421857681223676,
+      "grad_norm": 0.6464022397994995,
+      "learning_rate": 8.805836064156716e-06,
+      "loss": 0.0041,
+      "step": 837
+    },
+    {
+      "epoch": 0.743072489470184,
+      "grad_norm": 0.7758044004440308,
+      "learning_rate": 8.802729324908095e-06,
+      "loss": 0.013,
+      "step": 838
+    },
+    {
+      "epoch": 0.7439592108180004,
+      "grad_norm": 0.10477610677480698,
+      "learning_rate": 8.799619099215705e-06,
+      "loss": 0.0005,
+      "step": 839
+    },
+    {
+      "epoch": 0.7448459321658168,
+      "grad_norm": 1.6846563816070557,
+      "learning_rate": 8.796505389931082e-06,
+      "loss": 0.0206,
+      "step": 840
+    },
+    {
+      "epoch": 0.7457326535136334,
+      "grad_norm": 0.13585907220840454,
+      "learning_rate": 8.793388199908959e-06,
+      "loss": 0.0009,
+      "step": 841
+    },
+    {
+      "epoch": 0.7466193748614498,
+      "grad_norm": 0.3324284553527832,
+      "learning_rate": 8.79026753200726e-06,
+      "loss": 0.0037,
+      "step": 842
+    },
+    {
+      "epoch": 0.7475060962092662,
+      "grad_norm": 0.21177026629447937,
+      "learning_rate": 8.787143389087093e-06,
+      "loss": 0.0017,
+      "step": 843
+    },
+    {
+      "epoch": 0.7483928175570826,
+      "grad_norm": 0.45269775390625,
+      "learning_rate": 8.784015774012758e-06,
+      "loss": 0.0077,
+      "step": 844
+    },
+    {
+      "epoch": 0.7492795389048992,
+      "grad_norm": 0.2514367699623108,
+      "learning_rate": 8.780884689651738e-06,
+      "loss": 0.0032,
+      "step": 845
+    },
+    {
+      "epoch": 0.7501662602527156,
+      "grad_norm": 0.4557270109653473,
+      "learning_rate": 8.77775013887469e-06,
+      "loss": 0.0057,
+      "step": 846
+    },
+    {
+      "epoch": 0.751052981600532,
+      "grad_norm": 0.4305650293827057,
+      "learning_rate": 8.77461212455546e-06,
+      "loss": 0.0039,
+      "step": 847
+    },
+    {
+      "epoch": 0.7519397029483484,
+      "grad_norm": 0.3271408975124359,
+      "learning_rate": 8.771470649571056e-06,
+      "loss": 0.0032,
+      "step": 848
+    },
+    {
+      "epoch": 0.752826424296165,
+      "grad_norm": 0.8781595230102539,
+      "learning_rate": 8.768325716801673e-06,
+      "loss": 0.0094,
+      "step": 849
+    },
+    {
+      "epoch": 0.7537131456439814,
+      "grad_norm": 0.28247398138046265,
+      "learning_rate": 8.765177329130665e-06,
+      "loss": 0.0063,
+      "step": 850
+    },
+    {
+      "epoch": 0.7545998669917978,
+      "grad_norm": 0.5735777020454407,
+      "learning_rate": 8.76202548944456e-06,
+      "loss": 0.0037,
+      "step": 851
+    },
+    {
+      "epoch": 0.7554865883396142,
+      "grad_norm": 0.14849910140037537,
+      "learning_rate": 8.758870200633047e-06,
+      "loss": 0.0011,
+      "step": 852
+    },
+    {
+      "epoch": 0.7563733096874308,
+      "grad_norm": 0.23138736188411713,
+      "learning_rate": 8.755711465588981e-06,
+      "loss": 0.0062,
+      "step": 853
+    },
+    {
+      "epoch": 0.7572600310352472,
+      "grad_norm": 0.2745361924171448,
+      "learning_rate": 8.752549287208373e-06,
+      "loss": 0.0017,
+      "step": 854
+    },
+    {
+      "epoch": 0.7581467523830636,
+      "grad_norm": 0.31626009941101074,
+      "learning_rate": 8.749383668390393e-06,
+      "loss": 0.0026,
+      "step": 855
+    },
+    {
+      "epoch": 0.75903347373088,
+      "grad_norm": 0.29030266404151917,
+      "learning_rate": 8.746214612037365e-06,
+      "loss": 0.0055,
+      "step": 856
+    },
+    {
+      "epoch": 0.7599201950786966,
+      "grad_norm": 0.4959680736064911,
+      "learning_rate": 8.743042121054766e-06,
+      "loss": 0.0041,
+      "step": 857
+    },
+    {
+      "epoch": 0.760806916426513,
+      "grad_norm": 0.1988179087638855,
+      "learning_rate": 8.739866198351218e-06,
+      "loss": 0.0017,
+      "step": 858
+    },
+    {
+      "epoch": 0.7616936377743294,
+      "grad_norm": 0.06984910368919373,
+      "learning_rate": 8.736686846838491e-06,
+      "loss": 0.0005,
+      "step": 859
+    },
+    {
+      "epoch": 0.7625803591221458,
+      "grad_norm": 0.7530263662338257,
+      "learning_rate": 8.733504069431502e-06,
+      "loss": 0.0099,
+      "step": 860
+    },
+    {
+      "epoch": 0.7634670804699624,
+      "grad_norm": 0.47911015152931213,
+      "learning_rate": 8.730317869048306e-06,
+      "loss": 0.005,
+      "step": 861
+    },
+    {
+      "epoch": 0.7643538018177788,
+      "grad_norm": 0.7026636600494385,
+      "learning_rate": 8.727128248610096e-06,
+      "loss": 0.0083,
+      "step": 862
+    },
+    {
+      "epoch": 0.7652405231655952,
+      "grad_norm": 0.5120250582695007,
+      "learning_rate": 8.723935211041198e-06,
+      "loss": 0.006,
+      "step": 863
+    },
+    {
+      "epoch": 0.7661272445134116,
+      "grad_norm": 0.37591156363487244,
+      "learning_rate": 8.72073875926908e-06,
+      "loss": 0.003,
+      "step": 864
+    },
+    {
+      "epoch": 0.7670139658612282,
+      "grad_norm": 0.333332896232605,
+      "learning_rate": 8.717538896224333e-06,
+      "loss": 0.0045,
+      "step": 865
+    },
+    {
+      "epoch": 0.7679006872090446,
+      "grad_norm": 0.3260776996612549,
+      "learning_rate": 8.714335624840674e-06,
+      "loss": 0.003,
+      "step": 866
+    },
+    {
+      "epoch": 0.768787408556861,
+      "grad_norm": 0.44862043857574463,
+      "learning_rate": 8.711128948054947e-06,
+      "loss": 0.007,
+      "step": 867
+    },
+    {
+      "epoch": 0.7696741299046774,
+      "grad_norm": 0.8102560639381409,
+      "learning_rate": 8.707918868807123e-06,
+      "loss": 0.0022,
+      "step": 868
+    },
+    {
+      "epoch": 0.7705608512524939,
+      "grad_norm": 0.3125271797180176,
+      "learning_rate": 8.704705390040285e-06,
+      "loss": 0.0044,
+      "step": 869
+    },
+    {
+      "epoch": 0.7714475726003104,
+      "grad_norm": 0.35220733284950256,
+      "learning_rate": 8.701488514700638e-06,
+      "loss": 0.003,
+      "step": 870
+    },
+    {
+      "epoch": 0.7723342939481268,
+      "grad_norm": 0.32635432481765747,
+      "learning_rate": 8.6982682457375e-06,
+      "loss": 0.0012,
+      "step": 871
+    },
+    {
+      "epoch": 0.7732210152959432,
+      "grad_norm": 0.27862560749053955,
+      "learning_rate": 8.695044586103297e-06,
+      "loss": 0.0024,
+      "step": 872
+    },
+    {
+      "epoch": 0.7741077366437596,
+      "grad_norm": 0.11573431640863419,
+      "learning_rate": 8.691817538753566e-06,
+      "loss": 0.0005,
+      "step": 873
+    },
+    {
+      "epoch": 0.7749944579915762,
+      "grad_norm": 1.2635929584503174,
+      "learning_rate": 8.688587106646953e-06,
+      "loss": 0.0134,
+      "step": 874
+    },
+    {
+      "epoch": 0.7758811793393926,
+      "grad_norm": 0.4768292009830475,
+      "learning_rate": 8.685353292745203e-06,
+      "loss": 0.0062,
+      "step": 875
+    },
+    {
+      "epoch": 0.776767900687209,
+      "grad_norm": 0.44109392166137695,
+      "learning_rate": 8.682116100013163e-06,
+      "loss": 0.0058,
+      "step": 876
+    },
+    {
+      "epoch": 0.7776546220350254,
+      "grad_norm": 0.13942331075668335,
+      "learning_rate": 8.678875531418778e-06,
+      "loss": 0.0012,
+      "step": 877
+    },
+    {
+      "epoch": 0.778541343382842,
+      "grad_norm": 0.3853679597377777,
+      "learning_rate": 8.675631589933086e-06,
+      "loss": 0.0044,
+      "step": 878
+    },
+    {
+      "epoch": 0.7794280647306584,
+      "grad_norm": 0.7974981069564819,
+      "learning_rate": 8.672384278530222e-06,
+      "loss": 0.0141,
+      "step": 879
+    },
+    {
+      "epoch": 0.7803147860784748,
+      "grad_norm": 0.5448789000511169,
+      "learning_rate": 8.669133600187406e-06,
+      "loss": 0.0072,
+      "step": 880
+    },
+    {
+      "epoch": 0.7812015074262912,
+      "grad_norm": 0.6990935802459717,
+      "learning_rate": 8.665879557884946e-06,
+      "loss": 0.0113,
+      "step": 881
+    },
+    {
+      "epoch": 0.7820882287741078,
+      "grad_norm": 0.3424188494682312,
+      "learning_rate": 8.662622154606238e-06,
+      "loss": 0.0046,
+      "step": 882
+    },
+    {
+      "epoch": 0.7829749501219242,
+      "grad_norm": 0.22132602334022522,
+      "learning_rate": 8.659361393337752e-06,
+      "loss": 0.0023,
+      "step": 883
+    },
+    {
+      "epoch": 0.7838616714697406,
+      "grad_norm": 0.4831432104110718,
+      "learning_rate": 8.656097277069044e-06,
+      "loss": 0.0041,
+      "step": 884
+    },
+    {
+      "epoch": 0.784748392817557,
+      "grad_norm": 0.5113421082496643,
+      "learning_rate": 8.652829808792742e-06,
+      "loss": 0.0093,
+      "step": 885
+    },
+    {
+      "epoch": 0.7856351141653736,
+      "grad_norm": 0.36989861726760864,
+      "learning_rate": 8.64955899150455e-06,
+      "loss": 0.0039,
+      "step": 886
+    },
+    {
+      "epoch": 0.78652183551319,
+      "grad_norm": 0.3535897433757782,
+      "learning_rate": 8.646284828203236e-06,
+      "loss": 0.0124,
+      "step": 887
+    },
+    {
+      "epoch": 0.7874085568610064,
+      "grad_norm": 0.6427195072174072,
+      "learning_rate": 8.643007321890646e-06,
+      "loss": 0.005,
+      "step": 888
+    },
+    {
+      "epoch": 0.7882952782088228,
+      "grad_norm": 0.12736086547374725,
+      "learning_rate": 8.63972647557168e-06,
+      "loss": 0.0021,
+      "step": 889
+    },
+    {
+      "epoch": 0.7891819995566394,
+      "grad_norm": 0.1027836948633194,
+      "learning_rate": 8.63644229225431e-06,
+      "loss": 0.0017,
+      "step": 890
+    },
+    {
+      "epoch": 0.7900687209044558,
+      "grad_norm": 0.41020867228507996,
+      "learning_rate": 8.633154774949559e-06,
+      "loss": 0.0076,
+      "step": 891
+    },
+    {
+      "epoch": 0.7909554422522722,
+      "grad_norm": 0.2587355375289917,
+      "learning_rate": 8.629863926671514e-06,
+      "loss": 0.0027,
+      "step": 892
+    },
+    {
+      "epoch": 0.7918421636000886,
+      "grad_norm": 0.32279083132743835,
+      "learning_rate": 8.626569750437312e-06,
+      "loss": 0.0027,
+      "step": 893
+    },
+    {
+      "epoch": 0.7927288849479052,
+      "grad_norm": 0.26439985632896423,
+      "learning_rate": 8.623272249267139e-06,
+      "loss": 0.002,
+      "step": 894
+    },
+    {
+      "epoch": 0.7936156062957216,
+      "grad_norm": 0.4109850823879242,
+      "learning_rate": 8.619971426184236e-06,
+      "loss": 0.0094,
+      "step": 895
+    },
+    {
+      "epoch": 0.794502327643538,
+      "grad_norm": 0.2232225239276886,
+      "learning_rate": 8.616667284214882e-06,
+      "loss": 0.0037,
+      "step": 896
+    },
+    {
+      "epoch": 0.7953890489913544,
+      "grad_norm": 0.4170536696910858,
+      "learning_rate": 8.613359826388404e-06,
+      "loss": 0.0108,
+      "step": 897
+    },
+    {
+      "epoch": 0.796275770339171,
+      "grad_norm": 0.3371315598487854,
+      "learning_rate": 8.610049055737168e-06,
+      "loss": 0.0026,
+      "step": 898
+    },
+    {
+      "epoch": 0.7971624916869874,
+      "grad_norm": 0.6959087252616882,
+      "learning_rate": 8.606734975296578e-06,
+      "loss": 0.0085,
+      "step": 899
+    },
+    {
+      "epoch": 0.7980492130348038,
+      "grad_norm": 0.38720524311065674,
+      "learning_rate": 8.60341758810507e-06,
+      "loss": 0.0024,
+      "step": 900
+    },
+    {
+      "epoch": 0.7989359343826202,
+      "grad_norm": 0.39712467789649963,
+      "learning_rate": 8.600096897204113e-06,
+      "loss": 0.0038,
+      "step": 901
+    },
+    {
+      "epoch": 0.7998226557304368,
+      "grad_norm": 0.2283264547586441,
+      "learning_rate": 8.596772905638207e-06,
+      "loss": 0.0017,
+      "step": 902
+    },
+    {
+      "epoch": 0.8007093770782532,
+      "grad_norm": 0.13432735204696655,
+      "learning_rate": 8.593445616454873e-06,
+      "loss": 0.0018,
+      "step": 903
+    },
+    {
+      "epoch": 0.8015960984260696,
+      "grad_norm": 0.35728439688682556,
+      "learning_rate": 8.59011503270466e-06,
+      "loss": 0.0057,
+      "step": 904
+    },
+    {
+      "epoch": 0.802482819773886,
+      "grad_norm": 0.6101589798927307,
+      "learning_rate": 8.586781157441138e-06,
+      "loss": 0.0057,
+      "step": 905
+    },
+    {
+      "epoch": 0.8033695411217026,
+      "grad_norm": 0.2695213854312897,
+      "learning_rate": 8.58344399372089e-06,
+      "loss": 0.0025,
+      "step": 906
+    },
+    {
+      "epoch": 0.804256262469519,
+      "grad_norm": 0.37189701199531555,
+      "learning_rate": 8.580103544603517e-06,
+      "loss": 0.0043,
+      "step": 907
+    },
+    {
+      "epoch": 0.8051429838173354,
+      "grad_norm": 0.4200766682624817,
+      "learning_rate": 8.576759813151635e-06,
+      "loss": 0.0017,
+      "step": 908
+    },
+    {
+      "epoch": 0.8060297051651518,
+      "grad_norm": 0.31807032227516174,
+      "learning_rate": 8.573412802430863e-06,
+      "loss": 0.003,
+      "step": 909
+    },
+    {
+      "epoch": 0.8069164265129684,
+      "grad_norm": 0.6575264930725098,
+      "learning_rate": 8.57006251550983e-06,
+      "loss": 0.0048,
+      "step": 910
+    },
+    {
+      "epoch": 0.8078031478607848,
+      "grad_norm": 0.33519044518470764,
+      "learning_rate": 8.566708955460172e-06,
+      "loss": 0.0043,
+      "step": 911
+    },
+    {
+      "epoch": 0.8086898692086012,
+      "grad_norm": 0.7663727402687073,
+      "learning_rate": 8.56335212535652e-06,
+      "loss": 0.0068,
+      "step": 912
+    },
+    {
+      "epoch": 0.8095765905564176,
+      "grad_norm": 0.2573295831680298,
+      "learning_rate": 8.559992028276502e-06,
+      "loss": 0.0051,
+      "step": 913
+    },
+    {
+      "epoch": 0.810463311904234,
+      "grad_norm": 0.49369877576828003,
+      "learning_rate": 8.556628667300752e-06,
+      "loss": 0.0118,
+      "step": 914
+    },
+    {
+      "epoch": 0.8113500332520506,
+      "grad_norm": 0.10165636241436005,
+      "learning_rate": 8.553262045512882e-06,
+      "loss": 0.0004,
+      "step": 915
+    },
+    {
+      "epoch": 0.812236754599867,
+      "grad_norm": 0.7126047611236572,
+      "learning_rate": 8.549892165999505e-06,
+      "loss": 0.0069,
+      "step": 916
+    },
+    {
+      "epoch": 0.8131234759476834,
+      "grad_norm": 0.2403372973203659,
+      "learning_rate": 8.546519031850216e-06,
+      "loss": 0.0026,
+      "step": 917
+    },
+    {
+      "epoch": 0.8140101972954998,
+      "grad_norm": 0.5325596332550049,
+      "learning_rate": 8.543142646157594e-06,
+      "loss": 0.0066,
+      "step": 918
+    },
+    {
+      "epoch": 0.8148969186433164,
+      "grad_norm": 0.09692367166280746,
+      "learning_rate": 8.5397630120172e-06,
+      "loss": 0.001,
+      "step": 919
+    },
+    {
+      "epoch": 0.8157836399911328,
+      "grad_norm": 0.43505728244781494,
+      "learning_rate": 8.536380132527572e-06,
+      "loss": 0.0049,
+      "step": 920
+    },
+    {
+      "epoch": 0.8166703613389492,
+      "grad_norm": 1.0189909934997559,
+      "learning_rate": 8.53299401079022e-06,
+      "loss": 0.0202,
+      "step": 921
+    },
+    {
+      "epoch": 0.8175570826867656,
+      "grad_norm": 0.151530921459198,
+      "learning_rate": 8.52960464990964e-06,
+      "loss": 0.0016,
+      "step": 922
+    },
+    {
+      "epoch": 0.8184438040345822,
+      "grad_norm": 0.11746247112751007,
+      "learning_rate": 8.526212052993284e-06,
+      "loss": 0.0017,
+      "step": 923
+    },
+    {
+      "epoch": 0.8193305253823986,
+      "grad_norm": 0.25555944442749023,
+      "learning_rate": 8.522816223151573e-06,
+      "loss": 0.0018,
+      "step": 924
+    },
+    {
+      "epoch": 0.820217246730215,
+      "grad_norm": 0.13405092060565948,
+      "learning_rate": 8.519417163497898e-06,
+      "loss": 0.0024,
+      "step": 925
+    },
+    {
+      "epoch": 0.8211039680780314,
+      "grad_norm": 0.31290844082832336,
+      "learning_rate": 8.516014877148605e-06,
+      "loss": 0.0037,
+      "step": 926
+    },
+    {
+      "epoch": 0.821990689425848,
+      "grad_norm": 0.13099424540996552,
+      "learning_rate": 8.512609367223005e-06,
+      "loss": 0.0018,
+      "step": 927
+    },
+    {
+      "epoch": 0.8228774107736644,
+      "grad_norm": 0.18229612708091736,
+      "learning_rate": 8.509200636843354e-06,
+      "loss": 0.0025,
+      "step": 928
+    },
+    {
+      "epoch": 0.8237641321214808,
+      "grad_norm": 0.20748937129974365,
+      "learning_rate": 8.505788689134872e-06,
+      "loss": 0.0044,
+      "step": 929
+    },
+    {
+      "epoch": 0.8246508534692972,
+      "grad_norm": 0.28571903705596924,
+      "learning_rate": 8.502373527225725e-06,
+      "loss": 0.0035,
+      "step": 930
+    },
+    {
+      "epoch": 0.8255375748171138,
+      "grad_norm": 0.0889572873711586,
+      "learning_rate": 8.49895515424702e-06,
+      "loss": 0.0018,
+      "step": 931
+    },
+    {
+      "epoch": 0.8264242961649302,
+      "grad_norm": 0.16316168010234833,
+      "learning_rate": 8.495533573332814e-06,
+      "loss": 0.0017,
+      "step": 932
+    },
+    {
+      "epoch": 0.8273110175127466,
+      "grad_norm": 0.23393435776233673,
+      "learning_rate": 8.492108787620106e-06,
+      "loss": 0.0032,
+      "step": 933
+    },
+    {
+      "epoch": 0.828197738860563,
+      "grad_norm": 0.49376270174980164,
+      "learning_rate": 8.488680800248828e-06,
+      "loss": 0.0086,
+      "step": 934
+    },
+    {
+      "epoch": 0.8290844602083796,
+      "grad_norm": 0.3670783042907715,
+      "learning_rate": 8.485249614361852e-06,
+      "loss": 0.0024,
+      "step": 935
+    },
+    {
+      "epoch": 0.829971181556196,
+      "grad_norm": 0.26419225335121155,
+      "learning_rate": 8.481815233104982e-06,
+      "loss": 0.0016,
+      "step": 936
+    },
+    {
+      "epoch": 0.8308579029040124,
+      "grad_norm": 0.343707799911499,
+      "learning_rate": 8.478377659626949e-06,
+      "loss": 0.005,
+      "step": 937
+    },
+    {
+      "epoch": 0.8317446242518288,
+      "grad_norm": 0.035789605230093,
+      "learning_rate": 8.47493689707941e-06,
+      "loss": 0.0003,
+      "step": 938
+    },
+    {
+      "epoch": 0.8326313455996454,
+      "grad_norm": 0.5384289622306824,
+      "learning_rate": 8.471492948616953e-06,
+      "loss": 0.0126,
+      "step": 939
+    },
+    {
+      "epoch": 0.8335180669474618,
+      "grad_norm": 0.26499831676483154,
+      "learning_rate": 8.468045817397076e-06,
+      "loss": 0.0015,
+      "step": 940
+    },
+    {
+      "epoch": 0.8344047882952782,
+      "grad_norm": 0.5022488236427307,
+      "learning_rate": 8.464595506580207e-06,
+      "loss": 0.0058,
+      "step": 941
+    },
+    {
+      "epoch": 0.8352915096430946,
+      "grad_norm": 0.8323764801025391,
+      "learning_rate": 8.461142019329677e-06,
+      "loss": 0.0088,
+      "step": 942
+    },
+    {
+      "epoch": 0.8361782309909112,
+      "grad_norm": 0.3151646852493286,
+      "learning_rate": 8.45768535881174e-06,
+      "loss": 0.0044,
+      "step": 943
+    },
+    {
+      "epoch": 0.8370649523387276,
+      "grad_norm": 0.19980305433273315,
+      "learning_rate": 8.454225528195553e-06,
+      "loss": 0.001,
+      "step": 944
+    },
+    {
+      "epoch": 0.837951673686544,
+      "grad_norm": 0.3161185383796692,
+      "learning_rate": 8.45076253065318e-06,
+      "loss": 0.0045,
+      "step": 945
+    },
+    {
+      "epoch": 0.8388383950343604,
+      "grad_norm": 0.42954885959625244,
+      "learning_rate": 8.447296369359593e-06,
+      "loss": 0.0016,
+      "step": 946
+    },
+    {
+      "epoch": 0.839725116382177,
+      "grad_norm": 0.23368383944034576,
+      "learning_rate": 8.443827047492657e-06,
+      "loss": 0.0016,
+      "step": 947
+    },
+    {
+      "epoch": 0.8406118377299934,
+      "grad_norm": 0.14157956838607788,
+      "learning_rate": 8.44035456823314e-06,
+      "loss": 0.0008,
+      "step": 948
+    },
+    {
+      "epoch": 0.8414985590778098,
+      "grad_norm": 0.8546500205993652,
+      "learning_rate": 8.436878934764705e-06,
+      "loss": 0.0051,
+      "step": 949
+    },
+    {
+      "epoch": 0.8423852804256262,
+      "grad_norm": 0.16669033467769623,
+      "learning_rate": 8.433400150273907e-06,
+      "loss": 0.0008,
+      "step": 950
+    },
+    {
+      "epoch": 0.8432720017734427,
+      "grad_norm": 0.1262633204460144,
+      "learning_rate": 8.429918217950184e-06,
+      "loss": 0.0008,
+      "step": 951
+    },
+    {
+      "epoch": 0.8441587231212592,
+      "grad_norm": 0.38040781021118164,
+      "learning_rate": 8.426433140985867e-06,
+      "loss": 0.0023,
+      "step": 952
+    },
+    {
+      "epoch": 0.8450454444690756,
+      "grad_norm": 0.2027808576822281,
+      "learning_rate": 8.42294492257617e-06,
+      "loss": 0.0014,
+      "step": 953
+    },
+    {
+      "epoch": 0.845932165816892,
+      "grad_norm": 1.2989972829818726,
+      "learning_rate": 8.419453565919183e-06,
+      "loss": 0.007,
+      "step": 954
+    },
+    {
+      "epoch": 0.8468188871647085,
+      "grad_norm": 0.4618437588214874,
+      "learning_rate": 8.415959074215871e-06,
+      "loss": 0.0033,
+      "step": 955
+    },
+    {
+      "epoch": 0.847705608512525,
+      "grad_norm": 0.41583216190338135,
+      "learning_rate": 8.412461450670083e-06,
+      "loss": 0.0073,
+      "step": 956
+    },
+    {
+      "epoch": 0.8485923298603414,
+      "grad_norm": 0.26886528730392456,
+      "learning_rate": 8.408960698488531e-06,
+      "loss": 0.0009,
+      "step": 957
+    },
+    {
+      "epoch": 0.8494790512081578,
+      "grad_norm": 0.6726037859916687,
+      "learning_rate": 8.405456820880797e-06,
+      "loss": 0.0083,
+      "step": 958
+    },
+    {
+      "epoch": 0.8503657725559742,
+      "grad_norm": 0.9834551811218262,
+      "learning_rate": 8.401949821059331e-06,
+      "loss": 0.0086,
+      "step": 959
+    },
+    {
+      "epoch": 0.8512524939037908,
+      "grad_norm": 0.29446637630462646,
+      "learning_rate": 8.398439702239447e-06,
+      "loss": 0.0011,
+      "step": 960
+    },
+    {
+      "epoch": 0.8521392152516072,
+      "grad_norm": 0.6566726565361023,
+      "learning_rate": 8.39492646763931e-06,
+      "loss": 0.0047,
+      "step": 961
+    },
+    {
+      "epoch": 0.8530259365994236,
+      "grad_norm": 0.18082024157047272,
+      "learning_rate": 8.391410120479952e-06,
+      "loss": 0.002,
+      "step": 962
+    },
+    {
+      "epoch": 0.85391265794724,
+      "grad_norm": 0.038998160511255264,
+      "learning_rate": 8.38789066398525e-06,
+      "loss": 0.0004,
+      "step": 963
+    },
+    {
+      "epoch": 0.8547993792950566,
+      "grad_norm": 0.19278480112552643,
+      "learning_rate": 8.384368101381939e-06,
+      "loss": 0.0047,
+      "step": 964
+    },
+    {
+      "epoch": 0.855686100642873,
+      "grad_norm": 0.36077624559402466,
+      "learning_rate": 8.380842435899595e-06,
+      "loss": 0.0054,
+      "step": 965
+    },
+    {
+      "epoch": 0.8565728219906894,
+      "grad_norm": 0.6876051425933838,
+      "learning_rate": 8.377313670770647e-06,
+      "loss": 0.0052,
+      "step": 966
+    },
+    {
+      "epoch": 0.8574595433385058,
+      "grad_norm": 0.2629197835922241,
+      "learning_rate": 8.373781809230355e-06,
+      "loss": 0.0022,
+      "step": 967
+    },
+    {
+      "epoch": 0.8583462646863224,
+      "grad_norm": 0.4707011282444,
+      "learning_rate": 8.37024685451683e-06,
+      "loss": 0.0068,
+      "step": 968
+    },
+    {
+      "epoch": 0.8592329860341388,
+      "grad_norm": 0.639545202255249,
+      "learning_rate": 8.366708809871009e-06,
+      "loss": 0.0087,
+      "step": 969
+    },
+    {
+      "epoch": 0.8601197073819552,
+      "grad_norm": 0.47514888644218445,
+      "learning_rate": 8.363167678536667e-06,
+      "loss": 0.0027,
+      "step": 970
+    },
+    {
+      "epoch": 0.8610064287297716,
+      "grad_norm": 0.32375380396842957,
+      "learning_rate": 8.359623463760408e-06,
+      "loss": 0.0033,
+      "step": 971
+    },
+    {
+      "epoch": 0.8618931500775882,
+      "grad_norm": 0.1954675018787384,
+      "learning_rate": 8.356076168791663e-06,
+      "loss": 0.0025,
+      "step": 972
+    },
+    {
+      "epoch": 0.8627798714254046,
+      "grad_norm": 0.4773697853088379,
+      "learning_rate": 8.352525796882685e-06,
+      "loss": 0.008,
+      "step": 973
+    },
+    {
+      "epoch": 0.863666592773221,
+      "grad_norm": 0.5595993399620056,
+      "learning_rate": 8.34897235128855e-06,
+      "loss": 0.0074,
+      "step": 974
+    },
+    {
+      "epoch": 0.8645533141210374,
+      "grad_norm": 0.6411868333816528,
+      "learning_rate": 8.345415835267154e-06,
+      "loss": 0.0071,
+      "step": 975
+    },
+    {
+      "epoch": 0.865440035468854,
+      "grad_norm": 0.2740042209625244,
+      "learning_rate": 8.341856252079206e-06,
+      "loss": 0.004,
+      "step": 976
+    },
+    {
+      "epoch": 0.8663267568166704,
+      "grad_norm": 0.43187814950942993,
+      "learning_rate": 8.338293604988223e-06,
+      "loss": 0.0031,
+      "step": 977
+    },
+    {
+      "epoch": 0.8672134781644868,
+      "grad_norm": 0.24849332869052887,
+      "learning_rate": 8.33472789726054e-06,
+      "loss": 0.0035,
+      "step": 978
+    },
+    {
+      "epoch": 0.8681001995123032,
+      "grad_norm": 0.24358658492565155,
+      "learning_rate": 8.331159132165287e-06,
+      "loss": 0.0016,
+      "step": 979
+    },
+    {
+      "epoch": 0.8689869208601197,
+      "grad_norm": 0.553687572479248,
+      "learning_rate": 8.327587312974411e-06,
+      "loss": 0.0057,
+      "step": 980
+    },
+    {
+      "epoch": 0.8698736422079362,
+      "grad_norm": 0.23114646971225739,
+      "learning_rate": 8.324012442962644e-06,
+      "loss": 0.0026,
+      "step": 981
+    },
+    {
+      "epoch": 0.8707603635557526,
+      "grad_norm": 0.12544259428977966,
+      "learning_rate": 8.320434525407524e-06,
+      "loss": 0.0015,
+      "step": 982
+    },
+    {
+      "epoch": 0.871647084903569,
+      "grad_norm": 0.7338998317718506,
+      "learning_rate": 8.316853563589383e-06,
+      "loss": 0.0144,
+      "step": 983
+    },
+    {
+      "epoch": 0.8725338062513855,
+      "grad_norm": 0.38078662753105164,
+      "learning_rate": 8.313269560791343e-06,
+      "loss": 0.0033,
+      "step": 984
+    },
+    {
+      "epoch": 0.873420527599202,
+      "grad_norm": 0.37246477603912354,
+      "learning_rate": 8.309682520299312e-06,
+      "loss": 0.0047,
+      "step": 985
+    },
+    {
+      "epoch": 0.8743072489470184,
+      "grad_norm": 0.2513933777809143,
+      "learning_rate": 8.306092445401984e-06,
+      "loss": 0.0028,
+      "step": 986
+    },
+    {
+      "epoch": 0.8751939702948348,
+      "grad_norm": 0.7336315512657166,
+      "learning_rate": 8.302499339390836e-06,
+      "loss": 0.011,
+      "step": 987
+    },
+    {
+      "epoch": 0.8760806916426513,
+      "grad_norm": 0.07155678421258926,
+      "learning_rate": 8.298903205560123e-06,
+      "loss": 0.0008,
+      "step": 988
+    },
+    {
+      "epoch": 0.8769674129904678,
+      "grad_norm": 0.4643585681915283,
+      "learning_rate": 8.295304047206878e-06,
+      "loss": 0.003,
+      "step": 989
+    },
+    {
+      "epoch": 0.8778541343382842,
+      "grad_norm": 0.6717381477355957,
+      "learning_rate": 8.291701867630906e-06,
+      "loss": 0.0063,
+      "step": 990
+    },
+    {
+      "epoch": 0.8787408556861006,
+      "grad_norm": 0.491864949464798,
+      "learning_rate": 8.28809667013478e-06,
+      "loss": 0.0067,
+      "step": 991
+    },
+    {
+      "epoch": 0.8796275770339171,
+      "grad_norm": 0.6167854070663452,
+      "learning_rate": 8.284488458023841e-06,
+      "loss": 0.0048,
+      "step": 992
+    },
+    {
+      "epoch": 0.8805142983817336,
+      "grad_norm": 0.5936487317085266,
+      "learning_rate": 8.280877234606193e-06,
+      "loss": 0.0035,
+      "step": 993
+    },
+    {
+      "epoch": 0.88140101972955,
+      "grad_norm": 0.7591543793678284,
+      "learning_rate": 8.277263003192706e-06,
+      "loss": 0.0104,
+      "step": 994
+    },
+    {
+      "epoch": 0.8822877410773664,
+      "grad_norm": 0.43168148398399353,
+      "learning_rate": 8.273645767097001e-06,
+      "loss": 0.0035,
+      "step": 995
+    },
+    {
+      "epoch": 0.8831744624251829,
+      "grad_norm": 0.2748768627643585,
+      "learning_rate": 8.270025529635455e-06,
+      "loss": 0.0029,
+      "step": 996
+    },
+    {
+      "epoch": 0.8840611837729994,
+      "grad_norm": 0.3771921992301941,
+      "learning_rate": 8.266402294127203e-06,
+      "loss": 0.0061,
+      "step": 997
+    },
+    {
+      "epoch": 0.8849479051208158,
+      "grad_norm": 0.5574544072151184,
+      "learning_rate": 8.26277606389412e-06,
+      "loss": 0.0078,
+      "step": 998
+    },
+    {
+      "epoch": 0.8858346264686322,
+      "grad_norm": 0.36976248025894165,
+      "learning_rate": 8.259146842260834e-06,
+      "loss": 0.0059,
+      "step": 999
+    },
+    {
+      "epoch": 0.8867213478164487,
+      "grad_norm": 0.9293411374092102,
+      "learning_rate": 8.255514632554709e-06,
+      "loss": 0.0095,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8876080691642652,
+      "grad_norm": 0.5999060273170471,
+      "learning_rate": 8.251879438105854e-06,
+      "loss": 0.0062,
+      "step": 1001
+    },
+    {
+      "epoch": 0.8884947905120816,
+      "grad_norm": 0.6218252182006836,
+      "learning_rate": 8.24824126224711e-06,
+      "loss": 0.0092,
+      "step": 1002
+    },
+    {
+      "epoch": 0.889381511859898,
+      "grad_norm": 0.4173696041107178,
+      "learning_rate": 8.244600108314058e-06,
+      "loss": 0.0057,
+      "step": 1003
+    },
+    {
+      "epoch": 0.8902682332077144,
+      "grad_norm": 0.12720154225826263,
+      "learning_rate": 8.240955979645001e-06,
+      "loss": 0.0015,
+      "step": 1004
+    },
+    {
+      "epoch": 0.891154954555531,
+      "grad_norm": 0.8230817317962646,
+      "learning_rate": 8.237308879580974e-06,
+      "loss": 0.0071,
+      "step": 1005
+    },
+    {
+      "epoch": 0.8920416759033474,
+      "grad_norm": 0.7403504252433777,
+      "learning_rate": 8.23365881146574e-06,
+      "loss": 0.01,
+      "step": 1006
+    },
+    {
+      "epoch": 0.8929283972511638,
+      "grad_norm": 0.3044225871562958,
+      "learning_rate": 8.230005778645773e-06,
+      "loss": 0.0039,
+      "step": 1007
+    },
+    {
+      "epoch": 0.8938151185989802,
+      "grad_norm": 0.27469322085380554,
+      "learning_rate": 8.226349784470276e-06,
+      "loss": 0.0023,
+      "step": 1008
+    },
+    {
+      "epoch": 0.8947018399467968,
+      "grad_norm": 0.20195840299129486,
+      "learning_rate": 8.222690832291158e-06,
+      "loss": 0.0025,
+      "step": 1009
+    },
+    {
+      "epoch": 0.8955885612946132,
+      "grad_norm": 0.25450944900512695,
+      "learning_rate": 8.219028925463045e-06,
+      "loss": 0.0027,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8964752826424296,
+      "grad_norm": 0.6547467708587646,
+      "learning_rate": 8.215364067343272e-06,
+      "loss": 0.0064,
+      "step": 1011
+    },
+    {
+      "epoch": 0.897362003990246,
+      "grad_norm": 0.3546174466609955,
+      "learning_rate": 8.211696261291879e-06,
+      "loss": 0.0039,
+      "step": 1012
+    },
+    {
+      "epoch": 0.8982487253380625,
+      "grad_norm": 0.7712265253067017,
+      "learning_rate": 8.208025510671609e-06,
+      "loss": 0.005,
+      "step": 1013
+    },
+    {
+      "epoch": 0.899135446685879,
+      "grad_norm": 0.07138879597187042,
+      "learning_rate": 8.204351818847902e-06,
+      "loss": 0.0008,
+      "step": 1014
+    },
+    {
+      "epoch": 0.9000221680336954,
+      "grad_norm": 0.2987552583217621,
+      "learning_rate": 8.200675189188899e-06,
+      "loss": 0.0025,
+      "step": 1015
+    },
+    {
+      "epoch": 0.9009088893815118,
+      "grad_norm": 0.5162558555603027,
+      "learning_rate": 8.19699562506543e-06,
+      "loss": 0.0024,
+      "step": 1016
+    },
+    {
+      "epoch": 0.9017956107293283,
+      "grad_norm": 0.1842619776725769,
+      "learning_rate": 8.193313129851019e-06,
+      "loss": 0.002,
+      "step": 1017
+    },
+    {
+      "epoch": 0.9026823320771448,
+      "grad_norm": 0.4205833375453949,
+      "learning_rate": 8.189627706921876e-06,
+      "loss": 0.0058,
+      "step": 1018
+    },
+    {
+      "epoch": 0.9035690534249612,
+      "grad_norm": 0.9251869916915894,
+      "learning_rate": 8.185939359656895e-06,
+      "loss": 0.0108,
+      "step": 1019
+    },
+    {
+      "epoch": 0.9044557747727776,
+      "grad_norm": 0.5171855092048645,
+      "learning_rate": 8.18224809143765e-06,
+      "loss": 0.0018,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9053424961205941,
+      "grad_norm": 1.3423070907592773,
+      "learning_rate": 8.178553905648396e-06,
+      "loss": 0.0107,
+      "step": 1021
+    },
+    {
+      "epoch": 0.9062292174684106,
+      "grad_norm": 0.4659717381000519,
+      "learning_rate": 8.174856805676062e-06,
+      "loss": 0.0028,
+      "step": 1022
+    },
+    {
+      "epoch": 0.907115938816227,
+      "grad_norm": 0.1671382635831833,
+      "learning_rate": 8.171156794910245e-06,
+      "loss": 0.0014,
+      "step": 1023
+    },
+    {
+      "epoch": 0.9080026601640434,
+      "grad_norm": 0.921686589717865,
+      "learning_rate": 8.167453876743215e-06,
+      "loss": 0.0078,
+      "step": 1024
+    },
+    {
+      "epoch": 0.9088893815118599,
+      "grad_norm": 0.4190947413444519,
+      "learning_rate": 8.163748054569905e-06,
+      "loss": 0.0029,
+      "step": 1025
+    },
+    {
+      "epoch": 0.9097761028596764,
+      "grad_norm": 0.5790007710456848,
+      "learning_rate": 8.160039331787915e-06,
+      "loss": 0.0048,
+      "step": 1026
+    },
+    {
+      "epoch": 0.9106628242074928,
+      "grad_norm": 0.4046242833137512,
+      "learning_rate": 8.156327711797499e-06,
+      "loss": 0.0043,
+      "step": 1027
+    },
+    {
+      "epoch": 0.9115495455553092,
+      "grad_norm": 0.35926681756973267,
+      "learning_rate": 8.152613198001567e-06,
+      "loss": 0.0021,
+      "step": 1028
+    },
+    {
+      "epoch": 0.9124362669031257,
+      "grad_norm": 0.4275892376899719,
+      "learning_rate": 8.14889579380569e-06,
+      "loss": 0.005,
+      "step": 1029
+    },
+    {
+      "epoch": 0.9133229882509422,
+      "grad_norm": 0.5041569471359253,
+      "learning_rate": 8.14517550261808e-06,
+      "loss": 0.0032,
+      "step": 1030
+    },
+    {
+      "epoch": 0.9142097095987586,
+      "grad_norm": 0.7166852951049805,
+      "learning_rate": 8.141452327849596e-06,
+      "loss": 0.0091,
+      "step": 1031
+    },
+    {
+      "epoch": 0.915096430946575,
+      "grad_norm": 0.4294228255748749,
+      "learning_rate": 8.13772627291375e-06,
+      "loss": 0.0051,
+      "step": 1032
+    },
+    {
+      "epoch": 0.9159831522943915,
+      "grad_norm": 0.7514994144439697,
+      "learning_rate": 8.133997341226686e-06,
+      "loss": 0.014,
+      "step": 1033
+    },
+    {
+      "epoch": 0.916869873642208,
+      "grad_norm": 0.6899821162223816,
+      "learning_rate": 8.130265536207189e-06,
+      "loss": 0.0063,
+      "step": 1034
+    },
+    {
+      "epoch": 0.9177565949900244,
+      "grad_norm": 0.08081519603729248,
+      "learning_rate": 8.126530861276677e-06,
+      "loss": 0.0009,
+      "step": 1035
+    },
+    {
+      "epoch": 0.9186433163378408,
+      "grad_norm": 0.09478479623794556,
+      "learning_rate": 8.1227933198592e-06,
+      "loss": 0.0018,
+      "step": 1036
+    },
+    {
+      "epoch": 0.9195300376856573,
+      "grad_norm": 0.21334019303321838,
+      "learning_rate": 8.119052915381432e-06,
+      "loss": 0.0021,
+      "step": 1037
+    },
+    {
+      "epoch": 0.9204167590334738,
+      "grad_norm": 0.27267512679100037,
+      "learning_rate": 8.11530965127268e-06,
+      "loss": 0.0031,
+      "step": 1038
+    },
+    {
+      "epoch": 0.9213034803812902,
+      "grad_norm": 0.41256797313690186,
+      "learning_rate": 8.111563530964869e-06,
+      "loss": 0.004,
+      "step": 1039
+    },
+    {
+      "epoch": 0.9221902017291066,
+      "grad_norm": 0.9929311275482178,
+      "learning_rate": 8.107814557892539e-06,
+      "loss": 0.0042,
+      "step": 1040
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.3197706341743469,
+      "learning_rate": 8.104062735492852e-06,
+      "loss": 0.0071,
+      "step": 1041
+    },
+    {
+      "epoch": 0.9239636444247395,
+      "grad_norm": 1.6723576784133911,
+      "learning_rate": 8.100308067205576e-06,
+      "loss": 0.006,
+      "step": 1042
+    },
+    {
+      "epoch": 0.924850365772556,
+      "grad_norm": 0.22382718324661255,
+      "learning_rate": 8.096550556473092e-06,
+      "loss": 0.0019,
+      "step": 1043
+    },
+    {
+      "epoch": 0.9257370871203724,
+      "grad_norm": 0.2541332542896271,
+      "learning_rate": 8.092790206740386e-06,
+      "loss": 0.0021,
+      "step": 1044
+    },
+    {
+      "epoch": 0.9266238084681889,
+      "grad_norm": 0.22676189243793488,
+      "learning_rate": 8.089027021455047e-06,
+      "loss": 0.0048,
+      "step": 1045
+    },
+    {
+      "epoch": 0.9275105298160053,
+      "grad_norm": 0.14653460681438446,
+      "learning_rate": 8.085261004067264e-06,
+      "loss": 0.0012,
+      "step": 1046
+    },
+    {
+      "epoch": 0.9283972511638218,
+      "grad_norm": 0.5666484832763672,
+      "learning_rate": 8.081492158029824e-06,
+      "loss": 0.008,
+      "step": 1047
+    },
+    {
+      "epoch": 0.9292839725116382,
+      "grad_norm": 0.285400927066803,
+      "learning_rate": 8.077720486798102e-06,
+      "loss": 0.0032,
+      "step": 1048
+    },
+    {
+      "epoch": 0.9301706938594546,
+      "grad_norm": 0.5438045859336853,
+      "learning_rate": 8.07394599383007e-06,
+      "loss": 0.0024,
+      "step": 1049
+    },
+    {
+      "epoch": 0.9310574152072711,
+      "grad_norm": 0.37798741459846497,
+      "learning_rate": 8.070168682586283e-06,
+      "loss": 0.003,
+      "step": 1050
+    },
+    {
+      "epoch": 0.9319441365550876,
+      "grad_norm": 0.6178186535835266,
+      "learning_rate": 8.066388556529881e-06,
+      "loss": 0.0088,
+      "step": 1051
+    },
+    {
+      "epoch": 0.932830857902904,
+      "grad_norm": 0.39296531677246094,
+      "learning_rate": 8.062605619126585e-06,
+      "loss": 0.0044,
+      "step": 1052
+    },
+    {
+      "epoch": 0.9337175792507204,
+      "grad_norm": 0.47325724363327026,
+      "learning_rate": 8.05881987384469e-06,
+      "loss": 0.0043,
+      "step": 1053
+    },
+    {
+      "epoch": 0.9346043005985369,
+      "grad_norm": 0.7773491144180298,
+      "learning_rate": 8.055031324155072e-06,
+      "loss": 0.0097,
+      "step": 1054
+    },
+    {
+      "epoch": 0.9354910219463534,
+      "grad_norm": 0.5702770352363586,
+      "learning_rate": 8.05123997353117e-06,
+      "loss": 0.0024,
+      "step": 1055
+    },
+    {
+      "epoch": 0.9363777432941698,
+      "grad_norm": 0.20904433727264404,
+      "learning_rate": 8.047445825449001e-06,
+      "loss": 0.0009,
+      "step": 1056
+    },
+    {
+      "epoch": 0.9372644646419862,
+      "grad_norm": 0.293757826089859,
+      "learning_rate": 8.04364888338714e-06,
+      "loss": 0.0044,
+      "step": 1057
+    },
+    {
+      "epoch": 0.9381511859898027,
+      "grad_norm": 0.6664373874664307,
+      "learning_rate": 8.039849150826721e-06,
+      "loss": 0.0084,
+      "step": 1058
+    },
+    {
+      "epoch": 0.9390379073376192,
+      "grad_norm": 0.615222692489624,
+      "learning_rate": 8.036046631251444e-06,
+      "loss": 0.0038,
+      "step": 1059
+    },
+    {
+      "epoch": 0.9399246286854356,
+      "grad_norm": 0.6374018788337708,
+      "learning_rate": 8.03224132814756e-06,
+      "loss": 0.0052,
+      "step": 1060
+    },
+    {
+      "epoch": 0.940811350033252,
+      "grad_norm": 0.6782588362693787,
+      "learning_rate": 8.028433245003871e-06,
+      "loss": 0.0059,
+      "step": 1061
+    },
+    {
+      "epoch": 0.9416980713810685,
+      "grad_norm": 0.3152410686016083,
+      "learning_rate": 8.02462238531173e-06,
+      "loss": 0.0016,
+      "step": 1062
+    },
+    {
+      "epoch": 0.942584792728885,
+      "grad_norm": 0.793630063533783,
+      "learning_rate": 8.020808752565034e-06,
+      "loss": 0.009,
+      "step": 1063
+    },
+    {
+      "epoch": 0.9434715140767014,
+      "grad_norm": 0.3483654260635376,
+      "learning_rate": 8.016992350260227e-06,
+      "loss": 0.0029,
+      "step": 1064
+    },
+    {
+      "epoch": 0.9443582354245178,
+      "grad_norm": 0.8813765048980713,
+      "learning_rate": 8.013173181896283e-06,
+      "loss": 0.0039,
+      "step": 1065
+    },
+    {
+      "epoch": 0.9452449567723343,
+      "grad_norm": 0.2730615437030792,
+      "learning_rate": 8.009351250974721e-06,
+      "loss": 0.0032,
+      "step": 1066
+    },
+    {
+      "epoch": 0.9461316781201508,
+      "grad_norm": 0.3211769461631775,
+      "learning_rate": 8.00552656099959e-06,
+      "loss": 0.0022,
+      "step": 1067
+    },
+    {
+      "epoch": 0.9470183994679672,
+      "grad_norm": 0.5229982733726501,
+      "learning_rate": 8.001699115477464e-06,
+      "loss": 0.006,
+      "step": 1068
+    },
+    {
+      "epoch": 0.9479051208157836,
+      "grad_norm": 0.6488707661628723,
+      "learning_rate": 7.997868917917453e-06,
+      "loss": 0.0033,
+      "step": 1069
+    },
+    {
+      "epoch": 0.9487918421636001,
+      "grad_norm": 0.07874885201454163,
+      "learning_rate": 7.994035971831178e-06,
+      "loss": 0.0003,
+      "step": 1070
+    },
+    {
+      "epoch": 0.9496785635114166,
+      "grad_norm": 0.3380604684352875,
+      "learning_rate": 7.990200280732792e-06,
+      "loss": 0.0021,
+      "step": 1071
+    },
+    {
+      "epoch": 0.950565284859233,
+      "grad_norm": 0.48102006316185,
+      "learning_rate": 7.986361848138954e-06,
+      "loss": 0.0038,
+      "step": 1072
+    },
+    {
+      "epoch": 0.9514520062070494,
+      "grad_norm": 0.8021492958068848,
+      "learning_rate": 7.982520677568846e-06,
+      "loss": 0.0049,
+      "step": 1073
+    },
+    {
+      "epoch": 0.9523387275548659,
+      "grad_norm": 0.5544074177742004,
+      "learning_rate": 7.978676772544153e-06,
+      "loss": 0.0081,
+      "step": 1074
+    },
+    {
+      "epoch": 0.9532254489026823,
+      "grad_norm": 0.6286907196044922,
+      "learning_rate": 7.97483013658907e-06,
+      "loss": 0.0104,
+      "step": 1075
+    },
+    {
+      "epoch": 0.9541121702504988,
+      "grad_norm": 0.636128842830658,
+      "learning_rate": 7.970980773230296e-06,
+      "loss": 0.0027,
+      "step": 1076
+    },
+    {
+      "epoch": 0.9549988915983152,
+      "grad_norm": 0.018983036279678345,
+      "learning_rate": 7.96712868599703e-06,
+      "loss": 0.0002,
+      "step": 1077
+    },
+    {
+      "epoch": 0.9558856129461317,
+      "grad_norm": 0.3683958053588867,
+      "learning_rate": 7.96327387842097e-06,
+      "loss": 0.0024,
+      "step": 1078
+    },
+    {
+      "epoch": 0.9567723342939481,
+      "grad_norm": 0.36459797620773315,
+      "learning_rate": 7.959416354036303e-06,
+      "loss": 0.0014,
+      "step": 1079
+    },
+    {
+      "epoch": 0.9576590556417646,
+      "grad_norm": 0.45369142293930054,
+      "learning_rate": 7.955556116379715e-06,
+      "loss": 0.0044,
+      "step": 1080
+    },
+    {
+      "epoch": 0.958545776989581,
+      "grad_norm": 0.4924830198287964,
+      "learning_rate": 7.951693168990369e-06,
+      "loss": 0.0048,
+      "step": 1081
+    },
+    {
+      "epoch": 0.9594324983373975,
+      "grad_norm": 0.6706141233444214,
+      "learning_rate": 7.947827515409924e-06,
+      "loss": 0.0104,
+      "step": 1082
+    },
+    {
+      "epoch": 0.9603192196852139,
+      "grad_norm": 0.5923113226890564,
+      "learning_rate": 7.94395915918251e-06,
+      "loss": 0.0057,
+      "step": 1083
+    },
+    {
+      "epoch": 0.9612059410330304,
+      "grad_norm": 0.777499794960022,
+      "learning_rate": 7.940088103854741e-06,
+      "loss": 0.005,
+      "step": 1084
+    },
+    {
+      "epoch": 0.9620926623808468,
+      "grad_norm": 0.21997174620628357,
+      "learning_rate": 7.936214352975706e-06,
+      "loss": 0.0029,
+      "step": 1085
+    },
+    {
+      "epoch": 0.9629793837286633,
+      "grad_norm": 0.6433026790618896,
+      "learning_rate": 7.93233791009696e-06,
+      "loss": 0.0052,
+      "step": 1086
+    },
+    {
+      "epoch": 0.9638661050764797,
+      "grad_norm": 0.14001475274562836,
+      "learning_rate": 7.928458778772533e-06,
+      "loss": 0.0019,
+      "step": 1087
+    },
+    {
+      "epoch": 0.9647528264242962,
+      "grad_norm": 0.4639449417591095,
+      "learning_rate": 7.92457696255891e-06,
+      "loss": 0.0047,
+      "step": 1088
+    },
+    {
+      "epoch": 0.9656395477721126,
+      "grad_norm": 0.623169481754303,
+      "learning_rate": 7.920692465015052e-06,
+      "loss": 0.0053,
+      "step": 1089
+    },
+    {
+      "epoch": 0.9665262691199291,
+      "grad_norm": 0.5239494442939758,
+      "learning_rate": 7.916805289702363e-06,
+      "loss": 0.011,
+      "step": 1090
+    },
+    {
+      "epoch": 0.9674129904677455,
+      "grad_norm": 0.3540083169937134,
+      "learning_rate": 7.912915440184711e-06,
+      "loss": 0.004,
+      "step": 1091
+    },
+    {
+      "epoch": 0.968299711815562,
+      "grad_norm": 0.3683808147907257,
+      "learning_rate": 7.909022920028416e-06,
+      "loss": 0.0069,
+      "step": 1092
+    },
+    {
+      "epoch": 0.9691864331633784,
+      "grad_norm": 0.18447741866111755,
+      "learning_rate": 7.905127732802242e-06,
+      "loss": 0.0028,
+      "step": 1093
+    },
+    {
+      "epoch": 0.9700731545111948,
+      "grad_norm": 0.19582514464855194,
+      "learning_rate": 7.901229882077403e-06,
+      "loss": 0.0018,
+      "step": 1094
+    },
+    {
+      "epoch": 0.9709598758590113,
+      "grad_norm": 0.48264968395233154,
+      "learning_rate": 7.89732937142755e-06,
+      "loss": 0.0058,
+      "step": 1095
+    },
+    {
+      "epoch": 0.9718465972068278,
+      "grad_norm": 0.8256710767745972,
+      "learning_rate": 7.893426204428777e-06,
+      "loss": 0.0083,
+      "step": 1096
+    },
+    {
+      "epoch": 0.9727333185546442,
+      "grad_norm": 0.6322904825210571,
+      "learning_rate": 7.88952038465961e-06,
+      "loss": 0.0118,
+      "step": 1097
+    },
+    {
+      "epoch": 0.9736200399024606,
+      "grad_norm": 0.21434572339057922,
+      "learning_rate": 7.885611915701012e-06,
+      "loss": 0.0029,
+      "step": 1098
+    },
+    {
+      "epoch": 0.9745067612502771,
+      "grad_norm": 0.38644617795944214,
+      "learning_rate": 7.88170080113637e-06,
+      "loss": 0.0037,
+      "step": 1099
+    },
+    {
+      "epoch": 0.9753934825980936,
+      "grad_norm": 0.2619270384311676,
+      "learning_rate": 7.877787044551497e-06,
+      "loss": 0.0048,
+      "step": 1100
+    },
+    {
+      "epoch": 0.97628020394591,
+      "grad_norm": 0.5454851984977722,
+      "learning_rate": 7.873870649534631e-06,
+      "loss": 0.0052,
+      "step": 1101
+    },
+    {
+      "epoch": 0.9771669252937264,
+      "grad_norm": 0.46398743987083435,
+      "learning_rate": 7.869951619676428e-06,
+      "loss": 0.0067,
+      "step": 1102
+    },
+    {
+      "epoch": 0.9780536466415429,
+      "grad_norm": 0.2236352264881134,
+      "learning_rate": 7.866029958569956e-06,
+      "loss": 0.003,
+      "step": 1103
+    },
+    {
+      "epoch": 0.9789403679893594,
+      "grad_norm": 0.20091237127780914,
+      "learning_rate": 7.862105669810703e-06,
+      "loss": 0.0032,
+      "step": 1104
+    },
+    {
+      "epoch": 0.9798270893371758,
+      "grad_norm": 0.6972124576568604,
+      "learning_rate": 7.858178756996557e-06,
+      "loss": 0.0105,
+      "step": 1105
+    },
+    {
+      "epoch": 0.9807138106849922,
+      "grad_norm": 0.42576393485069275,
+      "learning_rate": 7.854249223727823e-06,
+      "loss": 0.0029,
+      "step": 1106
+    },
+    {
+      "epoch": 0.9816005320328087,
+      "grad_norm": 0.27041757106781006,
+      "learning_rate": 7.850317073607193e-06,
+      "loss": 0.0022,
+      "step": 1107
+    },
+    {
+      "epoch": 0.9824872533806251,
+      "grad_norm": 0.3191399574279785,
+      "learning_rate": 7.846382310239777e-06,
+      "loss": 0.0033,
+      "step": 1108
+    },
+    {
+      "epoch": 0.9833739747284416,
+      "grad_norm": 0.24137769639492035,
+      "learning_rate": 7.842444937233066e-06,
+      "loss": 0.0058,
+      "step": 1109
+    },
+    {
+      "epoch": 0.984260696076258,
+      "grad_norm": 0.5599633455276489,
+      "learning_rate": 7.838504958196947e-06,
+      "loss": 0.0052,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9851474174240745,
+      "grad_norm": 0.436379611492157,
+      "learning_rate": 7.834562376743701e-06,
+      "loss": 0.0088,
+      "step": 1111
+    },
+    {
+      "epoch": 0.986034138771891,
+      "grad_norm": 0.4467504918575287,
+      "learning_rate": 7.83061719648799e-06,
+      "loss": 0.006,
+      "step": 1112
+    },
+    {
+      "epoch": 0.9869208601197074,
+      "grad_norm": 0.4567229449748993,
+      "learning_rate": 7.826669421046864e-06,
+      "loss": 0.0061,
+      "step": 1113
+    },
+    {
+      "epoch": 0.9878075814675238,
+      "grad_norm": 0.2151392251253128,
+      "learning_rate": 7.822719054039744e-06,
+      "loss": 0.0024,
+      "step": 1114
+    },
+    {
+      "epoch": 0.9886943028153403,
+      "grad_norm": 0.24533632397651672,
+      "learning_rate": 7.818766099088437e-06,
+      "loss": 0.005,
+      "step": 1115
+    },
+    {
+      "epoch": 0.9895810241631567,
+      "grad_norm": 0.1394466906785965,
+      "learning_rate": 7.814810559817115e-06,
+      "loss": 0.0012,
+      "step": 1116
+    },
+    {
+      "epoch": 0.9904677455109732,
+      "grad_norm": 0.20075161755084991,
+      "learning_rate": 7.810852439852321e-06,
+      "loss": 0.0014,
+      "step": 1117
+    },
+    {
+      "epoch": 0.9913544668587896,
+      "grad_norm": 0.030413122847676277,
+      "learning_rate": 7.806891742822965e-06,
+      "loss": 0.0002,
+      "step": 1118
+    },
+    {
+      "epoch": 0.9922411882066061,
+      "grad_norm": 0.2860347628593445,
+      "learning_rate": 7.802928472360323e-06,
+      "loss": 0.0024,
+      "step": 1119
+    },
+    {
+      "epoch": 0.9931279095544225,
+      "grad_norm": 0.0985894724726677,
+      "learning_rate": 7.798962632098024e-06,
+      "loss": 0.0008,
+      "step": 1120
+    },
+    {
+      "epoch": 0.994014630902239,
+      "grad_norm": 0.1537179946899414,
+      "learning_rate": 7.794994225672057e-06,
+      "loss": 0.0012,
+      "step": 1121
+    },
+    {
+      "epoch": 0.9949013522500554,
+      "grad_norm": 0.2673646807670593,
+      "learning_rate": 7.791023256720765e-06,
+      "loss": 0.002,
+      "step": 1122
+    },
+    {
+      "epoch": 0.9957880735978719,
+      "grad_norm": 0.8631387948989868,
+      "learning_rate": 7.787049728884835e-06,
+      "loss": 0.0041,
+      "step": 1123
+    },
+    {
+      "epoch": 0.9966747949456883,
+      "grad_norm": 0.2717569172382355,
+      "learning_rate": 7.783073645807307e-06,
+      "loss": 0.0026,
+      "step": 1124
+    },
+    {
+      "epoch": 0.9975615162935048,
+      "grad_norm": 0.4493410885334015,
+      "learning_rate": 7.779095011133555e-06,
+      "loss": 0.002,
+      "step": 1125
+    },
+    {
+      "epoch": 0.9984482376413212,
+      "grad_norm": 0.20788197219371796,
+      "learning_rate": 7.775113828511303e-06,
+      "loss": 0.0008,
+      "step": 1126
+    },
+    {
+      "epoch": 0.9993349589891377,
+      "grad_norm": 0.1764083057641983,
+      "learning_rate": 7.771130101590602e-06,
+      "loss": 0.001,
+      "step": 1127
+    },
+    {
+      "epoch": 1.0008867213478165,
+      "grad_norm": 0.8892470002174377,
+      "learning_rate": 7.767143834023842e-06,
+      "loss": 0.0098,
+      "step": 1128
+    },
+    {
+      "epoch": 1.0008867213478165,
+      "eval_loss": 0.04110870510339737,
+      "eval_runtime": 60.9693,
+      "eval_samples_per_second": 3.149,
+      "eval_steps_per_second": 0.787,
+      "step": 1128
+    },
+    {
+      "epoch": 1.0017734426956328,
+      "grad_norm": 1.0256974697113037,
+      "learning_rate": 7.76315502946574e-06,
+      "loss": 0.0036,
+      "step": 1129
+    },
+    {
+      "epoch": 1.0026601640434494,
+      "grad_norm": 0.16955800354480743,
+      "learning_rate": 7.759163691573332e-06,
+      "loss": 0.0013,
+      "step": 1130
+    },
+    {
+      "epoch": 1.0035468853912657,
+      "grad_norm": 0.10564973205327988,
+      "learning_rate": 7.755169824005989e-06,
+      "loss": 0.0006,
+      "step": 1131
+    },
+    {
+      "epoch": 1.0044336067390822,
+      "grad_norm": 0.2520776689052582,
+      "learning_rate": 7.751173430425393e-06,
+      "loss": 0.001,
+      "step": 1132
+    },
+    {
+      "epoch": 1.0053203280868988,
+      "grad_norm": 0.9520653486251831,
+      "learning_rate": 7.747174514495547e-06,
+      "loss": 0.0093,
+      "step": 1133
+    },
+    {
+      "epoch": 1.006207049434715,
+      "grad_norm": 0.41113871335983276,
+      "learning_rate": 7.743173079882763e-06,
+      "loss": 0.0018,
+      "step": 1134
+    },
+    {
+      "epoch": 1.0070937707825316,
+      "grad_norm": 0.10978075861930847,
+      "learning_rate": 7.73916913025566e-06,
+      "loss": 0.0005,
+      "step": 1135
+    },
+    {
+      "epoch": 1.0079804921303481,
+      "grad_norm": 0.3824840188026428,
+      "learning_rate": 7.73516266928517e-06,
+      "loss": 0.0017,
+      "step": 1136
+    },
+    {
+      "epoch": 1.0088672134781644,
+      "grad_norm": 0.1943664401769638,
+      "learning_rate": 7.73115370064452e-06,
+      "loss": 0.0008,
+      "step": 1137
+    },
+    {
+      "epoch": 1.009753934825981,
+      "grad_norm": 0.08970417082309723,
+      "learning_rate": 7.727142228009244e-06,
+      "loss": 0.0003,
+      "step": 1138
+    },
+    {
+      "epoch": 1.0106406561737973,
+      "grad_norm": 0.0703229010105133,
+      "learning_rate": 7.723128255057162e-06,
+      "loss": 0.0005,
+      "step": 1139
+    },
+    {
+      "epoch": 1.0115273775216138,
+      "grad_norm": 0.28951114416122437,
+      "learning_rate": 7.719111785468394e-06,
+      "loss": 0.0008,
+      "step": 1140
+    },
+    {
+      "epoch": 1.0124140988694303,
+      "grad_norm": 0.13246096670627594,
+      "learning_rate": 7.715092822925346e-06,
+      "loss": 0.0008,
+      "step": 1141
+    },
+    {
+      "epoch": 1.0133008202172467,
+      "grad_norm": 1.2307990789413452,
+      "learning_rate": 7.71107137111271e-06,
+      "loss": 0.0081,
+      "step": 1142
+    },
+    {
+      "epoch": 1.0141875415650632,
+      "grad_norm": 0.3421829640865326,
+      "learning_rate": 7.707047433717464e-06,
+      "loss": 0.0031,
+      "step": 1143
+    },
+    {
+      "epoch": 1.0150742629128797,
+      "grad_norm": 0.6147262454032898,
+      "learning_rate": 7.703021014428855e-06,
+      "loss": 0.0048,
+      "step": 1144
+    },
+    {
+      "epoch": 1.015960984260696,
+      "grad_norm": 0.38976818323135376,
+      "learning_rate": 7.698992116938418e-06,
+      "loss": 0.0036,
+      "step": 1145
+    },
+    {
+      "epoch": 1.0168477056085126,
+      "grad_norm": 0.6530481576919556,
+      "learning_rate": 7.694960744939948e-06,
+      "loss": 0.003,
+      "step": 1146
+    },
+    {
+      "epoch": 1.0177344269563289,
+      "grad_norm": 0.4899827539920807,
+      "learning_rate": 7.690926902129519e-06,
+      "loss": 0.0024,
+      "step": 1147
+    },
+    {
+      "epoch": 1.0186211483041454,
+      "grad_norm": 1.1396379470825195,
+      "learning_rate": 7.686890592205462e-06,
+      "loss": 0.0008,
+      "step": 1148
+    },
+    {
+      "epoch": 1.019507869651962,
+      "grad_norm": 0.3091973662376404,
+      "learning_rate": 7.682851818868377e-06,
+      "loss": 0.0059,
+      "step": 1149
+    },
+    {
+      "epoch": 1.0203945909997783,
+      "grad_norm": 0.3940574526786804,
+      "learning_rate": 7.678810585821118e-06,
+      "loss": 0.0028,
+      "step": 1150
+    },
+    {
+      "epoch": 1.0212813123475948,
+      "grad_norm": 1.2961633205413818,
+      "learning_rate": 7.674766896768794e-06,
+      "loss": 0.0024,
+      "step": 1151
+    },
+    {
+      "epoch": 1.0221680336954113,
+      "grad_norm": 0.48337897658348083,
+      "learning_rate": 7.67072075541877e-06,
+      "loss": 0.0047,
+      "step": 1152
+    },
+    {
+      "epoch": 1.0230547550432276,
+      "grad_norm": 0.578425407409668,
+      "learning_rate": 7.666672165480653e-06,
+      "loss": 0.0042,
+      "step": 1153
+    },
+    {
+      "epoch": 1.0239414763910442,
+      "grad_norm": 0.25309690833091736,
+      "learning_rate": 7.6626211306663e-06,
+      "loss": 0.0009,
+      "step": 1154
+    },
+    {
+      "epoch": 1.0248281977388605,
+      "grad_norm": 1.7819604873657227,
+      "learning_rate": 7.65856765468981e-06,
+      "loss": 0.0062,
+      "step": 1155
+    },
+    {
+      "epoch": 1.025714919086677,
+      "grad_norm": 0.3901142477989197,
+      "learning_rate": 7.654511741267514e-06,
+      "loss": 0.0042,
+      "step": 1156
+    },
+    {
+      "epoch": 1.0266016404344935,
+      "grad_norm": 0.10478702187538147,
+      "learning_rate": 7.650453394117983e-06,
+      "loss": 0.0006,
+      "step": 1157
+    },
+    {
+      "epoch": 1.0274883617823098,
+      "grad_norm": 0.10292093455791473,
+      "learning_rate": 7.646392616962019e-06,
+      "loss": 0.0006,
+      "step": 1158
+    },
+    {
+      "epoch": 1.0283750831301264,
+      "grad_norm": 0.02415108121931553,
+      "learning_rate": 7.642329413522653e-06,
+      "loss": 0.0002,
+      "step": 1159
+    },
+    {
+      "epoch": 1.029261804477943,
+      "grad_norm": 0.3483298122882843,
+      "learning_rate": 7.638263787525134e-06,
+      "loss": 0.0028,
+      "step": 1160
+    },
+    {
+      "epoch": 1.0301485258257592,
+      "grad_norm": 0.5602533221244812,
+      "learning_rate": 7.63419574269694e-06,
+      "loss": 0.0067,
+      "step": 1161
+    },
+    {
+      "epoch": 1.0310352471735758,
+      "grad_norm": 0.5920149087905884,
+      "learning_rate": 7.63012528276776e-06,
+      "loss": 0.0074,
+      "step": 1162
+    },
+    {
+      "epoch": 1.031921968521392,
+      "grad_norm": 0.24287919700145721,
+      "learning_rate": 7.6260524114695e-06,
+      "loss": 0.0017,
+      "step": 1163
+    },
+    {
+      "epoch": 1.0328086898692086,
+      "grad_norm": 3.347308397293091,
+      "learning_rate": 7.6219771325362825e-06,
+      "loss": 0.0081,
+      "step": 1164
+    },
+    {
+      "epoch": 1.0336954112170251,
+      "grad_norm": 0.49304619431495667,
+      "learning_rate": 7.617899449704427e-06,
+      "loss": 0.0052,
+      "step": 1165
+    },
+    {
+      "epoch": 1.0345821325648414,
+      "grad_norm": 0.29841598868370056,
+      "learning_rate": 7.613819366712464e-06,
+      "loss": 0.0019,
+      "step": 1166
+    },
+    {
+      "epoch": 1.035468853912658,
+      "grad_norm": 0.08790603280067444,
+      "learning_rate": 7.60973688730112e-06,
+      "loss": 0.0007,
+      "step": 1167
+    },
+    {
+      "epoch": 1.0363555752604743,
+      "grad_norm": 0.1790074110031128,
+      "learning_rate": 7.605652015213326e-06,
+      "loss": 0.0018,
+      "step": 1168
+    },
+    {
+      "epoch": 1.0372422966082908,
+      "grad_norm": 0.1106707751750946,
+      "learning_rate": 7.601564754194198e-06,
+      "loss": 0.0008,
+      "step": 1169
+    },
+    {
+      "epoch": 1.0381290179561073,
+      "grad_norm": 0.31249484419822693,
+      "learning_rate": 7.597475107991046e-06,
+      "loss": 0.004,
+      "step": 1170
+    },
+    {
+      "epoch": 1.0390157393039237,
+      "grad_norm": 0.1908506155014038,
+      "learning_rate": 7.593383080353369e-06,
+      "loss": 0.0013,
+      "step": 1171
+    },
+    {
+      "epoch": 1.0399024606517402,
+      "grad_norm": 0.7766817212104797,
+      "learning_rate": 7.5892886750328446e-06,
+      "loss": 0.007,
+      "step": 1172
+    },
+    {
+      "epoch": 1.0407891819995567,
+      "grad_norm": 0.8299360871315002,
+      "learning_rate": 7.585191895783335e-06,
+      "loss": 0.0056,
+      "step": 1173
+    },
+    {
+      "epoch": 1.041675903347373,
+      "grad_norm": 0.23749642074108124,
+      "learning_rate": 7.5810927463608766e-06,
+      "loss": 0.0017,
+      "step": 1174
+    },
+    {
+      "epoch": 1.0425626246951896,
+      "grad_norm": 0.10918860882520676,
+      "learning_rate": 7.576991230523678e-06,
+      "loss": 0.0014,
+      "step": 1175
+    },
+    {
+      "epoch": 1.043449346043006,
+      "grad_norm": 0.030476948246359825,
+      "learning_rate": 7.572887352032119e-06,
+      "loss": 0.0002,
+      "step": 1176
+    },
+    {
+      "epoch": 1.0443360673908224,
+      "grad_norm": 0.2967153787612915,
+      "learning_rate": 7.568781114648744e-06,
+      "loss": 0.0025,
+      "step": 1177
+    },
+    {
+      "epoch": 1.045222788738639,
+      "grad_norm": 0.2687462270259857,
+      "learning_rate": 7.5646725221382635e-06,
+      "loss": 0.0015,
+      "step": 1178
+    },
+    {
+      "epoch": 1.0461095100864553,
+      "grad_norm": 0.2979382276535034,
+      "learning_rate": 7.560561578267542e-06,
+      "loss": 0.0035,
+      "step": 1179
+    },
+    {
+      "epoch": 1.0469962314342718,
+      "grad_norm": 0.18216152489185333,
+      "learning_rate": 7.556448286805605e-06,
+      "loss": 0.0011,
+      "step": 1180
+    },
+    {
+      "epoch": 1.0478829527820883,
+      "grad_norm": 0.26704421639442444,
+      "learning_rate": 7.552332651523626e-06,
+      "loss": 0.0021,
+      "step": 1181
+    },
+    {
+      "epoch": 1.0487696741299046,
+      "grad_norm": 0.37852510809898376,
+      "learning_rate": 7.54821467619493e-06,
+      "loss": 0.0052,
+      "step": 1182
+    },
+    {
+      "epoch": 1.0496563954777212,
+      "grad_norm": 0.02862503007054329,
+      "learning_rate": 7.5440943645949845e-06,
+      "loss": 0.0002,
+      "step": 1183
+    },
+    {
+      "epoch": 1.0505431168255375,
+      "grad_norm": 0.4329982101917267,
+      "learning_rate": 7.539971720501407e-06,
+      "loss": 0.0032,
+      "step": 1184
+    },
+    {
+      "epoch": 1.051429838173354,
+      "grad_norm": 0.05105764791369438,
+      "learning_rate": 7.535846747693944e-06,
+      "loss": 0.0003,
+      "step": 1185
+    },
+    {
+      "epoch": 1.0523165595211705,
+      "grad_norm": 0.22195594012737274,
+      "learning_rate": 7.531719449954479e-06,
+      "loss": 0.0016,
+      "step": 1186
+    },
+    {
+      "epoch": 1.0532032808689868,
+      "grad_norm": 0.49873724579811096,
+      "learning_rate": 7.527589831067032e-06,
+      "loss": 0.01,
+      "step": 1187
+    },
+    {
+      "epoch": 1.0540900022168034,
+      "grad_norm": 0.09682756662368774,
+      "learning_rate": 7.523457894817745e-06,
+      "loss": 0.0004,
+      "step": 1188
+    },
+    {
+      "epoch": 1.05497672356462,
+      "grad_norm": 0.06749970465898514,
+      "learning_rate": 7.519323644994892e-06,
+      "loss": 0.0004,
+      "step": 1189
+    },
+    {
+      "epoch": 1.0558634449124362,
+      "grad_norm": 0.06560672074556351,
+      "learning_rate": 7.515187085388858e-06,
+      "loss": 0.0006,
+      "step": 1190
+    },
+    {
+      "epoch": 1.0567501662602528,
+      "grad_norm": 4.2813286781311035,
+      "learning_rate": 7.511048219792154e-06,
+      "loss": 0.0045,
+      "step": 1191
+    },
+    {
+      "epoch": 1.057636887608069,
+      "grad_norm": 0.16461029648780823,
+      "learning_rate": 7.506907051999402e-06,
+      "loss": 0.0009,
+      "step": 1192
+    },
+    {
+      "epoch": 1.0585236089558856,
+      "grad_norm": 0.5340306758880615,
+      "learning_rate": 7.502763585807336e-06,
+      "loss": 0.0053,
+      "step": 1193
+    },
+    {
+      "epoch": 1.0594103303037021,
+      "grad_norm": 0.6085299253463745,
+      "learning_rate": 7.498617825014795e-06,
+      "loss": 0.004,
+      "step": 1194
+    },
+    {
+      "epoch": 1.0602970516515184,
+      "grad_norm": 0.5654335021972656,
+      "learning_rate": 7.494469773422726e-06,
+      "loss": 0.0029,
+      "step": 1195
+    },
+    {
+      "epoch": 1.061183772999335,
+      "grad_norm": 0.13969686627388,
+      "learning_rate": 7.49031943483417e-06,
+      "loss": 0.0008,
+      "step": 1196
+    },
+    {
+      "epoch": 1.0620704943471515,
+      "grad_norm": 0.521077573299408,
+      "learning_rate": 7.48616681305427e-06,
+      "loss": 0.0032,
+      "step": 1197
+    },
+    {
+      "epoch": 1.0629572156949678,
+      "grad_norm": 0.5144971013069153,
+      "learning_rate": 7.4820119118902615e-06,
+      "loss": 0.0085,
+      "step": 1198
+    },
+    {
+      "epoch": 1.0638439370427843,
+      "grad_norm": 0.04687139019370079,
+      "learning_rate": 7.477854735151466e-06,
+      "loss": 0.0003,
+      "step": 1199
+    },
+    {
+      "epoch": 1.0647306583906007,
+      "grad_norm": 5.557599067687988,
+      "learning_rate": 7.473695286649296e-06,
+      "loss": 0.0046,
+      "step": 1200
+    },
+    {
+      "epoch": 1.0656173797384172,
+      "grad_norm": 0.12024810910224915,
+      "learning_rate": 7.469533570197245e-06,
+      "loss": 0.0006,
+      "step": 1201
+    },
+    {
+      "epoch": 1.0665041010862337,
+      "grad_norm": 0.14419999718666077,
+      "learning_rate": 7.4653695896108855e-06,
+      "loss": 0.0011,
+      "step": 1202
+    },
+    {
+      "epoch": 1.06739082243405,
+      "grad_norm": 0.2992253303527832,
+      "learning_rate": 7.461203348707866e-06,
+      "loss": 0.0027,
+      "step": 1203
+    },
+    {
+      "epoch": 1.0682775437818666,
+      "grad_norm": 0.058901526033878326,
+      "learning_rate": 7.457034851307907e-06,
+      "loss": 0.0007,
+      "step": 1204
+    },
+    {
+      "epoch": 1.069164265129683,
+      "grad_norm": 0.1338261067867279,
+      "learning_rate": 7.452864101232798e-06,
+      "loss": 0.0016,
+      "step": 1205
+    },
+    {
+      "epoch": 1.0700509864774994,
+      "grad_norm": 0.18810372054576874,
+      "learning_rate": 7.448691102306396e-06,
+      "loss": 0.0011,
+      "step": 1206
+    },
+    {
+      "epoch": 1.070937707825316,
+      "grad_norm": 0.5619723200798035,
+      "learning_rate": 7.444515858354615e-06,
+      "loss": 0.0016,
+      "step": 1207
+    },
+    {
+      "epoch": 1.0718244291731323,
+      "grad_norm": 0.10859323292970657,
+      "learning_rate": 7.4403383732054325e-06,
+      "loss": 0.0012,
+      "step": 1208
+    },
+    {
+      "epoch": 1.0727111505209488,
+      "grad_norm": 0.16468924283981323,
+      "learning_rate": 7.436158650688877e-06,
+      "loss": 0.0009,
+      "step": 1209
+    },
+    {
+      "epoch": 1.0735978718687653,
+      "grad_norm": 0.2525959312915802,
+      "learning_rate": 7.431976694637031e-06,
+      "loss": 0.0022,
+      "step": 1210
+    },
+    {
+      "epoch": 1.0744845932165816,
+      "grad_norm": 0.2525767385959625,
+      "learning_rate": 7.427792508884022e-06,
+      "loss": 0.0013,
+      "step": 1211
+    },
+    {
+      "epoch": 1.0753713145643982,
+      "grad_norm": 0.330912709236145,
+      "learning_rate": 7.423606097266024e-06,
+      "loss": 0.0039,
+      "step": 1212
+    },
+    {
+      "epoch": 1.0762580359122147,
+      "grad_norm": 0.36091116070747375,
+      "learning_rate": 7.41941746362125e-06,
+      "loss": 0.0018,
+      "step": 1213
+    },
+    {
+      "epoch": 1.077144757260031,
+      "grad_norm": 0.4603426456451416,
+      "learning_rate": 7.415226611789952e-06,
+      "loss": 0.0027,
+      "step": 1214
+    },
+    {
+      "epoch": 1.0780314786078475,
+      "grad_norm": 0.04140254855155945,
+      "learning_rate": 7.4110335456144156e-06,
+      "loss": 0.0001,
+      "step": 1215
+    },
+    {
+      "epoch": 1.0789181999556638,
+      "grad_norm": 0.13625720143318176,
+      "learning_rate": 7.406838268938954e-06,
+      "loss": 0.0008,
+      "step": 1216
+    },
+    {
+      "epoch": 1.0798049213034804,
+      "grad_norm": 0.23671278357505798,
+      "learning_rate": 7.402640785609911e-06,
+      "loss": 0.0017,
+      "step": 1217
+    },
+    {
+      "epoch": 1.080691642651297,
+      "grad_norm": 0.15733811259269714,
+      "learning_rate": 7.398441099475649e-06,
+      "loss": 0.001,
+      "step": 1218
+    },
+    {
+      "epoch": 1.0815783639991132,
+      "grad_norm": 1.2790669202804565,
+      "learning_rate": 7.394239214386555e-06,
+      "loss": 0.0119,
+      "step": 1219
+    },
+    {
+      "epoch": 1.0824650853469298,
+      "grad_norm": 1.2404475212097168,
+      "learning_rate": 7.390035134195027e-06,
+      "loss": 0.0009,
+      "step": 1220
+    },
+    {
+      "epoch": 1.083351806694746,
+      "grad_norm": 0.176893413066864,
+      "learning_rate": 7.385828862755478e-06,
+      "loss": 0.0012,
+      "step": 1221
+    },
+    {
+      "epoch": 1.0842385280425626,
+      "grad_norm": 0.244059219956398,
+      "learning_rate": 7.381620403924333e-06,
+      "loss": 0.0067,
+      "step": 1222
+    },
+    {
+      "epoch": 1.0851252493903791,
+      "grad_norm": 0.16676774621009827,
+      "learning_rate": 7.377409761560016e-06,
+      "loss": 0.0005,
+      "step": 1223
+    },
+    {
+      "epoch": 1.0860119707381954,
+      "grad_norm": 0.7514981627464294,
+      "learning_rate": 7.373196939522959e-06,
+      "loss": 0.0057,
+      "step": 1224
+    },
+    {
+      "epoch": 1.086898692086012,
+      "grad_norm": 0.6947178840637207,
+      "learning_rate": 7.368981941675587e-06,
+      "loss": 0.0031,
+      "step": 1225
+    },
+    {
+      "epoch": 1.0877854134338285,
+      "grad_norm": 0.024032553657889366,
+      "learning_rate": 7.364764771882324e-06,
+      "loss": 0.0001,
+      "step": 1226
+    },
+    {
+      "epoch": 1.0886721347816448,
+      "grad_norm": 0.17391324043273926,
+      "learning_rate": 7.360545434009585e-06,
+      "loss": 0.0006,
+      "step": 1227
+    },
+    {
+      "epoch": 1.0895588561294614,
+      "grad_norm": 0.5214890837669373,
+      "learning_rate": 7.356323931925769e-06,
+      "loss": 0.0057,
+      "step": 1228
+    },
+    {
+      "epoch": 1.0904455774772779,
+      "grad_norm": 0.3625035583972931,
+      "learning_rate": 7.352100269501264e-06,
+      "loss": 0.0026,
+      "step": 1229
+    },
+    {
+      "epoch": 1.0913322988250942,
+      "grad_norm": 0.3312781751155853,
+      "learning_rate": 7.3478744506084355e-06,
+      "loss": 0.0045,
+      "step": 1230
+    },
+    {
+      "epoch": 1.0922190201729107,
+      "grad_norm": 0.26414018869400024,
+      "learning_rate": 7.343646479121626e-06,
+      "loss": 0.0039,
+      "step": 1231
+    },
+    {
+      "epoch": 1.093105741520727,
+      "grad_norm": 0.09360918402671814,
+      "learning_rate": 7.339416358917156e-06,
+      "loss": 0.0005,
+      "step": 1232
+    },
+    {
+      "epoch": 1.0939924628685436,
+      "grad_norm": 0.014107626862823963,
+      "learning_rate": 7.33518409387331e-06,
+      "loss": 0.0001,
+      "step": 1233
+    },
+    {
+      "epoch": 1.09487918421636,
+      "grad_norm": 0.2544485330581665,
+      "learning_rate": 7.3309496878703415e-06,
+      "loss": 0.0056,
+      "step": 1234
+    },
+    {
+      "epoch": 1.0957659055641764,
+      "grad_norm": 0.6181140542030334,
+      "learning_rate": 7.326713144790467e-06,
+      "loss": 0.001,
+      "step": 1235
+    },
+    {
+      "epoch": 1.096652626911993,
+      "grad_norm": 0.27269816398620605,
+      "learning_rate": 7.322474468517865e-06,
+      "loss": 0.0029,
+      "step": 1236
+    },
+    {
+      "epoch": 1.0975393482598093,
+      "grad_norm": 0.44560036063194275,
+      "learning_rate": 7.318233662938664e-06,
+      "loss": 0.0009,
+      "step": 1237
+    },
+    {
+      "epoch": 1.0984260696076258,
+      "grad_norm": 0.22260509431362152,
+      "learning_rate": 7.3139907319409496e-06,
+      "loss": 0.0012,
+      "step": 1238
+    },
+    {
+      "epoch": 1.0993127909554423,
+      "grad_norm": 0.2584804594516754,
+      "learning_rate": 7.309745679414751e-06,
+      "loss": 0.0014,
+      "step": 1239
+    },
+    {
+      "epoch": 1.1001995123032586,
+      "grad_norm": 0.2714986205101013,
+      "learning_rate": 7.305498509252052e-06,
+      "loss": 0.0044,
+      "step": 1240
+    },
+    {
+      "epoch": 1.1010862336510752,
+      "grad_norm": 0.6800200343132019,
+      "learning_rate": 7.301249225346764e-06,
+      "loss": 0.0045,
+      "step": 1241
+    },
+    {
+      "epoch": 1.1019729549988917,
+      "grad_norm": 0.061328668147325516,
+      "learning_rate": 7.2969978315947485e-06,
+      "loss": 0.0004,
+      "step": 1242
+    },
+    {
+      "epoch": 1.102859676346708,
+      "grad_norm": 0.23961175978183746,
+      "learning_rate": 7.292744331893796e-06,
+      "loss": 0.0009,
+      "step": 1243
+    },
+    {
+      "epoch": 1.1037463976945245,
+      "grad_norm": 0.36839234828948975,
+      "learning_rate": 7.288488730143629e-06,
+      "loss": 0.0025,
+      "step": 1244
+    },
+    {
+      "epoch": 1.1046331190423408,
+      "grad_norm": 0.47886547446250916,
+      "learning_rate": 7.284231030245896e-06,
+      "loss": 0.0051,
+      "step": 1245
+    },
+    {
+      "epoch": 1.1055198403901574,
+      "grad_norm": 0.4376900792121887,
+      "learning_rate": 7.279971236104171e-06,
+      "loss": 0.0064,
+      "step": 1246
+    },
+    {
+      "epoch": 1.106406561737974,
+      "grad_norm": 0.09138593822717667,
+      "learning_rate": 7.275709351623945e-06,
+      "loss": 0.0007,
+      "step": 1247
+    },
+    {
+      "epoch": 1.1072932830857902,
+      "grad_norm": 0.40806150436401367,
+      "learning_rate": 7.271445380712629e-06,
+      "loss": 0.0043,
+      "step": 1248
+    },
+    {
+      "epoch": 1.1081800044336068,
+      "grad_norm": 0.11947987228631973,
+      "learning_rate": 7.267179327279545e-06,
+      "loss": 0.001,
+      "step": 1249
+    },
+    {
+      "epoch": 1.1090667257814233,
+      "grad_norm": 0.11575794219970703,
+      "learning_rate": 7.262911195235925e-06,
+      "loss": 0.0009,
+      "step": 1250
+    },
+    {
+      "epoch": 1.1099534471292396,
+      "grad_norm": 0.1301448494195938,
+      "learning_rate": 7.258640988494906e-06,
+      "loss": 0.0011,
+      "step": 1251
+    },
+    {
+      "epoch": 1.1108401684770561,
+      "grad_norm": 0.308881938457489,
+      "learning_rate": 7.254368710971529e-06,
+      "loss": 0.0018,
+      "step": 1252
+    },
+    {
+      "epoch": 1.1117268898248724,
+      "grad_norm": 0.5369164943695068,
+      "learning_rate": 7.2500943665827285e-06,
+      "loss": 0.0033,
+      "step": 1253
+    },
+    {
+      "epoch": 1.112613611172689,
+      "grad_norm": 0.0686551108956337,
+      "learning_rate": 7.245817959247341e-06,
+      "loss": 0.0005,
+      "step": 1254
+    },
+    {
+      "epoch": 1.1135003325205055,
+      "grad_norm": 0.33711907267570496,
+      "learning_rate": 7.241539492886089e-06,
+      "loss": 0.0018,
+      "step": 1255
+    },
+    {
+      "epoch": 1.1143870538683218,
+      "grad_norm": 0.3586072325706482,
+      "learning_rate": 7.237258971421587e-06,
+      "loss": 0.0061,
+      "step": 1256
+    },
+    {
+      "epoch": 1.1152737752161384,
+      "grad_norm": 0.036523330956697464,
+      "learning_rate": 7.232976398778329e-06,
+      "loss": 0.0002,
+      "step": 1257
+    },
+    {
+      "epoch": 1.1161604965639547,
+      "grad_norm": 0.23500844836235046,
+      "learning_rate": 7.2286917788826926e-06,
+      "loss": 0.0016,
+      "step": 1258
+    },
+    {
+      "epoch": 1.1170472179117712,
+      "grad_norm": 0.3606179654598236,
+      "learning_rate": 7.224405115662931e-06,
+      "loss": 0.002,
+      "step": 1259
+    },
+    {
+      "epoch": 1.1179339392595877,
+      "grad_norm": 0.14782433211803436,
+      "learning_rate": 7.220116413049173e-06,
+      "loss": 0.0014,
+      "step": 1260
+    },
+    {
+      "epoch": 1.118820660607404,
+      "grad_norm": 0.4738283157348633,
+      "learning_rate": 7.2158256749734155e-06,
+      "loss": 0.0046,
+      "step": 1261
+    },
+    {
+      "epoch": 1.1197073819552206,
+      "grad_norm": 0.4056955873966217,
+      "learning_rate": 7.21153290536952e-06,
+      "loss": 0.0034,
+      "step": 1262
+    },
+    {
+      "epoch": 1.120594103303037,
+      "grad_norm": 0.08267973363399506,
+      "learning_rate": 7.207238108173216e-06,
+      "loss": 0.0004,
+      "step": 1263
+    },
+    {
+      "epoch": 1.1214808246508534,
+      "grad_norm": 0.31053709983825684,
+      "learning_rate": 7.202941287322084e-06,
+      "loss": 0.0046,
+      "step": 1264
+    },
+    {
+      "epoch": 1.12236754599867,
+      "grad_norm": 0.2635895609855652,
+      "learning_rate": 7.198642446755566e-06,
+      "loss": 0.0019,
+      "step": 1265
+    },
+    {
+      "epoch": 1.1232542673464865,
+      "grad_norm": 0.5915330648422241,
+      "learning_rate": 7.194341590414954e-06,
+      "loss": 0.0065,
+      "step": 1266
+    },
+    {
+      "epoch": 1.1241409886943028,
+      "grad_norm": 0.2291322946548462,
+      "learning_rate": 7.190038722243387e-06,
+      "loss": 0.0015,
+      "step": 1267
+    },
+    {
+      "epoch": 1.1250277100421193,
+      "grad_norm": 0.21318402886390686,
+      "learning_rate": 7.18573384618585e-06,
+      "loss": 0.0009,
+      "step": 1268
+    },
+    {
+      "epoch": 1.1259144313899356,
+      "grad_norm": 0.03609558939933777,
+      "learning_rate": 7.181426966189168e-06,
+      "loss": 0.0003,
+      "step": 1269
+    },
+    {
+      "epoch": 1.1268011527377522,
+      "grad_norm": 0.9004939198493958,
+      "learning_rate": 7.177118086202004e-06,
+      "loss": 0.0111,
+      "step": 1270
+    },
+    {
+      "epoch": 1.1276878740855687,
+      "grad_norm": 0.3457188010215759,
+      "learning_rate": 7.1728072101748546e-06,
+      "loss": 0.0012,
+      "step": 1271
+    },
+    {
+      "epoch": 1.128574595433385,
+      "grad_norm": 0.4781457185745239,
+      "learning_rate": 7.168494342060044e-06,
+      "loss": 0.0096,
+      "step": 1272
+    },
+    {
+      "epoch": 1.1294613167812015,
+      "grad_norm": 0.8124074935913086,
+      "learning_rate": 7.164179485811728e-06,
+      "loss": 0.0019,
+      "step": 1273
+    },
+    {
+      "epoch": 1.1303480381290179,
+      "grad_norm": 0.3004269599914551,
+      "learning_rate": 7.159862645385879e-06,
+      "loss": 0.0015,
+      "step": 1274
+    },
+    {
+      "epoch": 1.1312347594768344,
+      "grad_norm": 0.5909804105758667,
+      "learning_rate": 7.155543824740294e-06,
+      "loss": 0.0059,
+      "step": 1275
+    },
+    {
+      "epoch": 1.132121480824651,
+      "grad_norm": 0.09179907292127609,
+      "learning_rate": 7.151223027834581e-06,
+      "loss": 0.0006,
+      "step": 1276
+    },
+    {
+      "epoch": 1.1330082021724672,
+      "grad_norm": 0.11039888113737106,
+      "learning_rate": 7.146900258630163e-06,
+      "loss": 0.0005,
+      "step": 1277
+    },
+    {
+      "epoch": 1.1338949235202838,
+      "grad_norm": 0.10031022131443024,
+      "learning_rate": 7.1425755210902705e-06,
+      "loss": 0.0006,
+      "step": 1278
+    },
+    {
+      "epoch": 1.1347816448681003,
+      "grad_norm": 0.5158107876777649,
+      "learning_rate": 7.138248819179937e-06,
+      "loss": 0.0045,
+      "step": 1279
+    },
+    {
+      "epoch": 1.1356683662159166,
+      "grad_norm": 0.2349659502506256,
+      "learning_rate": 7.133920156866001e-06,
+      "loss": 0.0031,
+      "step": 1280
+    },
+    {
+      "epoch": 1.1365550875637331,
+      "grad_norm": 0.17994800209999084,
+      "learning_rate": 7.129589538117092e-06,
+      "loss": 0.001,
+      "step": 1281
+    },
+    {
+      "epoch": 1.1374418089115497,
+      "grad_norm": 0.3246804475784302,
+      "learning_rate": 7.125256966903639e-06,
+      "loss": 0.0019,
+      "step": 1282
+    },
+    {
+      "epoch": 1.138328530259366,
+      "grad_norm": 0.47080448269844055,
+      "learning_rate": 7.12092244719786e-06,
+      "loss": 0.0053,
+      "step": 1283
+    },
+    {
+      "epoch": 1.1392152516071825,
+      "grad_norm": 0.10333869606256485,
+      "learning_rate": 7.116585982973756e-06,
+      "loss": 0.0006,
+      "step": 1284
+    },
+    {
+      "epoch": 1.1401019729549988,
+      "grad_norm": 0.34819576144218445,
+      "learning_rate": 7.112247578207115e-06,
+      "loss": 0.0034,
+      "step": 1285
+    },
+    {
+      "epoch": 1.1409886943028154,
+      "grad_norm": 0.3301628828048706,
+      "learning_rate": 7.107907236875501e-06,
+      "loss": 0.0016,
+      "step": 1286
+    },
+    {
+      "epoch": 1.1418754156506319,
+      "grad_norm": 0.42407017946243286,
+      "learning_rate": 7.103564962958256e-06,
+      "loss": 0.0078,
+      "step": 1287
+    },
+    {
+      "epoch": 1.1427621369984482,
+      "grad_norm": 0.592100977897644,
+      "learning_rate": 7.099220760436492e-06,
+      "loss": 0.0069,
+      "step": 1288
+    },
+    {
+      "epoch": 1.1436488583462647,
+      "grad_norm": 0.6017158627510071,
+      "learning_rate": 7.094874633293091e-06,
+      "loss": 0.0044,
+      "step": 1289
+    },
+    {
+      "epoch": 1.144535579694081,
+      "grad_norm": 0.025082416832447052,
+      "learning_rate": 7.090526585512696e-06,
+      "loss": 0.0003,
+      "step": 1290
+    },
+    {
+      "epoch": 1.1454223010418976,
+      "grad_norm": 0.1530878245830536,
+      "learning_rate": 7.086176621081715e-06,
+      "loss": 0.0012,
+      "step": 1291
+    },
+    {
+      "epoch": 1.146309022389714,
+      "grad_norm": 0.09143998473882675,
+      "learning_rate": 7.0818247439883115e-06,
+      "loss": 0.0007,
+      "step": 1292
+    },
+    {
+      "epoch": 1.1471957437375304,
+      "grad_norm": 0.11759691685438156,
+      "learning_rate": 7.077470958222402e-06,
+      "loss": 0.0005,
+      "step": 1293
+    },
+    {
+      "epoch": 1.148082465085347,
+      "grad_norm": 0.6441949605941772,
+      "learning_rate": 7.073115267775654e-06,
+      "loss": 0.0047,
+      "step": 1294
+    },
+    {
+      "epoch": 1.1489691864331633,
+      "grad_norm": 0.01792193576693535,
+      "learning_rate": 7.06875767664148e-06,
+      "loss": 0.0001,
+      "step": 1295
+    },
+    {
+      "epoch": 1.1498559077809798,
+      "grad_norm": 0.26153331995010376,
+      "learning_rate": 7.064398188815038e-06,
+      "loss": 0.0017,
+      "step": 1296
+    },
+    {
+      "epoch": 1.1507426291287963,
+      "grad_norm": 0.48967376351356506,
+      "learning_rate": 7.060036808293221e-06,
+      "loss": 0.0045,
+      "step": 1297
+    },
+    {
+      "epoch": 1.1516293504766126,
+      "grad_norm": 0.11976376920938492,
+      "learning_rate": 7.055673539074657e-06,
+      "loss": 0.0009,
+      "step": 1298
+    },
+    {
+      "epoch": 1.1525160718244292,
+      "grad_norm": 0.13650448620319366,
+      "learning_rate": 7.051308385159712e-06,
+      "loss": 0.0008,
+      "step": 1299
+    },
+    {
+      "epoch": 1.1534027931722457,
+      "grad_norm": 0.1300816833972931,
+      "learning_rate": 7.0469413505504745e-06,
+      "loss": 0.001,
+      "step": 1300
+    },
+    {
+      "epoch": 1.154289514520062,
+      "grad_norm": 0.1258881539106369,
+      "learning_rate": 7.042572439250756e-06,
+      "loss": 0.0005,
+      "step": 1301
+    },
+    {
+      "epoch": 1.1551762358678785,
+      "grad_norm": 0.0720561072230339,
+      "learning_rate": 7.038201655266093e-06,
+      "loss": 0.0004,
+      "step": 1302
+    },
+    {
+      "epoch": 1.156062957215695,
+      "grad_norm": 0.08578696846961975,
+      "learning_rate": 7.033829002603738e-06,
+      "loss": 0.0004,
+      "step": 1303
+    },
+    {
+      "epoch": 1.1569496785635114,
+      "grad_norm": 0.2591750919818878,
+      "learning_rate": 7.029454485272654e-06,
+      "loss": 0.0014,
+      "step": 1304
+    },
+    {
+      "epoch": 1.157836399911328,
+      "grad_norm": 0.13878989219665527,
+      "learning_rate": 7.0250781072835165e-06,
+      "loss": 0.0009,
+      "step": 1305
+    },
+    {
+      "epoch": 1.1587231212591442,
+      "grad_norm": 0.13020068407058716,
+      "learning_rate": 7.020699872648703e-06,
+      "loss": 0.0005,
+      "step": 1306
+    },
+    {
+      "epoch": 1.1596098426069608,
+      "grad_norm": 0.0145335141569376,
+      "learning_rate": 7.0163197853822975e-06,
+      "loss": 0.0001,
+      "step": 1307
+    },
+    {
+      "epoch": 1.1604965639547773,
+      "grad_norm": 0.04610709846019745,
+      "learning_rate": 7.0119378495000824e-06,
+      "loss": 0.0002,
+      "step": 1308
+    },
+    {
+      "epoch": 1.1613832853025936,
+      "grad_norm": 0.027545208111405373,
+      "learning_rate": 7.007554069019532e-06,
+      "loss": 0.0002,
+      "step": 1309
+    },
+    {
+      "epoch": 1.1622700066504101,
+      "grad_norm": 0.8998842239379883,
+      "learning_rate": 7.003168447959814e-06,
+      "loss": 0.0093,
+      "step": 1310
+    },
+    {
+      "epoch": 1.1631567279982264,
+      "grad_norm": 0.5461855530738831,
+      "learning_rate": 6.99878099034178e-06,
+      "loss": 0.0028,
+      "step": 1311
+    },
+    {
+      "epoch": 1.164043449346043,
+      "grad_norm": 0.3846938908100128,
+      "learning_rate": 6.994391700187975e-06,
+      "loss": 0.0042,
+      "step": 1312
+    },
+    {
+      "epoch": 1.1649301706938595,
+      "grad_norm": 0.08041378855705261,
+      "learning_rate": 6.9900005815226095e-06,
+      "loss": 0.0004,
+      "step": 1313
+    },
+    {
+      "epoch": 1.1658168920416758,
+      "grad_norm": 0.2825201153755188,
+      "learning_rate": 6.9856076383715845e-06,
+      "loss": 0.0043,
+      "step": 1314
+    },
+    {
+      "epoch": 1.1667036133894924,
+      "grad_norm": 0.22830918431282043,
+      "learning_rate": 6.981212874762463e-06,
+      "loss": 0.003,
+      "step": 1315
+    },
+    {
+      "epoch": 1.1675903347373089,
+      "grad_norm": 0.37838220596313477,
+      "learning_rate": 6.976816294724484e-06,
+      "loss": 0.0045,
+      "step": 1316
+    },
+    {
+      "epoch": 1.1684770560851252,
+      "grad_norm": 0.2836399972438812,
+      "learning_rate": 6.9724179022885505e-06,
+      "loss": 0.0018,
+      "step": 1317
+    },
+    {
+      "epoch": 1.1693637774329417,
+      "grad_norm": 0.39304614067077637,
+      "learning_rate": 6.968017701487223e-06,
+      "loss": 0.0048,
+      "step": 1318
+    },
+    {
+      "epoch": 1.1702504987807583,
+      "grad_norm": 0.14998720586299896,
+      "learning_rate": 6.963615696354726e-06,
+      "loss": 0.0008,
+      "step": 1319
+    },
+    {
+      "epoch": 1.1711372201285746,
+      "grad_norm": 0.10552586615085602,
+      "learning_rate": 6.9592118909269335e-06,
+      "loss": 0.0005,
+      "step": 1320
+    },
+    {
+      "epoch": 1.172023941476391,
+      "grad_norm": 0.06915909796953201,
+      "learning_rate": 6.954806289241374e-06,
+      "loss": 0.0004,
+      "step": 1321
+    },
+    {
+      "epoch": 1.1729106628242074,
+      "grad_norm": 0.18020567297935486,
+      "learning_rate": 6.950398895337216e-06,
+      "loss": 0.0011,
+      "step": 1322
+    },
+    {
+      "epoch": 1.173797384172024,
+      "grad_norm": 0.10368692129850388,
+      "learning_rate": 6.945989713255281e-06,
+      "loss": 0.0004,
+      "step": 1323
+    },
+    {
+      "epoch": 1.1746841055198405,
+      "grad_norm": 0.9406795501708984,
+      "learning_rate": 6.941578747038024e-06,
+      "loss": 0.0038,
+      "step": 1324
+    },
+    {
+      "epoch": 1.1755708268676568,
+      "grad_norm": 0.23332077264785767,
+      "learning_rate": 6.937166000729534e-06,
+      "loss": 0.0024,
+      "step": 1325
+    },
+    {
+      "epoch": 1.1764575482154733,
+      "grad_norm": 0.0631319135427475,
+      "learning_rate": 6.932751478375537e-06,
+      "loss": 0.0005,
+      "step": 1326
+    },
+    {
+      "epoch": 1.1773442695632896,
+      "grad_norm": 0.07518002390861511,
+      "learning_rate": 6.928335184023384e-06,
+      "loss": 0.0003,
+      "step": 1327
+    },
+    {
+      "epoch": 1.1782309909111062,
+      "grad_norm": 0.30457931756973267,
+      "learning_rate": 6.923917121722051e-06,
+      "loss": 0.0024,
+      "step": 1328
+    },
+    {
+      "epoch": 1.1791177122589227,
+      "grad_norm": 0.10226471722126007,
+      "learning_rate": 6.919497295522137e-06,
+      "loss": 0.0006,
+      "step": 1329
+    },
+    {
+      "epoch": 1.180004433606739,
+      "grad_norm": 0.11103340983390808,
+      "learning_rate": 6.9150757094758554e-06,
+      "loss": 0.0007,
+      "step": 1330
+    },
+    {
+      "epoch": 1.1808911549545555,
+      "grad_norm": 0.08289068937301636,
+      "learning_rate": 6.910652367637035e-06,
+      "loss": 0.0005,
+      "step": 1331
+    },
+    {
+      "epoch": 1.181777876302372,
+      "grad_norm": 0.24665898084640503,
+      "learning_rate": 6.906227274061114e-06,
+      "loss": 0.0012,
+      "step": 1332
+    },
+    {
+      "epoch": 1.1826645976501884,
+      "grad_norm": 0.3390233516693115,
+      "learning_rate": 6.901800432805135e-06,
+      "loss": 0.0034,
+      "step": 1333
+    },
+    {
+      "epoch": 1.183551318998005,
+      "grad_norm": 0.3979106843471527,
+      "learning_rate": 6.897371847927745e-06,
+      "loss": 0.0036,
+      "step": 1334
+    },
+    {
+      "epoch": 1.1844380403458212,
+      "grad_norm": 0.15676546096801758,
+      "learning_rate": 6.892941523489189e-06,
+      "loss": 0.0015,
+      "step": 1335
+    },
+    {
+      "epoch": 1.1853247616936378,
+      "grad_norm": 0.17191384732723236,
+      "learning_rate": 6.888509463551307e-06,
+      "loss": 0.0018,
+      "step": 1336
+    },
+    {
+      "epoch": 1.1862114830414543,
+      "grad_norm": 0.06105402857065201,
+      "learning_rate": 6.8840756721775285e-06,
+      "loss": 0.0005,
+      "step": 1337
+    },
+    {
+      "epoch": 1.1870982043892706,
+      "grad_norm": 0.22978165745735168,
+      "learning_rate": 6.879640153432875e-06,
+      "loss": 0.0035,
+      "step": 1338
+    },
+    {
+      "epoch": 1.1879849257370871,
+      "grad_norm": 0.28052595257759094,
+      "learning_rate": 6.875202911383945e-06,
+      "loss": 0.0018,
+      "step": 1339
+    },
+    {
+      "epoch": 1.1888716470849037,
+      "grad_norm": 0.5201903581619263,
+      "learning_rate": 6.870763950098922e-06,
+      "loss": 0.0032,
+      "step": 1340
+    },
+    {
+      "epoch": 1.18975836843272,
+      "grad_norm": 0.13970860838890076,
+      "learning_rate": 6.866323273647564e-06,
+      "loss": 0.0006,
+      "step": 1341
+    },
+    {
+      "epoch": 1.1906450897805365,
+      "grad_norm": 0.0728214755654335,
+      "learning_rate": 6.8618808861012025e-06,
+      "loss": 0.0002,
+      "step": 1342
+    },
+    {
+      "epoch": 1.1915318111283528,
+      "grad_norm": 0.16536439955234528,
+      "learning_rate": 6.8574367915327345e-06,
+      "loss": 0.0006,
+      "step": 1343
+    },
+    {
+      "epoch": 1.1924185324761694,
+      "grad_norm": 0.48193657398223877,
+      "learning_rate": 6.852990994016627e-06,
+      "loss": 0.0034,
+      "step": 1344
+    },
+    {
+      "epoch": 1.193305253823986,
+      "grad_norm": 0.45254138112068176,
+      "learning_rate": 6.848543497628905e-06,
+      "loss": 0.0036,
+      "step": 1345
+    },
+    {
+      "epoch": 1.1941919751718022,
+      "grad_norm": 0.11150401830673218,
+      "learning_rate": 6.844094306447152e-06,
+      "loss": 0.0008,
+      "step": 1346
+    },
+    {
+      "epoch": 1.1950786965196187,
+      "grad_norm": 0.23752820491790771,
+      "learning_rate": 6.839643424550506e-06,
+      "loss": 0.0023,
+      "step": 1347
+    },
+    {
+      "epoch": 1.195965417867435,
+      "grad_norm": 0.012399467639625072,
+      "learning_rate": 6.835190856019654e-06,
+      "loss": 0.0001,
+      "step": 1348
+    },
+    {
+      "epoch": 1.1968521392152516,
+      "grad_norm": 0.33054450154304504,
+      "learning_rate": 6.8307366049368285e-06,
+      "loss": 0.0026,
+      "step": 1349
+    },
+    {
+      "epoch": 1.197738860563068,
+      "grad_norm": 0.714648425579071,
+      "learning_rate": 6.826280675385807e-06,
+      "loss": 0.0108,
+      "step": 1350
+    },
+    {
+      "epoch": 1.1986255819108844,
+      "grad_norm": 0.22860702872276306,
+      "learning_rate": 6.821823071451905e-06,
+      "loss": 0.0023,
+      "step": 1351
+    },
+    {
+      "epoch": 1.199512303258701,
+      "grad_norm": 0.6131689548492432,
+      "learning_rate": 6.817363797221971e-06,
+      "loss": 0.0072,
+      "step": 1352
+    },
+    {
+      "epoch": 1.2003990246065175,
+      "grad_norm": 0.2593003213405609,
+      "learning_rate": 6.812902856784388e-06,
+      "loss": 0.0014,
+      "step": 1353
+    },
+    {
+      "epoch": 1.2012857459543338,
+      "grad_norm": 0.177522674202919,
+      "learning_rate": 6.808440254229066e-06,
+      "loss": 0.0011,
+      "step": 1354
+    },
+    {
+      "epoch": 1.2021724673021503,
+      "grad_norm": 0.2457665652036667,
+      "learning_rate": 6.803975993647437e-06,
+      "loss": 0.0034,
+      "step": 1355
+    },
+    {
+      "epoch": 1.2030591886499669,
+      "grad_norm": 0.11679854243993759,
+      "learning_rate": 6.7995100791324545e-06,
+      "loss": 0.0012,
+      "step": 1356
+    },
+    {
+      "epoch": 1.2039459099977832,
+      "grad_norm": 0.07699250429868698,
+      "learning_rate": 6.7950425147785885e-06,
+      "loss": 0.0007,
+      "step": 1357
+    },
+    {
+      "epoch": 1.2048326313455997,
+      "grad_norm": 0.5148756504058838,
+      "learning_rate": 6.79057330468182e-06,
+      "loss": 0.0024,
+      "step": 1358
+    },
+    {
+      "epoch": 1.205719352693416,
+      "grad_norm": 0.1937873661518097,
+      "learning_rate": 6.786102452939642e-06,
+      "loss": 0.0033,
+      "step": 1359
+    },
+    {
+      "epoch": 1.2066060740412325,
+      "grad_norm": 0.362888365983963,
+      "learning_rate": 6.78162996365105e-06,
+      "loss": 0.0035,
+      "step": 1360
+    },
+    {
+      "epoch": 1.207492795389049,
+      "grad_norm": 0.3451005816459656,
+      "learning_rate": 6.7771558409165405e-06,
+      "loss": 0.0029,
+      "step": 1361
+    },
+    {
+      "epoch": 1.2083795167368654,
+      "grad_norm": 0.1482471525669098,
+      "learning_rate": 6.7726800888381105e-06,
+      "loss": 0.0009,
+      "step": 1362
+    },
+    {
+      "epoch": 1.209266238084682,
+      "grad_norm": 0.11421269923448563,
+      "learning_rate": 6.768202711519248e-06,
+      "loss": 0.0015,
+      "step": 1363
+    },
+    {
+      "epoch": 1.2101529594324982,
+      "grad_norm": 0.18156251311302185,
+      "learning_rate": 6.763723713064933e-06,
+      "loss": 0.0015,
+      "step": 1364
+    },
+    {
+      "epoch": 1.2110396807803148,
+      "grad_norm": 0.07586735486984253,
+      "learning_rate": 6.759243097581629e-06,
+      "loss": 0.0013,
+      "step": 1365
+    },
+    {
+      "epoch": 1.2119264021281313,
+      "grad_norm": 0.2878761887550354,
+      "learning_rate": 6.754760869177285e-06,
+      "loss": 0.0031,
+      "step": 1366
+    },
+    {
+      "epoch": 1.2128131234759476,
+      "grad_norm": 0.6761088967323303,
+      "learning_rate": 6.750277031961328e-06,
+      "loss": 0.0055,
+      "step": 1367
+    },
+    {
+      "epoch": 1.2136998448237641,
+      "grad_norm": 0.13870671391487122,
+      "learning_rate": 6.745791590044659e-06,
+      "loss": 0.0012,
+      "step": 1368
+    },
+    {
+      "epoch": 1.2145865661715807,
+      "grad_norm": 0.0526190847158432,
+      "learning_rate": 6.741304547539652e-06,
+      "loss": 0.0003,
+      "step": 1369
+    },
+    {
+      "epoch": 1.215473287519397,
+      "grad_norm": 0.6361402273178101,
+      "learning_rate": 6.7368159085601456e-06,
+      "loss": 0.0045,
+      "step": 1370
+    },
+    {
+      "epoch": 1.2163600088672135,
+      "grad_norm": 0.09560070186853409,
+      "learning_rate": 6.7323256772214455e-06,
+      "loss": 0.0006,
+      "step": 1371
+    },
+    {
+      "epoch": 1.21724673021503,
+      "grad_norm": 0.13184469938278198,
+      "learning_rate": 6.727833857640316e-06,
+      "loss": 0.0008,
+      "step": 1372
+    },
+    {
+      "epoch": 1.2181334515628464,
+      "grad_norm": 0.2193000614643097,
+      "learning_rate": 6.723340453934974e-06,
+      "loss": 0.0028,
+      "step": 1373
+    },
+    {
+      "epoch": 1.219020172910663,
+      "grad_norm": 0.13727962970733643,
+      "learning_rate": 6.718845470225095e-06,
+      "loss": 0.0008,
+      "step": 1374
+    },
+    {
+      "epoch": 1.2199068942584792,
+      "grad_norm": 0.042113423347473145,
+      "learning_rate": 6.7143489106318e-06,
+      "loss": 0.0003,
+      "step": 1375
+    },
+    {
+      "epoch": 1.2207936156062957,
+      "grad_norm": 0.1653071641921997,
+      "learning_rate": 6.709850779277653e-06,
+      "loss": 0.0012,
+      "step": 1376
+    },
+    {
+      "epoch": 1.2216803369541123,
+      "grad_norm": 0.21689732372760773,
+      "learning_rate": 6.705351080286664e-06,
+      "loss": 0.0012,
+      "step": 1377
+    },
+    {
+      "epoch": 1.2225670583019286,
+      "grad_norm": 0.19632472097873688,
+      "learning_rate": 6.700849817784274e-06,
+      "loss": 0.0017,
+      "step": 1378
+    },
+    {
+      "epoch": 1.223453779649745,
+      "grad_norm": 0.020970387384295464,
+      "learning_rate": 6.696346995897363e-06,
+      "loss": 0.0002,
+      "step": 1379
+    },
+    {
+      "epoch": 1.2243405009975614,
+      "grad_norm": 0.07651633024215698,
+      "learning_rate": 6.691842618754235e-06,
+      "loss": 0.0004,
+      "step": 1380
+    },
+    {
+      "epoch": 1.225227222345378,
+      "grad_norm": 0.1662818342447281,
+      "learning_rate": 6.687336690484626e-06,
+      "loss": 0.0011,
+      "step": 1381
+    },
+    {
+      "epoch": 1.2261139436931945,
+      "grad_norm": 0.19208109378814697,
+      "learning_rate": 6.682829215219692e-06,
+      "loss": 0.0012,
+      "step": 1382
+    },
+    {
+      "epoch": 1.2270006650410108,
+      "grad_norm": 0.09741144627332687,
+      "learning_rate": 6.678320197092004e-06,
+      "loss": 0.0006,
+      "step": 1383
+    },
+    {
+      "epoch": 1.2278873863888273,
+      "grad_norm": 0.17250119149684906,
+      "learning_rate": 6.673809640235552e-06,
+      "loss": 0.0012,
+      "step": 1384
+    },
+    {
+      "epoch": 1.2287741077366436,
+      "grad_norm": 0.13515061140060425,
+      "learning_rate": 6.6692975487857355e-06,
+      "loss": 0.0009,
+      "step": 1385
+    },
+    {
+      "epoch": 1.2296608290844602,
+      "grad_norm": 0.2506014406681061,
+      "learning_rate": 6.664783926879359e-06,
+      "loss": 0.0029,
+      "step": 1386
+    },
+    {
+      "epoch": 1.2305475504322767,
+      "grad_norm": 0.1211366206407547,
+      "learning_rate": 6.660268778654631e-06,
+      "loss": 0.0004,
+      "step": 1387
+    },
+    {
+      "epoch": 1.231434271780093,
+      "grad_norm": 0.20620831847190857,
+      "learning_rate": 6.655752108251161e-06,
+      "loss": 0.0008,
+      "step": 1388
+    },
+    {
+      "epoch": 1.2323209931279095,
+      "grad_norm": 0.785988450050354,
+      "learning_rate": 6.651233919809955e-06,
+      "loss": 0.0068,
+      "step": 1389
+    },
+    {
+      "epoch": 1.233207714475726,
+      "grad_norm": 0.35752174258232117,
+      "learning_rate": 6.6467142174734055e-06,
+      "loss": 0.0022,
+      "step": 1390
+    },
+    {
+      "epoch": 1.2340944358235424,
+      "grad_norm": 0.9221594929695129,
+      "learning_rate": 6.642193005385298e-06,
+      "loss": 0.0031,
+      "step": 1391
+    },
+    {
+      "epoch": 1.234981157171359,
+      "grad_norm": 0.4186972677707672,
+      "learning_rate": 6.6376702876908e-06,
+      "loss": 0.0023,
+      "step": 1392
+    },
+    {
+      "epoch": 1.2358678785191755,
+      "grad_norm": 0.18855422735214233,
+      "learning_rate": 6.63314606853646e-06,
+      "loss": 0.0005,
+      "step": 1393
+    },
+    {
+      "epoch": 1.2367545998669918,
+      "grad_norm": 0.047466155141592026,
+      "learning_rate": 6.628620352070201e-06,
+      "loss": 0.0003,
+      "step": 1394
+    },
+    {
+      "epoch": 1.2376413212148083,
+      "grad_norm": 0.17096386849880219,
+      "learning_rate": 6.624093142441323e-06,
+      "loss": 0.0006,
+      "step": 1395
+    },
+    {
+      "epoch": 1.2385280425626246,
+      "grad_norm": 0.038220904767513275,
+      "learning_rate": 6.619564443800494e-06,
+      "loss": 0.0003,
+      "step": 1396
+    },
+    {
+      "epoch": 1.2394147639104411,
+      "grad_norm": 0.006028464064002037,
+      "learning_rate": 6.615034260299742e-06,
+      "loss": 0.0,
+      "step": 1397
+    },
+    {
+      "epoch": 1.2403014852582577,
+      "grad_norm": 0.48989105224609375,
+      "learning_rate": 6.610502596092464e-06,
+      "loss": 0.004,
+      "step": 1398
+    },
+    {
+      "epoch": 1.241188206606074,
+      "grad_norm": 0.06377265602350235,
+      "learning_rate": 6.605969455333407e-06,
+      "loss": 0.0002,
+      "step": 1399
+    },
+    {
+      "epoch": 1.2420749279538905,
+      "grad_norm": 0.513083815574646,
+      "learning_rate": 6.601434842178678e-06,
+      "loss": 0.0088,
+      "step": 1400
+    },
+    {
+      "epoch": 1.2429616493017068,
+      "grad_norm": 0.038441240787506104,
+      "learning_rate": 6.596898760785731e-06,
+      "loss": 0.0002,
+      "step": 1401
+    },
+    {
+      "epoch": 1.2438483706495234,
+      "grad_norm": 1.2739007472991943,
+      "learning_rate": 6.592361215313365e-06,
+      "loss": 0.0077,
+      "step": 1402
+    },
+    {
+      "epoch": 1.24473509199734,
+      "grad_norm": 0.3760228157043457,
+      "learning_rate": 6.587822209921726e-06,
+      "loss": 0.005,
+      "step": 1403
+    },
+    {
+      "epoch": 1.2456218133451562,
+      "grad_norm": 0.2797979414463043,
+      "learning_rate": 6.583281748772291e-06,
+      "loss": 0.0008,
+      "step": 1404
+    },
+    {
+      "epoch": 1.2465085346929727,
+      "grad_norm": 0.30799371004104614,
+      "learning_rate": 6.5787398360278795e-06,
+      "loss": 0.0036,
+      "step": 1405
+    },
+    {
+      "epoch": 1.2473952560407893,
+      "grad_norm": 0.0893583670258522,
+      "learning_rate": 6.5741964758526375e-06,
+      "loss": 0.0005,
+      "step": 1406
+    },
+    {
+      "epoch": 1.2482819773886056,
+      "grad_norm": 0.5812922716140747,
+      "learning_rate": 6.569651672412038e-06,
+      "loss": 0.0091,
+      "step": 1407
+    },
+    {
+      "epoch": 1.2491686987364221,
+      "grad_norm": 0.4839387834072113,
+      "learning_rate": 6.56510542987288e-06,
+      "loss": 0.0051,
+      "step": 1408
+    },
+    {
+      "epoch": 1.2500554200842386,
+      "grad_norm": 0.04253406077623367,
+      "learning_rate": 6.560557752403277e-06,
+      "loss": 0.0005,
+      "step": 1409
+    },
+    {
+      "epoch": 1.250942141432055,
+      "grad_norm": 0.3183040916919708,
+      "learning_rate": 6.556008644172664e-06,
+      "loss": 0.0042,
+      "step": 1410
+    },
+    {
+      "epoch": 1.2518288627798715,
+      "grad_norm": 0.05635031685233116,
+      "learning_rate": 6.551458109351785e-06,
+      "loss": 0.0004,
+      "step": 1411
+    },
+    {
+      "epoch": 1.2527155841276878,
+      "grad_norm": 0.35388946533203125,
+      "learning_rate": 6.546906152112689e-06,
+      "loss": 0.0101,
+      "step": 1412
+    },
+    {
+      "epoch": 1.2536023054755043,
+      "grad_norm": 0.26416826248168945,
+      "learning_rate": 6.542352776628733e-06,
+      "loss": 0.0027,
+      "step": 1413
+    },
+    {
+      "epoch": 1.2544890268233209,
+      "grad_norm": 0.06187260150909424,
+      "learning_rate": 6.537797987074574e-06,
+      "loss": 0.0006,
+      "step": 1414
+    },
+    {
+      "epoch": 1.2553757481711372,
+      "grad_norm": 0.399342805147171,
+      "learning_rate": 6.533241787626165e-06,
+      "loss": 0.0059,
+      "step": 1415
+    },
+    {
+      "epoch": 1.2562624695189537,
+      "grad_norm": 0.34618619084358215,
+      "learning_rate": 6.5286841824607515e-06,
+      "loss": 0.0024,
+      "step": 1416
+    },
+    {
+      "epoch": 1.25714919086677,
+      "grad_norm": 0.08505801856517792,
+      "learning_rate": 6.524125175756866e-06,
+      "loss": 0.001,
+      "step": 1417
+    },
+    {
+      "epoch": 1.2580359122145865,
+      "grad_norm": 0.28683024644851685,
+      "learning_rate": 6.519564771694328e-06,
+      "loss": 0.0032,
+      "step": 1418
+    },
+    {
+      "epoch": 1.258922633562403,
+      "grad_norm": 0.1888272911310196,
+      "learning_rate": 6.515002974454239e-06,
+      "loss": 0.0019,
+      "step": 1419
+    },
+    {
+      "epoch": 1.2598093549102194,
+      "grad_norm": 0.10655654221773148,
+      "learning_rate": 6.510439788218975e-06,
+      "loss": 0.001,
+      "step": 1420
+    },
+    {
+      "epoch": 1.260696076258036,
+      "grad_norm": 0.16792140901088715,
+      "learning_rate": 6.505875217172188e-06,
+      "loss": 0.0017,
+      "step": 1421
+    },
+    {
+      "epoch": 1.2615827976058522,
+      "grad_norm": 0.5658694505691528,
+      "learning_rate": 6.501309265498797e-06,
+      "loss": 0.0057,
+      "step": 1422
+    },
+    {
+      "epoch": 1.2624695189536688,
+      "grad_norm": 0.3947063386440277,
+      "learning_rate": 6.496741937384992e-06,
+      "loss": 0.003,
+      "step": 1423
+    },
+    {
+      "epoch": 1.2633562403014853,
+      "grad_norm": 0.23458246886730194,
+      "learning_rate": 6.492173237018217e-06,
+      "loss": 0.0049,
+      "step": 1424
+    },
+    {
+      "epoch": 1.2642429616493018,
+      "grad_norm": 0.2951560616493225,
+      "learning_rate": 6.48760316858718e-06,
+      "loss": 0.0042,
+      "step": 1425
+    },
+    {
+      "epoch": 1.2651296829971181,
+      "grad_norm": 0.10598162561655045,
+      "learning_rate": 6.483031736281843e-06,
+      "loss": 0.0009,
+      "step": 1426
+    },
+    {
+      "epoch": 1.2660164043449347,
+      "grad_norm": 0.2090543806552887,
+      "learning_rate": 6.478458944293416e-06,
+      "loss": 0.0011,
+      "step": 1427
+    },
+    {
+      "epoch": 1.266903125692751,
+      "grad_norm": 0.18498195707798004,
+      "learning_rate": 6.473884796814357e-06,
+      "loss": 0.0019,
+      "step": 1428
+    },
+    {
+      "epoch": 1.2677898470405675,
+      "grad_norm": 0.3451266288757324,
+      "learning_rate": 6.469309298038367e-06,
+      "loss": 0.0038,
+      "step": 1429
+    },
+    {
+      "epoch": 1.268676568388384,
+      "grad_norm": 0.35533058643341064,
+      "learning_rate": 6.4647324521603845e-06,
+      "loss": 0.0062,
+      "step": 1430
+    },
+    {
+      "epoch": 1.2695632897362004,
+      "grad_norm": 0.1544898897409439,
+      "learning_rate": 6.4601542633765834e-06,
+      "loss": 0.0011,
+      "step": 1431
+    },
+    {
+      "epoch": 1.270450011084017,
+      "grad_norm": 0.4989369809627533,
+      "learning_rate": 6.455574735884373e-06,
+      "loss": 0.0028,
+      "step": 1432
+    },
+    {
+      "epoch": 1.2713367324318332,
+      "grad_norm": 0.13420936465263367,
+      "learning_rate": 6.450993873882383e-06,
+      "loss": 0.0007,
+      "step": 1433
+    },
+    {
+      "epoch": 1.2722234537796497,
+      "grad_norm": 0.046916741877794266,
+      "learning_rate": 6.44641168157047e-06,
+      "loss": 0.0004,
+      "step": 1434
+    },
+    {
+      "epoch": 1.2731101751274663,
+      "grad_norm": 0.7304254174232483,
+      "learning_rate": 6.44182816314971e-06,
+      "loss": 0.0039,
+      "step": 1435
+    },
+    {
+      "epoch": 1.2739968964752826,
+      "grad_norm": 0.979017972946167,
+      "learning_rate": 6.4372433228224e-06,
+      "loss": 0.0069,
+      "step": 1436
+    },
+    {
+      "epoch": 1.2748836178230991,
+      "grad_norm": 0.054384853690862656,
+      "learning_rate": 6.432657164792038e-06,
+      "loss": 0.0004,
+      "step": 1437
+    },
+    {
+      "epoch": 1.2757703391709154,
+      "grad_norm": 0.21733501553535461,
+      "learning_rate": 6.428069693263337e-06,
+      "loss": 0.0024,
+      "step": 1438
+    },
+    {
+      "epoch": 1.276657060518732,
+      "grad_norm": 0.3343576192855835,
+      "learning_rate": 6.423480912442216e-06,
+      "loss": 0.0031,
+      "step": 1439
+    },
+    {
+      "epoch": 1.2775437818665485,
+      "grad_norm": 0.2468477189540863,
+      "learning_rate": 6.418890826535791e-06,
+      "loss": 0.0018,
+      "step": 1440
+    },
+    {
+      "epoch": 1.278430503214365,
+      "grad_norm": 0.38270869851112366,
+      "learning_rate": 6.4142994397523735e-06,
+      "loss": 0.004,
+      "step": 1441
+    },
+    {
+      "epoch": 1.2793172245621813,
+      "grad_norm": 0.03215061128139496,
+      "learning_rate": 6.409706756301473e-06,
+      "loss": 0.0003,
+      "step": 1442
+    },
+    {
+      "epoch": 1.2802039459099979,
+      "grad_norm": 0.06875225156545639,
+      "learning_rate": 6.405112780393781e-06,
+      "loss": 0.0005,
+      "step": 1443
+    },
+    {
+      "epoch": 1.2810906672578142,
+      "grad_norm": 0.21331429481506348,
+      "learning_rate": 6.400517516241178e-06,
+      "loss": 0.0007,
+      "step": 1444
+    },
+    {
+      "epoch": 1.2819773886056307,
+      "grad_norm": 0.21490368247032166,
+      "learning_rate": 6.395920968056725e-06,
+      "loss": 0.0005,
+      "step": 1445
+    },
+    {
+      "epoch": 1.2828641099534472,
+      "grad_norm": 0.09645006060600281,
+      "learning_rate": 6.39132314005466e-06,
+      "loss": 0.0007,
+      "step": 1446
+    },
+    {
+      "epoch": 1.2837508313012636,
+      "grad_norm": 0.3721943795681,
+      "learning_rate": 6.386724036450396e-06,
+      "loss": 0.0065,
+      "step": 1447
+    },
+    {
+      "epoch": 1.28463755264908,
+      "grad_norm": 0.22285206615924835,
+      "learning_rate": 6.382123661460511e-06,
+      "loss": 0.0015,
+      "step": 1448
+    },
+    {
+      "epoch": 1.2855242739968964,
+      "grad_norm": 0.08951204270124435,
+      "learning_rate": 6.377522019302756e-06,
+      "loss": 0.0007,
+      "step": 1449
+    },
+    {
+      "epoch": 1.286410995344713,
+      "grad_norm": 0.23946380615234375,
+      "learning_rate": 6.372919114196035e-06,
+      "loss": 0.0013,
+      "step": 1450
+    },
+    {
+      "epoch": 1.2872977166925295,
+      "grad_norm": 0.05512690171599388,
+      "learning_rate": 6.368314950360416e-06,
+      "loss": 0.0004,
+      "step": 1451
+    },
+    {
+      "epoch": 1.2881844380403458,
+      "grad_norm": 0.43557584285736084,
+      "learning_rate": 6.3637095320171185e-06,
+      "loss": 0.0027,
+      "step": 1452
+    },
+    {
+      "epoch": 1.2890711593881623,
+      "grad_norm": 0.27949899435043335,
+      "learning_rate": 6.359102863388515e-06,
+      "loss": 0.0013,
+      "step": 1453
+    },
+    {
+      "epoch": 1.2899578807359786,
+      "grad_norm": 0.18375392258167267,
+      "learning_rate": 6.35449494869812e-06,
+      "loss": 0.001,
+      "step": 1454
+    },
+    {
+      "epoch": 1.2908446020837951,
+      "grad_norm": 0.1614723652601242,
+      "learning_rate": 6.349885792170595e-06,
+      "loss": 0.0009,
+      "step": 1455
+    },
+    {
+      "epoch": 1.2917313234316117,
+      "grad_norm": 0.13602754473686218,
+      "learning_rate": 6.345275398031736e-06,
+      "loss": 0.0007,
+      "step": 1456
+    },
+    {
+      "epoch": 1.292618044779428,
+      "grad_norm": 0.22377264499664307,
+      "learning_rate": 6.340663770508476e-06,
+      "loss": 0.0014,
+      "step": 1457
+    },
+    {
+      "epoch": 1.2935047661272445,
+      "grad_norm": 0.7576924562454224,
+      "learning_rate": 6.33605091382888e-06,
+      "loss": 0.0027,
+      "step": 1458
+    },
+    {
+      "epoch": 1.2943914874750608,
+      "grad_norm": 0.019061360508203506,
+      "learning_rate": 6.331436832222133e-06,
+      "loss": 0.0001,
+      "step": 1459
+    },
+    {
+      "epoch": 1.2952782088228774,
+      "grad_norm": 0.2781587839126587,
+      "learning_rate": 6.3268215299185545e-06,
+      "loss": 0.0011,
+      "step": 1460
+    },
+    {
+      "epoch": 1.296164930170694,
+      "grad_norm": 0.0837785005569458,
+      "learning_rate": 6.3222050111495715e-06,
+      "loss": 0.0004,
+      "step": 1461
+    },
+    {
+      "epoch": 1.2970516515185104,
+      "grad_norm": 0.37472015619277954,
+      "learning_rate": 6.317587280147737e-06,
+      "loss": 0.0069,
+      "step": 1462
+    },
+    {
+      "epoch": 1.2979383728663267,
+      "grad_norm": 0.09563448280096054,
+      "learning_rate": 6.312968341146705e-06,
+      "loss": 0.0004,
+      "step": 1463
+    },
+    {
+      "epoch": 1.2988250942141433,
+      "grad_norm": 0.029751967638731003,
+      "learning_rate": 6.308348198381244e-06,
+      "loss": 0.0002,
+      "step": 1464
+    },
+    {
+      "epoch": 1.2997118155619596,
+      "grad_norm": 0.5495652556419373,
+      "learning_rate": 6.303726856087227e-06,
+      "loss": 0.0031,
+      "step": 1465
+    },
+    {
+      "epoch": 1.3005985369097761,
+      "grad_norm": 0.20881287753582,
+      "learning_rate": 6.299104318501619e-06,
+      "loss": 0.0007,
+      "step": 1466
+    },
+    {
+      "epoch": 1.3014852582575926,
+      "grad_norm": 0.5076218843460083,
+      "learning_rate": 6.294480589862491e-06,
+      "loss": 0.0035,
+      "step": 1467
+    },
+    {
+      "epoch": 1.302371979605409,
+      "grad_norm": 0.3442370593547821,
+      "learning_rate": 6.289855674408995e-06,
+      "loss": 0.0034,
+      "step": 1468
+    },
+    {
+      "epoch": 1.3032587009532255,
+      "grad_norm": 0.9615652561187744,
+      "learning_rate": 6.285229576381382e-06,
+      "loss": 0.0082,
+      "step": 1469
+    },
+    {
+      "epoch": 1.3041454223010418,
+      "grad_norm": 0.042115919291973114,
+      "learning_rate": 6.280602300020982e-06,
+      "loss": 0.0002,
+      "step": 1470
+    },
+    {
+      "epoch": 1.3050321436488583,
+      "grad_norm": 0.15722385048866272,
+      "learning_rate": 6.275973849570201e-06,
+      "loss": 0.0008,
+      "step": 1471
+    },
+    {
+      "epoch": 1.3059188649966749,
+      "grad_norm": 0.011659053154289722,
+      "learning_rate": 6.27134422927253e-06,
+      "loss": 0.0001,
+      "step": 1472
+    },
+    {
+      "epoch": 1.3068055863444912,
+      "grad_norm": 0.05210238695144653,
+      "learning_rate": 6.266713443372527e-06,
+      "loss": 0.0003,
+      "step": 1473
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.6101747751235962,
+      "learning_rate": 6.262081496115822e-06,
+      "loss": 0.0065,
+      "step": 1474
+    },
+    {
+      "epoch": 1.308579029040124,
+      "grad_norm": 0.5870680212974548,
+      "learning_rate": 6.257448391749104e-06,
+      "loss": 0.0044,
+      "step": 1475
+    },
+    {
+      "epoch": 1.3094657503879406,
+      "grad_norm": 0.03537415340542793,
+      "learning_rate": 6.252814134520131e-06,
+      "loss": 0.0003,
+      "step": 1476
+    },
+    {
+      "epoch": 1.310352471735757,
+      "grad_norm": 0.23479506373405457,
+      "learning_rate": 6.2481787286777116e-06,
+      "loss": 0.0007,
+      "step": 1477
+    },
+    {
+      "epoch": 1.3112391930835736,
+      "grad_norm": 0.6235648989677429,
+      "learning_rate": 6.24354217847171e-06,
+      "loss": 0.0072,
+      "step": 1478
+    },
+    {
+      "epoch": 1.31212591443139,
+      "grad_norm": 0.5917138457298279,
+      "learning_rate": 6.238904488153039e-06,
+      "loss": 0.0114,
+      "step": 1479
+    },
+    {
+      "epoch": 1.3130126357792065,
+      "grad_norm": 0.3008297085762024,
+      "learning_rate": 6.234265661973656e-06,
+      "loss": 0.0026,
+      "step": 1480
+    },
+    {
+      "epoch": 1.3138993571270228,
+      "grad_norm": 0.2608731687068939,
+      "learning_rate": 6.229625704186564e-06,
+      "loss": 0.0043,
+      "step": 1481
+    },
+    {
+      "epoch": 1.3147860784748393,
+      "grad_norm": 0.46215519309043884,
+      "learning_rate": 6.2249846190457964e-06,
+      "loss": 0.0085,
+      "step": 1482
+    },
+    {
+      "epoch": 1.3156727998226558,
+      "grad_norm": 0.10965245962142944,
+      "learning_rate": 6.220342410806426e-06,
+      "loss": 0.0006,
+      "step": 1483
+    },
+    {
+      "epoch": 1.3165595211704721,
+      "grad_norm": 0.10256587713956833,
+      "learning_rate": 6.215699083724553e-06,
+      "loss": 0.0007,
+      "step": 1484
+    },
+    {
+      "epoch": 1.3174462425182887,
+      "grad_norm": 0.03453338146209717,
+      "learning_rate": 6.211054642057301e-06,
+      "loss": 0.0004,
+      "step": 1485
+    },
+    {
+      "epoch": 1.318332963866105,
+      "grad_norm": 0.14215219020843506,
+      "learning_rate": 6.206409090062823e-06,
+      "loss": 0.0011,
+      "step": 1486
+    },
+    {
+      "epoch": 1.3192196852139215,
+      "grad_norm": 0.30575263500213623,
+      "learning_rate": 6.201762432000283e-06,
+      "loss": 0.002,
+      "step": 1487
+    },
+    {
+      "epoch": 1.320106406561738,
+      "grad_norm": 0.19290179014205933,
+      "learning_rate": 6.197114672129859e-06,
+      "loss": 0.0011,
+      "step": 1488
+    },
+    {
+      "epoch": 1.3209931279095544,
+      "grad_norm": 0.10399499535560608,
+      "learning_rate": 6.192465814712744e-06,
+      "loss": 0.0008,
+      "step": 1489
+    },
+    {
+      "epoch": 1.321879849257371,
+      "grad_norm": 0.20446741580963135,
+      "learning_rate": 6.187815864011133e-06,
+      "loss": 0.0012,
+      "step": 1490
+    },
+    {
+      "epoch": 1.3227665706051872,
+      "grad_norm": 0.07639744877815247,
+      "learning_rate": 6.183164824288226e-06,
+      "loss": 0.001,
+      "step": 1491
+    },
+    {
+      "epoch": 1.3236532919530037,
+      "grad_norm": 0.04390874132514,
+      "learning_rate": 6.17851269980822e-06,
+      "loss": 0.0006,
+      "step": 1492
+    },
+    {
+      "epoch": 1.3245400133008203,
+      "grad_norm": 0.20111678540706635,
+      "learning_rate": 6.173859494836309e-06,
+      "loss": 0.0012,
+      "step": 1493
+    },
+    {
+      "epoch": 1.3254267346486366,
+      "grad_norm": 0.1883375197649002,
+      "learning_rate": 6.169205213638671e-06,
+      "loss": 0.0029,
+      "step": 1494
+    },
+    {
+      "epoch": 1.3263134559964531,
+      "grad_norm": 0.1244998425245285,
+      "learning_rate": 6.1645498604824785e-06,
+      "loss": 0.0004,
+      "step": 1495
+    },
+    {
+      "epoch": 1.3272001773442694,
+      "grad_norm": 0.11627326160669327,
+      "learning_rate": 6.1598934396358826e-06,
+      "loss": 0.0012,
+      "step": 1496
+    },
+    {
+      "epoch": 1.328086898692086,
+      "grad_norm": 0.12732315063476562,
+      "learning_rate": 6.155235955368014e-06,
+      "loss": 0.0008,
+      "step": 1497
+    },
+    {
+      "epoch": 1.3289736200399025,
+      "grad_norm": 0.18036852777004242,
+      "learning_rate": 6.1505774119489805e-06,
+      "loss": 0.0009,
+      "step": 1498
+    },
+    {
+      "epoch": 1.329860341387719,
+      "grad_norm": 0.2986014783382416,
+      "learning_rate": 6.145917813649856e-06,
+      "loss": 0.0022,
+      "step": 1499
+    },
+    {
+      "epoch": 1.3307470627355353,
+      "grad_norm": 0.11444338411092758,
+      "learning_rate": 6.141257164742689e-06,
+      "loss": 0.0009,
+      "step": 1500
+    },
+    {
+      "epoch": 1.3316337840833519,
+      "grad_norm": 0.3222125470638275,
+      "learning_rate": 6.1365954695004815e-06,
+      "loss": 0.0024,
+      "step": 1501
+    },
+    {
+      "epoch": 1.3325205054311682,
+      "grad_norm": 0.17182210087776184,
+      "learning_rate": 6.131932732197205e-06,
+      "loss": 0.0005,
+      "step": 1502
+    },
+    {
+      "epoch": 1.3334072267789847,
+      "grad_norm": 0.5611296892166138,
+      "learning_rate": 6.127268957107777e-06,
+      "loss": 0.0045,
+      "step": 1503
+    },
+    {
+      "epoch": 1.3342939481268012,
+      "grad_norm": 0.11059463024139404,
+      "learning_rate": 6.122604148508076e-06,
+      "loss": 0.0005,
+      "step": 1504
+    },
+    {
+      "epoch": 1.3351806694746176,
+      "grad_norm": 0.4216887354850769,
+      "learning_rate": 6.117938310674918e-06,
+      "loss": 0.0023,
+      "step": 1505
+    },
+    {
+      "epoch": 1.336067390822434,
+      "grad_norm": 0.04834161698818207,
+      "learning_rate": 6.113271447886071e-06,
+      "loss": 0.0002,
+      "step": 1506
+    },
+    {
+      "epoch": 1.3369541121702504,
+      "grad_norm": 0.044341325759887695,
+      "learning_rate": 6.108603564420239e-06,
+      "loss": 0.0002,
+      "step": 1507
+    },
+    {
+      "epoch": 1.337840833518067,
+      "grad_norm": 0.5602015256881714,
+      "learning_rate": 6.103934664557061e-06,
+      "loss": 0.0045,
+      "step": 1508
+    },
+    {
+      "epoch": 1.3387275548658835,
+      "grad_norm": 0.10628864914178848,
+      "learning_rate": 6.09926475257711e-06,
+      "loss": 0.0005,
+      "step": 1509
+    },
+    {
+      "epoch": 1.3396142762136998,
+      "grad_norm": 0.47171446681022644,
+      "learning_rate": 6.0945938327618855e-06,
+      "loss": 0.0074,
+      "step": 1510
+    },
+    {
+      "epoch": 1.3405009975615163,
+      "grad_norm": 0.41396036744117737,
+      "learning_rate": 6.089921909393812e-06,
+      "loss": 0.0027,
+      "step": 1511
+    },
+    {
+      "epoch": 1.3413877189093326,
+      "grad_norm": 0.5207869410514832,
+      "learning_rate": 6.085248986756233e-06,
+      "loss": 0.0069,
+      "step": 1512
+    },
+    {
+      "epoch": 1.3422744402571491,
+      "grad_norm": 0.057785842567682266,
+      "learning_rate": 6.08057506913341e-06,
+      "loss": 0.0004,
+      "step": 1513
+    },
+    {
+      "epoch": 1.3431611616049657,
+      "grad_norm": 0.6101086735725403,
+      "learning_rate": 6.0759001608105126e-06,
+      "loss": 0.0066,
+      "step": 1514
+    },
+    {
+      "epoch": 1.3440478829527822,
+      "grad_norm": 0.4098615050315857,
+      "learning_rate": 6.071224266073625e-06,
+      "loss": 0.0055,
+      "step": 1515
+    },
+    {
+      "epoch": 1.3449346043005985,
+      "grad_norm": 0.33896300196647644,
+      "learning_rate": 6.066547389209731e-06,
+      "loss": 0.0023,
+      "step": 1516
+    },
+    {
+      "epoch": 1.345821325648415,
+      "grad_norm": 0.49446535110473633,
+      "learning_rate": 6.061869534506715e-06,
+      "loss": 0.0073,
+      "step": 1517
+    },
+    {
+      "epoch": 1.3467080469962314,
+      "grad_norm": 0.15998847782611847,
+      "learning_rate": 6.0571907062533604e-06,
+      "loss": 0.0008,
+      "step": 1518
+    },
+    {
+      "epoch": 1.347594768344048,
+      "grad_norm": 0.2155783772468567,
+      "learning_rate": 6.052510908739342e-06,
+      "loss": 0.0019,
+      "step": 1519
+    },
+    {
+      "epoch": 1.3484814896918644,
+      "grad_norm": 0.41385430097579956,
+      "learning_rate": 6.047830146255221e-06,
+      "loss": 0.0053,
+      "step": 1520
+    },
+    {
+      "epoch": 1.3493682110396807,
+      "grad_norm": 0.11063788831233978,
+      "learning_rate": 6.04314842309245e-06,
+      "loss": 0.0013,
+      "step": 1521
+    },
+    {
+      "epoch": 1.3502549323874973,
+      "grad_norm": 0.09609077125787735,
+      "learning_rate": 6.038465743543353e-06,
+      "loss": 0.0007,
+      "step": 1522
+    },
+    {
+      "epoch": 1.3511416537353136,
+      "grad_norm": 0.12723086774349213,
+      "learning_rate": 6.033782111901138e-06,
+      "loss": 0.0008,
+      "step": 1523
+    },
+    {
+      "epoch": 1.3520283750831301,
+      "grad_norm": 1.1818429231643677,
+      "learning_rate": 6.029097532459881e-06,
+      "loss": 0.0073,
+      "step": 1524
+    },
+    {
+      "epoch": 1.3529150964309467,
+      "grad_norm": 0.09564555436372757,
+      "learning_rate": 6.024412009514533e-06,
+      "loss": 0.001,
+      "step": 1525
+    },
+    {
+      "epoch": 1.353801817778763,
+      "grad_norm": 0.13707663118839264,
+      "learning_rate": 6.019725547360905e-06,
+      "loss": 0.0011,
+      "step": 1526
+    },
+    {
+      "epoch": 1.3546885391265795,
+      "grad_norm": 0.5464258790016174,
+      "learning_rate": 6.01503815029567e-06,
+      "loss": 0.0058,
+      "step": 1527
+    },
+    {
+      "epoch": 1.3555752604743958,
+      "grad_norm": 0.29684606194496155,
+      "learning_rate": 6.01034982261636e-06,
+      "loss": 0.0018,
+      "step": 1528
+    },
+    {
+      "epoch": 1.3564619818222123,
+      "grad_norm": 0.03409336507320404,
+      "learning_rate": 6.005660568621359e-06,
+      "loss": 0.0004,
+      "step": 1529
+    },
+    {
+      "epoch": 1.3573487031700289,
+      "grad_norm": 0.14027540385723114,
+      "learning_rate": 6.000970392609899e-06,
+      "loss": 0.0012,
+      "step": 1530
+    },
+    {
+      "epoch": 1.3582354245178454,
+      "grad_norm": 0.4142434597015381,
+      "learning_rate": 5.996279298882058e-06,
+      "loss": 0.0038,
+      "step": 1531
+    },
+    {
+      "epoch": 1.3591221458656617,
+      "grad_norm": 0.06057576835155487,
+      "learning_rate": 5.991587291738759e-06,
+      "loss": 0.0007,
+      "step": 1532
+    },
+    {
+      "epoch": 1.3600088672134782,
+      "grad_norm": 0.3078071177005768,
+      "learning_rate": 5.986894375481757e-06,
+      "loss": 0.0019,
+      "step": 1533
+    },
+    {
+      "epoch": 1.3608955885612946,
+      "grad_norm": 0.31704726815223694,
+      "learning_rate": 5.982200554413643e-06,
+      "loss": 0.0019,
+      "step": 1534
+    },
+    {
+      "epoch": 1.361782309909111,
+      "grad_norm": 0.08941329270601273,
+      "learning_rate": 5.977505832837839e-06,
+      "loss": 0.0005,
+      "step": 1535
+    },
+    {
+      "epoch": 1.3626690312569276,
+      "grad_norm": 0.5131560564041138,
+      "learning_rate": 5.9728102150585885e-06,
+      "loss": 0.0019,
+      "step": 1536
+    },
+    {
+      "epoch": 1.363555752604744,
+      "grad_norm": 0.11163872480392456,
+      "learning_rate": 5.968113705380961e-06,
+      "loss": 0.0007,
+      "step": 1537
+    },
+    {
+      "epoch": 1.3644424739525605,
+      "grad_norm": 0.2688218355178833,
+      "learning_rate": 5.963416308110843e-06,
+      "loss": 0.0017,
+      "step": 1538
+    },
+    {
+      "epoch": 1.3653291953003768,
+      "grad_norm": 0.1860375553369522,
+      "learning_rate": 5.9587180275549324e-06,
+      "loss": 0.0027,
+      "step": 1539
+    },
+    {
+      "epoch": 1.3662159166481933,
+      "grad_norm": 0.57248455286026,
+      "learning_rate": 5.954018868020737e-06,
+      "loss": 0.006,
+      "step": 1540
+    },
+    {
+      "epoch": 1.3671026379960098,
+      "grad_norm": 0.47507718205451965,
+      "learning_rate": 5.9493188338165745e-06,
+      "loss": 0.0038,
+      "step": 1541
+    },
+    {
+      "epoch": 1.3679893593438262,
+      "grad_norm": 0.04093353450298309,
+      "learning_rate": 5.94461792925156e-06,
+      "loss": 0.0003,
+      "step": 1542
+    },
+    {
+      "epoch": 1.3688760806916427,
+      "grad_norm": 0.2782546281814575,
+      "learning_rate": 5.939916158635608e-06,
+      "loss": 0.0015,
+      "step": 1543
+    },
+    {
+      "epoch": 1.369762802039459,
+      "grad_norm": 0.444394588470459,
+      "learning_rate": 5.935213526279428e-06,
+      "loss": 0.0028,
+      "step": 1544
+    },
+    {
+      "epoch": 1.3706495233872755,
+      "grad_norm": 0.6129445433616638,
+      "learning_rate": 5.93051003649452e-06,
+      "loss": 0.0055,
+      "step": 1545
+    },
+    {
+      "epoch": 1.371536244735092,
+      "grad_norm": 0.3048208951950073,
+      "learning_rate": 5.925805693593167e-06,
+      "loss": 0.0016,
+      "step": 1546
+    },
+    {
+      "epoch": 1.3724229660829084,
+      "grad_norm": 0.013988547958433628,
+      "learning_rate": 5.921100501888437e-06,
+      "loss": 0.0001,
+      "step": 1547
+    },
+    {
+      "epoch": 1.373309687430725,
+      "grad_norm": 0.05953359231352806,
+      "learning_rate": 5.916394465694175e-06,
+      "loss": 0.0004,
+      "step": 1548
+    },
+    {
+      "epoch": 1.3741964087785412,
+      "grad_norm": 0.12987588346004486,
+      "learning_rate": 5.911687589325003e-06,
+      "loss": 0.0006,
+      "step": 1549
+    },
+    {
+      "epoch": 1.3750831301263577,
+      "grad_norm": 0.18447338044643402,
+      "learning_rate": 5.906979877096306e-06,
+      "loss": 0.0009,
+      "step": 1550
+    },
+    {
+      "epoch": 1.3759698514741743,
+      "grad_norm": 0.05198747292160988,
+      "learning_rate": 5.902271333324245e-06,
+      "loss": 0.0001,
+      "step": 1551
+    },
+    {
+      "epoch": 1.3768565728219908,
+      "grad_norm": 0.3160160183906555,
+      "learning_rate": 5.897561962325737e-06,
+      "loss": 0.0011,
+      "step": 1552
+    },
+    {
+      "epoch": 1.3777432941698071,
+      "grad_norm": 0.2100251019001007,
+      "learning_rate": 5.892851768418458e-06,
+      "loss": 0.0007,
+      "step": 1553
+    },
+    {
+      "epoch": 1.3786300155176237,
+      "grad_norm": 0.06642001122236252,
+      "learning_rate": 5.88814075592084e-06,
+      "loss": 0.0005,
+      "step": 1554
+    },
+    {
+      "epoch": 1.37951673686544,
+      "grad_norm": 0.08171337842941284,
+      "learning_rate": 5.883428929152067e-06,
+      "loss": 0.0003,
+      "step": 1555
+    },
+    {
+      "epoch": 1.3804034582132565,
+      "grad_norm": 0.4593549072742462,
+      "learning_rate": 5.878716292432065e-06,
+      "loss": 0.0023,
+      "step": 1556
+    },
+    {
+      "epoch": 1.381290179561073,
+      "grad_norm": 0.2839002013206482,
+      "learning_rate": 5.874002850081506e-06,
+      "loss": 0.0015,
+      "step": 1557
+    },
+    {
+      "epoch": 1.3821769009088893,
+      "grad_norm": 0.49203285574913025,
+      "learning_rate": 5.8692886064218015e-06,
+      "loss": 0.0011,
+      "step": 1558
+    },
+    {
+      "epoch": 1.3830636222567059,
+      "grad_norm": 0.018628446385264397,
+      "learning_rate": 5.864573565775092e-06,
+      "loss": 0.0001,
+      "step": 1559
+    },
+    {
+      "epoch": 1.3839503436045222,
+      "grad_norm": 0.716225266456604,
+      "learning_rate": 5.859857732464258e-06,
+      "loss": 0.0057,
+      "step": 1560
+    },
+    {
+      "epoch": 1.3848370649523387,
+      "grad_norm": 0.082828588783741,
+      "learning_rate": 5.855141110812896e-06,
+      "loss": 0.0002,
+      "step": 1561
+    },
+    {
+      "epoch": 1.3857237863001552,
+      "grad_norm": 0.5517600178718567,
+      "learning_rate": 5.850423705145334e-06,
+      "loss": 0.0026,
+      "step": 1562
+    },
+    {
+      "epoch": 1.3866105076479716,
+      "grad_norm": 0.3616451025009155,
+      "learning_rate": 5.845705519786613e-06,
+      "loss": 0.0007,
+      "step": 1563
+    },
+    {
+      "epoch": 1.387497228995788,
+      "grad_norm": 0.39561301469802856,
+      "learning_rate": 5.8409865590624955e-06,
+      "loss": 0.004,
+      "step": 1564
+    },
+    {
+      "epoch": 1.3883839503436044,
+      "grad_norm": 0.3806546926498413,
+      "learning_rate": 5.836266827299446e-06,
+      "loss": 0.0043,
+      "step": 1565
+    },
+    {
+      "epoch": 1.389270671691421,
+      "grad_norm": 0.005371665582060814,
+      "learning_rate": 5.831546328824643e-06,
+      "loss": 0.0,
+      "step": 1566
+    },
+    {
+      "epoch": 1.3901573930392375,
+      "grad_norm": 1.6570253372192383,
+      "learning_rate": 5.826825067965967e-06,
+      "loss": 0.0176,
+      "step": 1567
+    },
+    {
+      "epoch": 1.391044114387054,
+      "grad_norm": 0.19763243198394775,
+      "learning_rate": 5.822103049051993e-06,
+      "loss": 0.0009,
+      "step": 1568
+    },
+    {
+      "epoch": 1.3919308357348703,
+      "grad_norm": 1.0521655082702637,
+      "learning_rate": 5.817380276411995e-06,
+      "loss": 0.0014,
+      "step": 1569
+    },
+    {
+      "epoch": 1.3928175570826868,
+      "grad_norm": 0.4388374984264374,
+      "learning_rate": 5.812656754375937e-06,
+      "loss": 0.0145,
+      "step": 1570
+    },
+    {
+      "epoch": 1.3937042784305032,
+      "grad_norm": 4.534085750579834,
+      "learning_rate": 5.8079324872744706e-06,
+      "loss": 0.0201,
+      "step": 1571
+    },
+    {
+      "epoch": 1.3945909997783197,
+      "grad_norm": 0.01597362942993641,
+      "learning_rate": 5.803207479438929e-06,
+      "loss": 0.0001,
+      "step": 1572
+    },
+    {
+      "epoch": 1.3954777211261362,
+      "grad_norm": 0.13055969774723053,
+      "learning_rate": 5.798481735201327e-06,
+      "loss": 0.0009,
+      "step": 1573
+    },
+    {
+      "epoch": 1.3963644424739525,
+      "grad_norm": 0.13696056604385376,
+      "learning_rate": 5.793755258894354e-06,
+      "loss": 0.0003,
+      "step": 1574
+    },
+    {
+      "epoch": 1.397251163821769,
+      "grad_norm": 0.11396967619657516,
+      "learning_rate": 5.789028054851366e-06,
+      "loss": 0.0005,
+      "step": 1575
+    },
+    {
+      "epoch": 1.3981378851695854,
+      "grad_norm": 0.059259578585624695,
+      "learning_rate": 5.784300127406393e-06,
+      "loss": 0.0003,
+      "step": 1576
+    },
+    {
+      "epoch": 1.399024606517402,
+      "grad_norm": 0.2292395532131195,
+      "learning_rate": 5.779571480894122e-06,
+      "loss": 0.0013,
+      "step": 1577
+    },
+    {
+      "epoch": 1.3999113278652184,
+      "grad_norm": 0.06261775642633438,
+      "learning_rate": 5.774842119649905e-06,
+      "loss": 0.0004,
+      "step": 1578
+    },
+    {
+      "epoch": 1.4007980492130347,
+      "grad_norm": 0.27105244994163513,
+      "learning_rate": 5.770112048009747e-06,
+      "loss": 0.0025,
+      "step": 1579
+    },
+    {
+      "epoch": 1.4016847705608513,
+      "grad_norm": 0.12989850342273712,
+      "learning_rate": 5.765381270310302e-06,
+      "loss": 0.0005,
+      "step": 1580
+    },
+    {
+      "epoch": 1.4025714919086676,
+      "grad_norm": 0.43756383657455444,
+      "learning_rate": 5.7606497908888755e-06,
+      "loss": 0.004,
+      "step": 1581
+    },
+    {
+      "epoch": 1.4034582132564841,
+      "grad_norm": 0.27995091676712036,
+      "learning_rate": 5.755917614083412e-06,
+      "loss": 0.0026,
+      "step": 1582
+    },
+    {
+      "epoch": 1.4043449346043007,
+      "grad_norm": 0.15632747113704681,
+      "learning_rate": 5.751184744232499e-06,
+      "loss": 0.0009,
+      "step": 1583
+    },
+    {
+      "epoch": 1.405231655952117,
+      "grad_norm": 1.14409339427948,
+      "learning_rate": 5.746451185675358e-06,
+      "loss": 0.004,
+      "step": 1584
+    },
+    {
+      "epoch": 1.4061183772999335,
+      "grad_norm": 0.19565550982952118,
+      "learning_rate": 5.7417169427518394e-06,
+      "loss": 0.0016,
+      "step": 1585
+    },
+    {
+      "epoch": 1.4070050986477498,
+      "grad_norm": 0.43583473563194275,
+      "learning_rate": 5.736982019802427e-06,
+      "loss": 0.0039,
+      "step": 1586
+    },
+    {
+      "epoch": 1.4078918199955663,
+      "grad_norm": 0.3083972632884979,
+      "learning_rate": 5.732246421168223e-06,
+      "loss": 0.0021,
+      "step": 1587
+    },
+    {
+      "epoch": 1.4087785413433829,
+      "grad_norm": 0.4371331036090851,
+      "learning_rate": 5.727510151190949e-06,
+      "loss": 0.0037,
+      "step": 1588
+    },
+    {
+      "epoch": 1.4096652626911994,
+      "grad_norm": 0.24955230951309204,
+      "learning_rate": 5.722773214212946e-06,
+      "loss": 0.0046,
+      "step": 1589
+    },
+    {
+      "epoch": 1.4105519840390157,
+      "grad_norm": 0.18335281312465668,
+      "learning_rate": 5.718035614577164e-06,
+      "loss": 0.001,
+      "step": 1590
+    },
+    {
+      "epoch": 1.4114387053868322,
+      "grad_norm": 0.4750370979309082,
+      "learning_rate": 5.713297356627161e-06,
+      "loss": 0.0046,
+      "step": 1591
+    },
+    {
+      "epoch": 1.4123254267346486,
+      "grad_norm": 0.1709708869457245,
+      "learning_rate": 5.708558444707097e-06,
+      "loss": 0.0012,
+      "step": 1592
+    },
+    {
+      "epoch": 1.413212148082465,
+      "grad_norm": 0.6528603434562683,
+      "learning_rate": 5.7038188831617345e-06,
+      "loss": 0.0042,
+      "step": 1593
+    },
+    {
+      "epoch": 1.4140988694302816,
+      "grad_norm": 0.25174710154533386,
+      "learning_rate": 5.699078676336429e-06,
+      "loss": 0.0017,
+      "step": 1594
+    },
+    {
+      "epoch": 1.414985590778098,
+      "grad_norm": 0.07705551385879517,
+      "learning_rate": 5.69433782857713e-06,
+      "loss": 0.0004,
+      "step": 1595
+    },
+    {
+      "epoch": 1.4158723121259145,
+      "grad_norm": 0.10054202377796173,
+      "learning_rate": 5.68959634423037e-06,
+      "loss": 0.0005,
+      "step": 1596
+    },
+    {
+      "epoch": 1.4167590334737308,
+      "grad_norm": 0.09865143895149231,
+      "learning_rate": 5.684854227643274e-06,
+      "loss": 0.0006,
+      "step": 1597
+    },
+    {
+      "epoch": 1.4176457548215473,
+      "grad_norm": 0.2448788434267044,
+      "learning_rate": 5.680111483163535e-06,
+      "loss": 0.0023,
+      "step": 1598
+    },
+    {
+      "epoch": 1.4185324761693638,
+      "grad_norm": 0.14839304983615875,
+      "learning_rate": 5.675368115139432e-06,
+      "loss": 0.001,
+      "step": 1599
+    },
+    {
+      "epoch": 1.4194191975171802,
+      "grad_norm": 0.3813326060771942,
+      "learning_rate": 5.67062412791981e-06,
+      "loss": 0.0071,
+      "step": 1600
+    },
+    {
+      "epoch": 1.4203059188649967,
+      "grad_norm": 0.3784043490886688,
+      "learning_rate": 5.665879525854083e-06,
+      "loss": 0.0021,
+      "step": 1601
+    },
+    {
+      "epoch": 1.421192640212813,
+      "grad_norm": 0.7230820655822754,
+      "learning_rate": 5.661134313292232e-06,
+      "loss": 0.0088,
+      "step": 1602
+    },
+    {
+      "epoch": 1.4220793615606295,
+      "grad_norm": 0.3506825566291809,
+      "learning_rate": 5.656388494584788e-06,
+      "loss": 0.0071,
+      "step": 1603
+    },
+    {
+      "epoch": 1.422966082908446,
+      "grad_norm": 0.1567530781030655,
+      "learning_rate": 5.651642074082849e-06,
+      "loss": 0.0038,
+      "step": 1604
+    },
+    {
+      "epoch": 1.4238528042562626,
+      "grad_norm": 0.20204611122608185,
+      "learning_rate": 5.646895056138059e-06,
+      "loss": 0.002,
+      "step": 1605
+    },
+    {
+      "epoch": 1.424739525604079,
+      "grad_norm": 0.5711863040924072,
+      "learning_rate": 5.64214744510261e-06,
+      "loss": 0.0041,
+      "step": 1606
+    },
+    {
+      "epoch": 1.4256262469518954,
+      "grad_norm": 0.1815640926361084,
+      "learning_rate": 5.63739924532924e-06,
+      "loss": 0.0015,
+      "step": 1607
+    },
+    {
+      "epoch": 1.4265129682997117,
+      "grad_norm": 0.037703752517700195,
+      "learning_rate": 5.632650461171225e-06,
+      "loss": 0.0005,
+      "step": 1608
+    },
+    {
+      "epoch": 1.4273996896475283,
+      "grad_norm": 0.10812222957611084,
+      "learning_rate": 5.6279010969823775e-06,
+      "loss": 0.0009,
+      "step": 1609
+    },
+    {
+      "epoch": 1.4282864109953448,
+      "grad_norm": 0.13056260347366333,
+      "learning_rate": 5.623151157117038e-06,
+      "loss": 0.0007,
+      "step": 1610
+    },
+    {
+      "epoch": 1.4291731323431611,
+      "grad_norm": 0.17778554558753967,
+      "learning_rate": 5.6184006459300845e-06,
+      "loss": 0.0006,
+      "step": 1611
+    },
+    {
+      "epoch": 1.4300598536909777,
+      "grad_norm": 0.024510536342859268,
+      "learning_rate": 5.613649567776905e-06,
+      "loss": 0.0002,
+      "step": 1612
+    },
+    {
+      "epoch": 1.430946575038794,
+      "grad_norm": 0.1636805534362793,
+      "learning_rate": 5.60889792701342e-06,
+      "loss": 0.0008,
+      "step": 1613
+    },
+    {
+      "epoch": 1.4318332963866105,
+      "grad_norm": 0.1205761656165123,
+      "learning_rate": 5.60414572799606e-06,
+      "loss": 0.0005,
+      "step": 1614
+    },
+    {
+      "epoch": 1.432720017734427,
+      "grad_norm": 0.04635613411664963,
+      "learning_rate": 5.599392975081766e-06,
+      "loss": 0.0002,
+      "step": 1615
+    },
+    {
+      "epoch": 1.4336067390822433,
+      "grad_norm": 0.3327246308326721,
+      "learning_rate": 5.594639672627991e-06,
+      "loss": 0.0019,
+      "step": 1616
+    },
+    {
+      "epoch": 1.4344934604300599,
+      "grad_norm": 0.3695579767227173,
+      "learning_rate": 5.589885824992687e-06,
+      "loss": 0.0035,
+      "step": 1617
+    },
+    {
+      "epoch": 1.4353801817778762,
+      "grad_norm": 0.18341560661792755,
+      "learning_rate": 5.585131436534312e-06,
+      "loss": 0.0013,
+      "step": 1618
+    },
+    {
+      "epoch": 1.4362669031256927,
+      "grad_norm": 0.10450714826583862,
+      "learning_rate": 5.580376511611813e-06,
+      "loss": 0.0008,
+      "step": 1619
+    },
+    {
+      "epoch": 1.4371536244735093,
+      "grad_norm": 0.29582592844963074,
+      "learning_rate": 5.575621054584633e-06,
+      "loss": 0.0049,
+      "step": 1620
+    },
+    {
+      "epoch": 1.4380403458213258,
+      "grad_norm": 0.05713328719139099,
+      "learning_rate": 5.570865069812703e-06,
+      "loss": 0.0003,
+      "step": 1621
+    },
+    {
+      "epoch": 1.438927067169142,
+      "grad_norm": 0.027085937559604645,
+      "learning_rate": 5.5661085616564344e-06,
+      "loss": 0.0002,
+      "step": 1622
+    },
+    {
+      "epoch": 1.4398137885169586,
+      "grad_norm": 0.33440035581588745,
+      "learning_rate": 5.561351534476726e-06,
+      "loss": 0.002,
+      "step": 1623
+    },
+    {
+      "epoch": 1.440700509864775,
+      "grad_norm": 0.37347376346588135,
+      "learning_rate": 5.55659399263494e-06,
+      "loss": 0.0012,
+      "step": 1624
+    },
+    {
+      "epoch": 1.4415872312125915,
+      "grad_norm": 0.05196278169751167,
+      "learning_rate": 5.551835940492924e-06,
+      "loss": 0.0004,
+      "step": 1625
+    },
+    {
+      "epoch": 1.442473952560408,
+      "grad_norm": 0.354519248008728,
+      "learning_rate": 5.547077382412985e-06,
+      "loss": 0.0037,
+      "step": 1626
+    },
+    {
+      "epoch": 1.4433606739082243,
+      "grad_norm": 0.321370393037796,
+      "learning_rate": 5.542318322757895e-06,
+      "loss": 0.0016,
+      "step": 1627
+    },
+    {
+      "epoch": 1.4442473952560408,
+      "grad_norm": 0.035019587725400925,
+      "learning_rate": 5.537558765890888e-06,
+      "loss": 0.0002,
+      "step": 1628
+    },
+    {
+      "epoch": 1.4451341166038572,
+      "grad_norm": 0.7996811270713806,
+      "learning_rate": 5.532798716175652e-06,
+      "loss": 0.002,
+      "step": 1629
+    },
+    {
+      "epoch": 1.4460208379516737,
+      "grad_norm": 0.07523198425769806,
+      "learning_rate": 5.52803817797633e-06,
+      "loss": 0.0004,
+      "step": 1630
+    },
+    {
+      "epoch": 1.4469075592994902,
+      "grad_norm": 0.08954586833715439,
+      "learning_rate": 5.523277155657507e-06,
+      "loss": 0.0005,
+      "step": 1631
+    },
+    {
+      "epoch": 1.4477942806473065,
+      "grad_norm": 0.010750265792012215,
+      "learning_rate": 5.51851565358422e-06,
+      "loss": 0.0001,
+      "step": 1632
+    },
+    {
+      "epoch": 1.448681001995123,
+      "grad_norm": 0.138992577791214,
+      "learning_rate": 5.513753676121936e-06,
+      "loss": 0.0007,
+      "step": 1633
+    },
+    {
+      "epoch": 1.4495677233429394,
+      "grad_norm": 0.3091992437839508,
+      "learning_rate": 5.508991227636568e-06,
+      "loss": 0.0036,
+      "step": 1634
+    },
+    {
+      "epoch": 1.450454444690756,
+      "grad_norm": 0.12356047332286835,
+      "learning_rate": 5.50422831249445e-06,
+      "loss": 0.0002,
+      "step": 1635
+    },
+    {
+      "epoch": 1.4513411660385724,
+      "grad_norm": 0.24620208144187927,
+      "learning_rate": 5.499464935062355e-06,
+      "loss": 0.0011,
+      "step": 1636
+    },
+    {
+      "epoch": 1.4522278873863887,
+      "grad_norm": 0.07541129738092422,
+      "learning_rate": 5.494701099707475e-06,
+      "loss": 0.0002,
+      "step": 1637
+    },
+    {
+      "epoch": 1.4531146087342053,
+      "grad_norm": 0.13778525590896606,
+      "learning_rate": 5.489936810797416e-06,
+      "loss": 0.0004,
+      "step": 1638
+    },
+    {
+      "epoch": 1.4540013300820216,
+      "grad_norm": 0.014556895941495895,
+      "learning_rate": 5.48517207270021e-06,
+      "loss": 0.0001,
+      "step": 1639
+    },
+    {
+      "epoch": 1.4548880514298381,
+      "grad_norm": 0.43046942353248596,
+      "learning_rate": 5.480406889784293e-06,
+      "loss": 0.0014,
+      "step": 1640
+    },
+    {
+      "epoch": 1.4557747727776547,
+      "grad_norm": 0.30873122811317444,
+      "learning_rate": 5.4756412664185166e-06,
+      "loss": 0.0057,
+      "step": 1641
+    },
+    {
+      "epoch": 1.4566614941254712,
+      "grad_norm": 0.07226885110139847,
+      "learning_rate": 5.470875206972126e-06,
+      "loss": 0.0002,
+      "step": 1642
+    },
+    {
+      "epoch": 1.4575482154732875,
+      "grad_norm": 0.4386361539363861,
+      "learning_rate": 5.466108715814775e-06,
+      "loss": 0.0013,
+      "step": 1643
+    },
+    {
+      "epoch": 1.458434936821104,
+      "grad_norm": 0.9503544569015503,
+      "learning_rate": 5.46134179731651e-06,
+      "loss": 0.0019,
+      "step": 1644
+    },
+    {
+      "epoch": 1.4593216581689203,
+      "grad_norm": 0.9516314268112183,
+      "learning_rate": 5.456574455847767e-06,
+      "loss": 0.0039,
+      "step": 1645
+    },
+    {
+      "epoch": 1.4602083795167369,
+      "grad_norm": 0.33284255862236023,
+      "learning_rate": 5.451806695779376e-06,
+      "loss": 0.0016,
+      "step": 1646
+    },
+    {
+      "epoch": 1.4610951008645534,
+      "grad_norm": 0.8509582281112671,
+      "learning_rate": 5.447038521482542e-06,
+      "loss": 0.0038,
+      "step": 1647
+    },
+    {
+      "epoch": 1.4619818222123697,
+      "grad_norm": 0.07178089022636414,
+      "learning_rate": 5.442269937328856e-06,
+      "loss": 0.0001,
+      "step": 1648
+    },
+    {
+      "epoch": 1.4628685435601863,
+      "grad_norm": 0.3110867440700531,
+      "learning_rate": 5.4375009476902855e-06,
+      "loss": 0.0022,
+      "step": 1649
+    },
+    {
+      "epoch": 1.4637552649080026,
+      "grad_norm": 0.058392178267240524,
+      "learning_rate": 5.432731556939163e-06,
+      "loss": 0.0003,
+      "step": 1650
+    },
+    {
+      "epoch": 1.464641986255819,
+      "grad_norm": 0.5997623205184937,
+      "learning_rate": 5.427961769448199e-06,
+      "loss": 0.0034,
+      "step": 1651
+    },
+    {
+      "epoch": 1.4655287076036356,
+      "grad_norm": 0.7097246646881104,
+      "learning_rate": 5.423191589590457e-06,
+      "loss": 0.0043,
+      "step": 1652
+    },
+    {
+      "epoch": 1.466415428951452,
+      "grad_norm": 0.7380861639976501,
+      "learning_rate": 5.418421021739369e-06,
+      "loss": 0.008,
+      "step": 1653
+    },
+    {
+      "epoch": 1.4673021502992685,
+      "grad_norm": 0.6223729252815247,
+      "learning_rate": 5.413650070268716e-06,
+      "loss": 0.0023,
+      "step": 1654
+    },
+    {
+      "epoch": 1.4681888716470848,
+      "grad_norm": 0.269790381193161,
+      "learning_rate": 5.4088787395526365e-06,
+      "loss": 0.001,
+      "step": 1655
+    },
+    {
+      "epoch": 1.4690755929949013,
+      "grad_norm": 0.8038837313652039,
+      "learning_rate": 5.40410703396561e-06,
+      "loss": 0.0044,
+      "step": 1656
+    },
+    {
+      "epoch": 1.4699623143427178,
+      "grad_norm": 0.15287820994853973,
+      "learning_rate": 5.3993349578824675e-06,
+      "loss": 0.0008,
+      "step": 1657
+    },
+    {
+      "epoch": 1.4708490356905344,
+      "grad_norm": 0.4003780484199524,
+      "learning_rate": 5.394562515678374e-06,
+      "loss": 0.0023,
+      "step": 1658
+    },
+    {
+      "epoch": 1.4717357570383507,
+      "grad_norm": 0.04636048153042793,
+      "learning_rate": 5.389789711728832e-06,
+      "loss": 0.0002,
+      "step": 1659
+    },
+    {
+      "epoch": 1.4726224783861672,
+      "grad_norm": 0.23046471178531647,
+      "learning_rate": 5.385016550409676e-06,
+      "loss": 0.0011,
+      "step": 1660
+    },
+    {
+      "epoch": 1.4735091997339835,
+      "grad_norm": 0.5971014499664307,
+      "learning_rate": 5.380243036097067e-06,
+      "loss": 0.002,
+      "step": 1661
+    },
+    {
+      "epoch": 1.4743959210818,
+      "grad_norm": 0.6471933126449585,
+      "learning_rate": 5.375469173167491e-06,
+      "loss": 0.005,
+      "step": 1662
+    },
+    {
+      "epoch": 1.4752826424296166,
+      "grad_norm": 0.03686193749308586,
+      "learning_rate": 5.370694965997753e-06,
+      "loss": 0.0002,
+      "step": 1663
+    },
+    {
+      "epoch": 1.476169363777433,
+      "grad_norm": 0.24532069265842438,
+      "learning_rate": 5.365920418964973e-06,
+      "loss": 0.0019,
+      "step": 1664
+    },
+    {
+      "epoch": 1.4770560851252494,
+      "grad_norm": 0.8025774359703064,
+      "learning_rate": 5.361145536446582e-06,
+      "loss": 0.0054,
+      "step": 1665
+    },
+    {
+      "epoch": 1.4779428064730658,
+      "grad_norm": 0.2576480209827423,
+      "learning_rate": 5.356370322820325e-06,
+      "loss": 0.0014,
+      "step": 1666
+    },
+    {
+      "epoch": 1.4788295278208823,
+      "grad_norm": 1.0776647329330444,
+      "learning_rate": 5.351594782464239e-06,
+      "loss": 0.0042,
+      "step": 1667
+    },
+    {
+      "epoch": 1.4797162491686988,
+      "grad_norm": 0.6167517304420471,
+      "learning_rate": 5.34681891975667e-06,
+      "loss": 0.0015,
+      "step": 1668
+    },
+    {
+      "epoch": 1.4806029705165151,
+      "grad_norm": 0.6090970635414124,
+      "learning_rate": 5.342042739076257e-06,
+      "loss": 0.0022,
+      "step": 1669
+    },
+    {
+      "epoch": 1.4814896918643317,
+      "grad_norm": 0.044063255190849304,
+      "learning_rate": 5.337266244801927e-06,
+      "loss": 0.0002,
+      "step": 1670
+    },
+    {
+      "epoch": 1.482376413212148,
+      "grad_norm": 0.03505855053663254,
+      "learning_rate": 5.332489441312901e-06,
+      "loss": 0.0002,
+      "step": 1671
+    },
+    {
+      "epoch": 1.4832631345599645,
+      "grad_norm": 0.5114758014678955,
+      "learning_rate": 5.327712332988678e-06,
+      "loss": 0.0024,
+      "step": 1672
+    },
+    {
+      "epoch": 1.484149855907781,
+      "grad_norm": 0.08527572453022003,
+      "learning_rate": 5.322934924209039e-06,
+      "loss": 0.0005,
+      "step": 1673
+    },
+    {
+      "epoch": 1.4850365772555973,
+      "grad_norm": 0.12833604216575623,
+      "learning_rate": 5.318157219354041e-06,
+      "loss": 0.0004,
+      "step": 1674
+    },
+    {
+      "epoch": 1.4859232986034139,
+      "grad_norm": 0.045691631734371185,
+      "learning_rate": 5.313379222804008e-06,
+      "loss": 0.0004,
+      "step": 1675
+    },
+    {
+      "epoch": 1.4868100199512302,
+      "grad_norm": 0.2374916523694992,
+      "learning_rate": 5.308600938939538e-06,
+      "loss": 0.0008,
+      "step": 1676
+    },
+    {
+      "epoch": 1.4876967412990467,
+      "grad_norm": 0.047486886382102966,
+      "learning_rate": 5.303822372141489e-06,
+      "loss": 0.0003,
+      "step": 1677
+    },
+    {
+      "epoch": 1.4885834626468633,
+      "grad_norm": 0.08337344229221344,
+      "learning_rate": 5.2990435267909764e-06,
+      "loss": 0.0005,
+      "step": 1678
+    },
+    {
+      "epoch": 1.4894701839946798,
+      "grad_norm": 0.037117332220077515,
+      "learning_rate": 5.294264407269376e-06,
+      "loss": 0.0002,
+      "step": 1679
+    },
+    {
+      "epoch": 1.490356905342496,
+      "grad_norm": 1.1246777772903442,
+      "learning_rate": 5.2894850179583095e-06,
+      "loss": 0.0053,
+      "step": 1680
+    },
+    {
+      "epoch": 1.4912436266903126,
+      "grad_norm": 0.23861926794052124,
+      "learning_rate": 5.284705363239651e-06,
+      "loss": 0.0007,
+      "step": 1681
+    },
+    {
+      "epoch": 1.492130348038129,
+      "grad_norm": 0.34302806854248047,
+      "learning_rate": 5.279925447495513e-06,
+      "loss": 0.0018,
+      "step": 1682
+    },
+    {
+      "epoch": 1.4930170693859455,
+      "grad_norm": 0.00627083471044898,
+      "learning_rate": 5.275145275108252e-06,
+      "loss": 0.0,
+      "step": 1683
+    },
+    {
+      "epoch": 1.493903790733762,
+      "grad_norm": 0.0319259911775589,
+      "learning_rate": 5.270364850460456e-06,
+      "loss": 0.0001,
+      "step": 1684
+    },
+    {
+      "epoch": 1.4947905120815783,
+      "grad_norm": 0.26929083466529846,
+      "learning_rate": 5.26558417793495e-06,
+      "loss": 0.0014,
+      "step": 1685
+    },
+    {
+      "epoch": 1.4956772334293948,
+      "grad_norm": 0.4781963527202606,
+      "learning_rate": 5.2608032619147765e-06,
+      "loss": 0.0027,
+      "step": 1686
+    },
+    {
+      "epoch": 1.4965639547772112,
+      "grad_norm": 0.6385494470596313,
+      "learning_rate": 5.2560221067832095e-06,
+      "loss": 0.0107,
+      "step": 1687
+    },
+    {
+      "epoch": 1.4974506761250277,
+      "grad_norm": 0.07988189905881882,
+      "learning_rate": 5.251240716923741e-06,
+      "loss": 0.0004,
+      "step": 1688
+    },
+    {
+      "epoch": 1.4983373974728442,
+      "grad_norm": 0.02995728887617588,
+      "learning_rate": 5.246459096720075e-06,
+      "loss": 0.0002,
+      "step": 1689
+    },
+    {
+      "epoch": 1.4992241188206605,
+      "grad_norm": 0.2245960831642151,
+      "learning_rate": 5.2416772505561275e-06,
+      "loss": 0.0008,
+      "step": 1690
+    },
+    {
+      "epoch": 1.500110840168477,
+      "grad_norm": 0.3143225908279419,
+      "learning_rate": 5.236895182816027e-06,
+      "loss": 0.0021,
+      "step": 1691
+    },
+    {
+      "epoch": 1.5009975615162934,
+      "grad_norm": 0.03064090944826603,
+      "learning_rate": 5.232112897884097e-06,
+      "loss": 0.0001,
+      "step": 1692
+    },
+    {
+      "epoch": 1.5009975615162934,
+      "eval_loss": 0.01623193733394146,
+      "eval_runtime": 61.9901,
+      "eval_samples_per_second": 3.097,
+      "eval_steps_per_second": 0.774,
+      "step": 1692
+    },
+    {
+      "epoch": 1.50188428286411,
+      "grad_norm": 0.5175110101699829,
+      "learning_rate": 5.2273304001448656e-06,
+      "loss": 0.0029,
+      "step": 1693
+    },
+    {
+      "epoch": 1.5027710042119264,
+      "grad_norm": 0.60247802734375,
+      "learning_rate": 5.222547693983053e-06,
+      "loss": 0.0043,
+      "step": 1694
+    },
+    {
+      "epoch": 1.503657725559743,
+      "grad_norm": 0.5946943759918213,
+      "learning_rate": 5.217764783783574e-06,
+      "loss": 0.0049,
+      "step": 1695
+    },
+    {
+      "epoch": 1.5045444469075593,
+      "grad_norm": 0.13281765580177307,
+      "learning_rate": 5.212981673931528e-06,
+      "loss": 0.0005,
+      "step": 1696
+    },
+    {
+      "epoch": 1.5054311682553756,
+      "grad_norm": 0.4556334316730499,
+      "learning_rate": 5.208198368812197e-06,
+      "loss": 0.0031,
+      "step": 1697
+    },
+    {
+      "epoch": 1.5063178896031921,
+      "grad_norm": 0.5918658971786499,
+      "learning_rate": 5.203414872811042e-06,
+      "loss": 0.0014,
+      "step": 1698
+    },
+    {
+      "epoch": 1.5072046109510087,
+      "grad_norm": 0.30193421244621277,
+      "learning_rate": 5.198631190313702e-06,
+      "loss": 0.0006,
+      "step": 1699
+    },
+    {
+      "epoch": 1.5080913322988252,
+      "grad_norm": 0.3599817752838135,
+      "learning_rate": 5.193847325705983e-06,
+      "loss": 0.0066,
+      "step": 1700
+    },
+    {
+      "epoch": 1.5089780536466415,
+      "grad_norm": 0.35227012634277344,
+      "learning_rate": 5.189063283373861e-06,
+      "loss": 0.002,
+      "step": 1701
+    },
+    {
+      "epoch": 1.509864774994458,
+      "grad_norm": 0.35682663321495056,
+      "learning_rate": 5.184279067703473e-06,
+      "loss": 0.0015,
+      "step": 1702
+    },
+    {
+      "epoch": 1.5107514963422743,
+      "grad_norm": 0.03233305737376213,
+      "learning_rate": 5.179494683081116e-06,
+      "loss": 0.0002,
+      "step": 1703
+    },
+    {
+      "epoch": 1.5116382176900909,
+      "grad_norm": 0.15405134856700897,
+      "learning_rate": 5.17471013389324e-06,
+      "loss": 0.0011,
+      "step": 1704
+    },
+    {
+      "epoch": 1.5125249390379074,
+      "grad_norm": 0.4507070481777191,
+      "learning_rate": 5.16992542452645e-06,
+      "loss": 0.0033,
+      "step": 1705
+    },
+    {
+      "epoch": 1.513411660385724,
+      "grad_norm": 0.4773896634578705,
+      "learning_rate": 5.165140559367496e-06,
+      "loss": 0.0038,
+      "step": 1706
+    },
+    {
+      "epoch": 1.5142983817335403,
+      "grad_norm": 0.19090554118156433,
+      "learning_rate": 5.160355542803265e-06,
+      "loss": 0.0006,
+      "step": 1707
+    },
+    {
+      "epoch": 1.5151851030813566,
+      "grad_norm": 0.07626492530107498,
+      "learning_rate": 5.155570379220792e-06,
+      "loss": 0.0007,
+      "step": 1708
+    },
+    {
+      "epoch": 1.516071824429173,
+      "grad_norm": 0.20484142005443573,
+      "learning_rate": 5.150785073007243e-06,
+      "loss": 0.0018,
+      "step": 1709
+    },
+    {
+      "epoch": 1.5169585457769896,
+      "grad_norm": 0.1442720890045166,
+      "learning_rate": 5.145999628549913e-06,
+      "loss": 0.0005,
+      "step": 1710
+    },
+    {
+      "epoch": 1.5178452671248062,
+      "grad_norm": 0.24547992646694183,
+      "learning_rate": 5.141214050236225e-06,
+      "loss": 0.0012,
+      "step": 1711
+    },
+    {
+      "epoch": 1.5187319884726225,
+      "grad_norm": 0.10670553892850876,
+      "learning_rate": 5.136428342453726e-06,
+      "loss": 0.0008,
+      "step": 1712
+    },
+    {
+      "epoch": 1.5196187098204388,
+      "grad_norm": 0.4111662209033966,
+      "learning_rate": 5.131642509590081e-06,
+      "loss": 0.0069,
+      "step": 1713
+    },
+    {
+      "epoch": 1.5205054311682553,
+      "grad_norm": 0.13540194928646088,
+      "learning_rate": 5.1268565560330685e-06,
+      "loss": 0.0008,
+      "step": 1714
+    },
+    {
+      "epoch": 1.5213921525160718,
+      "grad_norm": 0.30105480551719666,
+      "learning_rate": 5.1220704861705775e-06,
+      "loss": 0.0019,
+      "step": 1715
+    },
+    {
+      "epoch": 1.5222788738638884,
+      "grad_norm": 0.11186516284942627,
+      "learning_rate": 5.117284304390606e-06,
+      "loss": 0.0006,
+      "step": 1716
+    },
+    {
+      "epoch": 1.5231655952117047,
+      "grad_norm": 0.44437602162361145,
+      "learning_rate": 5.112498015081254e-06,
+      "loss": 0.0028,
+      "step": 1717
+    },
+    {
+      "epoch": 1.5240523165595212,
+      "grad_norm": 0.1276712268590927,
+      "learning_rate": 5.107711622630716e-06,
+      "loss": 0.0008,
+      "step": 1718
+    },
+    {
+      "epoch": 1.5249390379073375,
+      "grad_norm": 0.9235378503799438,
+      "learning_rate": 5.102925131427289e-06,
+      "loss": 0.0036,
+      "step": 1719
+    },
+    {
+      "epoch": 1.525825759255154,
+      "grad_norm": 0.08117254823446274,
+      "learning_rate": 5.098138545859354e-06,
+      "loss": 0.0005,
+      "step": 1720
+    },
+    {
+      "epoch": 1.5267124806029706,
+      "grad_norm": 0.3608892858028412,
+      "learning_rate": 5.093351870315379e-06,
+      "loss": 0.0008,
+      "step": 1721
+    },
+    {
+      "epoch": 1.527599201950787,
+      "grad_norm": 0.3133014440536499,
+      "learning_rate": 5.088565109183917e-06,
+      "loss": 0.0014,
+      "step": 1722
+    },
+    {
+      "epoch": 1.5284859232986034,
+      "grad_norm": 0.5285730361938477,
+      "learning_rate": 5.083778266853598e-06,
+      "loss": 0.0017,
+      "step": 1723
+    },
+    {
+      "epoch": 1.5293726446464198,
+      "grad_norm": 0.4431508779525757,
+      "learning_rate": 5.078991347713127e-06,
+      "loss": 0.0073,
+      "step": 1724
+    },
+    {
+      "epoch": 1.5302593659942363,
+      "grad_norm": 0.27122896909713745,
+      "learning_rate": 5.074204356151281e-06,
+      "loss": 0.0044,
+      "step": 1725
+    },
+    {
+      "epoch": 1.5311460873420528,
+      "grad_norm": 0.14935316145420074,
+      "learning_rate": 5.069417296556899e-06,
+      "loss": 0.0011,
+      "step": 1726
+    },
+    {
+      "epoch": 1.5320328086898694,
+      "grad_norm": 0.14171618223190308,
+      "learning_rate": 5.064630173318887e-06,
+      "loss": 0.0005,
+      "step": 1727
+    },
+    {
+      "epoch": 1.5329195300376857,
+      "grad_norm": 0.13760340213775635,
+      "learning_rate": 5.059842990826206e-06,
+      "loss": 0.0011,
+      "step": 1728
+    },
+    {
+      "epoch": 1.533806251385502,
+      "grad_norm": 0.35050904750823975,
+      "learning_rate": 5.055055753467875e-06,
+      "loss": 0.0028,
+      "step": 1729
+    },
+    {
+      "epoch": 1.5346929727333185,
+      "grad_norm": 0.4986531138420105,
+      "learning_rate": 5.050268465632957e-06,
+      "loss": 0.0044,
+      "step": 1730
+    },
+    {
+      "epoch": 1.535579694081135,
+      "grad_norm": 0.020403970032930374,
+      "learning_rate": 5.045481131710568e-06,
+      "loss": 0.0002,
+      "step": 1731
+    },
+    {
+      "epoch": 1.5364664154289516,
+      "grad_norm": 0.6366745829582214,
+      "learning_rate": 5.040693756089865e-06,
+      "loss": 0.0076,
+      "step": 1732
+    },
+    {
+      "epoch": 1.5373531367767679,
+      "grad_norm": 0.2623487114906311,
+      "learning_rate": 5.035906343160038e-06,
+      "loss": 0.0017,
+      "step": 1733
+    },
+    {
+      "epoch": 1.5382398581245842,
+      "grad_norm": 0.19334276020526886,
+      "learning_rate": 5.031118897310318e-06,
+      "loss": 0.0039,
+      "step": 1734
+    },
+    {
+      "epoch": 1.5391265794724007,
+      "grad_norm": 0.5103495121002197,
+      "learning_rate": 5.026331422929962e-06,
+      "loss": 0.0043,
+      "step": 1735
+    },
+    {
+      "epoch": 1.5400133008202173,
+      "grad_norm": 0.03877022862434387,
+      "learning_rate": 5.021543924408254e-06,
+      "loss": 0.0003,
+      "step": 1736
+    },
+    {
+      "epoch": 1.5409000221680338,
+      "grad_norm": 0.2172171175479889,
+      "learning_rate": 5.016756406134501e-06,
+      "loss": 0.0049,
+      "step": 1737
+    },
+    {
+      "epoch": 1.54178674351585,
+      "grad_norm": 0.1811998188495636,
+      "learning_rate": 5.011968872498028e-06,
+      "loss": 0.0013,
+      "step": 1738
+    },
+    {
+      "epoch": 1.5426734648636666,
+      "grad_norm": 0.45922988653182983,
+      "learning_rate": 5.007181327888173e-06,
+      "loss": 0.0036,
+      "step": 1739
+    },
+    {
+      "epoch": 1.543560186211483,
+      "grad_norm": 0.04768901318311691,
+      "learning_rate": 5.002393776694284e-06,
+      "loss": 0.0003,
+      "step": 1740
+    },
+    {
+      "epoch": 1.5444469075592995,
+      "grad_norm": 0.011605698615312576,
+      "learning_rate": 4.997606223305717e-06,
+      "loss": 0.0001,
+      "step": 1741
+    },
+    {
+      "epoch": 1.545333628907116,
+      "grad_norm": 0.08056286722421646,
+      "learning_rate": 4.9928186721118285e-06,
+      "loss": 0.0006,
+      "step": 1742
+    },
+    {
+      "epoch": 1.5462203502549325,
+      "grad_norm": 0.05473679304122925,
+      "learning_rate": 4.988031127501974e-06,
+      "loss": 0.0003,
+      "step": 1743
+    },
+    {
+      "epoch": 1.5471070716027489,
+      "grad_norm": 0.11067911982536316,
+      "learning_rate": 4.983243593865501e-06,
+      "loss": 0.0007,
+      "step": 1744
+    },
+    {
+      "epoch": 1.5479937929505652,
+      "grad_norm": 0.19086086750030518,
+      "learning_rate": 4.978456075591747e-06,
+      "loss": 0.0027,
+      "step": 1745
+    },
+    {
+      "epoch": 1.5488805142983817,
+      "grad_norm": 0.3276596963405609,
+      "learning_rate": 4.97366857707004e-06,
+      "loss": 0.0067,
+      "step": 1746
+    },
+    {
+      "epoch": 1.5497672356461982,
+      "grad_norm": 0.16259552538394928,
+      "learning_rate": 4.968881102689684e-06,
+      "loss": 0.0009,
+      "step": 1747
+    },
+    {
+      "epoch": 1.5506539569940148,
+      "grad_norm": 0.14583039283752441,
+      "learning_rate": 4.964093656839963e-06,
+      "loss": 0.001,
+      "step": 1748
+    },
+    {
+      "epoch": 1.551540678341831,
+      "grad_norm": 0.0421629436314106,
+      "learning_rate": 4.959306243910137e-06,
+      "loss": 0.0004,
+      "step": 1749
+    },
+    {
+      "epoch": 1.5524273996896474,
+      "grad_norm": 0.29653817415237427,
+      "learning_rate": 4.954518868289434e-06,
+      "loss": 0.0028,
+      "step": 1750
+    },
+    {
+      "epoch": 1.553314121037464,
+      "grad_norm": 0.13450933992862701,
+      "learning_rate": 4.949731534367045e-06,
+      "loss": 0.0008,
+      "step": 1751
+    },
+    {
+      "epoch": 1.5542008423852804,
+      "grad_norm": 0.20457084476947784,
+      "learning_rate": 4.944944246532127e-06,
+      "loss": 0.0028,
+      "step": 1752
+    },
+    {
+      "epoch": 1.555087563733097,
+      "grad_norm": 0.598363995552063,
+      "learning_rate": 4.940157009173795e-06,
+      "loss": 0.0066,
+      "step": 1753
+    },
+    {
+      "epoch": 1.5559742850809133,
+      "grad_norm": 0.04115546494722366,
+      "learning_rate": 4.935369826681115e-06,
+      "loss": 0.0004,
+      "step": 1754
+    },
+    {
+      "epoch": 1.5568610064287298,
+      "grad_norm": 0.027022339403629303,
+      "learning_rate": 4.930582703443101e-06,
+      "loss": 0.0002,
+      "step": 1755
+    },
+    {
+      "epoch": 1.5577477277765461,
+      "grad_norm": 0.2436944842338562,
+      "learning_rate": 4.92579564384872e-06,
+      "loss": 0.0013,
+      "step": 1756
+    },
+    {
+      "epoch": 1.5586344491243627,
+      "grad_norm": 0.1759563386440277,
+      "learning_rate": 4.9210086522868736e-06,
+      "loss": 0.0014,
+      "step": 1757
+    },
+    {
+      "epoch": 1.5595211704721792,
+      "grad_norm": 0.0894976332783699,
+      "learning_rate": 4.916221733146405e-06,
+      "loss": 0.0003,
+      "step": 1758
+    },
+    {
+      "epoch": 1.5604078918199957,
+      "grad_norm": 0.09848005324602127,
+      "learning_rate": 4.911434890816084e-06,
+      "loss": 0.0006,
+      "step": 1759
+    },
+    {
+      "epoch": 1.561294613167812,
+      "grad_norm": 0.05898863077163696,
+      "learning_rate": 4.906648129684623e-06,
+      "loss": 0.0003,
+      "step": 1760
+    },
+    {
+      "epoch": 1.5621813345156284,
+      "grad_norm": 0.46867799758911133,
+      "learning_rate": 4.901861454140649e-06,
+      "loss": 0.007,
+      "step": 1761
+    },
+    {
+      "epoch": 1.5630680558634449,
+      "grad_norm": 0.30307456851005554,
+      "learning_rate": 4.897074868572711e-06,
+      "loss": 0.0012,
+      "step": 1762
+    },
+    {
+      "epoch": 1.5639547772112614,
+      "grad_norm": 0.2749338746070862,
+      "learning_rate": 4.892288377369285e-06,
+      "loss": 0.0015,
+      "step": 1763
+    },
+    {
+      "epoch": 1.564841498559078,
+      "grad_norm": 0.20320548117160797,
+      "learning_rate": 4.887501984918748e-06,
+      "loss": 0.0007,
+      "step": 1764
+    },
+    {
+      "epoch": 1.5657282199068943,
+      "grad_norm": 0.14280276000499725,
+      "learning_rate": 4.882715695609396e-06,
+      "loss": 0.0008,
+      "step": 1765
+    },
+    {
+      "epoch": 1.5666149412547106,
+      "grad_norm": 0.05326784774661064,
+      "learning_rate": 4.877929513829424e-06,
+      "loss": 0.0003,
+      "step": 1766
+    },
+    {
+      "epoch": 1.567501662602527,
+      "grad_norm": 0.4620476961135864,
+      "learning_rate": 4.873143443966933e-06,
+      "loss": 0.0022,
+      "step": 1767
+    },
+    {
+      "epoch": 1.5683883839503436,
+      "grad_norm": 0.8573052287101746,
+      "learning_rate": 4.868357490409921e-06,
+      "loss": 0.0051,
+      "step": 1768
+    },
+    {
+      "epoch": 1.5692751052981602,
+      "grad_norm": 0.24559076130390167,
+      "learning_rate": 4.863571657546275e-06,
+      "loss": 0.0008,
+      "step": 1769
+    },
+    {
+      "epoch": 1.5701618266459765,
+      "grad_norm": 0.09770537912845612,
+      "learning_rate": 4.858785949763776e-06,
+      "loss": 0.0004,
+      "step": 1770
+    },
+    {
+      "epoch": 1.5710485479937928,
+      "grad_norm": 0.06460510939359665,
+      "learning_rate": 4.854000371450089e-06,
+      "loss": 0.0002,
+      "step": 1771
+    },
+    {
+      "epoch": 1.5719352693416093,
+      "grad_norm": 0.08998130261898041,
+      "learning_rate": 4.849214926992759e-06,
+      "loss": 0.0005,
+      "step": 1772
+    },
+    {
+      "epoch": 1.5728219906894259,
+      "grad_norm": 0.215066060423851,
+      "learning_rate": 4.8444296207792085e-06,
+      "loss": 0.0008,
+      "step": 1773
+    },
+    {
+      "epoch": 1.5737087120372424,
+      "grad_norm": 0.11160604655742645,
+      "learning_rate": 4.839644457196737e-06,
+      "loss": 0.0007,
+      "step": 1774
+    },
+    {
+      "epoch": 1.5745954333850587,
+      "grad_norm": 0.24873779714107513,
+      "learning_rate": 4.834859440632508e-06,
+      "loss": 0.0004,
+      "step": 1775
+    },
+    {
+      "epoch": 1.5754821547328752,
+      "grad_norm": 0.6680960655212402,
+      "learning_rate": 4.83007457547355e-06,
+      "loss": 0.0077,
+      "step": 1776
+    },
+    {
+      "epoch": 1.5763688760806915,
+      "grad_norm": 0.024391671642661095,
+      "learning_rate": 4.8252898661067605e-06,
+      "loss": 0.0002,
+      "step": 1777
+    },
+    {
+      "epoch": 1.577255597428508,
+      "grad_norm": 0.4404471814632416,
+      "learning_rate": 4.820505316918887e-06,
+      "loss": 0.0059,
+      "step": 1778
+    },
+    {
+      "epoch": 1.5781423187763246,
+      "grad_norm": 0.3777647316455841,
+      "learning_rate": 4.81572093229653e-06,
+      "loss": 0.0058,
+      "step": 1779
+    },
+    {
+      "epoch": 1.5790290401241411,
+      "grad_norm": 0.3575120270252228,
+      "learning_rate": 4.81093671662614e-06,
+      "loss": 0.0014,
+      "step": 1780
+    },
+    {
+      "epoch": 1.5799157614719574,
+      "grad_norm": 0.1214371770620346,
+      "learning_rate": 4.8061526742940186e-06,
+      "loss": 0.0006,
+      "step": 1781
+    },
+    {
+      "epoch": 1.5808024828197738,
+      "grad_norm": 0.42392250895500183,
+      "learning_rate": 4.8013688096863e-06,
+      "loss": 0.0013,
+      "step": 1782
+    },
+    {
+      "epoch": 1.5816892041675903,
+      "grad_norm": 0.13642321527004242,
+      "learning_rate": 4.796585127188958e-06,
+      "loss": 0.0007,
+      "step": 1783
+    },
+    {
+      "epoch": 1.5825759255154068,
+      "grad_norm": 0.11742696911096573,
+      "learning_rate": 4.791801631187804e-06,
+      "loss": 0.0006,
+      "step": 1784
+    },
+    {
+      "epoch": 1.5834626468632234,
+      "grad_norm": 0.03347081318497658,
+      "learning_rate": 4.787018326068474e-06,
+      "loss": 0.0002,
+      "step": 1785
+    },
+    {
+      "epoch": 1.5843493682110397,
+      "grad_norm": 0.3243880569934845,
+      "learning_rate": 4.7822352162164275e-06,
+      "loss": 0.0026,
+      "step": 1786
+    },
+    {
+      "epoch": 1.585236089558856,
+      "grad_norm": 0.4610430896282196,
+      "learning_rate": 4.777452306016947e-06,
+      "loss": 0.0033,
+      "step": 1787
+    },
+    {
+      "epoch": 1.5861228109066725,
+      "grad_norm": 0.9628936052322388,
+      "learning_rate": 4.772669599855136e-06,
+      "loss": 0.011,
+      "step": 1788
+    },
+    {
+      "epoch": 1.587009532254489,
+      "grad_norm": 0.1590164303779602,
+      "learning_rate": 4.767887102115904e-06,
+      "loss": 0.0006,
+      "step": 1789
+    },
+    {
+      "epoch": 1.5878962536023056,
+      "grad_norm": 0.07788955420255661,
+      "learning_rate": 4.763104817183974e-06,
+      "loss": 0.0004,
+      "step": 1790
+    },
+    {
+      "epoch": 1.5887829749501219,
+      "grad_norm": 0.05542571097612381,
+      "learning_rate": 4.758322749443874e-06,
+      "loss": 0.0004,
+      "step": 1791
+    },
+    {
+      "epoch": 1.5896696962979384,
+      "grad_norm": 0.33823779225349426,
+      "learning_rate": 4.753540903279928e-06,
+      "loss": 0.0019,
+      "step": 1792
+    },
+    {
+      "epoch": 1.5905564176457547,
+      "grad_norm": 0.07747834920883179,
+      "learning_rate": 4.748759283076262e-06,
+      "loss": 0.0005,
+      "step": 1793
+    },
+    {
+      "epoch": 1.5914431389935713,
+      "grad_norm": 0.17952410876750946,
+      "learning_rate": 4.743977893216792e-06,
+      "loss": 0.0017,
+      "step": 1794
+    },
+    {
+      "epoch": 1.5923298603413878,
+      "grad_norm": 0.015386722050607204,
+      "learning_rate": 4.739196738085225e-06,
+      "loss": 0.0001,
+      "step": 1795
+    },
+    {
+      "epoch": 1.5932165816892043,
+      "grad_norm": 0.1653813123703003,
+      "learning_rate": 4.734415822065053e-06,
+      "loss": 0.0008,
+      "step": 1796
+    },
+    {
+      "epoch": 1.5941033030370206,
+      "grad_norm": 0.055652689188718796,
+      "learning_rate": 4.729635149539543e-06,
+      "loss": 0.0004,
+      "step": 1797
+    },
+    {
+      "epoch": 1.594990024384837,
+      "grad_norm": 0.09155047684907913,
+      "learning_rate": 4.7248547248917495e-06,
+      "loss": 0.001,
+      "step": 1798
+    },
+    {
+      "epoch": 1.5958767457326535,
+      "grad_norm": 0.4243689179420471,
+      "learning_rate": 4.7200745525044885e-06,
+      "loss": 0.0072,
+      "step": 1799
+    },
+    {
+      "epoch": 1.59676346708047,
+      "grad_norm": 0.2508626878261566,
+      "learning_rate": 4.715294636760352e-06,
+      "loss": 0.003,
+      "step": 1800
+    },
+    {
+      "epoch": 1.5976501884282865,
+      "grad_norm": 0.3041090965270996,
+      "learning_rate": 4.710514982041691e-06,
+      "loss": 0.0092,
+      "step": 1801
+    },
+    {
+      "epoch": 1.5985369097761029,
+      "grad_norm": 0.45713216066360474,
+      "learning_rate": 4.705735592730626e-06,
+      "loss": 0.0018,
+      "step": 1802
+    },
+    {
+      "epoch": 1.5994236311239192,
+      "grad_norm": 0.10100582242012024,
+      "learning_rate": 4.700956473209025e-06,
+      "loss": 0.001,
+      "step": 1803
+    },
+    {
+      "epoch": 1.6003103524717357,
+      "grad_norm": 0.0519060418009758,
+      "learning_rate": 4.696177627858511e-06,
+      "loss": 0.0003,
+      "step": 1804
+    },
+    {
+      "epoch": 1.6011970738195522,
+      "grad_norm": 1.5515557527542114,
+      "learning_rate": 4.6913990610604625e-06,
+      "loss": 0.0047,
+      "step": 1805
+    },
+    {
+      "epoch": 1.6020837951673688,
+      "grad_norm": 0.26683878898620605,
+      "learning_rate": 4.686620777195993e-06,
+      "loss": 0.0023,
+      "step": 1806
+    },
+    {
+      "epoch": 1.602970516515185,
+      "grad_norm": 0.23357915878295898,
+      "learning_rate": 4.681842780645961e-06,
+      "loss": 0.0015,
+      "step": 1807
+    },
+    {
+      "epoch": 1.6038572378630016,
+      "grad_norm": 0.40496546030044556,
+      "learning_rate": 4.677065075790961e-06,
+      "loss": 0.0038,
+      "step": 1808
+    },
+    {
+      "epoch": 1.604743959210818,
+      "grad_norm": 0.13487201929092407,
+      "learning_rate": 4.672287667011323e-06,
+      "loss": 0.0006,
+      "step": 1809
+    },
+    {
+      "epoch": 1.6056306805586344,
+      "grad_norm": 0.2809060513973236,
+      "learning_rate": 4.667510558687101e-06,
+      "loss": 0.0019,
+      "step": 1810
+    },
+    {
+      "epoch": 1.606517401906451,
+      "grad_norm": 0.08889652043581009,
+      "learning_rate": 4.662733755198073e-06,
+      "loss": 0.0007,
+      "step": 1811
+    },
+    {
+      "epoch": 1.6074041232542673,
+      "grad_norm": 0.25819990038871765,
+      "learning_rate": 4.657957260923745e-06,
+      "loss": 0.0037,
+      "step": 1812
+    },
+    {
+      "epoch": 1.6082908446020838,
+      "grad_norm": 0.09379664063453674,
+      "learning_rate": 4.653181080243332e-06,
+      "loss": 0.0009,
+      "step": 1813
+    },
+    {
+      "epoch": 1.6091775659499001,
+      "grad_norm": 0.06153678521513939,
+      "learning_rate": 4.648405217535763e-06,
+      "loss": 0.0005,
+      "step": 1814
+    },
+    {
+      "epoch": 1.6100642872977167,
+      "grad_norm": 0.1556420475244522,
+      "learning_rate": 4.643629677179677e-06,
+      "loss": 0.0022,
+      "step": 1815
+    },
+    {
+      "epoch": 1.6109510086455332,
+      "grad_norm": 0.29694435000419617,
+      "learning_rate": 4.6388544635534186e-06,
+      "loss": 0.002,
+      "step": 1816
+    },
+    {
+      "epoch": 1.6118377299933497,
+      "grad_norm": 0.09254162758588791,
+      "learning_rate": 4.634079581035029e-06,
+      "loss": 0.0008,
+      "step": 1817
+    },
+    {
+      "epoch": 1.612724451341166,
+      "grad_norm": 0.039080262184143066,
+      "learning_rate": 4.629305034002249e-06,
+      "loss": 0.0004,
+      "step": 1818
+    },
+    {
+      "epoch": 1.6136111726889824,
+      "grad_norm": 0.334262490272522,
+      "learning_rate": 4.62453082683251e-06,
+      "loss": 0.0021,
+      "step": 1819
+    },
+    {
+      "epoch": 1.6144978940367989,
+      "grad_norm": 0.07444017380475998,
+      "learning_rate": 4.619756963902935e-06,
+      "loss": 0.0004,
+      "step": 1820
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.4941510558128357,
+      "learning_rate": 4.614983449590326e-06,
+      "loss": 0.0143,
+      "step": 1821
+    },
+    {
+      "epoch": 1.616271336732432,
+      "grad_norm": 0.11141020804643631,
+      "learning_rate": 4.610210288271169e-06,
+      "loss": 0.0007,
+      "step": 1822
+    },
+    {
+      "epoch": 1.6171580580802483,
+      "grad_norm": 0.05057799816131592,
+      "learning_rate": 4.605437484321627e-06,
+      "loss": 0.0004,
+      "step": 1823
+    },
+    {
+      "epoch": 1.6180447794280646,
+      "grad_norm": 0.5352963209152222,
+      "learning_rate": 4.600665042117534e-06,
+      "loss": 0.0037,
+      "step": 1824
+    },
+    {
+      "epoch": 1.618931500775881,
+      "grad_norm": 0.4706566035747528,
+      "learning_rate": 4.59589296603439e-06,
+      "loss": 0.0059,
+      "step": 1825
+    },
+    {
+      "epoch": 1.6198182221236976,
+      "grad_norm": 0.09059499949216843,
+      "learning_rate": 4.591121260447365e-06,
+      "loss": 0.0006,
+      "step": 1826
+    },
+    {
+      "epoch": 1.6207049434715142,
+      "grad_norm": 0.13940253853797913,
+      "learning_rate": 4.586349929731285e-06,
+      "loss": 0.0014,
+      "step": 1827
+    },
+    {
+      "epoch": 1.6215916648193305,
+      "grad_norm": 0.11622869223356247,
+      "learning_rate": 4.581578978260634e-06,
+      "loss": 0.001,
+      "step": 1828
+    },
+    {
+      "epoch": 1.622478386167147,
+      "grad_norm": 0.20990999042987823,
+      "learning_rate": 4.5768084104095436e-06,
+      "loss": 0.0013,
+      "step": 1829
+    },
+    {
+      "epoch": 1.6233651075149633,
+      "grad_norm": 0.586199939250946,
+      "learning_rate": 4.5720382305518026e-06,
+      "loss": 0.004,
+      "step": 1830
+    },
+    {
+      "epoch": 1.6242518288627799,
+      "grad_norm": 0.2799239456653595,
+      "learning_rate": 4.5672684430608384e-06,
+      "loss": 0.0067,
+      "step": 1831
+    },
+    {
+      "epoch": 1.6251385502105964,
+      "grad_norm": 0.43799954652786255,
+      "learning_rate": 4.562499052309715e-06,
+      "loss": 0.0032,
+      "step": 1832
+    },
+    {
+      "epoch": 1.626025271558413,
+      "grad_norm": 0.02102653495967388,
+      "learning_rate": 4.5577300626711444e-06,
+      "loss": 0.0001,
+      "step": 1833
+    },
+    {
+      "epoch": 1.6269119929062292,
+      "grad_norm": 0.051001351326704025,
+      "learning_rate": 4.5529614785174606e-06,
+      "loss": 0.0004,
+      "step": 1834
+    },
+    {
+      "epoch": 1.6277987142540455,
+      "grad_norm": 0.23884452879428864,
+      "learning_rate": 4.548193304220627e-06,
+      "loss": 0.0014,
+      "step": 1835
+    },
+    {
+      "epoch": 1.628685435601862,
+      "grad_norm": 0.5026673674583435,
+      "learning_rate": 4.543425544152232e-06,
+      "loss": 0.0052,
+      "step": 1836
+    },
+    {
+      "epoch": 1.6295721569496786,
+      "grad_norm": 0.10448399931192398,
+      "learning_rate": 4.53865820268349e-06,
+      "loss": 0.0005,
+      "step": 1837
+    },
+    {
+      "epoch": 1.6304588782974951,
+      "grad_norm": 0.5924953818321228,
+      "learning_rate": 4.533891284185226e-06,
+      "loss": 0.0023,
+      "step": 1838
+    },
+    {
+      "epoch": 1.6313455996453115,
+      "grad_norm": 0.035826656967401505,
+      "learning_rate": 4.529124793027873e-06,
+      "loss": 0.0003,
+      "step": 1839
+    },
+    {
+      "epoch": 1.6322323209931278,
+      "grad_norm": 0.30904772877693176,
+      "learning_rate": 4.524358733581485e-06,
+      "loss": 0.0036,
+      "step": 1840
+    },
+    {
+      "epoch": 1.6331190423409443,
+      "grad_norm": 0.2722933888435364,
+      "learning_rate": 4.519593110215708e-06,
+      "loss": 0.0018,
+      "step": 1841
+    },
+    {
+      "epoch": 1.6340057636887608,
+      "grad_norm": 0.5684006214141846,
+      "learning_rate": 4.514827927299792e-06,
+      "loss": 0.0101,
+      "step": 1842
+    },
+    {
+      "epoch": 1.6348924850365774,
+      "grad_norm": 0.6988949775695801,
+      "learning_rate": 4.510063189202585e-06,
+      "loss": 0.0019,
+      "step": 1843
+    },
+    {
+      "epoch": 1.6357792063843937,
+      "grad_norm": 0.1531183421611786,
+      "learning_rate": 4.505298900292528e-06,
+      "loss": 0.001,
+      "step": 1844
+    },
+    {
+      "epoch": 1.6366659277322102,
+      "grad_norm": 0.4703754484653473,
+      "learning_rate": 4.5005350649376465e-06,
+      "loss": 0.0049,
+      "step": 1845
+    },
+    {
+      "epoch": 1.6375526490800265,
+      "grad_norm": 0.3406502902507782,
+      "learning_rate": 4.4957716875055505e-06,
+      "loss": 0.0036,
+      "step": 1846
+    },
+    {
+      "epoch": 1.638439370427843,
+      "grad_norm": 0.5264698266983032,
+      "learning_rate": 4.491008772363435e-06,
+      "loss": 0.0012,
+      "step": 1847
+    },
+    {
+      "epoch": 1.6393260917756596,
+      "grad_norm": 0.13831202685832977,
+      "learning_rate": 4.486246323878065e-06,
+      "loss": 0.0016,
+      "step": 1848
+    },
+    {
+      "epoch": 1.640212813123476,
+      "grad_norm": 0.06777837127447128,
+      "learning_rate": 4.481484346415783e-06,
+      "loss": 0.0006,
+      "step": 1849
+    },
+    {
+      "epoch": 1.6410995344712924,
+      "grad_norm": 0.13357211649417877,
+      "learning_rate": 4.476722844342493e-06,
+      "loss": 0.0006,
+      "step": 1850
+    },
+    {
+      "epoch": 1.6419862558191087,
+      "grad_norm": 0.3240200877189636,
+      "learning_rate": 4.4719618220236715e-06,
+      "loss": 0.0051,
+      "step": 1851
+    },
+    {
+      "epoch": 1.6428729771669253,
+      "grad_norm": 0.08127690851688385,
+      "learning_rate": 4.467201283824349e-06,
+      "loss": 0.0006,
+      "step": 1852
+    },
+    {
+      "epoch": 1.6437596985147418,
+      "grad_norm": 0.12601187825202942,
+      "learning_rate": 4.462441234109115e-06,
+      "loss": 0.0008,
+      "step": 1853
+    },
+    {
+      "epoch": 1.6446464198625583,
+      "grad_norm": 0.19904936850070953,
+      "learning_rate": 4.457681677242107e-06,
+      "loss": 0.0011,
+      "step": 1854
+    },
+    {
+      "epoch": 1.6455331412103746,
+      "grad_norm": 0.5217944383621216,
+      "learning_rate": 4.452922617587017e-06,
+      "loss": 0.0056,
+      "step": 1855
+    },
+    {
+      "epoch": 1.646419862558191,
+      "grad_norm": 0.21843411028385162,
+      "learning_rate": 4.448164059507078e-06,
+      "loss": 0.0014,
+      "step": 1856
+    },
+    {
+      "epoch": 1.6473065839060075,
+      "grad_norm": 0.1688118726015091,
+      "learning_rate": 4.44340600736506e-06,
+      "loss": 0.002,
+      "step": 1857
+    },
+    {
+      "epoch": 1.648193305253824,
+      "grad_norm": 0.12506018579006195,
+      "learning_rate": 4.438648465523277e-06,
+      "loss": 0.0014,
+      "step": 1858
+    },
+    {
+      "epoch": 1.6490800266016405,
+      "grad_norm": 0.139232337474823,
+      "learning_rate": 4.433891438343566e-06,
+      "loss": 0.0009,
+      "step": 1859
+    },
+    {
+      "epoch": 1.6499667479494569,
+      "grad_norm": 0.2715148627758026,
+      "learning_rate": 4.4291349301873e-06,
+      "loss": 0.0021,
+      "step": 1860
+    },
+    {
+      "epoch": 1.6508534692972732,
+      "grad_norm": 0.3670569360256195,
+      "learning_rate": 4.424378945415368e-06,
+      "loss": 0.0041,
+      "step": 1861
+    },
+    {
+      "epoch": 1.6517401906450897,
+      "grad_norm": 0.06340687721967697,
+      "learning_rate": 4.419623488388189e-06,
+      "loss": 0.0005,
+      "step": 1862
+    },
+    {
+      "epoch": 1.6526269119929062,
+      "grad_norm": 0.22658933699131012,
+      "learning_rate": 4.4148685634656905e-06,
+      "loss": 0.0017,
+      "step": 1863
+    },
+    {
+      "epoch": 1.6535136333407228,
+      "grad_norm": 0.33755749464035034,
+      "learning_rate": 4.410114175007313e-06,
+      "loss": 0.0033,
+      "step": 1864
+    },
+    {
+      "epoch": 1.654400354688539,
+      "grad_norm": 0.31697922945022583,
+      "learning_rate": 4.405360327372011e-06,
+      "loss": 0.0015,
+      "step": 1865
+    },
+    {
+      "epoch": 1.6552870760363556,
+      "grad_norm": 0.13214468955993652,
+      "learning_rate": 4.4006070249182345e-06,
+      "loss": 0.0007,
+      "step": 1866
+    },
+    {
+      "epoch": 1.656173797384172,
+      "grad_norm": 0.15995311737060547,
+      "learning_rate": 4.395854272003942e-06,
+      "loss": 0.0007,
+      "step": 1867
+    },
+    {
+      "epoch": 1.6570605187319885,
+      "grad_norm": 0.028402239084243774,
+      "learning_rate": 4.391102072986581e-06,
+      "loss": 0.0001,
+      "step": 1868
+    },
+    {
+      "epoch": 1.657947240079805,
+      "grad_norm": 0.054573215544223785,
+      "learning_rate": 4.386350432223096e-06,
+      "loss": 0.0004,
+      "step": 1869
+    },
+    {
+      "epoch": 1.6588339614276215,
+      "grad_norm": 0.1543371081352234,
+      "learning_rate": 4.381599354069919e-06,
+      "loss": 0.0009,
+      "step": 1870
+    },
+    {
+      "epoch": 1.6597206827754378,
+      "grad_norm": 0.21532535552978516,
+      "learning_rate": 4.3768488428829625e-06,
+      "loss": 0.0011,
+      "step": 1871
+    },
+    {
+      "epoch": 1.6606074041232541,
+      "grad_norm": 0.21318702399730682,
+      "learning_rate": 4.372098903017624e-06,
+      "loss": 0.0006,
+      "step": 1872
+    },
+    {
+      "epoch": 1.6614941254710707,
+      "grad_norm": 0.18713395297527313,
+      "learning_rate": 4.3673495388287764e-06,
+      "loss": 0.0011,
+      "step": 1873
+    },
+    {
+      "epoch": 1.6623808468188872,
+      "grad_norm": 0.3029296100139618,
+      "learning_rate": 4.3626007546707615e-06,
+      "loss": 0.0017,
+      "step": 1874
+    },
+    {
+      "epoch": 1.6632675681667037,
+      "grad_norm": 0.1091129258275032,
+      "learning_rate": 4.357852554897391e-06,
+      "loss": 0.0006,
+      "step": 1875
+    },
+    {
+      "epoch": 1.66415428951452,
+      "grad_norm": 0.12485738098621368,
+      "learning_rate": 4.353104943861943e-06,
+      "loss": 0.0008,
+      "step": 1876
+    },
+    {
+      "epoch": 1.6650410108623364,
+      "grad_norm": 0.3830507695674896,
+      "learning_rate": 4.348357925917153e-06,
+      "loss": 0.0022,
+      "step": 1877
+    },
+    {
+      "epoch": 1.665927732210153,
+      "grad_norm": 0.09489868581295013,
+      "learning_rate": 4.343611505415213e-06,
+      "loss": 0.0004,
+      "step": 1878
+    },
+    {
+      "epoch": 1.6668144535579694,
+      "grad_norm": 0.44516703486442566,
+      "learning_rate": 4.338865686707771e-06,
+      "loss": 0.0008,
+      "step": 1879
+    },
+    {
+      "epoch": 1.667701174905786,
+      "grad_norm": 0.6598436832427979,
+      "learning_rate": 4.334120474145918e-06,
+      "loss": 0.0062,
+      "step": 1880
+    },
+    {
+      "epoch": 1.6685878962536023,
+      "grad_norm": 0.015342209488153458,
+      "learning_rate": 4.329375872080191e-06,
+      "loss": 0.0001,
+      "step": 1881
+    },
+    {
+      "epoch": 1.6694746176014188,
+      "grad_norm": 0.002708827145397663,
+      "learning_rate": 4.324631884860569e-06,
+      "loss": 0.0,
+      "step": 1882
+    },
+    {
+      "epoch": 1.670361338949235,
+      "grad_norm": 0.09634935110807419,
+      "learning_rate": 4.3198885168364656e-06,
+      "loss": 0.0005,
+      "step": 1883
+    },
+    {
+      "epoch": 1.6712480602970516,
+      "grad_norm": 0.09328538179397583,
+      "learning_rate": 4.315145772356729e-06,
+      "loss": 0.0003,
+      "step": 1884
+    },
+    {
+      "epoch": 1.6721347816448682,
+      "grad_norm": 0.05238676443696022,
+      "learning_rate": 4.310403655769629e-06,
+      "loss": 0.0002,
+      "step": 1885
+    },
+    {
+      "epoch": 1.6730215029926847,
+      "grad_norm": 0.3598534166812897,
+      "learning_rate": 4.305662171422872e-06,
+      "loss": 0.0042,
+      "step": 1886
+    },
+    {
+      "epoch": 1.673908224340501,
+      "grad_norm": 0.11761850118637085,
+      "learning_rate": 4.3009213236635725e-06,
+      "loss": 0.0005,
+      "step": 1887
+    },
+    {
+      "epoch": 1.6747949456883173,
+      "grad_norm": 0.04603422060608864,
+      "learning_rate": 4.296181116838268e-06,
+      "loss": 0.0002,
+      "step": 1888
+    },
+    {
+      "epoch": 1.6756816670361339,
+      "grad_norm": 0.0496135875582695,
+      "learning_rate": 4.291441555292904e-06,
+      "loss": 0.0003,
+      "step": 1889
+    },
+    {
+      "epoch": 1.6765683883839504,
+      "grad_norm": 0.24192200601100922,
+      "learning_rate": 4.2867026433728405e-06,
+      "loss": 0.0006,
+      "step": 1890
+    },
+    {
+      "epoch": 1.677455109731767,
+      "grad_norm": 0.7403578162193298,
+      "learning_rate": 4.281964385422837e-06,
+      "loss": 0.0053,
+      "step": 1891
+    },
+    {
+      "epoch": 1.6783418310795832,
+      "grad_norm": 0.05213829502463341,
+      "learning_rate": 4.277226785787054e-06,
+      "loss": 0.0002,
+      "step": 1892
+    },
+    {
+      "epoch": 1.6792285524273995,
+      "grad_norm": 0.18640650808811188,
+      "learning_rate": 4.272489848809053e-06,
+      "loss": 0.0009,
+      "step": 1893
+    },
+    {
+      "epoch": 1.680115273775216,
+      "grad_norm": 0.07999394834041595,
+      "learning_rate": 4.2677535788317796e-06,
+      "loss": 0.0003,
+      "step": 1894
+    },
+    {
+      "epoch": 1.6810019951230326,
+      "grad_norm": 0.04001365602016449,
+      "learning_rate": 4.263017980197575e-06,
+      "loss": 0.0002,
+      "step": 1895
+    },
+    {
+      "epoch": 1.6818887164708491,
+      "grad_norm": 0.2568782567977905,
+      "learning_rate": 4.258283057248161e-06,
+      "loss": 0.0016,
+      "step": 1896
+    },
+    {
+      "epoch": 1.6827754378186655,
+      "grad_norm": 0.323002427816391,
+      "learning_rate": 4.253548814324644e-06,
+      "loss": 0.0023,
+      "step": 1897
+    },
+    {
+      "epoch": 1.683662159166482,
+      "grad_norm": 0.30060967803001404,
+      "learning_rate": 4.248815255767503e-06,
+      "loss": 0.0016,
+      "step": 1898
+    },
+    {
+      "epoch": 1.6845488805142983,
+      "grad_norm": 0.3999418616294861,
+      "learning_rate": 4.2440823859165884e-06,
+      "loss": 0.0017,
+      "step": 1899
+    },
+    {
+      "epoch": 1.6854356018621148,
+      "grad_norm": 0.049183495342731476,
+      "learning_rate": 4.239350209111125e-06,
+      "loss": 0.0001,
+      "step": 1900
+    },
+    {
+      "epoch": 1.6863223232099314,
+      "grad_norm": 0.33251315355300903,
+      "learning_rate": 4.234618729689699e-06,
+      "loss": 0.0067,
+      "step": 1901
+    },
+    {
+      "epoch": 1.6872090445577477,
+      "grad_norm": 0.3990643620491028,
+      "learning_rate": 4.229887951990255e-06,
+      "loss": 0.0063,
+      "step": 1902
+    },
+    {
+      "epoch": 1.6880957659055642,
+      "grad_norm": 0.07162373512983322,
+      "learning_rate": 4.225157880350095e-06,
+      "loss": 0.0002,
+      "step": 1903
+    },
+    {
+      "epoch": 1.6889824872533805,
+      "grad_norm": 0.231755331158638,
+      "learning_rate": 4.22042851910588e-06,
+      "loss": 0.0026,
+      "step": 1904
+    },
+    {
+      "epoch": 1.689869208601197,
+      "grad_norm": 0.12471719831228256,
+      "learning_rate": 4.2156998725936105e-06,
+      "loss": 0.0005,
+      "step": 1905
+    },
+    {
+      "epoch": 1.6907559299490136,
+      "grad_norm": 0.032764095813035965,
+      "learning_rate": 4.210971945148635e-06,
+      "loss": 0.0002,
+      "step": 1906
+    },
+    {
+      "epoch": 1.6916426512968301,
+      "grad_norm": 0.1930425763130188,
+      "learning_rate": 4.206244741105649e-06,
+      "loss": 0.0043,
+      "step": 1907
+    },
+    {
+      "epoch": 1.6925293726446464,
+      "grad_norm": 0.24028733372688293,
+      "learning_rate": 4.201518264798674e-06,
+      "loss": 0.0008,
+      "step": 1908
+    },
+    {
+      "epoch": 1.6934160939924627,
+      "grad_norm": 0.12100367993116379,
+      "learning_rate": 4.196792520561072e-06,
+      "loss": 0.0007,
+      "step": 1909
+    },
+    {
+      "epoch": 1.6943028153402793,
+      "grad_norm": 0.05092811957001686,
+      "learning_rate": 4.19206751272553e-06,
+      "loss": 0.0002,
+      "step": 1910
+    },
+    {
+      "epoch": 1.6951895366880958,
+      "grad_norm": 0.06047261506319046,
+      "learning_rate": 4.187343245624065e-06,
+      "loss": 0.0003,
+      "step": 1911
+    },
+    {
+      "epoch": 1.6960762580359123,
+      "grad_norm": 0.005897665396332741,
+      "learning_rate": 4.182619723588007e-06,
+      "loss": 0.0,
+      "step": 1912
+    },
+    {
+      "epoch": 1.6969629793837286,
+      "grad_norm": 0.21145783364772797,
+      "learning_rate": 4.177896950948008e-06,
+      "loss": 0.0014,
+      "step": 1913
+    },
+    {
+      "epoch": 1.697849700731545,
+      "grad_norm": 0.0033418999519199133,
+      "learning_rate": 4.173174932034035e-06,
+      "loss": 0.0,
+      "step": 1914
+    },
+    {
+      "epoch": 1.6987364220793615,
+      "grad_norm": 0.2090112268924713,
+      "learning_rate": 4.168453671175358e-06,
+      "loss": 0.0008,
+      "step": 1915
+    },
+    {
+      "epoch": 1.699623143427178,
+      "grad_norm": 0.042710330337285995,
+      "learning_rate": 4.163733172700555e-06,
+      "loss": 0.0003,
+      "step": 1916
+    },
+    {
+      "epoch": 1.7005098647749946,
+      "grad_norm": 0.2604596018791199,
+      "learning_rate": 4.159013440937506e-06,
+      "loss": 0.0009,
+      "step": 1917
+    },
+    {
+      "epoch": 1.7013965861228109,
+      "grad_norm": 0.07750516384840012,
+      "learning_rate": 4.1542944802133876e-06,
+      "loss": 0.0002,
+      "step": 1918
+    },
+    {
+      "epoch": 1.7022833074706274,
+      "grad_norm": 0.06373310834169388,
+      "learning_rate": 4.149576294854668e-06,
+      "loss": 0.0002,
+      "step": 1919
+    },
+    {
+      "epoch": 1.7031700288184437,
+      "grad_norm": 0.5672020316123962,
+      "learning_rate": 4.1448588891871055e-06,
+      "loss": 0.0034,
+      "step": 1920
+    },
+    {
+      "epoch": 1.7040567501662602,
+      "grad_norm": 0.11927378922700882,
+      "learning_rate": 4.140142267535744e-06,
+      "loss": 0.0006,
+      "step": 1921
+    },
+    {
+      "epoch": 1.7049434715140768,
+      "grad_norm": 0.3349062204360962,
+      "learning_rate": 4.135426434224909e-06,
+      "loss": 0.0025,
+      "step": 1922
+    },
+    {
+      "epoch": 1.7058301928618933,
+      "grad_norm": 0.38301882147789,
+      "learning_rate": 4.130711393578201e-06,
+      "loss": 0.0005,
+      "step": 1923
+    },
+    {
+      "epoch": 1.7067169142097096,
+      "grad_norm": 0.16485494375228882,
+      "learning_rate": 4.125997149918494e-06,
+      "loss": 0.0011,
+      "step": 1924
+    },
+    {
+      "epoch": 1.707603635557526,
+      "grad_norm": 0.13596734404563904,
+      "learning_rate": 4.1212837075679366e-06,
+      "loss": 0.001,
+      "step": 1925
+    },
+    {
+      "epoch": 1.7084903569053425,
+      "grad_norm": 0.08278538286685944,
+      "learning_rate": 4.116571070847935e-06,
+      "loss": 0.0003,
+      "step": 1926
+    },
+    {
+      "epoch": 1.709377078253159,
+      "grad_norm": 0.34995728731155396,
+      "learning_rate": 4.111859244079159e-06,
+      "loss": 0.0035,
+      "step": 1927
+    },
+    {
+      "epoch": 1.7102637996009755,
+      "grad_norm": 0.025332273915410042,
+      "learning_rate": 4.107148231581543e-06,
+      "loss": 0.0001,
+      "step": 1928
+    },
+    {
+      "epoch": 1.7111505209487918,
+      "grad_norm": 0.05006008595228195,
+      "learning_rate": 4.102438037674265e-06,
+      "loss": 0.0002,
+      "step": 1929
+    },
+    {
+      "epoch": 1.7120372422966081,
+      "grad_norm": 0.05314438045024872,
+      "learning_rate": 4.097728666675757e-06,
+      "loss": 0.0002,
+      "step": 1930
+    },
+    {
+      "epoch": 1.7129239636444247,
+      "grad_norm": 0.5813071727752686,
+      "learning_rate": 4.093020122903694e-06,
+      "loss": 0.0022,
+      "step": 1931
+    },
+    {
+      "epoch": 1.7138106849922412,
+      "grad_norm": 0.2786867916584015,
+      "learning_rate": 4.088312410674999e-06,
+      "loss": 0.001,
+      "step": 1932
+    },
+    {
+      "epoch": 1.7146974063400577,
+      "grad_norm": 0.26664793491363525,
+      "learning_rate": 4.0836055343058265e-06,
+      "loss": 0.0006,
+      "step": 1933
+    },
+    {
+      "epoch": 1.715584127687874,
+      "grad_norm": 0.0041130343452095985,
+      "learning_rate": 4.078899498111563e-06,
+      "loss": 0.0,
+      "step": 1934
+    },
+    {
+      "epoch": 1.7164708490356906,
+      "grad_norm": 0.18312929570674896,
+      "learning_rate": 4.074194306406834e-06,
+      "loss": 0.0006,
+      "step": 1935
+    },
+    {
+      "epoch": 1.717357570383507,
+      "grad_norm": 0.41918063163757324,
+      "learning_rate": 4.069489963505482e-06,
+      "loss": 0.004,
+      "step": 1936
+    },
+    {
+      "epoch": 1.7182442917313234,
+      "grad_norm": 0.30160436034202576,
+      "learning_rate": 4.0647864737205735e-06,
+      "loss": 0.0044,
+      "step": 1937
+    },
+    {
+      "epoch": 1.71913101307914,
+      "grad_norm": 0.7079899907112122,
+      "learning_rate": 4.060083841364392e-06,
+      "loss": 0.0154,
+      "step": 1938
+    },
+    {
+      "epoch": 1.7200177344269565,
+      "grad_norm": 0.5802380442619324,
+      "learning_rate": 4.055382070748441e-06,
+      "loss": 0.006,
+      "step": 1939
+    },
+    {
+      "epoch": 1.7209044557747728,
+      "grad_norm": 0.022900115698575974,
+      "learning_rate": 4.050681166183427e-06,
+      "loss": 0.0001,
+      "step": 1940
+    },
+    {
+      "epoch": 1.7217911771225891,
+      "grad_norm": 0.04548019915819168,
+      "learning_rate": 4.045981131979263e-06,
+      "loss": 0.0003,
+      "step": 1941
+    },
+    {
+      "epoch": 1.7226778984704056,
+      "grad_norm": 0.09870709478855133,
+      "learning_rate": 4.041281972445069e-06,
+      "loss": 0.0007,
+      "step": 1942
+    },
+    {
+      "epoch": 1.7235646198182222,
+      "grad_norm": 0.07183930277824402,
+      "learning_rate": 4.036583691889159e-06,
+      "loss": 0.0004,
+      "step": 1943
+    },
+    {
+      "epoch": 1.7244513411660387,
+      "grad_norm": 0.055653832852840424,
+      "learning_rate": 4.0318862946190396e-06,
+      "loss": 0.0002,
+      "step": 1944
+    },
+    {
+      "epoch": 1.725338062513855,
+      "grad_norm": 0.019786400720477104,
+      "learning_rate": 4.027189784941412e-06,
+      "loss": 0.0001,
+      "step": 1945
+    },
+    {
+      "epoch": 1.7262247838616713,
+      "grad_norm": 0.08150695264339447,
+      "learning_rate": 4.0224941671621625e-06,
+      "loss": 0.0004,
+      "step": 1946
+    },
+    {
+      "epoch": 1.7271115052094879,
+      "grad_norm": 0.4023008346557617,
+      "learning_rate": 4.017799445586358e-06,
+      "loss": 0.0025,
+      "step": 1947
+    },
+    {
+      "epoch": 1.7279982265573044,
+      "grad_norm": 0.04819345846772194,
+      "learning_rate": 4.013105624518244e-06,
+      "loss": 0.0003,
+      "step": 1948
+    },
+    {
+      "epoch": 1.728884947905121,
+      "grad_norm": 0.1354270726442337,
+      "learning_rate": 4.0084127082612414e-06,
+      "loss": 0.0005,
+      "step": 1949
+    },
+    {
+      "epoch": 1.7297716692529372,
+      "grad_norm": 0.3102273941040039,
+      "learning_rate": 4.003720701117943e-06,
+      "loss": 0.0016,
+      "step": 1950
+    },
+    {
+      "epoch": 1.7306583906007535,
+      "grad_norm": 0.07029831409454346,
+      "learning_rate": 3.999029607390103e-06,
+      "loss": 0.0003,
+      "step": 1951
+    },
+    {
+      "epoch": 1.73154511194857,
+      "grad_norm": 0.06340515613555908,
+      "learning_rate": 3.994339431378642e-06,
+      "loss": 0.0001,
+      "step": 1952
+    },
+    {
+      "epoch": 1.7324318332963866,
+      "grad_norm": 0.026167303323745728,
+      "learning_rate": 3.989650177383641e-06,
+      "loss": 0.0002,
+      "step": 1953
+    },
+    {
+      "epoch": 1.7333185546442031,
+      "grad_norm": 0.019675275310873985,
+      "learning_rate": 3.9849618497043316e-06,
+      "loss": 0.0001,
+      "step": 1954
+    },
+    {
+      "epoch": 1.7342052759920195,
+      "grad_norm": 0.7322620153427124,
+      "learning_rate": 3.980274452639097e-06,
+      "loss": 0.0028,
+      "step": 1955
+    },
+    {
+      "epoch": 1.735091997339836,
+      "grad_norm": 0.01168559119105339,
+      "learning_rate": 3.975587990485468e-06,
+      "loss": 0.0001,
+      "step": 1956
+    },
+    {
+      "epoch": 1.7359787186876523,
+      "grad_norm": 0.06696490198373795,
+      "learning_rate": 3.97090246754012e-06,
+      "loss": 0.0005,
+      "step": 1957
+    },
+    {
+      "epoch": 1.7368654400354688,
+      "grad_norm": 0.3530293107032776,
+      "learning_rate": 3.966217888098866e-06,
+      "loss": 0.0013,
+      "step": 1958
+    },
+    {
+      "epoch": 1.7377521613832854,
+      "grad_norm": 0.5210719108581543,
+      "learning_rate": 3.961534256456648e-06,
+      "loss": 0.0061,
+      "step": 1959
+    },
+    {
+      "epoch": 1.738638882731102,
+      "grad_norm": 0.28391262888908386,
+      "learning_rate": 3.9568515769075525e-06,
+      "loss": 0.0019,
+      "step": 1960
+    },
+    {
+      "epoch": 1.7395256040789182,
+      "grad_norm": 0.2902699112892151,
+      "learning_rate": 3.95216985374478e-06,
+      "loss": 0.002,
+      "step": 1961
+    },
+    {
+      "epoch": 1.7404123254267345,
+      "grad_norm": 0.1008397787809372,
+      "learning_rate": 3.94748909126066e-06,
+      "loss": 0.0009,
+      "step": 1962
+    },
+    {
+      "epoch": 1.741299046774551,
+      "grad_norm": 0.1551593691110611,
+      "learning_rate": 3.94280929374664e-06,
+      "loss": 0.0004,
+      "step": 1963
+    },
+    {
+      "epoch": 1.7421857681223676,
+      "grad_norm": 0.3223740756511688,
+      "learning_rate": 3.938130465493286e-06,
+      "loss": 0.0028,
+      "step": 1964
+    },
+    {
+      "epoch": 1.7430724894701841,
+      "grad_norm": 0.03544246777892113,
+      "learning_rate": 3.933452610790271e-06,
+      "loss": 0.0003,
+      "step": 1965
+    },
+    {
+      "epoch": 1.7439592108180004,
+      "grad_norm": 0.04058455675840378,
+      "learning_rate": 3.928775733926376e-06,
+      "loss": 0.0002,
+      "step": 1966
+    },
+    {
+      "epoch": 1.7448459321658167,
+      "grad_norm": 0.18618978559970856,
+      "learning_rate": 3.924099839189488e-06,
+      "loss": 0.001,
+      "step": 1967
+    },
+    {
+      "epoch": 1.7457326535136333,
+      "grad_norm": 0.4316830039024353,
+      "learning_rate": 3.9194249308665915e-06,
+      "loss": 0.0041,
+      "step": 1968
+    },
+    {
+      "epoch": 1.7466193748614498,
+      "grad_norm": 0.32625532150268555,
+      "learning_rate": 3.914751013243769e-06,
+      "loss": 0.0018,
+      "step": 1969
+    },
+    {
+      "epoch": 1.7475060962092663,
+      "grad_norm": 0.18915483355522156,
+      "learning_rate": 3.91007809060619e-06,
+      "loss": 0.0011,
+      "step": 1970
+    },
+    {
+      "epoch": 1.7483928175570826,
+      "grad_norm": 0.44998809695243835,
+      "learning_rate": 3.905406167238116e-06,
+      "loss": 0.0061,
+      "step": 1971
+    },
+    {
+      "epoch": 1.7492795389048992,
+      "grad_norm": 0.22313672304153442,
+      "learning_rate": 3.9007352474228925e-06,
+      "loss": 0.0019,
+      "step": 1972
+    },
+    {
+      "epoch": 1.7501662602527155,
+      "grad_norm": 0.018498016521334648,
+      "learning_rate": 3.896065335442941e-06,
+      "loss": 0.0001,
+      "step": 1973
+    },
+    {
+      "epoch": 1.751052981600532,
+      "grad_norm": 0.4142586886882782,
+      "learning_rate": 3.891396435579763e-06,
+      "loss": 0.0029,
+      "step": 1974
+    },
+    {
+      "epoch": 1.7519397029483486,
+      "grad_norm": 0.2485865354537964,
+      "learning_rate": 3.886728552113931e-06,
+      "loss": 0.0008,
+      "step": 1975
+    },
+    {
+      "epoch": 1.752826424296165,
+      "grad_norm": 0.2031048834323883,
+      "learning_rate": 3.882061689325084e-06,
+      "loss": 0.0007,
+      "step": 1976
+    },
+    {
+      "epoch": 1.7537131456439814,
+      "grad_norm": 0.23484554886817932,
+      "learning_rate": 3.877395851491926e-06,
+      "loss": 0.0009,
+      "step": 1977
+    },
+    {
+      "epoch": 1.7545998669917977,
+      "grad_norm": 0.13688446581363678,
+      "learning_rate": 3.872731042892225e-06,
+      "loss": 0.0007,
+      "step": 1978
+    },
+    {
+      "epoch": 1.7554865883396142,
+      "grad_norm": 0.15276920795440674,
+      "learning_rate": 3.868067267802798e-06,
+      "loss": 0.001,
+      "step": 1979
+    },
+    {
+      "epoch": 1.7563733096874308,
+      "grad_norm": 0.4296773672103882,
+      "learning_rate": 3.8634045304995185e-06,
+      "loss": 0.0086,
+      "step": 1980
+    },
+    {
+      "epoch": 1.7572600310352473,
+      "grad_norm": 0.19526557624340057,
+      "learning_rate": 3.858742835257313e-06,
+      "loss": 0.0011,
+      "step": 1981
+    },
+    {
+      "epoch": 1.7581467523830636,
+      "grad_norm": 0.5239609479904175,
+      "learning_rate": 3.854082186350145e-06,
+      "loss": 0.0028,
+      "step": 1982
+    },
+    {
+      "epoch": 1.75903347373088,
+      "grad_norm": 0.13047945499420166,
+      "learning_rate": 3.849422588051022e-06,
+      "loss": 0.0007,
+      "step": 1983
+    },
+    {
+      "epoch": 1.7599201950786965,
+      "grad_norm": 0.2438761293888092,
+      "learning_rate": 3.844764044631986e-06,
+      "loss": 0.0023,
+      "step": 1984
+    },
+    {
+      "epoch": 1.760806916426513,
+      "grad_norm": 0.034038957208395004,
+      "learning_rate": 3.840106560364119e-06,
+      "loss": 0.0001,
+      "step": 1985
+    },
+    {
+      "epoch": 1.7616936377743295,
+      "grad_norm": 0.11754358559846878,
+      "learning_rate": 3.835450139517524e-06,
+      "loss": 0.0004,
+      "step": 1986
+    },
+    {
+      "epoch": 1.7625803591221458,
+      "grad_norm": 0.49829450249671936,
+      "learning_rate": 3.83079478636133e-06,
+      "loss": 0.0027,
+      "step": 1987
+    },
+    {
+      "epoch": 1.7634670804699624,
+      "grad_norm": 0.04286176338791847,
+      "learning_rate": 3.826140505163694e-06,
+      "loss": 0.0002,
+      "step": 1988
+    },
+    {
+      "epoch": 1.7643538018177787,
+      "grad_norm": 0.1739022135734558,
+      "learning_rate": 3.821487300191782e-06,
+      "loss": 0.0007,
+      "step": 1989
+    },
+    {
+      "epoch": 1.7652405231655952,
+      "grad_norm": 0.07864267379045486,
+      "learning_rate": 3.816835175711775e-06,
+      "loss": 0.0005,
+      "step": 1990
+    },
+    {
+      "epoch": 1.7661272445134117,
+      "grad_norm": 0.05060047656297684,
+      "learning_rate": 3.812184135988868e-06,
+      "loss": 0.0001,
+      "step": 1991
+    },
+    {
+      "epoch": 1.7670139658612283,
+      "grad_norm": 0.30648404359817505,
+      "learning_rate": 3.8075341852872583e-06,
+      "loss": 0.001,
+      "step": 1992
+    },
+    {
+      "epoch": 1.7679006872090446,
+      "grad_norm": 0.15932850539684296,
+      "learning_rate": 3.802885327870143e-06,
+      "loss": 0.0016,
+      "step": 1993
+    },
+    {
+      "epoch": 1.768787408556861,
+      "grad_norm": 0.021853476762771606,
+      "learning_rate": 3.7982375679997186e-06,
+      "loss": 0.0002,
+      "step": 1994
+    },
+    {
+      "epoch": 1.7696741299046774,
+      "grad_norm": 0.0491185262799263,
+      "learning_rate": 3.7935909099371784e-06,
+      "loss": 0.0002,
+      "step": 1995
+    },
+    {
+      "epoch": 1.770560851252494,
+      "grad_norm": 0.08880337327718735,
+      "learning_rate": 3.7889453579426993e-06,
+      "loss": 0.0005,
+      "step": 1996
+    },
+    {
+      "epoch": 1.7714475726003105,
+      "grad_norm": 0.0074866595678031445,
+      "learning_rate": 3.7843009162754503e-06,
+      "loss": 0.0001,
+      "step": 1997
+    },
+    {
+      "epoch": 1.7723342939481268,
+      "grad_norm": 0.43862977623939514,
+      "learning_rate": 3.7796575891935748e-06,
+      "loss": 0.0025,
+      "step": 1998
+    },
+    {
+      "epoch": 1.7732210152959431,
+      "grad_norm": 0.1807900071144104,
+      "learning_rate": 3.7750153809542052e-06,
+      "loss": 0.0005,
+      "step": 1999
+    },
+    {
+      "epoch": 1.7741077366437596,
+      "grad_norm": 0.1964527666568756,
+      "learning_rate": 3.7703742958134383e-06,
+      "loss": 0.001,
+      "step": 2000
+    },
+    {
+      "epoch": 1.7749944579915762,
+      "grad_norm": 0.6390202641487122,
+      "learning_rate": 3.765734338026343e-06,
+      "loss": 0.0026,
+      "step": 2001
+    },
+    {
+      "epoch": 1.7758811793393927,
+      "grad_norm": 0.2554784417152405,
+      "learning_rate": 3.761095511846962e-06,
+      "loss": 0.0028,
+      "step": 2002
+    },
+    {
+      "epoch": 1.776767900687209,
+      "grad_norm": 0.42926931381225586,
+      "learning_rate": 3.756457821528292e-06,
+      "loss": 0.0014,
+      "step": 2003
+    },
+    {
+      "epoch": 1.7776546220350253,
+      "grad_norm": 0.44991567730903625,
+      "learning_rate": 3.7518212713222905e-06,
+      "loss": 0.0023,
+      "step": 2004
+    },
+    {
+      "epoch": 1.7785413433828419,
+      "grad_norm": 0.2629556655883789,
+      "learning_rate": 3.7471858654798697e-06,
+      "loss": 0.0007,
+      "step": 2005
+    },
+    {
+      "epoch": 1.7794280647306584,
+      "grad_norm": 0.003782880026847124,
+      "learning_rate": 3.742551608250896e-06,
+      "loss": 0.0,
+      "step": 2006
+    },
+    {
+      "epoch": 1.780314786078475,
+      "grad_norm": 0.17819075286388397,
+      "learning_rate": 3.73791850388418e-06,
+      "loss": 0.0009,
+      "step": 2007
+    },
+    {
+      "epoch": 1.7812015074262912,
+      "grad_norm": 0.5463308095932007,
+      "learning_rate": 3.733286556627473e-06,
+      "loss": 0.003,
+      "step": 2008
+    },
+    {
+      "epoch": 1.7820882287741078,
+      "grad_norm": 0.07021505385637283,
+      "learning_rate": 3.7286557707274707e-06,
+      "loss": 0.0002,
+      "step": 2009
+    },
+    {
+      "epoch": 1.782974950121924,
+      "grad_norm": 0.3599243760108948,
+      "learning_rate": 3.7240261504298002e-06,
+      "loss": 0.0067,
+      "step": 2010
+    },
+    {
+      "epoch": 1.7838616714697406,
+      "grad_norm": 0.04074014723300934,
+      "learning_rate": 3.7193976999790216e-06,
+      "loss": 0.0002,
+      "step": 2011
+    },
+    {
+      "epoch": 1.7847483928175571,
+      "grad_norm": 0.08518295735120773,
+      "learning_rate": 3.714770423618618e-06,
+      "loss": 0.0002,
+      "step": 2012
+    },
+    {
+      "epoch": 1.7856351141653737,
+      "grad_norm": 0.10783907026052475,
+      "learning_rate": 3.7101443255910062e-06,
+      "loss": 0.0003,
+      "step": 2013
+    },
+    {
+      "epoch": 1.78652183551319,
+      "grad_norm": 0.31293633580207825,
+      "learning_rate": 3.7055194101375127e-06,
+      "loss": 0.0023,
+      "step": 2014
+    },
+    {
+      "epoch": 1.7874085568610063,
+      "grad_norm": 0.5668357014656067,
+      "learning_rate": 3.700895681498381e-06,
+      "loss": 0.0053,
+      "step": 2015
+    },
+    {
+      "epoch": 1.7882952782088228,
+      "grad_norm": 0.01921240985393524,
+      "learning_rate": 3.696273143912775e-06,
+      "loss": 0.0001,
+      "step": 2016
+    },
+    {
+      "epoch": 1.7891819995566394,
+      "grad_norm": 0.018799765035510063,
+      "learning_rate": 3.6916518016187573e-06,
+      "loss": 0.0001,
+      "step": 2017
+    },
+    {
+      "epoch": 1.790068720904456,
+      "grad_norm": 0.11832908540964127,
+      "learning_rate": 3.6870316588532966e-06,
+      "loss": 0.0004,
+      "step": 2018
+    },
+    {
+      "epoch": 1.7909554422522722,
+      "grad_norm": 0.10770263522863388,
+      "learning_rate": 3.6824127198522652e-06,
+      "loss": 0.0005,
+      "step": 2019
+    },
+    {
+      "epoch": 1.7918421636000885,
+      "grad_norm": 0.01235107984393835,
+      "learning_rate": 3.6777949888504293e-06,
+      "loss": 0.0,
+      "step": 2020
+    },
+    {
+      "epoch": 1.792728884947905,
+      "grad_norm": 0.006794508080929518,
+      "learning_rate": 3.6731784700814476e-06,
+      "loss": 0.0,
+      "step": 2021
+    },
+    {
+      "epoch": 1.7936156062957216,
+      "grad_norm": 0.01359777431935072,
+      "learning_rate": 3.668563167777868e-06,
+      "loss": 0.0001,
+      "step": 2022
+    },
+    {
+      "epoch": 1.7945023276435381,
+      "grad_norm": 0.13341398537158966,
+      "learning_rate": 3.6639490861711224e-06,
+      "loss": 0.0005,
+      "step": 2023
+    },
+    {
+      "epoch": 1.7953890489913544,
+      "grad_norm": 0.022680554538965225,
+      "learning_rate": 3.659336229491525e-06,
+      "loss": 0.0001,
+      "step": 2024
+    },
+    {
+      "epoch": 1.796275770339171,
+      "grad_norm": 0.06998679041862488,
+      "learning_rate": 3.654724601968266e-06,
+      "loss": 0.0002,
+      "step": 2025
+    },
+    {
+      "epoch": 1.7971624916869873,
+      "grad_norm": 0.20700500905513763,
+      "learning_rate": 3.6501142078294053e-06,
+      "loss": 0.003,
+      "step": 2026
+    },
+    {
+      "epoch": 1.7980492130348038,
+      "grad_norm": 0.07539916038513184,
+      "learning_rate": 3.6455050513018807e-06,
+      "loss": 0.0003,
+      "step": 2027
+    },
+    {
+      "epoch": 1.7989359343826203,
+      "grad_norm": 0.45003005862236023,
+      "learning_rate": 3.6408971366114874e-06,
+      "loss": 0.0026,
+      "step": 2028
+    },
+    {
+      "epoch": 1.7998226557304369,
+      "grad_norm": 0.022108634933829308,
+      "learning_rate": 3.6362904679828814e-06,
+      "loss": 0.0001,
+      "step": 2029
+    },
+    {
+      "epoch": 1.8007093770782532,
+      "grad_norm": 0.24113281071186066,
+      "learning_rate": 3.6316850496395863e-06,
+      "loss": 0.0008,
+      "step": 2030
+    },
+    {
+      "epoch": 1.8015960984260695,
+      "grad_norm": 0.10309310257434845,
+      "learning_rate": 3.6270808858039677e-06,
+      "loss": 0.0003,
+      "step": 2031
+    },
+    {
+      "epoch": 1.802482819773886,
+      "grad_norm": 0.404070645570755,
+      "learning_rate": 3.6224779806972472e-06,
+      "loss": 0.0009,
+      "step": 2032
+    },
+    {
+      "epoch": 1.8033695411217026,
+      "grad_norm": 0.25489571690559387,
+      "learning_rate": 3.617876338539489e-06,
+      "loss": 0.0014,
+      "step": 2033
+    },
+    {
+      "epoch": 1.804256262469519,
+      "grad_norm": 0.9427079558372498,
+      "learning_rate": 3.6132759635496057e-06,
+      "loss": 0.005,
+      "step": 2034
+    },
+    {
+      "epoch": 1.8051429838173354,
+      "grad_norm": 0.02262832224369049,
+      "learning_rate": 3.6086768599453416e-06,
+      "loss": 0.0001,
+      "step": 2035
+    },
+    {
+      "epoch": 1.8060297051651517,
+      "grad_norm": 0.33301153779029846,
+      "learning_rate": 3.6040790319432756e-06,
+      "loss": 0.0031,
+      "step": 2036
+    },
+    {
+      "epoch": 1.8069164265129682,
+      "grad_norm": 0.20669648051261902,
+      "learning_rate": 3.5994824837588235e-06,
+      "loss": 0.0021,
+      "step": 2037
+    },
+    {
+      "epoch": 1.8078031478607848,
+      "grad_norm": 0.12887296080589294,
+      "learning_rate": 3.594887219606221e-06,
+      "loss": 0.0007,
+      "step": 2038
+    },
+    {
+      "epoch": 1.8086898692086013,
+      "grad_norm": 0.017986498773097992,
+      "learning_rate": 3.5902932436985293e-06,
+      "loss": 0.0002,
+      "step": 2039
+    },
+    {
+      "epoch": 1.8095765905564176,
+      "grad_norm": 0.11879166960716248,
+      "learning_rate": 3.585700560247626e-06,
+      "loss": 0.0005,
+      "step": 2040
+    },
+    {
+      "epoch": 1.810463311904234,
+      "grad_norm": 0.011947376653552055,
+      "learning_rate": 3.58110917346421e-06,
+      "loss": 0.0001,
+      "step": 2041
+    },
+    {
+      "epoch": 1.8113500332520505,
+      "grad_norm": 0.427824467420578,
+      "learning_rate": 3.5765190875577855e-06,
+      "loss": 0.0025,
+      "step": 2042
+    },
+    {
+      "epoch": 1.812236754599867,
+      "grad_norm": 0.5713347792625427,
+      "learning_rate": 3.571930306736663e-06,
+      "loss": 0.0046,
+      "step": 2043
+    },
+    {
+      "epoch": 1.8131234759476835,
+      "grad_norm": 0.021777095273137093,
+      "learning_rate": 3.567342835207964e-06,
+      "loss": 0.0001,
+      "step": 2044
+    },
+    {
+      "epoch": 1.8140101972954998,
+      "grad_norm": 0.026599332690238953,
+      "learning_rate": 3.562756677177602e-06,
+      "loss": 0.0001,
+      "step": 2045
+    },
+    {
+      "epoch": 1.8148969186433164,
+      "grad_norm": 0.10613416880369186,
+      "learning_rate": 3.5581718368502906e-06,
+      "loss": 0.0005,
+      "step": 2046
+    },
+    {
+      "epoch": 1.8157836399911327,
+      "grad_norm": 0.5284315347671509,
+      "learning_rate": 3.5535883184295316e-06,
+      "loss": 0.0016,
+      "step": 2047
+    },
+    {
+      "epoch": 1.8166703613389492,
+      "grad_norm": 0.087979257106781,
+      "learning_rate": 3.549006126117619e-06,
+      "loss": 0.0005,
+      "step": 2048
+    },
+    {
+      "epoch": 1.8175570826867657,
+      "grad_norm": 0.006895984057337046,
+      "learning_rate": 3.54442526411563e-06,
+      "loss": 0.0001,
+      "step": 2049
+    },
+    {
+      "epoch": 1.8184438040345823,
+      "grad_norm": 0.15001380443572998,
+      "learning_rate": 3.5398457366234174e-06,
+      "loss": 0.0002,
+      "step": 2050
+    },
+    {
+      "epoch": 1.8193305253823986,
+      "grad_norm": 0.011246156878769398,
+      "learning_rate": 3.535267547839617e-06,
+      "loss": 0.0001,
+      "step": 2051
+    },
+    {
+      "epoch": 1.820217246730215,
+      "grad_norm": 0.006199950352311134,
+      "learning_rate": 3.5306907019616356e-06,
+      "loss": 0.0,
+      "step": 2052
+    },
+    {
+      "epoch": 1.8211039680780314,
+      "grad_norm": 0.013676552101969719,
+      "learning_rate": 3.5261152031856457e-06,
+      "loss": 0.0001,
+      "step": 2053
+    },
+    {
+      "epoch": 1.821990689425848,
+      "grad_norm": 0.10605849325656891,
+      "learning_rate": 3.5215410557065856e-06,
+      "loss": 0.0003,
+      "step": 2054
+    },
+    {
+      "epoch": 1.8228774107736645,
+      "grad_norm": 0.05242261290550232,
+      "learning_rate": 3.516968263718159e-06,
+      "loss": 0.0001,
+      "step": 2055
+    },
+    {
+      "epoch": 1.8237641321214808,
+      "grad_norm": 0.12605375051498413,
+      "learning_rate": 3.512396831412822e-06,
+      "loss": 0.0007,
+      "step": 2056
+    },
+    {
+      "epoch": 1.8246508534692971,
+      "grad_norm": 0.004360876511782408,
+      "learning_rate": 3.507826762981784e-06,
+      "loss": 0.0,
+      "step": 2057
+    },
+    {
+      "epoch": 1.8255375748171137,
+      "grad_norm": 0.4643060564994812,
+      "learning_rate": 3.50325806261501e-06,
+      "loss": 0.0056,
+      "step": 2058
+    },
+    {
+      "epoch": 1.8264242961649302,
+      "grad_norm": 0.06710820645093918,
+      "learning_rate": 3.498690734501204e-06,
+      "loss": 0.0003,
+      "step": 2059
+    },
+    {
+      "epoch": 1.8273110175127467,
+      "grad_norm": 0.36204835772514343,
+      "learning_rate": 3.4941247828278142e-06,
+      "loss": 0.0048,
+      "step": 2060
+    },
+    {
+      "epoch": 1.828197738860563,
+      "grad_norm": 0.143352210521698,
+      "learning_rate": 3.4895602117810256e-06,
+      "loss": 0.0004,
+      "step": 2061
+    },
+    {
+      "epoch": 1.8290844602083796,
+      "grad_norm": 0.15158715844154358,
+      "learning_rate": 3.484997025545762e-06,
+      "loss": 0.0008,
+      "step": 2062
+    },
+    {
+      "epoch": 1.8299711815561959,
+      "grad_norm": 0.0470881387591362,
+      "learning_rate": 3.480435228305673e-06,
+      "loss": 0.0002,
+      "step": 2063
+    },
+    {
+      "epoch": 1.8308579029040124,
+      "grad_norm": 0.15281467139720917,
+      "learning_rate": 3.4758748242431363e-06,
+      "loss": 0.001,
+      "step": 2064
+    },
+    {
+      "epoch": 1.831744624251829,
+      "grad_norm": 0.44346439838409424,
+      "learning_rate": 3.4713158175392493e-06,
+      "loss": 0.0027,
+      "step": 2065
+    },
+    {
+      "epoch": 1.8326313455996455,
+      "grad_norm": 0.09658630192279816,
+      "learning_rate": 3.466758212373836e-06,
+      "loss": 0.0005,
+      "step": 2066
+    },
+    {
+      "epoch": 1.8335180669474618,
+      "grad_norm": 0.19065168499946594,
+      "learning_rate": 3.4622020129254273e-06,
+      "loss": 0.0014,
+      "step": 2067
+    },
+    {
+      "epoch": 1.834404788295278,
+      "grad_norm": 0.39815741777420044,
+      "learning_rate": 3.4576472233712677e-06,
+      "loss": 0.0017,
+      "step": 2068
+    },
+    {
+      "epoch": 1.8352915096430946,
+      "grad_norm": 0.020850352942943573,
+      "learning_rate": 3.4530938478873134e-06,
+      "loss": 0.0001,
+      "step": 2069
+    },
+    {
+      "epoch": 1.8361782309909112,
+      "grad_norm": 0.0858813151717186,
+      "learning_rate": 3.448541890648217e-06,
+      "loss": 0.0005,
+      "step": 2070
+    },
+    {
+      "epoch": 1.8370649523387277,
+      "grad_norm": 0.016161078587174416,
+      "learning_rate": 3.4439913558273374e-06,
+      "loss": 0.0001,
+      "step": 2071
+    },
+    {
+      "epoch": 1.837951673686544,
+      "grad_norm": 0.08752348273992538,
+      "learning_rate": 3.439442247596724e-06,
+      "loss": 0.0004,
+      "step": 2072
+    },
+    {
+      "epoch": 1.8388383950343603,
+      "grad_norm": 0.07176404446363449,
+      "learning_rate": 3.4348945701271218e-06,
+      "loss": 0.0003,
+      "step": 2073
+    },
+    {
+      "epoch": 1.8397251163821768,
+      "grad_norm": 0.021734774112701416,
+      "learning_rate": 3.4303483275879633e-06,
+      "loss": 0.0001,
+      "step": 2074
+    },
+    {
+      "epoch": 1.8406118377299934,
+      "grad_norm": 0.13391822576522827,
+      "learning_rate": 3.425803524147364e-06,
+      "loss": 0.0006,
+      "step": 2075
+    },
+    {
+      "epoch": 1.84149855907781,
+      "grad_norm": 0.43593868613243103,
+      "learning_rate": 3.421260163972121e-06,
+      "loss": 0.0015,
+      "step": 2076
+    },
+    {
+      "epoch": 1.8423852804256262,
+      "grad_norm": 0.037805862724781036,
+      "learning_rate": 3.4167182512277105e-06,
+      "loss": 0.0001,
+      "step": 2077
+    },
+    {
+      "epoch": 1.8432720017734427,
+      "grad_norm": 0.019509362056851387,
+      "learning_rate": 3.412177790078277e-06,
+      "loss": 0.0001,
+      "step": 2078
+    },
+    {
+      "epoch": 1.844158723121259,
+      "grad_norm": 0.09484810382127762,
+      "learning_rate": 3.4076387846866353e-06,
+      "loss": 0.0006,
+      "step": 2079
+    },
+    {
+      "epoch": 1.8450454444690756,
+      "grad_norm": 0.046936165541410446,
+      "learning_rate": 3.403101239214271e-06,
+      "loss": 0.0002,
+      "step": 2080
+    },
+    {
+      "epoch": 1.8459321658168921,
+      "grad_norm": 0.11847790330648422,
+      "learning_rate": 3.3985651578213243e-06,
+      "loss": 0.0005,
+      "step": 2081
+    },
+    {
+      "epoch": 1.8468188871647087,
+      "grad_norm": 0.09800742566585541,
+      "learning_rate": 3.394030544666593e-06,
+      "loss": 0.0003,
+      "step": 2082
+    },
+    {
+      "epoch": 1.847705608512525,
+      "grad_norm": 0.29517197608947754,
+      "learning_rate": 3.3894974039075377e-06,
+      "loss": 0.0024,
+      "step": 2083
+    },
+    {
+      "epoch": 1.8485923298603413,
+      "grad_norm": 0.0667266771197319,
+      "learning_rate": 3.3849657397002588e-06,
+      "loss": 0.0002,
+      "step": 2084
+    },
+    {
+      "epoch": 1.8494790512081578,
+      "grad_norm": 0.3751622140407562,
+      "learning_rate": 3.3804355561995085e-06,
+      "loss": 0.0022,
+      "step": 2085
+    },
+    {
+      "epoch": 1.8503657725559743,
+      "grad_norm": 0.6706212162971497,
+      "learning_rate": 3.375906857558676e-06,
+      "loss": 0.0024,
+      "step": 2086
+    },
+    {
+      "epoch": 1.8512524939037909,
+      "grad_norm": 0.08441176265478134,
+      "learning_rate": 3.3713796479298e-06,
+      "loss": 0.0003,
+      "step": 2087
+    },
+    {
+      "epoch": 1.8521392152516072,
+      "grad_norm": 0.3854922652244568,
+      "learning_rate": 3.3668539314635428e-06,
+      "loss": 0.0024,
+      "step": 2088
+    },
+    {
+      "epoch": 1.8530259365994235,
+      "grad_norm": 0.40729087591171265,
+      "learning_rate": 3.3623297123092007e-06,
+      "loss": 0.0042,
+      "step": 2089
+    },
+    {
+      "epoch": 1.85391265794724,
+      "grad_norm": 0.05538596957921982,
+      "learning_rate": 3.357806994614703e-06,
+      "loss": 0.0003,
+      "step": 2090
+    },
+    {
+      "epoch": 1.8547993792950566,
+      "grad_norm": 0.6646614670753479,
+      "learning_rate": 3.353285782526596e-06,
+      "loss": 0.005,
+      "step": 2091
+    },
+    {
+      "epoch": 1.855686100642873,
+      "grad_norm": 0.0477665476500988,
+      "learning_rate": 3.348766080190046e-06,
+      "loss": 0.0001,
+      "step": 2092
+    },
+    {
+      "epoch": 1.8565728219906894,
+      "grad_norm": 0.1342734843492508,
+      "learning_rate": 3.3442478917488384e-06,
+      "loss": 0.0009,
+      "step": 2093
+    },
+    {
+      "epoch": 1.8574595433385057,
+      "grad_norm": 0.36796683073043823,
+      "learning_rate": 3.33973122134537e-06,
+      "loss": 0.0031,
+      "step": 2094
+    },
+    {
+      "epoch": 1.8583462646863222,
+      "grad_norm": 0.5031121969223022,
+      "learning_rate": 3.335216073120643e-06,
+      "loss": 0.0021,
+      "step": 2095
+    },
+    {
+      "epoch": 1.8592329860341388,
+      "grad_norm": 0.4082201421260834,
+      "learning_rate": 3.330702451214266e-06,
+      "loss": 0.0012,
+      "step": 2096
+    },
+    {
+      "epoch": 1.8601197073819553,
+      "grad_norm": 0.6267004013061523,
+      "learning_rate": 3.3261903597644497e-06,
+      "loss": 0.0019,
+      "step": 2097
+    },
+    {
+      "epoch": 1.8610064287297716,
+      "grad_norm": 0.12733691930770874,
+      "learning_rate": 3.321679802907997e-06,
+      "loss": 0.0004,
+      "step": 2098
+    },
+    {
+      "epoch": 1.8618931500775882,
+      "grad_norm": 0.030501484870910645,
+      "learning_rate": 3.31717078478031e-06,
+      "loss": 0.0001,
+      "step": 2099
+    },
+    {
+      "epoch": 1.8627798714254045,
+      "grad_norm": 0.41649121046066284,
+      "learning_rate": 3.3126633095153745e-06,
+      "loss": 0.003,
+      "step": 2100
+    },
+    {
+      "epoch": 1.863666592773221,
+      "grad_norm": 0.25532305240631104,
+      "learning_rate": 3.308157381245767e-06,
+      "loss": 0.0014,
+      "step": 2101
+    },
+    {
+      "epoch": 1.8645533141210375,
+      "grad_norm": 0.020190473645925522,
+      "learning_rate": 3.3036530041026405e-06,
+      "loss": 0.0001,
+      "step": 2102
+    },
+    {
+      "epoch": 1.865440035468854,
+      "grad_norm": 0.28363004326820374,
+      "learning_rate": 3.2991501822157266e-06,
+      "loss": 0.0005,
+      "step": 2103
+    },
+    {
+      "epoch": 1.8663267568166704,
+      "grad_norm": 0.008495116606354713,
+      "learning_rate": 3.2946489197133367e-06,
+      "loss": 0.0,
+      "step": 2104
+    },
+    {
+      "epoch": 1.8672134781644867,
+      "grad_norm": 0.5957363247871399,
+      "learning_rate": 3.2901492207223475e-06,
+      "loss": 0.0014,
+      "step": 2105
+    },
+    {
+      "epoch": 1.8681001995123032,
+      "grad_norm": 0.049815353006124496,
+      "learning_rate": 3.285651089368202e-06,
+      "loss": 0.0002,
+      "step": 2106
+    },
+    {
+      "epoch": 1.8689869208601197,
+      "grad_norm": 0.1587996631860733,
+      "learning_rate": 3.281154529774905e-06,
+      "loss": 0.001,
+      "step": 2107
+    },
+    {
+      "epoch": 1.8698736422079363,
+      "grad_norm": 0.058114007115364075,
+      "learning_rate": 3.276659546065027e-06,
+      "loss": 0.0002,
+      "step": 2108
+    },
+    {
+      "epoch": 1.8707603635557526,
+      "grad_norm": 0.010123222135007381,
+      "learning_rate": 3.2721661423596866e-06,
+      "loss": 0.0,
+      "step": 2109
+    },
+    {
+      "epoch": 1.871647084903569,
+      "grad_norm": 0.038756851106882095,
+      "learning_rate": 3.2676743227785545e-06,
+      "loss": 0.0002,
+      "step": 2110
+    },
+    {
+      "epoch": 1.8725338062513854,
+      "grad_norm": 0.0023471799213439226,
+      "learning_rate": 3.263184091439855e-06,
+      "loss": 0.0,
+      "step": 2111
+    },
+    {
+      "epoch": 1.873420527599202,
+      "grad_norm": 0.5632521510124207,
+      "learning_rate": 3.25869545246035e-06,
+      "loss": 0.0043,
+      "step": 2112
+    },
+    {
+      "epoch": 1.8743072489470185,
+      "grad_norm": 0.0417516715824604,
+      "learning_rate": 3.2542084099553426e-06,
+      "loss": 0.0002,
+      "step": 2113
+    },
+    {
+      "epoch": 1.8751939702948348,
+      "grad_norm": 0.30594179034233093,
+      "learning_rate": 3.249722968038673e-06,
+      "loss": 0.0008,
+      "step": 2114
+    },
+    {
+      "epoch": 1.8760806916426513,
+      "grad_norm": 0.009588775224983692,
+      "learning_rate": 3.245239130822716e-06,
+      "loss": 0.0001,
+      "step": 2115
+    },
+    {
+      "epoch": 1.8769674129904677,
+      "grad_norm": 0.3712175190448761,
+      "learning_rate": 3.2407569024183734e-06,
+      "loss": 0.0059,
+      "step": 2116
+    },
+    {
+      "epoch": 1.8778541343382842,
+      "grad_norm": 0.48841285705566406,
+      "learning_rate": 3.2362762869350685e-06,
+      "loss": 0.0025,
+      "step": 2117
+    },
+    {
+      "epoch": 1.8787408556861007,
+      "grad_norm": 0.19866853952407837,
+      "learning_rate": 3.231797288480753e-06,
+      "loss": 0.001,
+      "step": 2118
+    },
+    {
+      "epoch": 1.8796275770339173,
+      "grad_norm": 0.017454639077186584,
+      "learning_rate": 3.227319911161891e-06,
+      "loss": 0.0001,
+      "step": 2119
+    },
+    {
+      "epoch": 1.8805142983817336,
+      "grad_norm": 0.03350000083446503,
+      "learning_rate": 3.2228441590834608e-06,
+      "loss": 0.0002,
+      "step": 2120
+    },
+    {
+      "epoch": 1.8814010197295499,
+      "grad_norm": 1.9975967407226562,
+      "learning_rate": 3.218370036348952e-06,
+      "loss": 0.0062,
+      "step": 2121
+    },
+    {
+      "epoch": 1.8822877410773664,
+      "grad_norm": 0.011665020138025284,
+      "learning_rate": 3.21389754706036e-06,
+      "loss": 0.0,
+      "step": 2122
+    },
+    {
+      "epoch": 1.883174462425183,
+      "grad_norm": 0.7389943599700928,
+      "learning_rate": 3.2094266953181817e-06,
+      "loss": 0.0015,
+      "step": 2123
+    },
+    {
+      "epoch": 1.8840611837729995,
+      "grad_norm": 0.293914258480072,
+      "learning_rate": 3.204957485221413e-06,
+      "loss": 0.001,
+      "step": 2124
+    },
+    {
+      "epoch": 1.8849479051208158,
+      "grad_norm": 0.09404734522104263,
+      "learning_rate": 3.2004899208675468e-06,
+      "loss": 0.0004,
+      "step": 2125
+    },
+    {
+      "epoch": 1.885834626468632,
+      "grad_norm": 0.5290479063987732,
+      "learning_rate": 3.1960240063525648e-06,
+      "loss": 0.0008,
+      "step": 2126
+    },
+    {
+      "epoch": 1.8867213478164486,
+      "grad_norm": 0.06976180523633957,
+      "learning_rate": 3.1915597457709364e-06,
+      "loss": 0.0003,
+      "step": 2127
+    },
+    {
+      "epoch": 1.8876080691642652,
+      "grad_norm": 0.2480458766222,
+      "learning_rate": 3.187097143215612e-06,
+      "loss": 0.0029,
+      "step": 2128
+    },
+    {
+      "epoch": 1.8884947905120817,
+      "grad_norm": 0.45888465642929077,
+      "learning_rate": 3.18263620277803e-06,
+      "loss": 0.0034,
+      "step": 2129
+    },
+    {
+      "epoch": 1.889381511859898,
+      "grad_norm": 0.4611452519893646,
+      "learning_rate": 3.1781769285480977e-06,
+      "loss": 0.0034,
+      "step": 2130
+    },
+    {
+      "epoch": 1.8902682332077143,
+      "grad_norm": 0.011414594016969204,
+      "learning_rate": 3.1737193246141933e-06,
+      "loss": 0.0001,
+      "step": 2131
+    },
+    {
+      "epoch": 1.8911549545555308,
+      "grad_norm": 0.008071592077612877,
+      "learning_rate": 3.1692633950631723e-06,
+      "loss": 0.0001,
+      "step": 2132
+    },
+    {
+      "epoch": 1.8920416759033474,
+      "grad_norm": 0.09630648791790009,
+      "learning_rate": 3.164809143980348e-06,
+      "loss": 0.0005,
+      "step": 2133
+    },
+    {
+      "epoch": 1.892928397251164,
+      "grad_norm": 0.03766035661101341,
+      "learning_rate": 3.160356575449496e-06,
+      "loss": 0.0002,
+      "step": 2134
+    },
+    {
+      "epoch": 1.8938151185989802,
+      "grad_norm": 0.011658454313874245,
+      "learning_rate": 3.1559056935528486e-06,
+      "loss": 0.0001,
+      "step": 2135
+    },
+    {
+      "epoch": 1.8947018399467968,
+      "grad_norm": 0.24089200794696808,
+      "learning_rate": 3.1514565023710964e-06,
+      "loss": 0.0032,
+      "step": 2136
+    },
+    {
+      "epoch": 1.895588561294613,
+      "grad_norm": 0.134489506483078,
+      "learning_rate": 3.1470090059833752e-06,
+      "loss": 0.0003,
+      "step": 2137
+    },
+    {
+      "epoch": 1.8964752826424296,
+      "grad_norm": 0.2924594283103943,
+      "learning_rate": 3.1425632084672663e-06,
+      "loss": 0.0011,
+      "step": 2138
+    },
+    {
+      "epoch": 1.8973620039902461,
+      "grad_norm": 0.14403395354747772,
+      "learning_rate": 3.1381191138988e-06,
+      "loss": 0.0005,
+      "step": 2139
+    },
+    {
+      "epoch": 1.8982487253380627,
+      "grad_norm": 0.24188585579395294,
+      "learning_rate": 3.133676726352438e-06,
+      "loss": 0.0015,
+      "step": 2140
+    },
+    {
+      "epoch": 1.899135446685879,
+      "grad_norm": 0.31872034072875977,
+      "learning_rate": 3.129236049901081e-06,
+      "loss": 0.0027,
+      "step": 2141
+    },
+    {
+      "epoch": 1.9000221680336953,
+      "grad_norm": 0.2157914638519287,
+      "learning_rate": 3.124797088616056e-06,
+      "loss": 0.0016,
+      "step": 2142
+    },
+    {
+      "epoch": 1.9009088893815118,
+      "grad_norm": 0.24194389581680298,
+      "learning_rate": 3.120359846567127e-06,
+      "loss": 0.0004,
+      "step": 2143
+    },
+    {
+      "epoch": 1.9017956107293283,
+      "grad_norm": 0.052015915513038635,
+      "learning_rate": 3.1159243278224727e-06,
+      "loss": 0.0003,
+      "step": 2144
+    },
+    {
+      "epoch": 1.9026823320771449,
+      "grad_norm": 0.07202459126710892,
+      "learning_rate": 3.1114905364486934e-06,
+      "loss": 0.0005,
+      "step": 2145
+    },
+    {
+      "epoch": 1.9035690534249612,
+      "grad_norm": 0.12261311709880829,
+      "learning_rate": 3.107058476510812e-06,
+      "loss": 0.0009,
+      "step": 2146
+    },
+    {
+      "epoch": 1.9044557747727775,
+      "grad_norm": 0.12206902354955673,
+      "learning_rate": 3.1026281520722555e-06,
+      "loss": 0.0004,
+      "step": 2147
+    },
+    {
+      "epoch": 1.905342496120594,
+      "grad_norm": 0.054110489785671234,
+      "learning_rate": 3.098199567194867e-06,
+      "loss": 0.0003,
+      "step": 2148
+    },
+    {
+      "epoch": 1.9062292174684106,
+      "grad_norm": 0.10819480568170547,
+      "learning_rate": 3.0937727259388877e-06,
+      "loss": 0.0005,
+      "step": 2149
+    },
+    {
+      "epoch": 1.907115938816227,
+      "grad_norm": 0.2881719470024109,
+      "learning_rate": 3.0893476323629657e-06,
+      "loss": 0.0014,
+      "step": 2150
+    },
+    {
+      "epoch": 1.9080026601640434,
+      "grad_norm": 0.5649845004081726,
+      "learning_rate": 3.0849242905241462e-06,
+      "loss": 0.0016,
+      "step": 2151
+    },
+    {
+      "epoch": 1.90888938151186,
+      "grad_norm": 0.09853377938270569,
+      "learning_rate": 3.0805027044778647e-06,
+      "loss": 0.0005,
+      "step": 2152
+    },
+    {
+      "epoch": 1.9097761028596763,
+      "grad_norm": 0.25022897124290466,
+      "learning_rate": 3.0760828782779496e-06,
+      "loss": 0.001,
+      "step": 2153
+    },
+    {
+      "epoch": 1.9106628242074928,
+      "grad_norm": 0.47225239872932434,
+      "learning_rate": 3.071664815976618e-06,
+      "loss": 0.003,
+      "step": 2154
+    },
+    {
+      "epoch": 1.9115495455553093,
+      "grad_norm": 0.6766409277915955,
+      "learning_rate": 3.067248521624465e-06,
+      "loss": 0.0027,
+      "step": 2155
+    },
+    {
+      "epoch": 1.9124362669031258,
+      "grad_norm": 0.1587907373905182,
+      "learning_rate": 3.062833999270467e-06,
+      "loss": 0.0005,
+      "step": 2156
+    },
+    {
+      "epoch": 1.9133229882509422,
+      "grad_norm": 0.01313844695687294,
+      "learning_rate": 3.0584212529619777e-06,
+      "loss": 0.0001,
+      "step": 2157
+    },
+    {
+      "epoch": 1.9142097095987585,
+      "grad_norm": 0.0655895546078682,
+      "learning_rate": 3.0540102867447204e-06,
+      "loss": 0.0002,
+      "step": 2158
+    },
+    {
+      "epoch": 1.915096430946575,
+      "grad_norm": 0.4105219841003418,
+      "learning_rate": 3.049601104662784e-06,
+      "loss": 0.0025,
+      "step": 2159
+    },
+    {
+      "epoch": 1.9159831522943915,
+      "grad_norm": 0.0187697596848011,
+      "learning_rate": 3.045193710758628e-06,
+      "loss": 0.0001,
+      "step": 2160
+    },
+    {
+      "epoch": 1.916869873642208,
+      "grad_norm": 0.01989288628101349,
+      "learning_rate": 3.0407881090730677e-06,
+      "loss": 0.0001,
+      "step": 2161
+    },
+    {
+      "epoch": 1.9177565949900244,
+      "grad_norm": 0.01322174072265625,
+      "learning_rate": 3.0363843036452762e-06,
+      "loss": 0.0001,
+      "step": 2162
+    },
+    {
+      "epoch": 1.9186433163378407,
+      "grad_norm": 0.03816716745495796,
+      "learning_rate": 3.031982298512777e-06,
+      "loss": 0.0002,
+      "step": 2163
+    },
+    {
+      "epoch": 1.9195300376856572,
+      "grad_norm": 0.4447609484195709,
+      "learning_rate": 3.027582097711451e-06,
+      "loss": 0.0022,
+      "step": 2164
+    },
+    {
+      "epoch": 1.9204167590334738,
+      "grad_norm": 0.10088513046503067,
+      "learning_rate": 3.0231837052755176e-06,
+      "loss": 0.0004,
+      "step": 2165
+    },
+    {
+      "epoch": 1.9213034803812903,
+      "grad_norm": 0.3154107928276062,
+      "learning_rate": 3.0187871252375394e-06,
+      "loss": 0.006,
+      "step": 2166
+    },
+    {
+      "epoch": 1.9221902017291066,
+      "grad_norm": 0.03484093025326729,
+      "learning_rate": 3.0143923616284176e-06,
+      "loss": 0.0002,
+      "step": 2167
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.062375616282224655,
+      "learning_rate": 3.0099994184773918e-06,
+      "loss": 0.0003,
+      "step": 2168
+    },
+    {
+      "epoch": 1.9239636444247394,
+      "grad_norm": 0.1312718689441681,
+      "learning_rate": 3.0056082998120283e-06,
+      "loss": 0.0005,
+      "step": 2169
+    },
+    {
+      "epoch": 1.924850365772556,
+      "grad_norm": 0.030425092205405235,
+      "learning_rate": 3.0012190096582195e-06,
+      "loss": 0.0001,
+      "step": 2170
+    },
+    {
+      "epoch": 1.9257370871203725,
+      "grad_norm": 0.022313162684440613,
+      "learning_rate": 2.996831552040188e-06,
+      "loss": 0.0001,
+      "step": 2171
+    },
+    {
+      "epoch": 1.926623808468189,
+      "grad_norm": 0.49964627623558044,
+      "learning_rate": 2.9924459309804693e-06,
+      "loss": 0.0058,
+      "step": 2172
+    },
+    {
+      "epoch": 1.9275105298160053,
+      "grad_norm": 0.014140648767352104,
+      "learning_rate": 2.9880621504999196e-06,
+      "loss": 0.0001,
+      "step": 2173
+    },
+    {
+      "epoch": 1.9283972511638217,
+      "grad_norm": 0.29723238945007324,
+      "learning_rate": 2.9836802146177034e-06,
+      "loss": 0.0012,
+      "step": 2174
+    },
+    {
+      "epoch": 1.9292839725116382,
+      "grad_norm": 0.04823289066553116,
+      "learning_rate": 2.979300127351299e-06,
+      "loss": 0.0001,
+      "step": 2175
+    },
+    {
+      "epoch": 1.9301706938594547,
+      "grad_norm": 0.8130165338516235,
+      "learning_rate": 2.9749218927164864e-06,
+      "loss": 0.0102,
+      "step": 2176
+    },
+    {
+      "epoch": 1.9310574152072713,
+      "grad_norm": 0.0337294302880764,
+      "learning_rate": 2.9705455147273467e-06,
+      "loss": 0.0002,
+      "step": 2177
+    },
+    {
+      "epoch": 1.9319441365550876,
+      "grad_norm": 0.25490546226501465,
+      "learning_rate": 2.9661709973962626e-06,
+      "loss": 0.0006,
+      "step": 2178
+    },
+    {
+      "epoch": 1.9328308579029039,
+      "grad_norm": 0.02805035188794136,
+      "learning_rate": 2.961798344733907e-06,
+      "loss": 0.0001,
+      "step": 2179
+    },
+    {
+      "epoch": 1.9337175792507204,
+      "grad_norm": 0.1288713812828064,
+      "learning_rate": 2.9574275607492454e-06,
+      "loss": 0.0006,
+      "step": 2180
+    },
+    {
+      "epoch": 1.934604300598537,
+      "grad_norm": 0.654469907283783,
+      "learning_rate": 2.9530586494495267e-06,
+      "loss": 0.0095,
+      "step": 2181
+    },
+    {
+      "epoch": 1.9354910219463535,
+      "grad_norm": 0.045253507792949677,
+      "learning_rate": 2.948691614840289e-06,
+      "loss": 0.0002,
+      "step": 2182
+    },
+    {
+      "epoch": 1.9363777432941698,
+      "grad_norm": 0.0367266871035099,
+      "learning_rate": 2.9443264609253443e-06,
+      "loss": 0.0003,
+      "step": 2183
+    },
+    {
+      "epoch": 1.937264464641986,
+      "grad_norm": 0.014604181051254272,
+      "learning_rate": 2.9399631917067806e-06,
+      "loss": 0.0001,
+      "step": 2184
+    },
+    {
+      "epoch": 1.9381511859898026,
+      "grad_norm": 0.02937917225062847,
+      "learning_rate": 2.935601811184964e-06,
+      "loss": 0.0001,
+      "step": 2185
+    },
+    {
+      "epoch": 1.9390379073376192,
+      "grad_norm": 0.01375783421099186,
+      "learning_rate": 2.9312423233585214e-06,
+      "loss": 0.0001,
+      "step": 2186
+    },
+    {
+      "epoch": 1.9399246286854357,
+      "grad_norm": 0.25753626227378845,
+      "learning_rate": 2.926884732224349e-06,
+      "loss": 0.0019,
+      "step": 2187
+    },
+    {
+      "epoch": 1.940811350033252,
+      "grad_norm": 0.09537655860185623,
+      "learning_rate": 2.922529041777599e-06,
+      "loss": 0.0003,
+      "step": 2188
+    },
+    {
+      "epoch": 1.9416980713810685,
+      "grad_norm": 0.1624821424484253,
+      "learning_rate": 2.9181752560116894e-06,
+      "loss": 0.0012,
+      "step": 2189
+    },
+    {
+      "epoch": 1.9425847927288848,
+      "grad_norm": 0.021007006987929344,
+      "learning_rate": 2.913823378918287e-06,
+      "loss": 0.0001,
+      "step": 2190
+    },
+    {
+      "epoch": 1.9434715140767014,
+      "grad_norm": 0.268496572971344,
+      "learning_rate": 2.9094734144873037e-06,
+      "loss": 0.0018,
+      "step": 2191
+    },
+    {
+      "epoch": 1.944358235424518,
+      "grad_norm": 0.04218098148703575,
+      "learning_rate": 2.9051253667069107e-06,
+      "loss": 0.0002,
+      "step": 2192
+    },
+    {
+      "epoch": 1.9452449567723344,
+      "grad_norm": 0.07987596094608307,
+      "learning_rate": 2.900779239563508e-06,
+      "loss": 0.0003,
+      "step": 2193
+    },
+    {
+      "epoch": 1.9461316781201508,
+      "grad_norm": 0.25981444120407104,
+      "learning_rate": 2.896435037041746e-06,
+      "loss": 0.0022,
+      "step": 2194
+    },
+    {
+      "epoch": 1.947018399467967,
+      "grad_norm": 0.2508253753185272,
+      "learning_rate": 2.892092763124499e-06,
+      "loss": 0.0014,
+      "step": 2195
+    },
+    {
+      "epoch": 1.9479051208157836,
+      "grad_norm": 0.2507166862487793,
+      "learning_rate": 2.8877524217928866e-06,
+      "loss": 0.0016,
+      "step": 2196
+    },
+    {
+      "epoch": 1.9487918421636001,
+      "grad_norm": 0.15352725982666016,
+      "learning_rate": 2.8834140170262446e-06,
+      "loss": 0.0013,
+      "step": 2197
+    },
+    {
+      "epoch": 1.9496785635114167,
+      "grad_norm": 0.021937541663646698,
+      "learning_rate": 2.8790775528021415e-06,
+      "loss": 0.0002,
+      "step": 2198
+    },
+    {
+      "epoch": 1.950565284859233,
+      "grad_norm": 0.12473490834236145,
+      "learning_rate": 2.874743033096361e-06,
+      "loss": 0.0006,
+      "step": 2199
+    },
+    {
+      "epoch": 1.9514520062070493,
+      "grad_norm": 0.3052102327346802,
+      "learning_rate": 2.8704104618829105e-06,
+      "loss": 0.0024,
+      "step": 2200
+    },
+    {
+      "epoch": 1.9523387275548658,
+      "grad_norm": 0.21318025887012482,
+      "learning_rate": 2.866079843134002e-06,
+      "loss": 0.0013,
+      "step": 2201
+    },
+    {
+      "epoch": 1.9532254489026823,
+      "grad_norm": 0.02442754991352558,
+      "learning_rate": 2.8617511808200645e-06,
+      "loss": 0.0001,
+      "step": 2202
+    },
+    {
+      "epoch": 1.9541121702504989,
+      "grad_norm": 0.30121952295303345,
+      "learning_rate": 2.857424478909731e-06,
+      "loss": 0.0004,
+      "step": 2203
+    },
+    {
+      "epoch": 1.9549988915983152,
+      "grad_norm": 0.1846991777420044,
+      "learning_rate": 2.85309974136984e-06,
+      "loss": 0.0011,
+      "step": 2204
+    },
+    {
+      "epoch": 1.9558856129461317,
+      "grad_norm": 0.5073151588439941,
+      "learning_rate": 2.8487769721654197e-06,
+      "loss": 0.0019,
+      "step": 2205
+    },
+    {
+      "epoch": 1.956772334293948,
+      "grad_norm": 0.014602754265069962,
+      "learning_rate": 2.8444561752597088e-06,
+      "loss": 0.0001,
+      "step": 2206
+    },
+    {
+      "epoch": 1.9576590556417646,
+      "grad_norm": 0.3761642873287201,
+      "learning_rate": 2.840137354614122e-06,
+      "loss": 0.0011,
+      "step": 2207
+    },
+    {
+      "epoch": 1.958545776989581,
+      "grad_norm": 0.044096872210502625,
+      "learning_rate": 2.8358205141882735e-06,
+      "loss": 0.0003,
+      "step": 2208
+    },
+    {
+      "epoch": 1.9594324983373976,
+      "grad_norm": 0.026059670373797417,
+      "learning_rate": 2.8315056579399563e-06,
+      "loss": 0.0002,
+      "step": 2209
+    },
+    {
+      "epoch": 1.960319219685214,
+      "grad_norm": 0.07342034578323364,
+      "learning_rate": 2.8271927898251484e-06,
+      "loss": 0.0002,
+      "step": 2210
+    },
+    {
+      "epoch": 1.9612059410330303,
+      "grad_norm": 0.34168630838394165,
+      "learning_rate": 2.822881913797998e-06,
+      "loss": 0.0018,
+      "step": 2211
+    },
+    {
+      "epoch": 1.9620926623808468,
+      "grad_norm": 0.15169137716293335,
+      "learning_rate": 2.8185730338108336e-06,
+      "loss": 0.0007,
+      "step": 2212
+    },
+    {
+      "epoch": 1.9629793837286633,
+      "grad_norm": 0.09074265509843826,
+      "learning_rate": 2.8142661538141514e-06,
+      "loss": 0.0004,
+      "step": 2213
+    },
+    {
+      "epoch": 1.9638661050764799,
+      "grad_norm": 0.03273439034819603,
+      "learning_rate": 2.809961277756614e-06,
+      "loss": 0.0001,
+      "step": 2214
+    },
+    {
+      "epoch": 1.9647528264242962,
+      "grad_norm": 0.048764728009700775,
+      "learning_rate": 2.8056584095850494e-06,
+      "loss": 0.0003,
+      "step": 2215
+    },
+    {
+      "epoch": 1.9656395477721125,
+      "grad_norm": 0.002200677525252104,
+      "learning_rate": 2.801357553244434e-06,
+      "loss": 0.0,
+      "step": 2216
+    },
+    {
+      "epoch": 1.966526269119929,
+      "grad_norm": 0.004921832587569952,
+      "learning_rate": 2.7970587126779173e-06,
+      "loss": 0.0,
+      "step": 2217
+    },
+    {
+      "epoch": 1.9674129904677455,
+      "grad_norm": 0.4720174968242645,
+      "learning_rate": 2.792761891826785e-06,
+      "loss": 0.0042,
+      "step": 2218
+    },
+    {
+      "epoch": 1.968299711815562,
+      "grad_norm": 0.008707297965884209,
+      "learning_rate": 2.788467094630479e-06,
+      "loss": 0.0,
+      "step": 2219
+    },
+    {
+      "epoch": 1.9691864331633784,
+      "grad_norm": 0.006691179238259792,
+      "learning_rate": 2.7841743250265844e-06,
+      "loss": 0.0,
+      "step": 2220
+    },
+    {
+      "epoch": 1.9700731545111947,
+      "grad_norm": 0.008885095827281475,
+      "learning_rate": 2.779883586950828e-06,
+      "loss": 0.0,
+      "step": 2221
+    },
+    {
+      "epoch": 1.9709598758590112,
+      "grad_norm": 0.11954095214605331,
+      "learning_rate": 2.7755948843370706e-06,
+      "loss": 0.0006,
+      "step": 2222
+    },
+    {
+      "epoch": 1.9718465972068278,
+      "grad_norm": 0.0582672655582428,
+      "learning_rate": 2.771308221117309e-06,
+      "loss": 0.0001,
+      "step": 2223
+    },
+    {
+      "epoch": 1.9727333185546443,
+      "grad_norm": 0.005173410754650831,
+      "learning_rate": 2.7670236012216722e-06,
+      "loss": 0.0,
+      "step": 2224
+    },
+    {
+      "epoch": 1.9736200399024606,
+      "grad_norm": 0.0285594891756773,
+      "learning_rate": 2.7627410285784164e-06,
+      "loss": 0.0001,
+      "step": 2225
+    },
+    {
+      "epoch": 1.9745067612502771,
+      "grad_norm": 0.0014770013513043523,
+      "learning_rate": 2.75846050711391e-06,
+      "loss": 0.0,
+      "step": 2226
+    },
+    {
+      "epoch": 1.9753934825980934,
+      "grad_norm": 0.299988329410553,
+      "learning_rate": 2.754182040752661e-06,
+      "loss": 0.0012,
+      "step": 2227
+    },
+    {
+      "epoch": 1.97628020394591,
+      "grad_norm": 0.20655956864356995,
+      "learning_rate": 2.749905633417273e-06,
+      "loss": 0.0015,
+      "step": 2228
+    },
+    {
+      "epoch": 1.9771669252937265,
+      "grad_norm": 0.07857910543680191,
+      "learning_rate": 2.7456312890284755e-06,
+      "loss": 0.0002,
+      "step": 2229
+    },
+    {
+      "epoch": 1.978053646641543,
+      "grad_norm": 0.029224885627627373,
+      "learning_rate": 2.741359011505095e-06,
+      "loss": 0.0001,
+      "step": 2230
+    },
+    {
+      "epoch": 1.9789403679893594,
+      "grad_norm": 0.22522462904453278,
+      "learning_rate": 2.737088804764077e-06,
+      "loss": 0.0007,
+      "step": 2231
+    },
+    {
+      "epoch": 1.9798270893371757,
+      "grad_norm": 0.01182663906365633,
+      "learning_rate": 2.7328206727204565e-06,
+      "loss": 0.0001,
+      "step": 2232
+    },
+    {
+      "epoch": 1.9807138106849922,
+      "grad_norm": 0.06838423013687134,
+      "learning_rate": 2.7285546192873723e-06,
+      "loss": 0.0003,
+      "step": 2233
+    },
+    {
+      "epoch": 1.9816005320328087,
+      "grad_norm": 0.6740751266479492,
+      "learning_rate": 2.7242906483760562e-06,
+      "loss": 0.0011,
+      "step": 2234
+    },
+    {
+      "epoch": 1.9824872533806253,
+      "grad_norm": 0.061367567628622055,
+      "learning_rate": 2.7200287638958323e-06,
+      "loss": 0.0002,
+      "step": 2235
+    },
+    {
+      "epoch": 1.9833739747284416,
+      "grad_norm": 0.29687654972076416,
+      "learning_rate": 2.7157689697541057e-06,
+      "loss": 0.0014,
+      "step": 2236
+    },
+    {
+      "epoch": 1.9842606960762579,
+      "grad_norm": 0.3020949363708496,
+      "learning_rate": 2.7115112698563727e-06,
+      "loss": 0.0029,
+      "step": 2237
+    },
+    {
+      "epoch": 1.9851474174240744,
+      "grad_norm": 0.021518951281905174,
+      "learning_rate": 2.7072556681062044e-06,
+      "loss": 0.0001,
+      "step": 2238
+    },
+    {
+      "epoch": 1.986034138771891,
+      "grad_norm": 0.15249262750148773,
+      "learning_rate": 2.703002168405252e-06,
+      "loss": 0.0004,
+      "step": 2239
+    },
+    {
+      "epoch": 1.9869208601197075,
+      "grad_norm": 0.0033653369173407555,
+      "learning_rate": 2.698750774653237e-06,
+      "loss": 0.0,
+      "step": 2240
+    },
+    {
+      "epoch": 1.9878075814675238,
+      "grad_norm": 0.21170556545257568,
+      "learning_rate": 2.6945014907479495e-06,
+      "loss": 0.0003,
+      "step": 2241
+    },
+    {
+      "epoch": 1.9886943028153403,
+      "grad_norm": 0.0033403278794139624,
+      "learning_rate": 2.6902543205852496e-06,
+      "loss": 0.0,
+      "step": 2242
+    },
+    {
+      "epoch": 1.9895810241631566,
+      "grad_norm": 0.052868060767650604,
+      "learning_rate": 2.686009268059052e-06,
+      "loss": 0.0001,
+      "step": 2243
+    },
+    {
+      "epoch": 1.9904677455109732,
+      "grad_norm": 0.057781487703323364,
+      "learning_rate": 2.6817663370613367e-06,
+      "loss": 0.0002,
+      "step": 2244
+    },
+    {
+      "epoch": 1.9913544668587897,
+      "grad_norm": 0.13221290707588196,
+      "learning_rate": 2.677525531482135e-06,
+      "loss": 0.0005,
+      "step": 2245
+    },
+    {
+      "epoch": 1.9922411882066062,
+      "grad_norm": 0.6088939309120178,
+      "learning_rate": 2.6732868552095337e-06,
+      "loss": 0.002,
+      "step": 2246
+    },
+    {
+      "epoch": 1.9931279095544225,
+      "grad_norm": 0.16569292545318604,
+      "learning_rate": 2.669050312129658e-06,
+      "loss": 0.0005,
+      "step": 2247
+    },
+    {
+      "epoch": 1.9940146309022388,
+      "grad_norm": 0.0092005031183362,
+      "learning_rate": 2.664815906126691e-06,
+      "loss": 0.0,
+      "step": 2248
+    },
+    {
+      "epoch": 1.9949013522500554,
+      "grad_norm": 0.015410387888550758,
+      "learning_rate": 2.6605836410828443e-06,
+      "loss": 0.0,
+      "step": 2249
+    },
+    {
+      "epoch": 1.995788073597872,
+      "grad_norm": 0.18164335191249847,
+      "learning_rate": 2.6563535208783753e-06,
+      "loss": 0.0007,
+      "step": 2250
+    },
+    {
+      "epoch": 1.9966747949456884,
+      "grad_norm": 0.3885621428489685,
+      "learning_rate": 2.652125549391565e-06,
+      "loss": 0.0045,
+      "step": 2251
+    },
+    {
+      "epoch": 1.9975615162935048,
+      "grad_norm": 0.009692193940281868,
+      "learning_rate": 2.6478997304987383e-06,
+      "loss": 0.0,
+      "step": 2252
+    },
+    {
+      "epoch": 1.998448237641321,
+      "grad_norm": 0.008395759388804436,
+      "learning_rate": 2.6436760680742325e-06,
+      "loss": 0.0,
+      "step": 2253
+    },
+    {
+      "epoch": 1.9993349589891376,
+      "grad_norm": 0.5207878351211548,
+      "learning_rate": 2.639454565990417e-06,
+      "loss": 0.0015,
+      "step": 2254
+    },
+    {
+      "epoch": 2.0008867213478165,
+      "grad_norm": 0.31940406560897827,
+      "learning_rate": 2.6352352281176763e-06,
+      "loss": 0.0038,
+      "step": 2255
+    },
+    {
+      "epoch": 2.001773442695633,
+      "grad_norm": 0.0426461435854435,
+      "learning_rate": 2.6310180583244157e-06,
+      "loss": 0.0002,
+      "step": 2256
+    },
+    {
+      "epoch": 2.001773442695633,
+      "eval_loss": 0.03277111053466797,
+      "eval_runtime": 63.8702,
+      "eval_samples_per_second": 3.006,
+      "eval_steps_per_second": 0.752,
+      "step": 2256
+    },
+    {
+      "epoch": 2.002660164043449,
+      "grad_norm": 0.3251824378967285,
+      "learning_rate": 2.6268030604770435e-06,
+      "loss": 0.0043,
+      "step": 2257
+    },
+    {
+      "epoch": 2.0035468853912657,
+      "grad_norm": 0.010555939748883247,
+      "learning_rate": 2.622590238439985e-06,
+      "loss": 0.0001,
+      "step": 2258
+    },
+    {
+      "epoch": 2.004433606739082,
+      "grad_norm": 0.003281393786892295,
+      "learning_rate": 2.618379596075668e-06,
+      "loss": 0.0,
+      "step": 2259
+    },
+    {
+      "epoch": 2.0053203280868988,
+      "grad_norm": 0.23340563476085663,
+      "learning_rate": 2.6141711372445216e-06,
+      "loss": 0.0008,
+      "step": 2260
+    },
+    {
+      "epoch": 2.0062070494347153,
+      "grad_norm": 0.07923705130815506,
+      "learning_rate": 2.609964865804974e-06,
+      "loss": 0.0001,
+      "step": 2261
+    },
+    {
+      "epoch": 2.0070937707825314,
+      "grad_norm": 0.531326174736023,
+      "learning_rate": 2.605760785613447e-06,
+      "loss": 0.0067,
+      "step": 2262
+    },
+    {
+      "epoch": 2.007980492130348,
+      "grad_norm": 0.07579508423805237,
+      "learning_rate": 2.6015589005243517e-06,
+      "loss": 0.0002,
+      "step": 2263
+    },
+    {
+      "epoch": 2.0088672134781644,
+      "grad_norm": 0.009762165136635303,
+      "learning_rate": 2.59735921439009e-06,
+      "loss": 0.0001,
+      "step": 2264
+    },
+    {
+      "epoch": 2.009753934825981,
+      "grad_norm": 0.03709012642502785,
+      "learning_rate": 2.593161731061046e-06,
+      "loss": 0.0002,
+      "step": 2265
+    },
+    {
+      "epoch": 2.0106406561737975,
+      "grad_norm": 0.0150918485596776,
+      "learning_rate": 2.588966454385585e-06,
+      "loss": 0.0001,
+      "step": 2266
+    },
+    {
+      "epoch": 2.011527377521614,
+      "grad_norm": 0.006500888615846634,
+      "learning_rate": 2.584773388210049e-06,
+      "loss": 0.0,
+      "step": 2267
+    },
+    {
+      "epoch": 2.01241409886943,
+      "grad_norm": 0.03855859860777855,
+      "learning_rate": 2.5805825363787496e-06,
+      "loss": 0.0001,
+      "step": 2268
+    },
+    {
+      "epoch": 2.0133008202172467,
+      "grad_norm": 0.0074251629412174225,
+      "learning_rate": 2.5763939027339775e-06,
+      "loss": 0.0,
+      "step": 2269
+    },
+    {
+      "epoch": 2.014187541565063,
+      "grad_norm": 0.022057898342609406,
+      "learning_rate": 2.572207491115979e-06,
+      "loss": 0.0001,
+      "step": 2270
+    },
+    {
+      "epoch": 2.0150742629128797,
+      "grad_norm": 0.08520907163619995,
+      "learning_rate": 2.568023305362971e-06,
+      "loss": 0.0003,
+      "step": 2271
+    },
+    {
+      "epoch": 2.0159609842606963,
+      "grad_norm": 0.0048646205104887486,
+      "learning_rate": 2.5638413493111227e-06,
+      "loss": 0.0,
+      "step": 2272
+    },
+    {
+      "epoch": 2.0168477056085123,
+      "grad_norm": 0.031922418624162674,
+      "learning_rate": 2.5596616267945683e-06,
+      "loss": 0.0002,
+      "step": 2273
+    },
+    {
+      "epoch": 2.017734426956329,
+      "grad_norm": 0.043785277754068375,
+      "learning_rate": 2.5554841416453856e-06,
+      "loss": 0.0001,
+      "step": 2274
+    },
+    {
+      "epoch": 2.0186211483041454,
+      "grad_norm": 0.00607112143188715,
+      "learning_rate": 2.551308897693607e-06,
+      "loss": 0.0,
+      "step": 2275
+    },
+    {
+      "epoch": 2.019507869651962,
+      "grad_norm": 0.0038311337120831013,
+      "learning_rate": 2.547135898767202e-06,
+      "loss": 0.0,
+      "step": 2276
+    },
+    {
+      "epoch": 2.0203945909997785,
+      "grad_norm": 0.023988040164113045,
+      "learning_rate": 2.542965148692095e-06,
+      "loss": 0.0001,
+      "step": 2277
+    },
+    {
+      "epoch": 2.0212813123475946,
+      "grad_norm": 0.37009167671203613,
+      "learning_rate": 2.5387966512921357e-06,
+      "loss": 0.0024,
+      "step": 2278
+    },
+    {
+      "epoch": 2.022168033695411,
+      "grad_norm": 0.0046903216280043125,
+      "learning_rate": 2.534630410389116e-06,
+      "loss": 0.0,
+      "step": 2279
+    },
+    {
+      "epoch": 2.0230547550432276,
+      "grad_norm": 0.032938845455646515,
+      "learning_rate": 2.530466429802756e-06,
+      "loss": 0.0001,
+      "step": 2280
+    },
+    {
+      "epoch": 2.023941476391044,
+      "grad_norm": 0.04838014394044876,
+      "learning_rate": 2.5263047133507067e-06,
+      "loss": 0.0002,
+      "step": 2281
+    },
+    {
+      "epoch": 2.0248281977388607,
+      "grad_norm": 0.018787918612360954,
+      "learning_rate": 2.5221452648485367e-06,
+      "loss": 0.0001,
+      "step": 2282
+    },
+    {
+      "epoch": 2.0257149190866772,
+      "grad_norm": 0.009505733847618103,
+      "learning_rate": 2.5179880881097406e-06,
+      "loss": 0.0001,
+      "step": 2283
+    },
+    {
+      "epoch": 2.0266016404344933,
+      "grad_norm": 0.010092353448271751,
+      "learning_rate": 2.513833186945731e-06,
+      "loss": 0.0,
+      "step": 2284
+    },
+    {
+      "epoch": 2.02748836178231,
+      "grad_norm": 0.521740198135376,
+      "learning_rate": 2.509680565165831e-06,
+      "loss": 0.0035,
+      "step": 2285
+    },
+    {
+      "epoch": 2.0283750831301264,
+      "grad_norm": 0.0034733442589640617,
+      "learning_rate": 2.505530226577275e-06,
+      "loss": 0.0,
+      "step": 2286
+    },
+    {
+      "epoch": 2.029261804477943,
+      "grad_norm": 0.043323416262865067,
+      "learning_rate": 2.5013821749852063e-06,
+      "loss": 0.0002,
+      "step": 2287
+    },
+    {
+      "epoch": 2.0301485258257594,
+      "grad_norm": 0.8648737668991089,
+      "learning_rate": 2.4972364141926663e-06,
+      "loss": 0.0036,
+      "step": 2288
+    },
+    {
+      "epoch": 2.0310352471735755,
+      "grad_norm": 0.24954581260681152,
+      "learning_rate": 2.4930929480006e-06,
+      "loss": 0.0037,
+      "step": 2289
+    },
+    {
+      "epoch": 2.031921968521392,
+      "grad_norm": 0.26250550150871277,
+      "learning_rate": 2.4889517802078477e-06,
+      "loss": 0.001,
+      "step": 2290
+    },
+    {
+      "epoch": 2.0328086898692086,
+      "grad_norm": 0.03862825408577919,
+      "learning_rate": 2.484812914611144e-06,
+      "loss": 0.0002,
+      "step": 2291
+    },
+    {
+      "epoch": 2.033695411217025,
+      "grad_norm": 0.04618274047970772,
+      "learning_rate": 2.4806763550051115e-06,
+      "loss": 0.0002,
+      "step": 2292
+    },
+    {
+      "epoch": 2.0345821325648417,
+      "grad_norm": 0.0045454599894583225,
+      "learning_rate": 2.476542105182254e-06,
+      "loss": 0.0,
+      "step": 2293
+    },
+    {
+      "epoch": 2.0354688539126577,
+      "grad_norm": 0.060077376663684845,
+      "learning_rate": 2.4724101689329694e-06,
+      "loss": 0.0002,
+      "step": 2294
+    },
+    {
+      "epoch": 2.0363555752604743,
+      "grad_norm": 0.016459325328469276,
+      "learning_rate": 2.4682805500455214e-06,
+      "loss": 0.0001,
+      "step": 2295
+    },
+    {
+      "epoch": 2.037242296608291,
+      "grad_norm": 0.03847726806998253,
+      "learning_rate": 2.464153252306059e-06,
+      "loss": 0.0001,
+      "step": 2296
+    },
+    {
+      "epoch": 2.0381290179561073,
+      "grad_norm": 0.14360575377941132,
+      "learning_rate": 2.4600282794985923e-06,
+      "loss": 0.0006,
+      "step": 2297
+    },
+    {
+      "epoch": 2.039015739303924,
+      "grad_norm": 0.07891727983951569,
+      "learning_rate": 2.4559056354050154e-06,
+      "loss": 0.0003,
+      "step": 2298
+    },
+    {
+      "epoch": 2.0399024606517404,
+      "grad_norm": 0.01991586573421955,
+      "learning_rate": 2.451785323805072e-06,
+      "loss": 0.0001,
+      "step": 2299
+    },
+    {
+      "epoch": 2.0407891819995565,
+      "grad_norm": 0.32266005873680115,
+      "learning_rate": 2.4476673484763756e-06,
+      "loss": 0.0011,
+      "step": 2300
+    },
+    {
+      "epoch": 2.041675903347373,
+      "grad_norm": 0.0066023049876093864,
+      "learning_rate": 2.4435517131943965e-06,
+      "loss": 0.0,
+      "step": 2301
+    },
+    {
+      "epoch": 2.0425626246951896,
+      "grad_norm": 0.9420242309570312,
+      "learning_rate": 2.43943842173246e-06,
+      "loss": 0.0041,
+      "step": 2302
+    },
+    {
+      "epoch": 2.043449346043006,
+      "grad_norm": 0.09942077100276947,
+      "learning_rate": 2.435327477861739e-06,
+      "loss": 0.0006,
+      "step": 2303
+    },
+    {
+      "epoch": 2.0443360673908226,
+      "grad_norm": 0.12234442681074142,
+      "learning_rate": 2.431218885351257e-06,
+      "loss": 0.0005,
+      "step": 2304
+    },
+    {
+      "epoch": 2.0452227887386387,
+      "grad_norm": 0.19722875952720642,
+      "learning_rate": 2.4271126479678824e-06,
+      "loss": 0.0007,
+      "step": 2305
+    },
+    {
+      "epoch": 2.0461095100864553,
+      "grad_norm": 0.1235419362783432,
+      "learning_rate": 2.423008769476325e-06,
+      "loss": 0.0003,
+      "step": 2306
+    },
+    {
+      "epoch": 2.046996231434272,
+      "grad_norm": 0.012925482355058193,
+      "learning_rate": 2.418907253639124e-06,
+      "loss": 0.0001,
+      "step": 2307
+    },
+    {
+      "epoch": 2.0478829527820883,
+      "grad_norm": 0.012632215395569801,
+      "learning_rate": 2.4148081042166667e-06,
+      "loss": 0.0001,
+      "step": 2308
+    },
+    {
+      "epoch": 2.048769674129905,
+      "grad_norm": 0.003258196171373129,
+      "learning_rate": 2.4107113249671567e-06,
+      "loss": 0.0,
+      "step": 2309
+    },
+    {
+      "epoch": 2.049656395477721,
+      "grad_norm": 0.05988994985818863,
+      "learning_rate": 2.4066169196466326e-06,
+      "loss": 0.0003,
+      "step": 2310
+    },
+    {
+      "epoch": 2.0505431168255375,
+      "grad_norm": 0.003724791342392564,
+      "learning_rate": 2.4025248920089544e-06,
+      "loss": 0.0,
+      "step": 2311
+    },
+    {
+      "epoch": 2.051429838173354,
+      "grad_norm": 0.006751666311174631,
+      "learning_rate": 2.3984352458058045e-06,
+      "loss": 0.0,
+      "step": 2312
+    },
+    {
+      "epoch": 2.0523165595211705,
+      "grad_norm": 0.002435932168737054,
+      "learning_rate": 2.3943479847866764e-06,
+      "loss": 0.0,
+      "step": 2313
+    },
+    {
+      "epoch": 2.053203280868987,
+      "grad_norm": 0.02041725628077984,
+      "learning_rate": 2.3902631126988808e-06,
+      "loss": 0.0001,
+      "step": 2314
+    },
+    {
+      "epoch": 2.054090002216803,
+      "grad_norm": 0.004865054506808519,
+      "learning_rate": 2.3861806332875377e-06,
+      "loss": 0.0,
+      "step": 2315
+    },
+    {
+      "epoch": 2.0549767235646197,
+      "grad_norm": 0.060207922011613846,
+      "learning_rate": 2.382100550295574e-06,
+      "loss": 0.0002,
+      "step": 2316
+    },
+    {
+      "epoch": 2.055863444912436,
+      "grad_norm": 0.010751526802778244,
+      "learning_rate": 2.37802286746372e-06,
+      "loss": 0.0,
+      "step": 2317
+    },
+    {
+      "epoch": 2.0567501662602528,
+      "grad_norm": 0.014327171258628368,
+      "learning_rate": 2.373947588530499e-06,
+      "loss": 0.0001,
+      "step": 2318
+    },
+    {
+      "epoch": 2.0576368876080693,
+      "grad_norm": 0.006228189449757338,
+      "learning_rate": 2.369874717232242e-06,
+      "loss": 0.0,
+      "step": 2319
+    },
+    {
+      "epoch": 2.058523608955886,
+      "grad_norm": 0.2894449830055237,
+      "learning_rate": 2.365804257303062e-06,
+      "loss": 0.0045,
+      "step": 2320
+    },
+    {
+      "epoch": 2.059410330303702,
+      "grad_norm": 0.014009971171617508,
+      "learning_rate": 2.3617362124748664e-06,
+      "loss": 0.0001,
+      "step": 2321
+    },
+    {
+      "epoch": 2.0602970516515184,
+      "grad_norm": 0.014323338866233826,
+      "learning_rate": 2.3576705864773477e-06,
+      "loss": 0.0001,
+      "step": 2322
+    },
+    {
+      "epoch": 2.061183772999335,
+      "grad_norm": 0.0071457927115261555,
+      "learning_rate": 2.3536073830379814e-06,
+      "loss": 0.0,
+      "step": 2323
+    },
+    {
+      "epoch": 2.0620704943471515,
+      "grad_norm": 0.013069824315607548,
+      "learning_rate": 2.349546605882018e-06,
+      "loss": 0.0001,
+      "step": 2324
+    },
+    {
+      "epoch": 2.062957215694968,
+      "grad_norm": 0.07725962996482849,
+      "learning_rate": 2.345488258732488e-06,
+      "loss": 0.0004,
+      "step": 2325
+    },
+    {
+      "epoch": 2.063843937042784,
+      "grad_norm": 0.1443031281232834,
+      "learning_rate": 2.3414323453101924e-06,
+      "loss": 0.0003,
+      "step": 2326
+    },
+    {
+      "epoch": 2.0647306583906007,
+      "grad_norm": 0.009164338000118732,
+      "learning_rate": 2.3373788693337024e-06,
+      "loss": 0.0001,
+      "step": 2327
+    },
+    {
+      "epoch": 2.065617379738417,
+      "grad_norm": 0.016295766457915306,
+      "learning_rate": 2.333327834519348e-06,
+      "loss": 0.0001,
+      "step": 2328
+    },
+    {
+      "epoch": 2.0665041010862337,
+      "grad_norm": 0.10227362811565399,
+      "learning_rate": 2.3292792445812327e-06,
+      "loss": 0.0002,
+      "step": 2329
+    },
+    {
+      "epoch": 2.0673908224340503,
+      "grad_norm": 0.01212971843779087,
+      "learning_rate": 2.3252331032312075e-06,
+      "loss": 0.0,
+      "step": 2330
+    },
+    {
+      "epoch": 2.0682775437818663,
+      "grad_norm": 0.4432136118412018,
+      "learning_rate": 2.3211894141788856e-06,
+      "loss": 0.0078,
+      "step": 2331
+    },
+    {
+      "epoch": 2.069164265129683,
+      "grad_norm": 0.04203218221664429,
+      "learning_rate": 2.317148181131624e-06,
+      "loss": 0.0002,
+      "step": 2332
+    },
+    {
+      "epoch": 2.0700509864774994,
+      "grad_norm": 0.07127822935581207,
+      "learning_rate": 2.31310940779454e-06,
+      "loss": 0.0003,
+      "step": 2333
+    },
+    {
+      "epoch": 2.070937707825316,
+      "grad_norm": 0.3511061370372772,
+      "learning_rate": 2.3090730978704832e-06,
+      "loss": 0.0007,
+      "step": 2334
+    },
+    {
+      "epoch": 2.0718244291731325,
+      "grad_norm": 0.4882556200027466,
+      "learning_rate": 2.305039255060053e-06,
+      "loss": 0.0008,
+      "step": 2335
+    },
+    {
+      "epoch": 2.0727111505209486,
+      "grad_norm": 0.12589851021766663,
+      "learning_rate": 2.301007883061584e-06,
+      "loss": 0.0004,
+      "step": 2336
+    },
+    {
+      "epoch": 2.073597871868765,
+      "grad_norm": 0.02481425739824772,
+      "learning_rate": 2.296978985571145e-06,
+      "loss": 0.0002,
+      "step": 2337
+    },
+    {
+      "epoch": 2.0744845932165816,
+      "grad_norm": 0.012816986069083214,
+      "learning_rate": 2.2929525662825384e-06,
+      "loss": 0.0001,
+      "step": 2338
+    },
+    {
+      "epoch": 2.075371314564398,
+      "grad_norm": 0.02135675586760044,
+      "learning_rate": 2.28892862888729e-06,
+      "loss": 0.0001,
+      "step": 2339
+    },
+    {
+      "epoch": 2.0762580359122147,
+      "grad_norm": 0.00505865877494216,
+      "learning_rate": 2.284907177074655e-06,
+      "loss": 0.0,
+      "step": 2340
+    },
+    {
+      "epoch": 2.0771447572600312,
+      "grad_norm": 0.006438021082431078,
+      "learning_rate": 2.2808882145316075e-06,
+      "loss": 0.0,
+      "step": 2341
+    },
+    {
+      "epoch": 2.0780314786078473,
+      "grad_norm": 0.020976608619093895,
+      "learning_rate": 2.276871744942839e-06,
+      "loss": 0.0001,
+      "step": 2342
+    },
+    {
+      "epoch": 2.078918199955664,
+      "grad_norm": 0.06125025451183319,
+      "learning_rate": 2.2728577719907576e-06,
+      "loss": 0.0003,
+      "step": 2343
+    },
+    {
+      "epoch": 2.0798049213034804,
+      "grad_norm": 0.029785003513097763,
+      "learning_rate": 2.268846299355481e-06,
+      "loss": 0.0001,
+      "step": 2344
+    },
+    {
+      "epoch": 2.080691642651297,
+      "grad_norm": 0.002666131593286991,
+      "learning_rate": 2.2648373307148314e-06,
+      "loss": 0.0,
+      "step": 2345
+    },
+    {
+      "epoch": 2.0815783639991134,
+      "grad_norm": 0.31236812472343445,
+      "learning_rate": 2.2608308697443405e-06,
+      "loss": 0.0016,
+      "step": 2346
+    },
+    {
+      "epoch": 2.0824650853469295,
+      "grad_norm": 0.007688076235353947,
+      "learning_rate": 2.2568269201172382e-06,
+      "loss": 0.0001,
+      "step": 2347
+    },
+    {
+      "epoch": 2.083351806694746,
+      "grad_norm": 0.06622789800167084,
+      "learning_rate": 2.2528254855044546e-06,
+      "loss": 0.0002,
+      "step": 2348
+    },
+    {
+      "epoch": 2.0842385280425626,
+      "grad_norm": 0.18980400264263153,
+      "learning_rate": 2.248826569574606e-06,
+      "loss": 0.0019,
+      "step": 2349
+    },
+    {
+      "epoch": 2.085125249390379,
+      "grad_norm": 0.025342190638184547,
+      "learning_rate": 2.2448301759940126e-06,
+      "loss": 0.0001,
+      "step": 2350
+    },
+    {
+      "epoch": 2.0860119707381957,
+      "grad_norm": 0.025142235681414604,
+      "learning_rate": 2.2408363084266693e-06,
+      "loss": 0.0001,
+      "step": 2351
+    },
+    {
+      "epoch": 2.086898692086012,
+      "grad_norm": 0.3066583573818207,
+      "learning_rate": 2.236844970534265e-06,
+      "loss": 0.001,
+      "step": 2352
+    },
+    {
+      "epoch": 2.0877854134338283,
+      "grad_norm": 0.2630266547203064,
+      "learning_rate": 2.2328561659761576e-06,
+      "loss": 0.001,
+      "step": 2353
+    },
+    {
+      "epoch": 2.088672134781645,
+      "grad_norm": 0.006459026597440243,
+      "learning_rate": 2.228869898409399e-06,
+      "loss": 0.0,
+      "step": 2354
+    },
+    {
+      "epoch": 2.0895588561294614,
+      "grad_norm": 0.02460658550262451,
+      "learning_rate": 2.224886171488698e-06,
+      "loss": 0.0001,
+      "step": 2355
+    },
+    {
+      "epoch": 2.090445577477278,
+      "grad_norm": 0.00549986120313406,
+      "learning_rate": 2.2209049888664453e-06,
+      "loss": 0.0,
+      "step": 2356
+    },
+    {
+      "epoch": 2.0913322988250944,
+      "grad_norm": 0.035316918045282364,
+      "learning_rate": 2.216926354192695e-06,
+      "loss": 0.0001,
+      "step": 2357
+    },
+    {
+      "epoch": 2.0922190201729105,
+      "grad_norm": 0.026010386645793915,
+      "learning_rate": 2.212950271115167e-06,
+      "loss": 0.0001,
+      "step": 2358
+    },
+    {
+      "epoch": 2.093105741520727,
+      "grad_norm": 0.08127646893262863,
+      "learning_rate": 2.2089767432792375e-06,
+      "loss": 0.0001,
+      "step": 2359
+    },
+    {
+      "epoch": 2.0939924628685436,
+      "grad_norm": 0.11064977198839188,
+      "learning_rate": 2.205005774327944e-06,
+      "loss": 0.0003,
+      "step": 2360
+    },
+    {
+      "epoch": 2.09487918421636,
+      "grad_norm": 0.24248307943344116,
+      "learning_rate": 2.2010373679019773e-06,
+      "loss": 0.0005,
+      "step": 2361
+    },
+    {
+      "epoch": 2.0957659055641766,
+      "grad_norm": 0.037203628569841385,
+      "learning_rate": 2.1970715276396783e-06,
+      "loss": 0.0002,
+      "step": 2362
+    },
+    {
+      "epoch": 2.0966526269119927,
+      "grad_norm": 0.007871000096201897,
+      "learning_rate": 2.193108257177035e-06,
+      "loss": 0.0001,
+      "step": 2363
+    },
+    {
+      "epoch": 2.0975393482598093,
+      "grad_norm": 0.017412377521395683,
+      "learning_rate": 2.189147560147682e-06,
+      "loss": 0.0001,
+      "step": 2364
+    },
+    {
+      "epoch": 2.098426069607626,
+      "grad_norm": 0.008841924369335175,
+      "learning_rate": 2.185189440182887e-06,
+      "loss": 0.0,
+      "step": 2365
+    },
+    {
+      "epoch": 2.0993127909554423,
+      "grad_norm": 0.0007497490732930601,
+      "learning_rate": 2.181233900911564e-06,
+      "loss": 0.0,
+      "step": 2366
+    },
+    {
+      "epoch": 2.100199512303259,
+      "grad_norm": 0.010741816833615303,
+      "learning_rate": 2.177280945960255e-06,
+      "loss": 0.0,
+      "step": 2367
+    },
+    {
+      "epoch": 2.101086233651075,
+      "grad_norm": 0.010129451751708984,
+      "learning_rate": 2.1733305789531363e-06,
+      "loss": 0.0,
+      "step": 2368
+    },
+    {
+      "epoch": 2.1019729549988915,
+      "grad_norm": 0.03930528834462166,
+      "learning_rate": 2.16938280351201e-06,
+      "loss": 0.0001,
+      "step": 2369
+    },
+    {
+      "epoch": 2.102859676346708,
+      "grad_norm": 0.004623851738870144,
+      "learning_rate": 2.1654376232562985e-06,
+      "loss": 0.0,
+      "step": 2370
+    },
+    {
+      "epoch": 2.1037463976945245,
+      "grad_norm": 0.017552360892295837,
+      "learning_rate": 2.161495041803054e-06,
+      "loss": 0.0001,
+      "step": 2371
+    },
+    {
+      "epoch": 2.104633119042341,
+      "grad_norm": 0.23550549149513245,
+      "learning_rate": 2.1575550627669355e-06,
+      "loss": 0.0016,
+      "step": 2372
+    },
+    {
+      "epoch": 2.1055198403901576,
+      "grad_norm": 0.006746775936335325,
+      "learning_rate": 2.1536176897602245e-06,
+      "loss": 0.0,
+      "step": 2373
+    },
+    {
+      "epoch": 2.1064065617379737,
+      "grad_norm": 0.06297089904546738,
+      "learning_rate": 2.149682926392805e-06,
+      "loss": 0.0003,
+      "step": 2374
+    },
+    {
+      "epoch": 2.1072932830857902,
+      "grad_norm": 0.037527166306972504,
+      "learning_rate": 2.1457507762721792e-06,
+      "loss": 0.0001,
+      "step": 2375
+    },
+    {
+      "epoch": 2.1081800044336068,
+      "grad_norm": 0.17847929894924164,
+      "learning_rate": 2.141821243003443e-06,
+      "loss": 0.0003,
+      "step": 2376
+    },
+    {
+      "epoch": 2.1090667257814233,
+      "grad_norm": 0.04266642406582832,
+      "learning_rate": 2.1378943301893003e-06,
+      "loss": 0.0001,
+      "step": 2377
+    },
+    {
+      "epoch": 2.10995344712924,
+      "grad_norm": 0.22053800523281097,
+      "learning_rate": 2.133970041430044e-06,
+      "loss": 0.001,
+      "step": 2378
+    },
+    {
+      "epoch": 2.110840168477056,
+      "grad_norm": 0.0320192314684391,
+      "learning_rate": 2.130048380323575e-06,
+      "loss": 0.0001,
+      "step": 2379
+    },
+    {
+      "epoch": 2.1117268898248724,
+      "grad_norm": 0.0019540840294212103,
+      "learning_rate": 2.1261293504653707e-06,
+      "loss": 0.0,
+      "step": 2380
+    },
+    {
+      "epoch": 2.112613611172689,
+      "grad_norm": 0.0041855634190142155,
+      "learning_rate": 2.122212955448504e-06,
+      "loss": 0.0,
+      "step": 2381
+    },
+    {
+      "epoch": 2.1135003325205055,
+      "grad_norm": 0.028047163039445877,
+      "learning_rate": 2.118299198863631e-06,
+      "loss": 0.0001,
+      "step": 2382
+    },
+    {
+      "epoch": 2.114387053868322,
+      "grad_norm": 0.02840139903128147,
+      "learning_rate": 2.1143880842989895e-06,
+      "loss": 0.0001,
+      "step": 2383
+    },
+    {
+      "epoch": 2.115273775216138,
+      "grad_norm": 0.22898061573505402,
+      "learning_rate": 2.1104796153403906e-06,
+      "loss": 0.0012,
+      "step": 2384
+    },
+    {
+      "epoch": 2.1161604965639547,
+      "grad_norm": 0.0021610562689602375,
+      "learning_rate": 2.1065737955712244e-06,
+      "loss": 0.0,
+      "step": 2385
+    },
+    {
+      "epoch": 2.117047217911771,
+      "grad_norm": 0.003969789016991854,
+      "learning_rate": 2.102670628572451e-06,
+      "loss": 0.0,
+      "step": 2386
+    },
+    {
+      "epoch": 2.1179339392595877,
+      "grad_norm": 0.0022155435290187597,
+      "learning_rate": 2.0987701179225985e-06,
+      "loss": 0.0,
+      "step": 2387
+    },
+    {
+      "epoch": 2.1188206606074043,
+      "grad_norm": 0.008832978084683418,
+      "learning_rate": 2.0948722671977584e-06,
+      "loss": 0.0,
+      "step": 2388
+    },
+    {
+      "epoch": 2.1197073819552203,
+      "grad_norm": 0.027194837108254433,
+      "learning_rate": 2.0909770799715865e-06,
+      "loss": 0.0001,
+      "step": 2389
+    },
+    {
+      "epoch": 2.120594103303037,
+      "grad_norm": 0.0010967071866616607,
+      "learning_rate": 2.0870845598152907e-06,
+      "loss": 0.0,
+      "step": 2390
+    },
+    {
+      "epoch": 2.1214808246508534,
+      "grad_norm": 0.0053475042805075645,
+      "learning_rate": 2.0831947102976392e-06,
+      "loss": 0.0,
+      "step": 2391
+    },
+    {
+      "epoch": 2.12236754599867,
+      "grad_norm": 0.008405130356550217,
+      "learning_rate": 2.07930753498495e-06,
+      "loss": 0.0,
+      "step": 2392
+    },
+    {
+      "epoch": 2.1232542673464865,
+      "grad_norm": 0.005467743147164583,
+      "learning_rate": 2.0754230374410895e-06,
+      "loss": 0.0,
+      "step": 2393
+    },
+    {
+      "epoch": 2.124140988694303,
+      "grad_norm": 0.004836921580135822,
+      "learning_rate": 2.0715412212274698e-06,
+      "loss": 0.0,
+      "step": 2394
+    },
+    {
+      "epoch": 2.125027710042119,
+      "grad_norm": 0.001921136979945004,
+      "learning_rate": 2.0676620899030393e-06,
+      "loss": 0.0,
+      "step": 2395
+    },
+    {
+      "epoch": 2.1259144313899356,
+      "grad_norm": 0.054766010493040085,
+      "learning_rate": 2.063785647024295e-06,
+      "loss": 0.0002,
+      "step": 2396
+    },
+    {
+      "epoch": 2.126801152737752,
+      "grad_norm": 0.02580122835934162,
+      "learning_rate": 2.0599118961452593e-06,
+      "loss": 0.0001,
+      "step": 2397
+    },
+    {
+      "epoch": 2.1276878740855687,
+      "grad_norm": 0.01703774183988571,
+      "learning_rate": 2.0560408408174925e-06,
+      "loss": 0.0001,
+      "step": 2398
+    },
+    {
+      "epoch": 2.1285745954333852,
+      "grad_norm": 0.016674144193530083,
+      "learning_rate": 2.052172484590077e-06,
+      "loss": 0.0001,
+      "step": 2399
+    },
+    {
+      "epoch": 2.1294613167812013,
+      "grad_norm": 0.004239710979163647,
+      "learning_rate": 2.0483068310096327e-06,
+      "loss": 0.0,
+      "step": 2400
+    },
+    {
+      "epoch": 2.130348038129018,
+      "grad_norm": 0.019841166213154793,
+      "learning_rate": 2.0444438836202872e-06,
+      "loss": 0.0001,
+      "step": 2401
+    },
+    {
+      "epoch": 2.1312347594768344,
+      "grad_norm": 0.004029495641589165,
+      "learning_rate": 2.040583645963697e-06,
+      "loss": 0.0,
+      "step": 2402
+    },
+    {
+      "epoch": 2.132121480824651,
+      "grad_norm": 0.005214055068790913,
+      "learning_rate": 2.0367261215790306e-06,
+      "loss": 0.0,
+      "step": 2403
+    },
+    {
+      "epoch": 2.1330082021724674,
+      "grad_norm": 0.02571919932961464,
+      "learning_rate": 2.032871314002971e-06,
+      "loss": 0.0001,
+      "step": 2404
+    },
+    {
+      "epoch": 2.133894923520284,
+      "grad_norm": 0.01948658563196659,
+      "learning_rate": 2.0290192267697052e-06,
+      "loss": 0.0,
+      "step": 2405
+    },
+    {
+      "epoch": 2.1347816448681,
+      "grad_norm": 0.6353626251220703,
+      "learning_rate": 2.0251698634109314e-06,
+      "loss": 0.0012,
+      "step": 2406
+    },
+    {
+      "epoch": 2.1356683662159166,
+      "grad_norm": 0.024343444034457207,
+      "learning_rate": 2.0213232274558483e-06,
+      "loss": 0.0001,
+      "step": 2407
+    },
+    {
+      "epoch": 2.136555087563733,
+      "grad_norm": 0.023843051865696907,
+      "learning_rate": 2.0174793224311566e-06,
+      "loss": 0.0001,
+      "step": 2408
+    },
+    {
+      "epoch": 2.1374418089115497,
+      "grad_norm": 0.025098059326410294,
+      "learning_rate": 2.0136381518610455e-06,
+      "loss": 0.0,
+      "step": 2409
+    },
+    {
+      "epoch": 2.138328530259366,
+      "grad_norm": 0.00488665234297514,
+      "learning_rate": 2.0097997192672103e-06,
+      "loss": 0.0,
+      "step": 2410
+    },
+    {
+      "epoch": 2.1392152516071823,
+      "grad_norm": 0.003173389006406069,
+      "learning_rate": 2.005964028168823e-06,
+      "loss": 0.0,
+      "step": 2411
+    },
+    {
+      "epoch": 2.140101972954999,
+      "grad_norm": 0.0017301311017945409,
+      "learning_rate": 2.002131082082549e-06,
+      "loss": 0.0,
+      "step": 2412
+    },
+    {
+      "epoch": 2.1409886943028154,
+      "grad_norm": 0.4865742623806,
+      "learning_rate": 1.998300884522536e-06,
+      "loss": 0.001,
+      "step": 2413
+    },
+    {
+      "epoch": 2.141875415650632,
+      "grad_norm": 0.011351162567734718,
+      "learning_rate": 1.994473439000413e-06,
+      "loss": 0.0001,
+      "step": 2414
+    },
+    {
+      "epoch": 2.1427621369984484,
+      "grad_norm": 0.003956715576350689,
+      "learning_rate": 1.9906487490252807e-06,
+      "loss": 0.0,
+      "step": 2415
+    },
+    {
+      "epoch": 2.1436488583462645,
+      "grad_norm": 0.01261945441365242,
+      "learning_rate": 1.9868268181037186e-06,
+      "loss": 0.0,
+      "step": 2416
+    },
+    {
+      "epoch": 2.144535579694081,
+      "grad_norm": 0.0019356987904757261,
+      "learning_rate": 1.9830076497397753e-06,
+      "loss": 0.0,
+      "step": 2417
+    },
+    {
+      "epoch": 2.1454223010418976,
+      "grad_norm": 0.0014316060114651918,
+      "learning_rate": 1.979191247434966e-06,
+      "loss": 0.0,
+      "step": 2418
+    },
+    {
+      "epoch": 2.146309022389714,
+      "grad_norm": 0.0029126235749572515,
+      "learning_rate": 1.9753776146882724e-06,
+      "loss": 0.0,
+      "step": 2419
+    },
+    {
+      "epoch": 2.1471957437375306,
+      "grad_norm": 0.049275338649749756,
+      "learning_rate": 1.97156675499613e-06,
+      "loss": 0.0001,
+      "step": 2420
+    },
+    {
+      "epoch": 2.1480824650853467,
+      "grad_norm": 0.0023662822786718607,
+      "learning_rate": 1.9677586718524422e-06,
+      "loss": 0.0,
+      "step": 2421
+    },
+    {
+      "epoch": 2.1489691864331633,
+      "grad_norm": 0.001958158565685153,
+      "learning_rate": 1.963953368748557e-06,
+      "loss": 0.0,
+      "step": 2422
+    },
+    {
+      "epoch": 2.14985590778098,
+      "grad_norm": 0.07051583379507065,
+      "learning_rate": 1.960150849173279e-06,
+      "loss": 0.0002,
+      "step": 2423
+    },
+    {
+      "epoch": 2.1507426291287963,
+      "grad_norm": 0.005282245110720396,
+      "learning_rate": 1.9563511166128607e-06,
+      "loss": 0.0,
+      "step": 2424
+    },
+    {
+      "epoch": 2.151629350476613,
+      "grad_norm": 0.030105119571089745,
+      "learning_rate": 1.9525541745509997e-06,
+      "loss": 0.0001,
+      "step": 2425
+    },
+    {
+      "epoch": 2.1525160718244294,
+      "grad_norm": 0.001675124978646636,
+      "learning_rate": 1.94876002646883e-06,
+      "loss": 0.0,
+      "step": 2426
+    },
+    {
+      "epoch": 2.1534027931722455,
+      "grad_norm": 0.004662784747779369,
+      "learning_rate": 1.9449686758449298e-06,
+      "loss": 0.0,
+      "step": 2427
+    },
+    {
+      "epoch": 2.154289514520062,
+      "grad_norm": 0.033499255776405334,
+      "learning_rate": 1.941180126155311e-06,
+      "loss": 0.0001,
+      "step": 2428
+    },
+    {
+      "epoch": 2.1551762358678785,
+      "grad_norm": 0.009505156427621841,
+      "learning_rate": 1.937394380873418e-06,
+      "loss": 0.0,
+      "step": 2429
+    },
+    {
+      "epoch": 2.156062957215695,
+      "grad_norm": 0.012780265882611275,
+      "learning_rate": 1.9336114434701186e-06,
+      "loss": 0.0001,
+      "step": 2430
+    },
+    {
+      "epoch": 2.1569496785635116,
+      "grad_norm": 0.015024818480014801,
+      "learning_rate": 1.9298313174137177e-06,
+      "loss": 0.0001,
+      "step": 2431
+    },
+    {
+      "epoch": 2.1578363999113277,
+      "grad_norm": 0.2287774533033371,
+      "learning_rate": 1.92605400616993e-06,
+      "loss": 0.0004,
+      "step": 2432
+    },
+    {
+      "epoch": 2.1587231212591442,
+      "grad_norm": 0.0413081981241703,
+      "learning_rate": 1.9222795132018995e-06,
+      "loss": 0.0002,
+      "step": 2433
+    },
+    {
+      "epoch": 2.1596098426069608,
+      "grad_norm": 0.0011024032719433308,
+      "learning_rate": 1.918507841970177e-06,
+      "loss": 0.0,
+      "step": 2434
+    },
+    {
+      "epoch": 2.1604965639547773,
+      "grad_norm": 0.04966243356466293,
+      "learning_rate": 1.914738995932737e-06,
+      "loss": 0.0002,
+      "step": 2435
+    },
+    {
+      "epoch": 2.161383285302594,
+      "grad_norm": 0.29149574041366577,
+      "learning_rate": 1.9109729785449543e-06,
+      "loss": 0.0024,
+      "step": 2436
+    },
+    {
+      "epoch": 2.16227000665041,
+      "grad_norm": 0.004733963403850794,
+      "learning_rate": 1.9072097932596156e-06,
+      "loss": 0.0,
+      "step": 2437
+    },
+    {
+      "epoch": 2.1631567279982264,
+      "grad_norm": 0.07190831005573273,
+      "learning_rate": 1.9034494435269097e-06,
+      "loss": 0.0003,
+      "step": 2438
+    },
+    {
+      "epoch": 2.164043449346043,
+      "grad_norm": 0.0017673807451501489,
+      "learning_rate": 1.8996919327944252e-06,
+      "loss": 0.0,
+      "step": 2439
+    },
+    {
+      "epoch": 2.1649301706938595,
+      "grad_norm": 0.3513320982456207,
+      "learning_rate": 1.8959372645071506e-06,
+      "loss": 0.0016,
+      "step": 2440
+    },
+    {
+      "epoch": 2.165816892041676,
+      "grad_norm": 0.44481173157691956,
+      "learning_rate": 1.892185442107462e-06,
+      "loss": 0.0066,
+      "step": 2441
+    },
+    {
+      "epoch": 2.166703613389492,
+      "grad_norm": 0.015360888093709946,
+      "learning_rate": 1.888436469035132e-06,
+      "loss": 0.0,
+      "step": 2442
+    },
+    {
+      "epoch": 2.1675903347373087,
+      "grad_norm": 0.4169534146785736,
+      "learning_rate": 1.88469034872732e-06,
+      "loss": 0.0014,
+      "step": 2443
+    },
+    {
+      "epoch": 2.168477056085125,
+      "grad_norm": 0.012291036546230316,
+      "learning_rate": 1.8809470846185684e-06,
+      "loss": 0.0,
+      "step": 2444
+    },
+    {
+      "epoch": 2.1693637774329417,
+      "grad_norm": 0.19249936938285828,
+      "learning_rate": 1.8772066801408017e-06,
+      "loss": 0.0005,
+      "step": 2445
+    },
+    {
+      "epoch": 2.1702504987807583,
+      "grad_norm": 0.00133326998911798,
+      "learning_rate": 1.873469138723325e-06,
+      "loss": 0.0,
+      "step": 2446
+    },
+    {
+      "epoch": 2.171137220128575,
+      "grad_norm": 0.00433770939707756,
+      "learning_rate": 1.8697344637928123e-06,
+      "loss": 0.0,
+      "step": 2447
+    },
+    {
+      "epoch": 2.172023941476391,
+      "grad_norm": 0.016662631183862686,
+      "learning_rate": 1.8660026587733137e-06,
+      "loss": 0.0,
+      "step": 2448
+    },
+    {
+      "epoch": 2.1729106628242074,
+      "grad_norm": 0.045981135219335556,
+      "learning_rate": 1.8622737270862496e-06,
+      "loss": 0.0002,
+      "step": 2449
+    },
+    {
+      "epoch": 2.173797384172024,
+      "grad_norm": 0.008949820883572102,
+      "learning_rate": 1.8585476721504053e-06,
+      "loss": 0.0,
+      "step": 2450
+    },
+    {
+      "epoch": 2.1746841055198405,
+      "grad_norm": 0.0555342398583889,
+      "learning_rate": 1.8548244973819213e-06,
+      "loss": 0.0001,
+      "step": 2451
+    },
+    {
+      "epoch": 2.175570826867657,
+      "grad_norm": 0.020420093089342117,
+      "learning_rate": 1.8511042061943119e-06,
+      "loss": 0.0,
+      "step": 2452
+    },
+    {
+      "epoch": 2.176457548215473,
+      "grad_norm": 0.01802026480436325,
+      "learning_rate": 1.847386801998433e-06,
+      "loss": 0.0001,
+      "step": 2453
+    },
+    {
+      "epoch": 2.1773442695632896,
+      "grad_norm": 0.016103439033031464,
+      "learning_rate": 1.8436722882025043e-06,
+      "loss": 0.0,
+      "step": 2454
+    },
+    {
+      "epoch": 2.178230990911106,
+      "grad_norm": 0.46386483311653137,
+      "learning_rate": 1.8399606682120858e-06,
+      "loss": 0.0047,
+      "step": 2455
+    },
+    {
+      "epoch": 2.1791177122589227,
+      "grad_norm": 0.0012606150703504682,
+      "learning_rate": 1.8362519454300964e-06,
+      "loss": 0.0,
+      "step": 2456
+    },
+    {
+      "epoch": 2.1800044336067392,
+      "grad_norm": 0.005236682016402483,
+      "learning_rate": 1.8325461232567875e-06,
+      "loss": 0.0,
+      "step": 2457
+    },
+    {
+      "epoch": 2.1808911549545558,
+      "grad_norm": 0.04357163980603218,
+      "learning_rate": 1.828843205089757e-06,
+      "loss": 0.0001,
+      "step": 2458
+    },
+    {
+      "epoch": 2.181777876302372,
+      "grad_norm": 0.004069244954735041,
+      "learning_rate": 1.8251431943239396e-06,
+      "loss": 0.0,
+      "step": 2459
+    },
+    {
+      "epoch": 2.1826645976501884,
+      "grad_norm": 0.020366471260786057,
+      "learning_rate": 1.8214460943516055e-06,
+      "loss": 0.0001,
+      "step": 2460
+    },
+    {
+      "epoch": 2.183551318998005,
+      "grad_norm": 0.012451727874577045,
+      "learning_rate": 1.8177519085623513e-06,
+      "loss": 0.0001,
+      "step": 2461
+    },
+    {
+      "epoch": 2.1844380403458215,
+      "grad_norm": 0.006223964970558882,
+      "learning_rate": 1.8140606403431067e-06,
+      "loss": 0.0,
+      "step": 2462
+    },
+    {
+      "epoch": 2.1853247616936375,
+      "grad_norm": 0.04250934347510338,
+      "learning_rate": 1.8103722930781249e-06,
+      "loss": 0.0002,
+      "step": 2463
+    },
+    {
+      "epoch": 2.186211483041454,
+      "grad_norm": 0.02039632387459278,
+      "learning_rate": 1.8066868701489815e-06,
+      "loss": 0.0001,
+      "step": 2464
+    },
+    {
+      "epoch": 2.1870982043892706,
+      "grad_norm": 0.021646128967404366,
+      "learning_rate": 1.80300437493457e-06,
+      "loss": 0.0001,
+      "step": 2465
+    },
+    {
+      "epoch": 2.187984925737087,
+      "grad_norm": 0.003513685893267393,
+      "learning_rate": 1.7993248108111027e-06,
+      "loss": 0.0,
+      "step": 2466
+    },
+    {
+      "epoch": 2.1888716470849037,
+      "grad_norm": 0.0011019895318895578,
+      "learning_rate": 1.7956481811520987e-06,
+      "loss": 0.0,
+      "step": 2467
+    },
+    {
+      "epoch": 2.18975836843272,
+      "grad_norm": 0.05865537375211716,
+      "learning_rate": 1.791974489328392e-06,
+      "loss": 0.0003,
+      "step": 2468
+    },
+    {
+      "epoch": 2.1906450897805363,
+      "grad_norm": 0.023978404700756073,
+      "learning_rate": 1.788303738708121e-06,
+      "loss": 0.0001,
+      "step": 2469
+    },
+    {
+      "epoch": 2.191531811128353,
+      "grad_norm": 0.03099576011300087,
+      "learning_rate": 1.7846359326567274e-06,
+      "loss": 0.0001,
+      "step": 2470
+    },
+    {
+      "epoch": 2.1924185324761694,
+      "grad_norm": 0.10269967466592789,
+      "learning_rate": 1.780971074536957e-06,
+      "loss": 0.0002,
+      "step": 2471
+    },
+    {
+      "epoch": 2.193305253823986,
+      "grad_norm": 0.20377202332019806,
+      "learning_rate": 1.7773091677088427e-06,
+      "loss": 0.0008,
+      "step": 2472
+    },
+    {
+      "epoch": 2.1941919751718024,
+      "grad_norm": 0.005082233808934689,
+      "learning_rate": 1.7736502155297263e-06,
+      "loss": 0.0,
+      "step": 2473
+    },
+    {
+      "epoch": 2.1950786965196185,
+      "grad_norm": 0.09815764427185059,
+      "learning_rate": 1.769994221354227e-06,
+      "loss": 0.0005,
+      "step": 2474
+    },
+    {
+      "epoch": 2.195965417867435,
+      "grad_norm": 0.1481843888759613,
+      "learning_rate": 1.7663411885342625e-06,
+      "loss": 0.0002,
+      "step": 2475
+    },
+    {
+      "epoch": 2.1968521392152516,
+      "grad_norm": 0.007009209133684635,
+      "learning_rate": 1.7626911204190245e-06,
+      "loss": 0.0,
+      "step": 2476
+    },
+    {
+      "epoch": 2.197738860563068,
+      "grad_norm": 0.1537850797176361,
+      "learning_rate": 1.7590440203550009e-06,
+      "loss": 0.0005,
+      "step": 2477
+    },
+    {
+      "epoch": 2.1986255819108846,
+      "grad_norm": 0.32946306467056274,
+      "learning_rate": 1.7553998916859439e-06,
+      "loss": 0.0017,
+      "step": 2478
+    },
+    {
+      "epoch": 2.199512303258701,
+      "grad_norm": 0.002217859961092472,
+      "learning_rate": 1.75175873775289e-06,
+      "loss": 0.0,
+      "step": 2479
+    },
+    {
+      "epoch": 2.2003990246065173,
+      "grad_norm": 0.2838335633277893,
+      "learning_rate": 1.7481205618941472e-06,
+      "loss": 0.0008,
+      "step": 2480
+    },
+    {
+      "epoch": 2.201285745954334,
+      "grad_norm": 0.13207651674747467,
+      "learning_rate": 1.7444853674452932e-06,
+      "loss": 0.0005,
+      "step": 2481
+    },
+    {
+      "epoch": 2.2021724673021503,
+      "grad_norm": 0.0635002851486206,
+      "learning_rate": 1.7408531577391685e-06,
+      "loss": 0.0002,
+      "step": 2482
+    },
+    {
+      "epoch": 2.203059188649967,
+      "grad_norm": 0.306825190782547,
+      "learning_rate": 1.7372239361058813e-06,
+      "loss": 0.0008,
+      "step": 2483
+    },
+    {
+      "epoch": 2.2039459099977834,
+      "grad_norm": 0.0018865448655560613,
+      "learning_rate": 1.7335977058727983e-06,
+      "loss": 0.0,
+      "step": 2484
+    },
+    {
+      "epoch": 2.2048326313455995,
+      "grad_norm": 0.00354435364715755,
+      "learning_rate": 1.7299744703645466e-06,
+      "loss": 0.0,
+      "step": 2485
+    },
+    {
+      "epoch": 2.205719352693416,
+      "grad_norm": 0.19610072672367096,
+      "learning_rate": 1.726354232903002e-06,
+      "loss": 0.0005,
+      "step": 2486
+    },
+    {
+      "epoch": 2.2066060740412325,
+      "grad_norm": 0.0014539690455421805,
+      "learning_rate": 1.722736996807296e-06,
+      "loss": 0.0,
+      "step": 2487
+    },
+    {
+      "epoch": 2.207492795389049,
+      "grad_norm": 0.16978108882904053,
+      "learning_rate": 1.7191227653938081e-06,
+      "loss": 0.0005,
+      "step": 2488
+    },
+    {
+      "epoch": 2.2083795167368656,
+      "grad_norm": 0.06128441169857979,
+      "learning_rate": 1.715511541976161e-06,
+      "loss": 0.0003,
+      "step": 2489
+    },
+    {
+      "epoch": 2.2092662380846817,
+      "grad_norm": 0.0005087190656922758,
+      "learning_rate": 1.711903329865221e-06,
+      "loss": 0.0,
+      "step": 2490
+    },
+    {
+      "epoch": 2.2101529594324982,
+      "grad_norm": 0.08252158761024475,
+      "learning_rate": 1.7082981323690957e-06,
+      "loss": 0.0002,
+      "step": 2491
+    },
+    {
+      "epoch": 2.2110396807803148,
+      "grad_norm": 0.0013363012112677097,
+      "learning_rate": 1.7046959527931224e-06,
+      "loss": 0.0,
+      "step": 2492
+    },
+    {
+      "epoch": 2.2119264021281313,
+      "grad_norm": 0.009803982451558113,
+      "learning_rate": 1.7010967944398777e-06,
+      "loss": 0.0,
+      "step": 2493
+    },
+    {
+      "epoch": 2.212813123475948,
+      "grad_norm": 0.00928582064807415,
+      "learning_rate": 1.6975006606091655e-06,
+      "loss": 0.0,
+      "step": 2494
+    },
+    {
+      "epoch": 2.213699844823764,
+      "grad_norm": 0.04216178134083748,
+      "learning_rate": 1.6939075545980172e-06,
+      "loss": 0.0001,
+      "step": 2495
+    },
+    {
+      "epoch": 2.2145865661715805,
+      "grad_norm": 0.012317902408540249,
+      "learning_rate": 1.690317479700691e-06,
+      "loss": 0.0,
+      "step": 2496
+    },
+    {
+      "epoch": 2.215473287519397,
+      "grad_norm": 0.008212944492697716,
+      "learning_rate": 1.6867304392086575e-06,
+      "loss": 0.0,
+      "step": 2497
+    },
+    {
+      "epoch": 2.2163600088672135,
+      "grad_norm": 0.0018377496162429452,
+      "learning_rate": 1.6831464364106182e-06,
+      "loss": 0.0,
+      "step": 2498
+    },
+    {
+      "epoch": 2.21724673021503,
+      "grad_norm": 0.051457684487104416,
+      "learning_rate": 1.6795654745924772e-06,
+      "loss": 0.0002,
+      "step": 2499
+    },
+    {
+      "epoch": 2.2181334515628466,
+      "grad_norm": 0.0026265771593898535,
+      "learning_rate": 1.6759875570373597e-06,
+      "loss": 0.0,
+      "step": 2500
+    },
+    {
+      "epoch": 2.2190201729106627,
+      "grad_norm": 0.8870090842247009,
+      "learning_rate": 1.6724126870255908e-06,
+      "loss": 0.0042,
+      "step": 2501
+    },
+    {
+      "epoch": 2.219906894258479,
+      "grad_norm": 0.002598776947706938,
+      "learning_rate": 1.6688408678347134e-06,
+      "loss": 0.0,
+      "step": 2502
+    },
+    {
+      "epoch": 2.2207936156062957,
+      "grad_norm": 0.004365300294011831,
+      "learning_rate": 1.6652721027394619e-06,
+      "loss": 0.0,
+      "step": 2503
+    },
+    {
+      "epoch": 2.2216803369541123,
+      "grad_norm": 0.0315382219851017,
+      "learning_rate": 1.6617063950117773e-06,
+      "loss": 0.0001,
+      "step": 2504
+    },
+    {
+      "epoch": 2.222567058301929,
+      "grad_norm": 0.5729379653930664,
+      "learning_rate": 1.6581437479207952e-06,
+      "loss": 0.0028,
+      "step": 2505
+    },
+    {
+      "epoch": 2.223453779649745,
+      "grad_norm": 0.02651681751012802,
+      "learning_rate": 1.654584164732847e-06,
+      "loss": 0.0001,
+      "step": 2506
+    },
+    {
+      "epoch": 2.2243405009975614,
+      "grad_norm": 0.00810746569186449,
+      "learning_rate": 1.6510276487114508e-06,
+      "loss": 0.0,
+      "step": 2507
+    },
+    {
+      "epoch": 2.225227222345378,
+      "grad_norm": 0.027231650426983833,
+      "learning_rate": 1.6474742031173164e-06,
+      "loss": 0.0001,
+      "step": 2508
+    },
+    {
+      "epoch": 2.2261139436931945,
+      "grad_norm": 0.4568202495574951,
+      "learning_rate": 1.643923831208339e-06,
+      "loss": 0.0009,
+      "step": 2509
+    },
+    {
+      "epoch": 2.227000665041011,
+      "grad_norm": 0.12720799446105957,
+      "learning_rate": 1.640376536239594e-06,
+      "loss": 0.0003,
+      "step": 2510
+    },
+    {
+      "epoch": 2.227887386388827,
+      "grad_norm": 0.020172515884041786,
+      "learning_rate": 1.6368323214633331e-06,
+      "loss": 0.0001,
+      "step": 2511
+    },
+    {
+      "epoch": 2.2287741077366436,
+      "grad_norm": 0.10302843153476715,
+      "learning_rate": 1.6332911901289926e-06,
+      "loss": 0.0003,
+      "step": 2512
+    },
+    {
+      "epoch": 2.22966082908446,
+      "grad_norm": 0.0018525099148973823,
+      "learning_rate": 1.6297531454831718e-06,
+      "loss": 0.0,
+      "step": 2513
+    },
+    {
+      "epoch": 2.2305475504322767,
+      "grad_norm": 0.009870581328868866,
+      "learning_rate": 1.6262181907696456e-06,
+      "loss": 0.0,
+      "step": 2514
+    },
+    {
+      "epoch": 2.2314342717800932,
+      "grad_norm": 0.0018539935117587447,
+      "learning_rate": 1.622686329229355e-06,
+      "loss": 0.0,
+      "step": 2515
+    },
+    {
+      "epoch": 2.2323209931279093,
+      "grad_norm": 0.1914931684732437,
+      "learning_rate": 1.6191575641004049e-06,
+      "loss": 0.0006,
+      "step": 2516
+    },
+    {
+      "epoch": 2.233207714475726,
+      "grad_norm": 0.10301623493432999,
+      "learning_rate": 1.6156318986180636e-06,
+      "loss": 0.0002,
+      "step": 2517
+    },
+    {
+      "epoch": 2.2340944358235424,
+      "grad_norm": 0.013894562609493732,
+      "learning_rate": 1.6121093360147517e-06,
+      "loss": 0.0001,
+      "step": 2518
+    },
+    {
+      "epoch": 2.234981157171359,
+      "grad_norm": 0.021681584417819977,
+      "learning_rate": 1.60858987952005e-06,
+      "loss": 0.0001,
+      "step": 2519
+    },
+    {
+      "epoch": 2.2358678785191755,
+      "grad_norm": 0.0033847822342067957,
+      "learning_rate": 1.6050735323606904e-06,
+      "loss": 0.0,
+      "step": 2520
+    },
+    {
+      "epoch": 2.236754599866992,
+      "grad_norm": 0.006841412745416164,
+      "learning_rate": 1.6015602977605555e-06,
+      "loss": 0.0,
+      "step": 2521
+    },
+    {
+      "epoch": 2.237641321214808,
+      "grad_norm": 0.031053803861141205,
+      "learning_rate": 1.5980501789406673e-06,
+      "loss": 0.0001,
+      "step": 2522
+    },
+    {
+      "epoch": 2.2385280425626246,
+      "grad_norm": 0.004785686265677214,
+      "learning_rate": 1.594543179119204e-06,
+      "loss": 0.0,
+      "step": 2523
+    },
+    {
+      "epoch": 2.239414763910441,
+      "grad_norm": 0.014937586151063442,
+      "learning_rate": 1.5910393015114705e-06,
+      "loss": 0.0,
+      "step": 2524
+    },
+    {
+      "epoch": 2.2403014852582577,
+      "grad_norm": 0.22507883608341217,
+      "learning_rate": 1.5875385493299179e-06,
+      "loss": 0.0023,
+      "step": 2525
+    },
+    {
+      "epoch": 2.241188206606074,
+      "grad_norm": 0.04809168726205826,
+      "learning_rate": 1.5840409257841294e-06,
+      "loss": 0.0002,
+      "step": 2526
+    },
+    {
+      "epoch": 2.2420749279538903,
+      "grad_norm": 0.018942655995488167,
+      "learning_rate": 1.5805464340808201e-06,
+      "loss": 0.0,
+      "step": 2527
+    },
+    {
+      "epoch": 2.242961649301707,
+      "grad_norm": 0.009502418339252472,
+      "learning_rate": 1.5770550774238312e-06,
+      "loss": 0.0,
+      "step": 2528
+    },
+    {
+      "epoch": 2.2438483706495234,
+      "grad_norm": 0.008832428604364395,
+      "learning_rate": 1.5735668590141328e-06,
+      "loss": 0.0,
+      "step": 2529
+    },
+    {
+      "epoch": 2.24473509199734,
+      "grad_norm": 0.001664695329964161,
+      "learning_rate": 1.5700817820498165e-06,
+      "loss": 0.0,
+      "step": 2530
+    },
+    {
+      "epoch": 2.2456218133451564,
+      "grad_norm": 0.009838556870818138,
+      "learning_rate": 1.5665998497260959e-06,
+      "loss": 0.0,
+      "step": 2531
+    },
+    {
+      "epoch": 2.246508534692973,
+      "grad_norm": 0.04569023847579956,
+      "learning_rate": 1.5631210652352945e-06,
+      "loss": 0.0001,
+      "step": 2532
+    },
+    {
+      "epoch": 2.247395256040789,
+      "grad_norm": 0.005965916905552149,
+      "learning_rate": 1.5596454317668613e-06,
+      "loss": 0.0,
+      "step": 2533
+    },
+    {
+      "epoch": 2.2482819773886056,
+      "grad_norm": 0.010405507870018482,
+      "learning_rate": 1.5561729525073454e-06,
+      "loss": 0.0,
+      "step": 2534
+    },
+    {
+      "epoch": 2.249168698736422,
+      "grad_norm": 0.2523658573627472,
+      "learning_rate": 1.552703630640411e-06,
+      "loss": 0.0012,
+      "step": 2535
+    },
+    {
+      "epoch": 2.2500554200842386,
+      "grad_norm": 0.00390112167224288,
+      "learning_rate": 1.5492374693468204e-06,
+      "loss": 0.0,
+      "step": 2536
+    },
+    {
+      "epoch": 2.2509421414320547,
+      "grad_norm": 0.010760684497654438,
+      "learning_rate": 1.5457744718044493e-06,
+      "loss": 0.0,
+      "step": 2537
+    },
+    {
+      "epoch": 2.2518288627798713,
+      "grad_norm": 0.0015860433923080564,
+      "learning_rate": 1.5423146411882616e-06,
+      "loss": 0.0,
+      "step": 2538
+    },
+    {
+      "epoch": 2.252715584127688,
+      "grad_norm": 0.019439872354269028,
+      "learning_rate": 1.5388579806703241e-06,
+      "loss": 0.0001,
+      "step": 2539
+    },
+    {
+      "epoch": 2.2536023054755043,
+      "grad_norm": 0.017621183767914772,
+      "learning_rate": 1.5354044934197953e-06,
+      "loss": 0.0001,
+      "step": 2540
+    },
+    {
+      "epoch": 2.254489026823321,
+      "grad_norm": 0.03982497379183769,
+      "learning_rate": 1.5319541826029243e-06,
+      "loss": 0.0002,
+      "step": 2541
+    },
+    {
+      "epoch": 2.2553757481711374,
+      "grad_norm": 0.0021173080895096064,
+      "learning_rate": 1.5285070513830502e-06,
+      "loss": 0.0,
+      "step": 2542
+    },
+    {
+      "epoch": 2.2562624695189535,
+      "grad_norm": 0.008100396953523159,
+      "learning_rate": 1.525063102920591e-06,
+      "loss": 0.0,
+      "step": 2543
+    },
+    {
+      "epoch": 2.25714919086677,
+      "grad_norm": 0.014582299627363682,
+      "learning_rate": 1.521622340373053e-06,
+      "loss": 0.0001,
+      "step": 2544
+    },
+    {
+      "epoch": 2.2580359122145865,
+      "grad_norm": 0.06363100558519363,
+      "learning_rate": 1.518184766895019e-06,
+      "loss": 0.0001,
+      "step": 2545
+    },
+    {
+      "epoch": 2.258922633562403,
+      "grad_norm": 0.06486165523529053,
+      "learning_rate": 1.5147503856381478e-06,
+      "loss": 0.0002,
+      "step": 2546
+    },
+    {
+      "epoch": 2.2598093549102196,
+      "grad_norm": 0.008981321938335896,
+      "learning_rate": 1.5113191997511716e-06,
+      "loss": 0.0,
+      "step": 2547
+    },
+    {
+      "epoch": 2.2606960762580357,
+      "grad_norm": 0.17823724448680878,
+      "learning_rate": 1.507891212379896e-06,
+      "loss": 0.0006,
+      "step": 2548
+    },
+    {
+      "epoch": 2.2615827976058522,
+      "grad_norm": 0.02821299619972706,
+      "learning_rate": 1.504466426667187e-06,
+      "loss": 0.0001,
+      "step": 2549
+    },
+    {
+      "epoch": 2.2624695189536688,
+      "grad_norm": 0.0017538291867822409,
+      "learning_rate": 1.5010448457529814e-06,
+      "loss": 0.0,
+      "step": 2550
+    },
+    {
+      "epoch": 2.2633562403014853,
+      "grad_norm": 0.0058082821778953075,
+      "learning_rate": 1.4976264727742768e-06,
+      "loss": 0.0,
+      "step": 2551
+    },
+    {
+      "epoch": 2.264242961649302,
+      "grad_norm": 0.10759159922599792,
+      "learning_rate": 1.4942113108651291e-06,
+      "loss": 0.0003,
+      "step": 2552
+    },
+    {
+      "epoch": 2.2651296829971184,
+      "grad_norm": 0.058488212525844574,
+      "learning_rate": 1.490799363156646e-06,
+      "loss": 0.0002,
+      "step": 2553
+    },
+    {
+      "epoch": 2.2660164043449345,
+      "grad_norm": 0.011463595554232597,
+      "learning_rate": 1.4873906327769978e-06,
+      "loss": 0.0,
+      "step": 2554
+    },
+    {
+      "epoch": 2.266903125692751,
+      "grad_norm": 0.0010025546653196216,
+      "learning_rate": 1.4839851228513957e-06,
+      "loss": 0.0,
+      "step": 2555
+    },
+    {
+      "epoch": 2.2677898470405675,
+      "grad_norm": 0.002267697127535939,
+      "learning_rate": 1.4805828365021047e-06,
+      "loss": 0.0,
+      "step": 2556
+    },
+    {
+      "epoch": 2.268676568388384,
+      "grad_norm": 0.008600475266575813,
+      "learning_rate": 1.4771837768484276e-06,
+      "loss": 0.0,
+      "step": 2557
+    },
+    {
+      "epoch": 2.2695632897362006,
+      "grad_norm": 0.4348275363445282,
+      "learning_rate": 1.4737879470067184e-06,
+      "loss": 0.0024,
+      "step": 2558
+    },
+    {
+      "epoch": 2.2704500110840167,
+      "grad_norm": 0.0017552981153130531,
+      "learning_rate": 1.4703953500903613e-06,
+      "loss": 0.0,
+      "step": 2559
+    },
+    {
+      "epoch": 2.271336732431833,
+      "grad_norm": 0.0007721121073700488,
+      "learning_rate": 1.4670059892097793e-06,
+      "loss": 0.0,
+      "step": 2560
+    },
+    {
+      "epoch": 2.2722234537796497,
+      "grad_norm": 0.037658073008060455,
+      "learning_rate": 1.4636198674724305e-06,
+      "loss": 0.0001,
+      "step": 2561
+    },
+    {
+      "epoch": 2.2731101751274663,
+      "grad_norm": 0.017900565639138222,
+      "learning_rate": 1.4602369879828033e-06,
+      "loss": 0.0001,
+      "step": 2562
+    },
+    {
+      "epoch": 2.273996896475283,
+      "grad_norm": 0.025672055780887604,
+      "learning_rate": 1.4568573538424075e-06,
+      "loss": 0.0001,
+      "step": 2563
+    },
+    {
+      "epoch": 2.2748836178230993,
+      "grad_norm": 0.7328742742538452,
+      "learning_rate": 1.4534809681497846e-06,
+      "loss": 0.0042,
+      "step": 2564
+    },
+    {
+      "epoch": 2.2757703391709154,
+      "grad_norm": 0.02617855742573738,
+      "learning_rate": 1.4501078340004954e-06,
+      "loss": 0.0001,
+      "step": 2565
+    },
+    {
+      "epoch": 2.276657060518732,
+      "grad_norm": 0.014606747776269913,
+      "learning_rate": 1.446737954487118e-06,
+      "loss": 0.0001,
+      "step": 2566
+    },
+    {
+      "epoch": 2.2775437818665485,
+      "grad_norm": 0.0028215658385306597,
+      "learning_rate": 1.4433713326992488e-06,
+      "loss": 0.0,
+      "step": 2567
+    },
+    {
+      "epoch": 2.278430503214365,
+      "grad_norm": 0.06955783069133759,
+      "learning_rate": 1.4400079717234983e-06,
+      "loss": 0.0001,
+      "step": 2568
+    },
+    {
+      "epoch": 2.279317224562181,
+      "grad_norm": 0.0066056400537490845,
+      "learning_rate": 1.436647874643482e-06,
+      "loss": 0.0,
+      "step": 2569
+    },
+    {
+      "epoch": 2.2802039459099976,
+      "grad_norm": 0.0011473114136606455,
+      "learning_rate": 1.4332910445398285e-06,
+      "loss": 0.0,
+      "step": 2570
+    },
+    {
+      "epoch": 2.281090667257814,
+      "grad_norm": 0.07209219038486481,
+      "learning_rate": 1.4299374844901692e-06,
+      "loss": 0.0002,
+      "step": 2571
+    },
+    {
+      "epoch": 2.2819773886056307,
+      "grad_norm": 0.0020093126222491264,
+      "learning_rate": 1.426587197569137e-06,
+      "loss": 0.0,
+      "step": 2572
+    },
+    {
+      "epoch": 2.2828641099534472,
+      "grad_norm": 0.002489017089828849,
+      "learning_rate": 1.4232401868483669e-06,
+      "loss": 0.0,
+      "step": 2573
+    },
+    {
+      "epoch": 2.2837508313012638,
+      "grad_norm": 0.006025847978889942,
+      "learning_rate": 1.4198964553964823e-06,
+      "loss": 0.0,
+      "step": 2574
+    },
+    {
+      "epoch": 2.28463755264908,
+      "grad_norm": 0.00035993586061522365,
+      "learning_rate": 1.4165560062791117e-06,
+      "loss": 0.0,
+      "step": 2575
+    },
+    {
+      "epoch": 2.2855242739968964,
+      "grad_norm": 0.005870227701961994,
+      "learning_rate": 1.4132188425588633e-06,
+      "loss": 0.0,
+      "step": 2576
+    },
+    {
+      "epoch": 2.286410995344713,
+      "grad_norm": 0.009169611148536205,
+      "learning_rate": 1.4098849672953418e-06,
+      "loss": 0.0,
+      "step": 2577
+    },
+    {
+      "epoch": 2.2872977166925295,
+      "grad_norm": 0.008256441913545132,
+      "learning_rate": 1.4065543835451278e-06,
+      "loss": 0.0,
+      "step": 2578
+    },
+    {
+      "epoch": 2.288184438040346,
+      "grad_norm": 0.03790901228785515,
+      "learning_rate": 1.4032270943617953e-06,
+      "loss": 0.0001,
+      "step": 2579
+    },
+    {
+      "epoch": 2.289071159388162,
+      "grad_norm": 0.07731454819440842,
+      "learning_rate": 1.3999031027958876e-06,
+      "loss": 0.0003,
+      "step": 2580
+    },
+    {
+      "epoch": 2.2899578807359786,
+      "grad_norm": 0.018880745396018028,
+      "learning_rate": 1.3965824118949307e-06,
+      "loss": 0.0001,
+      "step": 2581
+    },
+    {
+      "epoch": 2.290844602083795,
+      "grad_norm": 0.40576595067977905,
+      "learning_rate": 1.393265024703422e-06,
+      "loss": 0.0011,
+      "step": 2582
+    },
+    {
+      "epoch": 2.2917313234316117,
+      "grad_norm": 0.002390408655628562,
+      "learning_rate": 1.3899509442628328e-06,
+      "loss": 0.0,
+      "step": 2583
+    },
+    {
+      "epoch": 2.292618044779428,
+      "grad_norm": 0.11626479774713516,
+      "learning_rate": 1.3866401736115975e-06,
+      "loss": 0.0004,
+      "step": 2584
+    },
+    {
+      "epoch": 2.2935047661272447,
+      "grad_norm": 0.0015793382190167904,
+      "learning_rate": 1.3833327157851194e-06,
+      "loss": 0.0,
+      "step": 2585
+    },
+    {
+      "epoch": 2.294391487475061,
+      "grad_norm": 0.012690646573901176,
+      "learning_rate": 1.380028573815766e-06,
+      "loss": 0.0,
+      "step": 2586
+    },
+    {
+      "epoch": 2.2952782088228774,
+      "grad_norm": 0.029712093994021416,
+      "learning_rate": 1.376727750732863e-06,
+      "loss": 0.0001,
+      "step": 2587
+    },
+    {
+      "epoch": 2.296164930170694,
+      "grad_norm": 0.002025979571044445,
+      "learning_rate": 1.3734302495626905e-06,
+      "loss": 0.0,
+      "step": 2588
+    },
+    {
+      "epoch": 2.2970516515185104,
+      "grad_norm": 0.012119759805500507,
+      "learning_rate": 1.370136073328487e-06,
+      "loss": 0.0,
+      "step": 2589
+    },
+    {
+      "epoch": 2.2979383728663265,
+      "grad_norm": 0.10140902549028397,
+      "learning_rate": 1.3668452250504417e-06,
+      "loss": 0.0002,
+      "step": 2590
+    },
+    {
+      "epoch": 2.298825094214143,
+      "grad_norm": 0.007041087839752436,
+      "learning_rate": 1.3635577077456913e-06,
+      "loss": 0.0,
+      "step": 2591
+    },
+    {
+      "epoch": 2.2997118155619596,
+      "grad_norm": 0.004575217142701149,
+      "learning_rate": 1.3602735244283204e-06,
+      "loss": 0.0,
+      "step": 2592
+    },
+    {
+      "epoch": 2.300598536909776,
+      "grad_norm": 0.01836671680212021,
+      "learning_rate": 1.3569926781093568e-06,
+      "loss": 0.0001,
+      "step": 2593
+    },
+    {
+      "epoch": 2.3014852582575926,
+      "grad_norm": 0.010562287643551826,
+      "learning_rate": 1.353715171796765e-06,
+      "loss": 0.0,
+      "step": 2594
+    },
+    {
+      "epoch": 2.302371979605409,
+      "grad_norm": 0.07208259403705597,
+      "learning_rate": 1.3504410084954521e-06,
+      "loss": 0.0004,
+      "step": 2595
+    },
+    {
+      "epoch": 2.3032587009532253,
+      "grad_norm": 0.0030105796176940203,
+      "learning_rate": 1.3471701912072581e-06,
+      "loss": 0.0,
+      "step": 2596
+    },
+    {
+      "epoch": 2.304145422301042,
+      "grad_norm": 0.001862941775470972,
+      "learning_rate": 1.3439027229309564e-06,
+      "loss": 0.0,
+      "step": 2597
+    },
+    {
+      "epoch": 2.3050321436488583,
+      "grad_norm": 0.02082369290292263,
+      "learning_rate": 1.3406386066622495e-06,
+      "loss": 0.0,
+      "step": 2598
+    },
+    {
+      "epoch": 2.305918864996675,
+      "grad_norm": 0.13240396976470947,
+      "learning_rate": 1.337377845393763e-06,
+      "loss": 0.0004,
+      "step": 2599
+    },
+    {
+      "epoch": 2.3068055863444914,
+      "grad_norm": 0.0022385409101843834,
+      "learning_rate": 1.334120442115055e-06,
+      "loss": 0.0,
+      "step": 2600
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.5025140643119812,
+      "learning_rate": 1.3308663998125954e-06,
+      "loss": 0.0013,
+      "step": 2601
+    },
+    {
+      "epoch": 2.308579029040124,
+      "grad_norm": 0.06153545901179314,
+      "learning_rate": 1.3276157214697805e-06,
+      "loss": 0.0002,
+      "step": 2602
+    },
+    {
+      "epoch": 2.3094657503879406,
+      "grad_norm": 0.007376537658274174,
+      "learning_rate": 1.324368410066914e-06,
+      "loss": 0.0,
+      "step": 2603
+    },
+    {
+      "epoch": 2.310352471735757,
+      "grad_norm": 0.04724806919693947,
+      "learning_rate": 1.321124468581224e-06,
+      "loss": 0.0002,
+      "step": 2604
+    },
+    {
+      "epoch": 2.3112391930835736,
+      "grad_norm": 0.10160135477781296,
+      "learning_rate": 1.3178838999868381e-06,
+      "loss": 0.0004,
+      "step": 2605
+    },
+    {
+      "epoch": 2.31212591443139,
+      "grad_norm": 0.0051342276856303215,
+      "learning_rate": 1.3146467072547975e-06,
+      "loss": 0.0,
+      "step": 2606
+    },
+    {
+      "epoch": 2.3130126357792062,
+      "grad_norm": 0.009772558696568012,
+      "learning_rate": 1.3114128933530473e-06,
+      "loss": 0.0001,
+      "step": 2607
+    },
+    {
+      "epoch": 2.3138993571270228,
+      "grad_norm": 0.04966363310813904,
+      "learning_rate": 1.308182461246435e-06,
+      "loss": 0.0001,
+      "step": 2608
+    },
+    {
+      "epoch": 2.3147860784748393,
+      "grad_norm": 0.0012651075376197696,
+      "learning_rate": 1.3049554138967052e-06,
+      "loss": 0.0,
+      "step": 2609
+    },
+    {
+      "epoch": 2.315672799822656,
+      "grad_norm": 0.48004817962646484,
+      "learning_rate": 1.3017317542625018e-06,
+      "loss": 0.0029,
+      "step": 2610
+    },
+    {
+      "epoch": 2.3165595211704724,
+      "grad_norm": 0.27828601002693176,
+      "learning_rate": 1.2985114852993624e-06,
+      "loss": 0.0004,
+      "step": 2611
+    },
+    {
+      "epoch": 2.3174462425182885,
+      "grad_norm": 0.3281623423099518,
+      "learning_rate": 1.2952946099597169e-06,
+      "loss": 0.0009,
+      "step": 2612
+    },
+    {
+      "epoch": 2.318332963866105,
+      "grad_norm": 0.01670690067112446,
+      "learning_rate": 1.2920811311928778e-06,
+      "loss": 0.0001,
+      "step": 2613
+    },
+    {
+      "epoch": 2.3192196852139215,
+      "grad_norm": 0.002526923082768917,
+      "learning_rate": 1.2888710519450543e-06,
+      "loss": 0.0,
+      "step": 2614
+    },
+    {
+      "epoch": 2.320106406561738,
+      "grad_norm": 0.027562735602259636,
+      "learning_rate": 1.2856643751593285e-06,
+      "loss": 0.0001,
+      "step": 2615
+    },
+    {
+      "epoch": 2.3209931279095546,
+      "grad_norm": 0.3966238498687744,
+      "learning_rate": 1.2824611037756686e-06,
+      "loss": 0.0032,
+      "step": 2616
+    },
+    {
+      "epoch": 2.321879849257371,
+      "grad_norm": 0.23547692596912384,
+      "learning_rate": 1.2792612407309195e-06,
+      "loss": 0.0006,
+      "step": 2617
+    },
+    {
+      "epoch": 2.322766570605187,
+      "grad_norm": 0.07173073291778564,
+      "learning_rate": 1.2760647889588008e-06,
+      "loss": 0.0002,
+      "step": 2618
+    },
+    {
+      "epoch": 2.3236532919530037,
+      "grad_norm": 0.000607346068136394,
+      "learning_rate": 1.2728717513899065e-06,
+      "loss": 0.0,
+      "step": 2619
+    },
+    {
+      "epoch": 2.3245400133008203,
+      "grad_norm": 0.03587135300040245,
+      "learning_rate": 1.2696821309516954e-06,
+      "loss": 0.0001,
+      "step": 2620
+    },
+    {
+      "epoch": 2.325426734648637,
+      "grad_norm": 0.01693054474890232,
+      "learning_rate": 1.2664959305684988e-06,
+      "loss": 0.0001,
+      "step": 2621
+    },
+    {
+      "epoch": 2.326313455996453,
+      "grad_norm": 0.16542497277259827,
+      "learning_rate": 1.2633131531615101e-06,
+      "loss": 0.0002,
+      "step": 2622
+    },
+    {
+      "epoch": 2.3272001773442694,
+      "grad_norm": 0.03863265737891197,
+      "learning_rate": 1.2601338016487852e-06,
+      "loss": 0.0001,
+      "step": 2623
+    },
+    {
+      "epoch": 2.328086898692086,
+      "grad_norm": 0.16595076024532318,
+      "learning_rate": 1.2569578789452353e-06,
+      "loss": 0.0002,
+      "step": 2624
+    },
+    {
+      "epoch": 2.3289736200399025,
+      "grad_norm": 0.004780843388289213,
+      "learning_rate": 1.253785387962636e-06,
+      "loss": 0.0,
+      "step": 2625
+    },
+    {
+      "epoch": 2.329860341387719,
+      "grad_norm": 0.5518388152122498,
+      "learning_rate": 1.2506163316096081e-06,
+      "loss": 0.0053,
+      "step": 2626
+    },
+    {
+      "epoch": 2.3307470627355356,
+      "grad_norm": 0.09815961122512817,
+      "learning_rate": 1.247450712791628e-06,
+      "loss": 0.0004,
+      "step": 2627
+    },
+    {
+      "epoch": 2.3316337840833516,
+      "grad_norm": 0.016248105093836784,
+      "learning_rate": 1.24428853441102e-06,
+      "loss": 0.0,
+      "step": 2628
+    },
+    {
+      "epoch": 2.332520505431168,
+      "grad_norm": 0.13292931020259857,
+      "learning_rate": 1.2411297993669546e-06,
+      "loss": 0.0004,
+      "step": 2629
+    },
+    {
+      "epoch": 2.3334072267789847,
+      "grad_norm": 0.45038118958473206,
+      "learning_rate": 1.2379745105554413e-06,
+      "loss": 0.0055,
+      "step": 2630
+    },
+    {
+      "epoch": 2.3342939481268012,
+      "grad_norm": 0.009716760367155075,
+      "learning_rate": 1.2348226708693356e-06,
+      "loss": 0.0,
+      "step": 2631
+    },
+    {
+      "epoch": 2.3351806694746178,
+      "grad_norm": 0.29626765847206116,
+      "learning_rate": 1.231674283198328e-06,
+      "loss": 0.001,
+      "step": 2632
+    },
+    {
+      "epoch": 2.336067390822434,
+      "grad_norm": 0.0017720351461321115,
+      "learning_rate": 1.2285293504289448e-06,
+      "loss": 0.0,
+      "step": 2633
+    },
+    {
+      "epoch": 2.3369541121702504,
+      "grad_norm": 0.0024435853119939566,
+      "learning_rate": 1.2253878754445414e-06,
+      "loss": 0.0,
+      "step": 2634
+    },
+    {
+      "epoch": 2.337840833518067,
+      "grad_norm": 0.04254931956529617,
+      "learning_rate": 1.2222498611253103e-06,
+      "loss": 0.0001,
+      "step": 2635
+    },
+    {
+      "epoch": 2.3387275548658835,
+      "grad_norm": 0.28088611364364624,
+      "learning_rate": 1.219115310348264e-06,
+      "loss": 0.0023,
+      "step": 2636
+    },
+    {
+      "epoch": 2.3396142762137,
+      "grad_norm": 0.000748008256778121,
+      "learning_rate": 1.215984225987244e-06,
+      "loss": 0.0,
+      "step": 2637
+    },
+    {
+      "epoch": 2.3405009975615165,
+      "grad_norm": 0.0008762306533753872,
+      "learning_rate": 1.212856610912908e-06,
+      "loss": 0.0,
+      "step": 2638
+    },
+    {
+      "epoch": 2.3413877189093326,
+      "grad_norm": 0.03883393481373787,
+      "learning_rate": 1.2097324679927431e-06,
+      "loss": 0.0001,
+      "step": 2639
+    },
+    {
+      "epoch": 2.342274440257149,
+      "grad_norm": 0.0028116086032241583,
+      "learning_rate": 1.2066118000910421e-06,
+      "loss": 0.0,
+      "step": 2640
+    },
+    {
+      "epoch": 2.3431611616049657,
+      "grad_norm": 0.02033114992082119,
+      "learning_rate": 1.2034946100689188e-06,
+      "loss": 0.0001,
+      "step": 2641
+    },
+    {
+      "epoch": 2.344047882952782,
+      "grad_norm": 0.036860495805740356,
+      "learning_rate": 1.200380900784296e-06,
+      "loss": 0.0002,
+      "step": 2642
+    },
+    {
+      "epoch": 2.3449346043005983,
+      "grad_norm": 0.00874006561934948,
+      "learning_rate": 1.197270675091905e-06,
+      "loss": 0.0,
+      "step": 2643
+    },
+    {
+      "epoch": 2.345821325648415,
+      "grad_norm": 0.005195686127990484,
+      "learning_rate": 1.1941639358432855e-06,
+      "loss": 0.0,
+      "step": 2644
+    },
+    {
+      "epoch": 2.3467080469962314,
+      "grad_norm": 0.005843705963343382,
+      "learning_rate": 1.1910606858867763e-06,
+      "loss": 0.0,
+      "step": 2645
+    },
+    {
+      "epoch": 2.347594768344048,
+      "grad_norm": 0.0016122246161103249,
+      "learning_rate": 1.1879609280675208e-06,
+      "loss": 0.0,
+      "step": 2646
+    },
+    {
+      "epoch": 2.3484814896918644,
+      "grad_norm": 0.01298630889505148,
+      "learning_rate": 1.1848646652274604e-06,
+      "loss": 0.0001,
+      "step": 2647
+    },
+    {
+      "epoch": 2.349368211039681,
+      "grad_norm": 0.004974827636033297,
+      "learning_rate": 1.1817719002053307e-06,
+      "loss": 0.0,
+      "step": 2648
+    },
+    {
+      "epoch": 2.350254932387497,
+      "grad_norm": 0.009422971867024899,
+      "learning_rate": 1.1786826358366621e-06,
+      "loss": 0.0,
+      "step": 2649
+    },
+    {
+      "epoch": 2.3511416537353136,
+      "grad_norm": 0.0029839402996003628,
+      "learning_rate": 1.1755968749537755e-06,
+      "loss": 0.0,
+      "step": 2650
+    },
+    {
+      "epoch": 2.35202837508313,
+      "grad_norm": 0.043002091348171234,
+      "learning_rate": 1.1725146203857768e-06,
+      "loss": 0.0001,
+      "step": 2651
+    },
+    {
+      "epoch": 2.3529150964309467,
+      "grad_norm": 0.010507393628358841,
+      "learning_rate": 1.1694358749585593e-06,
+      "loss": 0.0,
+      "step": 2652
+    },
+    {
+      "epoch": 2.353801817778763,
+      "grad_norm": 0.21527473628520966,
+      "learning_rate": 1.1663606414947998e-06,
+      "loss": 0.0005,
+      "step": 2653
+    },
+    {
+      "epoch": 2.3546885391265793,
+      "grad_norm": 0.021156540140509605,
+      "learning_rate": 1.1632889228139565e-06,
+      "loss": 0.0001,
+      "step": 2654
+    },
+    {
+      "epoch": 2.355575260474396,
+      "grad_norm": 0.010046728886663914,
+      "learning_rate": 1.1602207217322575e-06,
+      "loss": 0.0,
+      "step": 2655
+    },
+    {
+      "epoch": 2.3564619818222123,
+      "grad_norm": 0.0030884298030287027,
+      "learning_rate": 1.1571560410627181e-06,
+      "loss": 0.0,
+      "step": 2656
+    },
+    {
+      "epoch": 2.357348703170029,
+      "grad_norm": 0.02951064519584179,
+      "learning_rate": 1.154094883615115e-06,
+      "loss": 0.0001,
+      "step": 2657
+    },
+    {
+      "epoch": 2.3582354245178454,
+      "grad_norm": 0.028621038421988487,
+      "learning_rate": 1.1510372521960016e-06,
+      "loss": 0.0001,
+      "step": 2658
+    },
+    {
+      "epoch": 2.359122145865662,
+      "grad_norm": 0.09628434479236603,
+      "learning_rate": 1.1479831496086919e-06,
+      "loss": 0.0004,
+      "step": 2659
+    },
+    {
+      "epoch": 2.360008867213478,
+      "grad_norm": 0.007439910434186459,
+      "learning_rate": 1.1449325786532744e-06,
+      "loss": 0.0,
+      "step": 2660
+    },
+    {
+      "epoch": 2.3608955885612946,
+      "grad_norm": 0.12967965006828308,
+      "learning_rate": 1.1418855421265896e-06,
+      "loss": 0.0005,
+      "step": 2661
+    },
+    {
+      "epoch": 2.361782309909111,
+      "grad_norm": 0.2929927110671997,
+      "learning_rate": 1.1388420428222435e-06,
+      "loss": 0.0007,
+      "step": 2662
+    },
+    {
+      "epoch": 2.3626690312569276,
+      "grad_norm": 0.0033991162199527025,
+      "learning_rate": 1.1358020835305978e-06,
+      "loss": 0.0,
+      "step": 2663
+    },
+    {
+      "epoch": 2.363555752604744,
+      "grad_norm": 0.0073296381160616875,
+      "learning_rate": 1.1327656670387694e-06,
+      "loss": 0.0,
+      "step": 2664
+    },
+    {
+      "epoch": 2.3644424739525602,
+      "grad_norm": 0.017435623332858086,
+      "learning_rate": 1.1297327961306226e-06,
+      "loss": 0.0,
+      "step": 2665
+    },
+    {
+      "epoch": 2.3653291953003768,
+      "grad_norm": 0.002517417771741748,
+      "learning_rate": 1.1267034735867771e-06,
+      "loss": 0.0,
+      "step": 2666
+    },
+    {
+      "epoch": 2.3662159166481933,
+      "grad_norm": 0.0027054729871451855,
+      "learning_rate": 1.1236777021845957e-06,
+      "loss": 0.0,
+      "step": 2667
+    },
+    {
+      "epoch": 2.36710263799601,
+      "grad_norm": 0.10869599133729935,
+      "learning_rate": 1.1206554846981866e-06,
+      "loss": 0.0005,
+      "step": 2668
+    },
+    {
+      "epoch": 2.3679893593438264,
+      "grad_norm": 0.02760649099946022,
+      "learning_rate": 1.1176368238983997e-06,
+      "loss": 0.0001,
+      "step": 2669
+    },
+    {
+      "epoch": 2.3688760806916425,
+      "grad_norm": 0.008239630609750748,
+      "learning_rate": 1.1146217225528255e-06,
+      "loss": 0.0,
+      "step": 2670
+    },
+    {
+      "epoch": 2.369762802039459,
+      "grad_norm": 0.0009814087534323335,
+      "learning_rate": 1.111610183425786e-06,
+      "loss": 0.0,
+      "step": 2671
+    },
+    {
+      "epoch": 2.3706495233872755,
+      "grad_norm": 0.010030284523963928,
+      "learning_rate": 1.1086022092783422e-06,
+      "loss": 0.0,
+      "step": 2672
+    },
+    {
+      "epoch": 2.371536244735092,
+      "grad_norm": 0.0005799742066301405,
+      "learning_rate": 1.1055978028682858e-06,
+      "loss": 0.0,
+      "step": 2673
+    },
+    {
+      "epoch": 2.3724229660829086,
+      "grad_norm": 0.01730470359325409,
+      "learning_rate": 1.1025969669501357e-06,
+      "loss": 0.0001,
+      "step": 2674
+    },
+    {
+      "epoch": 2.3733096874307247,
+      "grad_norm": 0.14865989983081818,
+      "learning_rate": 1.09959970427514e-06,
+      "loss": 0.0004,
+      "step": 2675
+    },
+    {
+      "epoch": 2.374196408778541,
+      "grad_norm": 0.31325671076774597,
+      "learning_rate": 1.0966060175912658e-06,
+      "loss": 0.0007,
+      "step": 2676
+    },
+    {
+      "epoch": 2.3750831301263577,
+      "grad_norm": 0.01875937730073929,
+      "learning_rate": 1.0936159096432097e-06,
+      "loss": 0.0001,
+      "step": 2677
+    },
+    {
+      "epoch": 2.3759698514741743,
+      "grad_norm": 0.05218518152832985,
+      "learning_rate": 1.090629383172379e-06,
+      "loss": 0.0001,
+      "step": 2678
+    },
+    {
+      "epoch": 2.376856572821991,
+      "grad_norm": 0.03152287006378174,
+      "learning_rate": 1.0876464409169029e-06,
+      "loss": 0.0001,
+      "step": 2679
+    },
+    {
+      "epoch": 2.3777432941698073,
+      "grad_norm": 0.001231821603141725,
+      "learning_rate": 1.0846670856116193e-06,
+      "loss": 0.0,
+      "step": 2680
+    },
+    {
+      "epoch": 2.3786300155176234,
+      "grad_norm": 0.0006567314267158508,
+      "learning_rate": 1.0816913199880852e-06,
+      "loss": 0.0,
+      "step": 2681
+    },
+    {
+      "epoch": 2.37951673686544,
+      "grad_norm": 0.0030319185461848974,
+      "learning_rate": 1.0787191467745584e-06,
+      "loss": 0.0,
+      "step": 2682
+    },
+    {
+      "epoch": 2.3804034582132565,
+      "grad_norm": 0.010536277666687965,
+      "learning_rate": 1.075750568696008e-06,
+      "loss": 0.0,
+      "step": 2683
+    },
+    {
+      "epoch": 2.381290179561073,
+      "grad_norm": 0.07841084152460098,
+      "learning_rate": 1.0727855884741057e-06,
+      "loss": 0.0003,
+      "step": 2684
+    },
+    {
+      "epoch": 2.3821769009088896,
+      "grad_norm": 0.005802239757031202,
+      "learning_rate": 1.0698242088272253e-06,
+      "loss": 0.0,
+      "step": 2685
+    },
+    {
+      "epoch": 2.3830636222567056,
+      "grad_norm": 0.019029144197702408,
+      "learning_rate": 1.066866432470437e-06,
+      "loss": 0.0001,
+      "step": 2686
+    },
+    {
+      "epoch": 2.383950343604522,
+      "grad_norm": 0.0072951787151396275,
+      "learning_rate": 1.0639122621155102e-06,
+      "loss": 0.0,
+      "step": 2687
+    },
+    {
+      "epoch": 2.3848370649523387,
+      "grad_norm": 0.0025982209481298923,
+      "learning_rate": 1.0609617004709061e-06,
+      "loss": 0.0,
+      "step": 2688
+    },
+    {
+      "epoch": 2.3857237863001552,
+      "grad_norm": 0.001721179811283946,
+      "learning_rate": 1.058014750241781e-06,
+      "loss": 0.0,
+      "step": 2689
+    },
+    {
+      "epoch": 2.386610507647972,
+      "grad_norm": 0.006247865501791239,
+      "learning_rate": 1.055071414129973e-06,
+      "loss": 0.0,
+      "step": 2690
+    },
+    {
+      "epoch": 2.3874972289957883,
+      "grad_norm": 0.0005024709389545023,
+      "learning_rate": 1.0521316948340166e-06,
+      "loss": 0.0,
+      "step": 2691
+    },
+    {
+      "epoch": 2.3883839503436044,
+      "grad_norm": 0.011277176439762115,
+      "learning_rate": 1.0491955950491212e-06,
+      "loss": 0.0,
+      "step": 2692
+    },
+    {
+      "epoch": 2.389270671691421,
+      "grad_norm": 0.0024250997230410576,
+      "learning_rate": 1.0462631174671823e-06,
+      "loss": 0.0,
+      "step": 2693
+    },
+    {
+      "epoch": 2.3901573930392375,
+      "grad_norm": 0.10233346372842789,
+      "learning_rate": 1.0433342647767735e-06,
+      "loss": 0.0004,
+      "step": 2694
+    },
+    {
+      "epoch": 2.391044114387054,
+      "grad_norm": 0.0037988275289535522,
+      "learning_rate": 1.0404090396631477e-06,
+      "loss": 0.0,
+      "step": 2695
+    },
+    {
+      "epoch": 2.39193083573487,
+      "grad_norm": 0.003683975664898753,
+      "learning_rate": 1.0374874448082257e-06,
+      "loss": 0.0,
+      "step": 2696
+    },
+    {
+      "epoch": 2.3928175570826866,
+      "grad_norm": 0.371255487203598,
+      "learning_rate": 1.034569482890605e-06,
+      "loss": 0.0012,
+      "step": 2697
+    },
+    {
+      "epoch": 2.393704278430503,
+      "grad_norm": 0.2531975209712982,
+      "learning_rate": 1.0316551565855515e-06,
+      "loss": 0.0008,
+      "step": 2698
+    },
+    {
+      "epoch": 2.3945909997783197,
+      "grad_norm": 0.008249549195170403,
+      "learning_rate": 1.0287444685649968e-06,
+      "loss": 0.0,
+      "step": 2699
+    },
+    {
+      "epoch": 2.395477721126136,
+      "grad_norm": 0.004005208145827055,
+      "learning_rate": 1.0258374214975386e-06,
+      "loss": 0.0,
+      "step": 2700
+    },
+    {
+      "epoch": 2.3963644424739527,
+      "grad_norm": 0.010472718626260757,
+      "learning_rate": 1.022934018048432e-06,
+      "loss": 0.0,
+      "step": 2701
+    },
+    {
+      "epoch": 2.397251163821769,
+      "grad_norm": 0.027334703132510185,
+      "learning_rate": 1.0200342608795998e-06,
+      "loss": 0.0001,
+      "step": 2702
+    },
+    {
+      "epoch": 2.3981378851695854,
+      "grad_norm": 0.0051247309893369675,
+      "learning_rate": 1.0171381526496132e-06,
+      "loss": 0.0,
+      "step": 2703
+    },
+    {
+      "epoch": 2.399024606517402,
+      "grad_norm": 0.6095221638679504,
+      "learning_rate": 1.0142456960137032e-06,
+      "loss": 0.0016,
+      "step": 2704
+    },
+    {
+      "epoch": 2.3999113278652184,
+      "grad_norm": 0.1208229809999466,
+      "learning_rate": 1.0113568936237478e-06,
+      "loss": 0.0001,
+      "step": 2705
+    },
+    {
+      "epoch": 2.400798049213035,
+      "grad_norm": 0.0789174810051918,
+      "learning_rate": 1.0084717481282842e-06,
+      "loss": 0.0002,
+      "step": 2706
+    },
+    {
+      "epoch": 2.401684770560851,
+      "grad_norm": 0.02274036966264248,
+      "learning_rate": 1.0055902621724855e-06,
+      "loss": 0.0001,
+      "step": 2707
+    },
+    {
+      "epoch": 2.4025714919086676,
+      "grad_norm": 0.011838889680802822,
+      "learning_rate": 1.0027124383981768e-06,
+      "loss": 0.0,
+      "step": 2708
+    },
+    {
+      "epoch": 2.403458213256484,
+      "grad_norm": 0.11658446490764618,
+      "learning_rate": 9.998382794438243e-07,
+      "loss": 0.0004,
+      "step": 2709
+    },
+    {
+      "epoch": 2.4043449346043007,
+      "grad_norm": 0.002783681731671095,
+      "learning_rate": 9.969677879445344e-07,
+      "loss": 0.0,
+      "step": 2710
+    },
+    {
+      "epoch": 2.405231655952117,
+      "grad_norm": 0.0004409880784805864,
+      "learning_rate": 9.941009665320483e-07,
+      "loss": 0.0,
+      "step": 2711
+    },
+    {
+      "epoch": 2.4061183772999337,
+      "grad_norm": 0.23596900701522827,
+      "learning_rate": 9.912378178347448e-07,
+      "loss": 0.0012,
+      "step": 2712
+    },
+    {
+      "epoch": 2.40700509864775,
+      "grad_norm": 0.1281604915857315,
+      "learning_rate": 9.883783444776357e-07,
+      "loss": 0.0008,
+      "step": 2713
+    },
+    {
+      "epoch": 2.4078918199955663,
+      "grad_norm": 0.0167441014200449,
+      "learning_rate": 9.855225490823639e-07,
+      "loss": 0.0001,
+      "step": 2714
+    },
+    {
+      "epoch": 2.408778541343383,
+      "grad_norm": 0.009180407971143723,
+      "learning_rate": 9.826704342671945e-07,
+      "loss": 0.0,
+      "step": 2715
+    },
+    {
+      "epoch": 2.4096652626911994,
+      "grad_norm": 0.020933065563440323,
+      "learning_rate": 9.798220026470285e-07,
+      "loss": 0.0001,
+      "step": 2716
+    },
+    {
+      "epoch": 2.4105519840390155,
+      "grad_norm": 0.012441801838576794,
+      "learning_rate": 9.76977256833379e-07,
+      "loss": 0.0,
+      "step": 2717
+    },
+    {
+      "epoch": 2.411438705386832,
+      "grad_norm": 0.13534188270568848,
+      "learning_rate": 9.741361994343867e-07,
+      "loss": 0.0007,
+      "step": 2718
+    },
+    {
+      "epoch": 2.4123254267346486,
+      "grad_norm": 0.004485164303332567,
+      "learning_rate": 9.712988330548096e-07,
+      "loss": 0.0,
+      "step": 2719
+    },
+    {
+      "epoch": 2.413212148082465,
+      "grad_norm": 0.005443104077130556,
+      "learning_rate": 9.6846516029602e-07,
+      "loss": 0.0,
+      "step": 2720
+    },
+    {
+      "epoch": 2.4140988694302816,
+      "grad_norm": 0.0010421723127365112,
+      "learning_rate": 9.65635183756007e-07,
+      "loss": 0.0,
+      "step": 2721
+    },
+    {
+      "epoch": 2.414985590778098,
+      "grad_norm": 0.1741264909505844,
+      "learning_rate": 9.628089060293643e-07,
+      "loss": 0.0009,
+      "step": 2722
+    },
+    {
+      "epoch": 2.4158723121259142,
+      "grad_norm": 0.05585017427802086,
+      "learning_rate": 9.59986329707302e-07,
+      "loss": 0.0002,
+      "step": 2723
+    },
+    {
+      "epoch": 2.4167590334737308,
+      "grad_norm": 0.0028329435735940933,
+      "learning_rate": 9.57167457377633e-07,
+      "loss": 0.0,
+      "step": 2724
+    },
+    {
+      "epoch": 2.4176457548215473,
+      "grad_norm": 0.002247181022539735,
+      "learning_rate": 9.543522916247749e-07,
+      "loss": 0.0,
+      "step": 2725
+    },
+    {
+      "epoch": 2.418532476169364,
+      "grad_norm": 0.020004788413643837,
+      "learning_rate": 9.515408350297439e-07,
+      "loss": 0.0001,
+      "step": 2726
+    },
+    {
+      "epoch": 2.4194191975171804,
+      "grad_norm": 0.012717186473309994,
+      "learning_rate": 9.487330901701636e-07,
+      "loss": 0.0,
+      "step": 2727
+    },
+    {
+      "epoch": 2.4203059188649965,
+      "grad_norm": 0.012186716310679913,
+      "learning_rate": 9.459290596202453e-07,
+      "loss": 0.0,
+      "step": 2728
+    },
+    {
+      "epoch": 2.421192640212813,
+      "grad_norm": 0.003059729468077421,
+      "learning_rate": 9.431287459508004e-07,
+      "loss": 0.0,
+      "step": 2729
+    },
+    {
+      "epoch": 2.4220793615606295,
+      "grad_norm": 0.005862931255251169,
+      "learning_rate": 9.403321517292313e-07,
+      "loss": 0.0,
+      "step": 2730
+    },
+    {
+      "epoch": 2.422966082908446,
+      "grad_norm": 0.012926910072565079,
+      "learning_rate": 9.375392795195315e-07,
+      "loss": 0.0,
+      "step": 2731
+    },
+    {
+      "epoch": 2.4238528042562626,
+      "grad_norm": 0.009246543981134892,
+      "learning_rate": 9.347501318822783e-07,
+      "loss": 0.0,
+      "step": 2732
+    },
+    {
+      "epoch": 2.424739525604079,
+      "grad_norm": 0.033203281462192535,
+      "learning_rate": 9.319647113746383e-07,
+      "loss": 0.0001,
+      "step": 2733
+    },
+    {
+      "epoch": 2.425626246951895,
+      "grad_norm": 0.0029225186444818974,
+      "learning_rate": 9.291830205503588e-07,
+      "loss": 0.0,
+      "step": 2734
+    },
+    {
+      "epoch": 2.4265129682997117,
+      "grad_norm": 0.00041219202103093266,
+      "learning_rate": 9.264050619597697e-07,
+      "loss": 0.0,
+      "step": 2735
+    },
+    {
+      "epoch": 2.4273996896475283,
+      "grad_norm": 0.012741847895085812,
+      "learning_rate": 9.236308381497738e-07,
+      "loss": 0.0001,
+      "step": 2736
+    },
+    {
+      "epoch": 2.428286410995345,
+      "grad_norm": 0.04531732574105263,
+      "learning_rate": 9.208603516638587e-07,
+      "loss": 0.0002,
+      "step": 2737
+    },
+    {
+      "epoch": 2.4291731323431613,
+      "grad_norm": 0.006424774881452322,
+      "learning_rate": 9.180936050420763e-07,
+      "loss": 0.0,
+      "step": 2738
+    },
+    {
+      "epoch": 2.4300598536909774,
+      "grad_norm": 0.000581156462430954,
+      "learning_rate": 9.153306008210572e-07,
+      "loss": 0.0,
+      "step": 2739
+    },
+    {
+      "epoch": 2.430946575038794,
+      "grad_norm": 0.016396358609199524,
+      "learning_rate": 9.125713415339921e-07,
+      "loss": 0.0,
+      "step": 2740
+    },
+    {
+      "epoch": 2.4318332963866105,
+      "grad_norm": 0.0013209610478952527,
+      "learning_rate": 9.098158297106501e-07,
+      "loss": 0.0,
+      "step": 2741
+    },
+    {
+      "epoch": 2.432720017734427,
+      "grad_norm": 0.013338295742869377,
+      "learning_rate": 9.070640678773535e-07,
+      "loss": 0.0,
+      "step": 2742
+    },
+    {
+      "epoch": 2.4336067390822436,
+      "grad_norm": 0.002022724598646164,
+      "learning_rate": 9.043160585569926e-07,
+      "loss": 0.0,
+      "step": 2743
+    },
+    {
+      "epoch": 2.43449346043006,
+      "grad_norm": 0.3681130111217499,
+      "learning_rate": 9.015718042690152e-07,
+      "loss": 0.0006,
+      "step": 2744
+    },
+    {
+      "epoch": 2.435380181777876,
+      "grad_norm": 0.24407528340816498,
+      "learning_rate": 8.988313075294275e-07,
+      "loss": 0.0019,
+      "step": 2745
+    },
+    {
+      "epoch": 2.4362669031256927,
+      "grad_norm": 0.0003583901561796665,
+      "learning_rate": 8.960945708507907e-07,
+      "loss": 0.0,
+      "step": 2746
+    },
+    {
+      "epoch": 2.4371536244735093,
+      "grad_norm": 0.0018106335774064064,
+      "learning_rate": 8.93361596742216e-07,
+      "loss": 0.0,
+      "step": 2747
+    },
+    {
+      "epoch": 2.438040345821326,
+      "grad_norm": 1.3477489948272705,
+      "learning_rate": 8.906323877093681e-07,
+      "loss": 0.0048,
+      "step": 2748
+    },
+    {
+      "epoch": 2.438927067169142,
+      "grad_norm": 0.0024084593169391155,
+      "learning_rate": 8.879069462544593e-07,
+      "loss": 0.0,
+      "step": 2749
+    },
+    {
+      "epoch": 2.4398137885169584,
+      "grad_norm": 0.0646350234746933,
+      "learning_rate": 8.851852748762457e-07,
+      "loss": 0.0001,
+      "step": 2750
+    },
+    {
+      "epoch": 2.440700509864775,
+      "grad_norm": 0.13703598082065582,
+      "learning_rate": 8.824673760700298e-07,
+      "loss": 0.0002,
+      "step": 2751
+    },
+    {
+      "epoch": 2.4415872312125915,
+      "grad_norm": 0.0044349804520606995,
+      "learning_rate": 8.797532523276542e-07,
+      "loss": 0.0,
+      "step": 2752
+    },
+    {
+      "epoch": 2.442473952560408,
+      "grad_norm": 0.005696154665201902,
+      "learning_rate": 8.770429061374979e-07,
+      "loss": 0.0,
+      "step": 2753
+    },
+    {
+      "epoch": 2.4433606739082245,
+      "grad_norm": 0.0011021459940820932,
+      "learning_rate": 8.743363399844795e-07,
+      "loss": 0.0,
+      "step": 2754
+    },
+    {
+      "epoch": 2.4442473952560406,
+      "grad_norm": 0.051142916083335876,
+      "learning_rate": 8.716335563500511e-07,
+      "loss": 0.0001,
+      "step": 2755
+    },
+    {
+      "epoch": 2.445134116603857,
+      "grad_norm": 0.5838609933853149,
+      "learning_rate": 8.689345577121988e-07,
+      "loss": 0.0019,
+      "step": 2756
+    },
+    {
+      "epoch": 2.4460208379516737,
+      "grad_norm": 0.015151509083807468,
+      "learning_rate": 8.662393465454322e-07,
+      "loss": 0.0,
+      "step": 2757
+    },
+    {
+      "epoch": 2.44690755929949,
+      "grad_norm": 0.000398889125790447,
+      "learning_rate": 8.635479253207979e-07,
+      "loss": 0.0,
+      "step": 2758
+    },
+    {
+      "epoch": 2.4477942806473068,
+      "grad_norm": 0.08696360141038895,
+      "learning_rate": 8.608602965058594e-07,
+      "loss": 0.0003,
+      "step": 2759
+    },
+    {
+      "epoch": 2.448681001995123,
+      "grad_norm": 0.001075335661880672,
+      "learning_rate": 8.581764625647088e-07,
+      "loss": 0.0,
+      "step": 2760
+    },
+    {
+      "epoch": 2.4495677233429394,
+      "grad_norm": 0.025259824469685555,
+      "learning_rate": 8.554964259579524e-07,
+      "loss": 0.0001,
+      "step": 2761
+    },
+    {
+      "epoch": 2.450454444690756,
+      "grad_norm": 0.001714448444545269,
+      "learning_rate": 8.528201891427257e-07,
+      "loss": 0.0,
+      "step": 2762
+    },
+    {
+      "epoch": 2.4513411660385724,
+      "grad_norm": 0.005722302943468094,
+      "learning_rate": 8.501477545726688e-07,
+      "loss": 0.0,
+      "step": 2763
+    },
+    {
+      "epoch": 2.452227887386389,
+      "grad_norm": 0.016258642077445984,
+      "learning_rate": 8.474791246979436e-07,
+      "loss": 0.0001,
+      "step": 2764
+    },
+    {
+      "epoch": 2.4531146087342055,
+      "grad_norm": 0.005662992130964994,
+      "learning_rate": 8.448143019652199e-07,
+      "loss": 0.0,
+      "step": 2765
+    },
+    {
+      "epoch": 2.4540013300820216,
+      "grad_norm": 0.006093035452067852,
+      "learning_rate": 8.421532888176808e-07,
+      "loss": 0.0,
+      "step": 2766
+    },
+    {
+      "epoch": 2.454888051429838,
+      "grad_norm": 0.38589486479759216,
+      "learning_rate": 8.394960876950109e-07,
+      "loss": 0.003,
+      "step": 2767
+    },
+    {
+      "epoch": 2.4557747727776547,
+      "grad_norm": 0.007774090860038996,
+      "learning_rate": 8.368427010334052e-07,
+      "loss": 0.0,
+      "step": 2768
+    },
+    {
+      "epoch": 2.456661494125471,
+      "grad_norm": 0.00017777753237169236,
+      "learning_rate": 8.341931312655582e-07,
+      "loss": 0.0,
+      "step": 2769
+    },
+    {
+      "epoch": 2.4575482154732873,
+      "grad_norm": 0.04939320310950279,
+      "learning_rate": 8.315473808206676e-07,
+      "loss": 0.0001,
+      "step": 2770
+    },
+    {
+      "epoch": 2.458434936821104,
+      "grad_norm": 0.0007739551365375519,
+      "learning_rate": 8.289054521244266e-07,
+      "loss": 0.0,
+      "step": 2771
+    },
+    {
+      "epoch": 2.4593216581689203,
+      "grad_norm": 0.015925316140055656,
+      "learning_rate": 8.262673475990285e-07,
+      "loss": 0.0001,
+      "step": 2772
+    },
+    {
+      "epoch": 2.460208379516737,
+      "grad_norm": 0.1478816717863083,
+      "learning_rate": 8.236330696631545e-07,
+      "loss": 0.0006,
+      "step": 2773
+    },
+    {
+      "epoch": 2.4610951008645534,
+      "grad_norm": 0.009654730558395386,
+      "learning_rate": 8.210026207319827e-07,
+      "loss": 0.0,
+      "step": 2774
+    },
+    {
+      "epoch": 2.46198182221237,
+      "grad_norm": 0.018739847466349602,
+      "learning_rate": 8.183760032171789e-07,
+      "loss": 0.0001,
+      "step": 2775
+    },
+    {
+      "epoch": 2.462868543560186,
+      "grad_norm": 0.011451950296759605,
+      "learning_rate": 8.157532195268963e-07,
+      "loss": 0.0,
+      "step": 2776
+    },
+    {
+      "epoch": 2.4637552649080026,
+      "grad_norm": 0.10962649434804916,
+      "learning_rate": 8.131342720657737e-07,
+      "loss": 0.0005,
+      "step": 2777
+    },
+    {
+      "epoch": 2.464641986255819,
+      "grad_norm": 0.01172646600753069,
+      "learning_rate": 8.105191632349291e-07,
+      "loss": 0.0,
+      "step": 2778
+    },
+    {
+      "epoch": 2.4655287076036356,
+      "grad_norm": 0.0004526283300947398,
+      "learning_rate": 8.07907895431968e-07,
+      "loss": 0.0,
+      "step": 2779
+    },
+    {
+      "epoch": 2.466415428951452,
+      "grad_norm": 0.012956739403307438,
+      "learning_rate": 8.053004710509676e-07,
+      "loss": 0.0001,
+      "step": 2780
+    },
+    {
+      "epoch": 2.4673021502992682,
+      "grad_norm": 0.0058890231885015965,
+      "learning_rate": 8.026968924824868e-07,
+      "loss": 0.0,
+      "step": 2781
+    },
+    {
+      "epoch": 2.468188871647085,
+      "grad_norm": 0.11938003450632095,
+      "learning_rate": 8.000971621135512e-07,
+      "loss": 0.0004,
+      "step": 2782
+    },
+    {
+      "epoch": 2.4690755929949013,
+      "grad_norm": 0.0003406075993552804,
+      "learning_rate": 7.97501282327669e-07,
+      "loss": 0.0,
+      "step": 2783
+    },
+    {
+      "epoch": 2.469962314342718,
+      "grad_norm": 0.0017089827451854944,
+      "learning_rate": 7.949092555048077e-07,
+      "loss": 0.0,
+      "step": 2784
+    },
+    {
+      "epoch": 2.4708490356905344,
+      "grad_norm": 0.03521645814180374,
+      "learning_rate": 7.923210840214085e-07,
+      "loss": 0.0001,
+      "step": 2785
+    },
+    {
+      "epoch": 2.471735757038351,
+      "grad_norm": 0.0007434281869791448,
+      "learning_rate": 7.897367702503755e-07,
+      "loss": 0.0,
+      "step": 2786
+    },
+    {
+      "epoch": 2.472622478386167,
+      "grad_norm": 0.0026796984020620584,
+      "learning_rate": 7.871563165610779e-07,
+      "loss": 0.0,
+      "step": 2787
+    },
+    {
+      "epoch": 2.4735091997339835,
+      "grad_norm": 0.0020774805452674627,
+      "learning_rate": 7.845797253193427e-07,
+      "loss": 0.0,
+      "step": 2788
+    },
+    {
+      "epoch": 2.4743959210818,
+      "grad_norm": 0.015672534704208374,
+      "learning_rate": 7.820069988874585e-07,
+      "loss": 0.0,
+      "step": 2789
+    },
+    {
+      "epoch": 2.4752826424296166,
+      "grad_norm": 0.014547284692525864,
+      "learning_rate": 7.79438139624169e-07,
+      "loss": 0.0001,
+      "step": 2790
+    },
+    {
+      "epoch": 2.476169363777433,
+      "grad_norm": 0.15646186470985413,
+      "learning_rate": 7.768731498846749e-07,
+      "loss": 0.0006,
+      "step": 2791
+    },
+    {
+      "epoch": 2.477056085125249,
+      "grad_norm": 0.0015324123669415712,
+      "learning_rate": 7.743120320206232e-07,
+      "loss": 0.0,
+      "step": 2792
+    },
+    {
+      "epoch": 2.4779428064730658,
+      "grad_norm": 0.004034361802041531,
+      "learning_rate": 7.717547883801197e-07,
+      "loss": 0.0,
+      "step": 2793
+    },
+    {
+      "epoch": 2.4788295278208823,
+      "grad_norm": 0.000717060174793005,
+      "learning_rate": 7.692014213077109e-07,
+      "loss": 0.0,
+      "step": 2794
+    },
+    {
+      "epoch": 2.479716249168699,
+      "grad_norm": 0.060480207204818726,
+      "learning_rate": 7.666519331443928e-07,
+      "loss": 0.0002,
+      "step": 2795
+    },
+    {
+      "epoch": 2.4806029705165153,
+      "grad_norm": 0.002845141338184476,
+      "learning_rate": 7.641063262276039e-07,
+      "loss": 0.0,
+      "step": 2796
+    },
+    {
+      "epoch": 2.481489691864332,
+      "grad_norm": 0.11835973709821701,
+      "learning_rate": 7.615646028912243e-07,
+      "loss": 0.0002,
+      "step": 2797
+    },
+    {
+      "epoch": 2.482376413212148,
+      "grad_norm": 0.04561325162649155,
+      "learning_rate": 7.590267654655758e-07,
+      "loss": 0.0001,
+      "step": 2798
+    },
+    {
+      "epoch": 2.4832631345599645,
+      "grad_norm": 0.09880010783672333,
+      "learning_rate": 7.564928162774127e-07,
+      "loss": 0.0003,
+      "step": 2799
+    },
+    {
+      "epoch": 2.484149855907781,
+      "grad_norm": 0.4014248847961426,
+      "learning_rate": 7.539627576499287e-07,
+      "loss": 0.0002,
+      "step": 2800
+    },
+    {
+      "epoch": 2.4850365772555976,
+      "grad_norm": 0.00363646587356925,
+      "learning_rate": 7.514365919027483e-07,
+      "loss": 0.0,
+      "step": 2801
+    },
+    {
+      "epoch": 2.4859232986034137,
+      "grad_norm": 0.6536305546760559,
+      "learning_rate": 7.489143213519301e-07,
+      "loss": 0.0043,
+      "step": 2802
+    },
+    {
+      "epoch": 2.48681001995123,
+      "grad_norm": 0.00101925665512681,
+      "learning_rate": 7.463959483099547e-07,
+      "loss": 0.0,
+      "step": 2803
+    },
+    {
+      "epoch": 2.4876967412990467,
+      "grad_norm": 0.0003748614399228245,
+      "learning_rate": 7.438814750857387e-07,
+      "loss": 0.0,
+      "step": 2804
+    },
+    {
+      "epoch": 2.4885834626468633,
+      "grad_norm": 0.1380384862422943,
+      "learning_rate": 7.41370903984615e-07,
+      "loss": 0.0001,
+      "step": 2805
+    },
+    {
+      "epoch": 2.48947018399468,
+      "grad_norm": 0.028476646170020103,
+      "learning_rate": 7.388642373083438e-07,
+      "loss": 0.0001,
+      "step": 2806
+    },
+    {
+      "epoch": 2.4903569053424963,
+      "grad_norm": 0.01267907302826643,
+      "learning_rate": 7.363614773551015e-07,
+      "loss": 0.0,
+      "step": 2807
+    },
+    {
+      "epoch": 2.4912436266903124,
+      "grad_norm": 0.005699229426681995,
+      "learning_rate": 7.338626264194893e-07,
+      "loss": 0.0,
+      "step": 2808
+    },
+    {
+      "epoch": 2.492130348038129,
+      "grad_norm": 0.0008996912511065602,
+      "learning_rate": 7.313676867925168e-07,
+      "loss": 0.0,
+      "step": 2809
+    },
+    {
+      "epoch": 2.4930170693859455,
+      "grad_norm": 0.0016446348745375872,
+      "learning_rate": 7.288766607616132e-07,
+      "loss": 0.0,
+      "step": 2810
+    },
+    {
+      "epoch": 2.493903790733762,
+      "grad_norm": 0.0015573484124615788,
+      "learning_rate": 7.263895506106167e-07,
+      "loss": 0.0,
+      "step": 2811
+    },
+    {
+      "epoch": 2.4947905120815785,
+      "grad_norm": 0.7243852615356445,
+      "learning_rate": 7.239063586197786e-07,
+      "loss": 0.0028,
+      "step": 2812
+    },
+    {
+      "epoch": 2.4956772334293946,
+      "grad_norm": 0.030200662091374397,
+      "learning_rate": 7.21427087065752e-07,
+      "loss": 0.0001,
+      "step": 2813
+    },
+    {
+      "epoch": 2.496563954777211,
+      "grad_norm": 0.0018416463863104582,
+      "learning_rate": 7.18951738221601e-07,
+      "loss": 0.0,
+      "step": 2814
+    },
+    {
+      "epoch": 2.4974506761250277,
+      "grad_norm": 0.000541276705916971,
+      "learning_rate": 7.164803143567922e-07,
+      "loss": 0.0,
+      "step": 2815
+    },
+    {
+      "epoch": 2.4983373974728442,
+      "grad_norm": 0.0023916105274111032,
+      "learning_rate": 7.14012817737193e-07,
+      "loss": 0.0,
+      "step": 2816
+    },
+    {
+      "epoch": 2.4992241188206608,
+      "grad_norm": 0.0036319156643003225,
+      "learning_rate": 7.11549250625067e-07,
+      "loss": 0.0,
+      "step": 2817
+    },
+    {
+      "epoch": 2.5001108401684773,
+      "grad_norm": 0.020262181758880615,
+      "learning_rate": 7.090896152790838e-07,
+      "loss": 0.0001,
+      "step": 2818
+    },
+    {
+      "epoch": 2.5009975615162934,
+      "grad_norm": 0.0017561853164806962,
+      "learning_rate": 7.066339139542983e-07,
+      "loss": 0.0,
+      "step": 2819
+    },
+    {
+      "epoch": 2.50188428286411,
+      "grad_norm": 0.0007896814495325089,
+      "learning_rate": 7.041821489021639e-07,
+      "loss": 0.0,
+      "step": 2820
+    },
+    {
+      "epoch": 2.50188428286411,
+      "eval_loss": 0.02093465067446232,
+      "eval_runtime": 60.5435,
+      "eval_samples_per_second": 3.171,
+      "eval_steps_per_second": 0.793,
+      "step": 2820
+    },
+    {
+      "epoch": 2.5027710042119264,
+      "grad_norm": 0.2156534045934677,
+      "learning_rate": 7.017343223705253e-07,
+      "loss": 0.0009,
+      "step": 2821
+    },
+    {
+      "epoch": 2.503657725559743,
+      "grad_norm": 0.0018699497450143099,
+      "learning_rate": 6.992904366036152e-07,
+      "loss": 0.0,
+      "step": 2822
+    },
+    {
+      "epoch": 2.504544446907559,
+      "grad_norm": 0.012214919552206993,
+      "learning_rate": 6.968504938420534e-07,
+      "loss": 0.0,
+      "step": 2823
+    },
+    {
+      "epoch": 2.5054311682553756,
+      "grad_norm": 0.0025382922030985355,
+      "learning_rate": 6.944144963228433e-07,
+      "loss": 0.0,
+      "step": 2824
+    },
+    {
+      "epoch": 2.506317889603192,
+      "grad_norm": 0.004923071712255478,
+      "learning_rate": 6.919824462793728e-07,
+      "loss": 0.0,
+      "step": 2825
+    },
+    {
+      "epoch": 2.5072046109510087,
+      "grad_norm": 0.06211470812559128,
+      "learning_rate": 6.895543459414112e-07,
+      "loss": 0.0003,
+      "step": 2826
+    },
+    {
+      "epoch": 2.508091332298825,
+      "grad_norm": 0.0008407292189076543,
+      "learning_rate": 6.871301975351063e-07,
+      "loss": 0.0,
+      "step": 2827
+    },
+    {
+      "epoch": 2.5089780536466417,
+      "grad_norm": 0.007816070690751076,
+      "learning_rate": 6.847100032829784e-07,
+      "loss": 0.0,
+      "step": 2828
+    },
+    {
+      "epoch": 2.5098647749944583,
+      "grad_norm": 0.06127519533038139,
+      "learning_rate": 6.822937654039313e-07,
+      "loss": 0.0001,
+      "step": 2829
+    },
+    {
+      "epoch": 2.5107514963422743,
+      "grad_norm": 0.010569370351731777,
+      "learning_rate": 6.79881486113233e-07,
+      "loss": 0.0,
+      "step": 2830
+    },
+    {
+      "epoch": 2.511638217690091,
+      "grad_norm": 0.000980584416538477,
+      "learning_rate": 6.774731676225261e-07,
+      "loss": 0.0,
+      "step": 2831
+    },
+    {
+      "epoch": 2.5125249390379074,
+      "grad_norm": 0.0009631279390305281,
+      "learning_rate": 6.750688121398219e-07,
+      "loss": 0.0,
+      "step": 2832
+    },
+    {
+      "epoch": 2.513411660385724,
+      "grad_norm": 0.002405727282166481,
+      "learning_rate": 6.726684218694984e-07,
+      "loss": 0.0,
+      "step": 2833
+    },
+    {
+      "epoch": 2.51429838173354,
+      "grad_norm": 0.0005489909672178328,
+      "learning_rate": 6.702719990122958e-07,
+      "loss": 0.0,
+      "step": 2834
+    },
+    {
+      "epoch": 2.5151851030813566,
+      "grad_norm": 0.005501510575413704,
+      "learning_rate": 6.678795457653186e-07,
+      "loss": 0.0,
+      "step": 2835
+    },
+    {
+      "epoch": 2.516071824429173,
+      "grad_norm": 0.003498939797282219,
+      "learning_rate": 6.65491064322033e-07,
+      "loss": 0.0,
+      "step": 2836
+    },
+    {
+      "epoch": 2.5169585457769896,
+      "grad_norm": 0.12416200339794159,
+      "learning_rate": 6.631065568722633e-07,
+      "loss": 0.0002,
+      "step": 2837
+    },
+    {
+      "epoch": 2.517845267124806,
+      "grad_norm": 0.0013328570639714599,
+      "learning_rate": 6.607260256021864e-07,
+      "loss": 0.0,
+      "step": 2838
+    },
+    {
+      "epoch": 2.5187319884726227,
+      "grad_norm": 0.00046582092181779444,
+      "learning_rate": 6.583494726943401e-07,
+      "loss": 0.0,
+      "step": 2839
+    },
+    {
+      "epoch": 2.519618709820439,
+      "grad_norm": 0.13117870688438416,
+      "learning_rate": 6.559769003276101e-07,
+      "loss": 0.0002,
+      "step": 2840
+    },
+    {
+      "epoch": 2.5205054311682553,
+      "grad_norm": 0.006309604272246361,
+      "learning_rate": 6.536083106772351e-07,
+      "loss": 0.0,
+      "step": 2841
+    },
+    {
+      "epoch": 2.521392152516072,
+      "grad_norm": 0.0005661431350745261,
+      "learning_rate": 6.512437059147986e-07,
+      "loss": 0.0,
+      "step": 2842
+    },
+    {
+      "epoch": 2.5222788738638884,
+      "grad_norm": 0.0011204613838344812,
+      "learning_rate": 6.48883088208237e-07,
+      "loss": 0.0,
+      "step": 2843
+    },
+    {
+      "epoch": 2.5231655952117045,
+      "grad_norm": 0.0654761791229248,
+      "learning_rate": 6.465264597218252e-07,
+      "loss": 0.0002,
+      "step": 2844
+    },
+    {
+      "epoch": 2.524052316559521,
+      "grad_norm": 0.021184800192713737,
+      "learning_rate": 6.441738226161842e-07,
+      "loss": 0.0,
+      "step": 2845
+    },
+    {
+      "epoch": 2.5249390379073375,
+      "grad_norm": 0.3001684546470642,
+      "learning_rate": 6.418251790482738e-07,
+      "loss": 0.0029,
+      "step": 2846
+    },
+    {
+      "epoch": 2.525825759255154,
+      "grad_norm": 0.09699290245771408,
+      "learning_rate": 6.394805311713936e-07,
+      "loss": 0.0003,
+      "step": 2847
+    },
+    {
+      "epoch": 2.5267124806029706,
+      "grad_norm": 0.021082287654280663,
+      "learning_rate": 6.371398811351804e-07,
+      "loss": 0.0001,
+      "step": 2848
+    },
+    {
+      "epoch": 2.527599201950787,
+      "grad_norm": 0.015293094329535961,
+      "learning_rate": 6.348032310856022e-07,
+      "loss": 0.0,
+      "step": 2849
+    },
+    {
+      "epoch": 2.5284859232986037,
+      "grad_norm": 0.00025907374219968915,
+      "learning_rate": 6.324705831649636e-07,
+      "loss": 0.0,
+      "step": 2850
+    },
+    {
+      "epoch": 2.5293726446464198,
+      "grad_norm": 0.003451921045780182,
+      "learning_rate": 6.301419395118985e-07,
+      "loss": 0.0,
+      "step": 2851
+    },
+    {
+      "epoch": 2.5302593659942363,
+      "grad_norm": 0.2557615041732788,
+      "learning_rate": 6.278173022613693e-07,
+      "loss": 0.0012,
+      "step": 2852
+    },
+    {
+      "epoch": 2.531146087342053,
+      "grad_norm": 0.020587336272001266,
+      "learning_rate": 6.25496673544666e-07,
+      "loss": 0.0001,
+      "step": 2853
+    },
+    {
+      "epoch": 2.5320328086898694,
+      "grad_norm": 0.004580834414809942,
+      "learning_rate": 6.231800554894029e-07,
+      "loss": 0.0,
+      "step": 2854
+    },
+    {
+      "epoch": 2.5329195300376854,
+      "grad_norm": 0.004284840542823076,
+      "learning_rate": 6.208674502195167e-07,
+      "loss": 0.0,
+      "step": 2855
+    },
+    {
+      "epoch": 2.533806251385502,
+      "grad_norm": 0.0342896431684494,
+      "learning_rate": 6.185588598552655e-07,
+      "loss": 0.0001,
+      "step": 2856
+    },
+    {
+      "epoch": 2.5346929727333185,
+      "grad_norm": 0.002259077737107873,
+      "learning_rate": 6.162542865132277e-07,
+      "loss": 0.0,
+      "step": 2857
+    },
+    {
+      "epoch": 2.535579694081135,
+      "grad_norm": 0.7990791201591492,
+      "learning_rate": 6.139537323062972e-07,
+      "loss": 0.0023,
+      "step": 2858
+    },
+    {
+      "epoch": 2.5364664154289516,
+      "grad_norm": 0.0004105519037693739,
+      "learning_rate": 6.116571993436815e-07,
+      "loss": 0.0,
+      "step": 2859
+    },
+    {
+      "epoch": 2.537353136776768,
+      "grad_norm": 0.001817844226025045,
+      "learning_rate": 6.093646897309074e-07,
+      "loss": 0.0,
+      "step": 2860
+    },
+    {
+      "epoch": 2.538239858124584,
+      "grad_norm": 0.05438079684972763,
+      "learning_rate": 6.070762055698054e-07,
+      "loss": 0.0001,
+      "step": 2861
+    },
+    {
+      "epoch": 2.5391265794724007,
+      "grad_norm": 0.03450392559170723,
+      "learning_rate": 6.047917489585208e-07,
+      "loss": 0.0001,
+      "step": 2862
+    },
+    {
+      "epoch": 2.5400133008202173,
+      "grad_norm": 0.21090054512023926,
+      "learning_rate": 6.025113219915024e-07,
+      "loss": 0.0005,
+      "step": 2863
+    },
+    {
+      "epoch": 2.540900022168034,
+      "grad_norm": 0.022419922053813934,
+      "learning_rate": 6.002349267595092e-07,
+      "loss": 0.0,
+      "step": 2864
+    },
+    {
+      "epoch": 2.54178674351585,
+      "grad_norm": 0.0006861055735498667,
+      "learning_rate": 5.979625653495996e-07,
+      "loss": 0.0,
+      "step": 2865
+    },
+    {
+      "epoch": 2.5426734648636664,
+      "grad_norm": 0.0008977435645647347,
+      "learning_rate": 5.956942398451349e-07,
+      "loss": 0.0,
+      "step": 2866
+    },
+    {
+      "epoch": 2.543560186211483,
+      "grad_norm": 0.0007839131285436451,
+      "learning_rate": 5.934299523257769e-07,
+      "loss": 0.0,
+      "step": 2867
+    },
+    {
+      "epoch": 2.5444469075592995,
+      "grad_norm": 0.020759517326951027,
+      "learning_rate": 5.911697048674864e-07,
+      "loss": 0.0001,
+      "step": 2868
+    },
+    {
+      "epoch": 2.545333628907116,
+      "grad_norm": 0.028356509283185005,
+      "learning_rate": 5.889134995425155e-07,
+      "loss": 0.0001,
+      "step": 2869
+    },
+    {
+      "epoch": 2.5462203502549325,
+      "grad_norm": 0.001867869053967297,
+      "learning_rate": 5.866613384194153e-07,
+      "loss": 0.0,
+      "step": 2870
+    },
+    {
+      "epoch": 2.547107071602749,
+      "grad_norm": 0.029476385563611984,
+      "learning_rate": 5.844132235630273e-07,
+      "loss": 0.0002,
+      "step": 2871
+    },
+    {
+      "epoch": 2.547993792950565,
+      "grad_norm": 0.014416356571018696,
+      "learning_rate": 5.821691570344829e-07,
+      "loss": 0.0,
+      "step": 2872
+    },
+    {
+      "epoch": 2.5488805142983817,
+      "grad_norm": 0.0014337707543745637,
+      "learning_rate": 5.799291408912022e-07,
+      "loss": 0.0,
+      "step": 2873
+    },
+    {
+      "epoch": 2.5497672356461982,
+      "grad_norm": 0.029101770371198654,
+      "learning_rate": 5.776931771868932e-07,
+      "loss": 0.0001,
+      "step": 2874
+    },
+    {
+      "epoch": 2.5506539569940148,
+      "grad_norm": 0.2410978376865387,
+      "learning_rate": 5.754612679715443e-07,
+      "loss": 0.0012,
+      "step": 2875
+    },
+    {
+      "epoch": 2.551540678341831,
+      "grad_norm": 0.018282808363437653,
+      "learning_rate": 5.732334152914315e-07,
+      "loss": 0.0,
+      "step": 2876
+    },
+    {
+      "epoch": 2.5524273996896474,
+      "grad_norm": 0.00830524880439043,
+      "learning_rate": 5.710096211891087e-07,
+      "loss": 0.0,
+      "step": 2877
+    },
+    {
+      "epoch": 2.553314121037464,
+      "grad_norm": 0.2229091078042984,
+      "learning_rate": 5.687898877034104e-07,
+      "loss": 0.0003,
+      "step": 2878
+    },
+    {
+      "epoch": 2.5542008423852804,
+      "grad_norm": 0.12925313413143158,
+      "learning_rate": 5.665742168694483e-07,
+      "loss": 0.0003,
+      "step": 2879
+    },
+    {
+      "epoch": 2.555087563733097,
+      "grad_norm": 0.007797439582645893,
+      "learning_rate": 5.643626107186051e-07,
+      "loss": 0.0,
+      "step": 2880
+    },
+    {
+      "epoch": 2.5559742850809135,
+      "grad_norm": 0.12223513424396515,
+      "learning_rate": 5.621550712785445e-07,
+      "loss": 0.0001,
+      "step": 2881
+    },
+    {
+      "epoch": 2.55686100642873,
+      "grad_norm": 0.0494576133787632,
+      "learning_rate": 5.599516005731953e-07,
+      "loss": 0.0002,
+      "step": 2882
+    },
+    {
+      "epoch": 2.557747727776546,
+      "grad_norm": 0.010439654812216759,
+      "learning_rate": 5.577522006227598e-07,
+      "loss": 0.0,
+      "step": 2883
+    },
+    {
+      "epoch": 2.5586344491243627,
+      "grad_norm": 0.000774744083173573,
+      "learning_rate": 5.555568734437039e-07,
+      "loss": 0.0,
+      "step": 2884
+    },
+    {
+      "epoch": 2.559521170472179,
+      "grad_norm": 0.0021131117828190327,
+      "learning_rate": 5.533656210487654e-07,
+      "loss": 0.0,
+      "step": 2885
+    },
+    {
+      "epoch": 2.5604078918199957,
+      "grad_norm": 0.0041869510896503925,
+      "learning_rate": 5.511784454469404e-07,
+      "loss": 0.0,
+      "step": 2886
+    },
+    {
+      "epoch": 2.561294613167812,
+      "grad_norm": 0.010568622499704361,
+      "learning_rate": 5.489953486434913e-07,
+      "loss": 0.0,
+      "step": 2887
+    },
+    {
+      "epoch": 2.5621813345156284,
+      "grad_norm": 0.0009484928450547159,
+      "learning_rate": 5.46816332639939e-07,
+      "loss": 0.0,
+      "step": 2888
+    },
+    {
+      "epoch": 2.563068055863445,
+      "grad_norm": 0.0006807691534049809,
+      "learning_rate": 5.446413994340649e-07,
+      "loss": 0.0,
+      "step": 2889
+    },
+    {
+      "epoch": 2.5639547772112614,
+      "grad_norm": 0.7828634977340698,
+      "learning_rate": 5.424705510199036e-07,
+      "loss": 0.0015,
+      "step": 2890
+    },
+    {
+      "epoch": 2.564841498559078,
+      "grad_norm": 0.04002891480922699,
+      "learning_rate": 5.403037893877478e-07,
+      "loss": 0.0001,
+      "step": 2891
+    },
+    {
+      "epoch": 2.5657282199068945,
+      "grad_norm": 0.2507193982601166,
+      "learning_rate": 5.381411165241429e-07,
+      "loss": 0.0018,
+      "step": 2892
+    },
+    {
+      "epoch": 2.5666149412547106,
+      "grad_norm": 0.0006097395671531558,
+      "learning_rate": 5.359825344118858e-07,
+      "loss": 0.0,
+      "step": 2893
+    },
+    {
+      "epoch": 2.567501662602527,
+      "grad_norm": 0.048260822892189026,
+      "learning_rate": 5.338280450300187e-07,
+      "loss": 0.0001,
+      "step": 2894
+    },
+    {
+      "epoch": 2.5683883839503436,
+      "grad_norm": 0.029657047241926193,
+      "learning_rate": 5.316776503538401e-07,
+      "loss": 0.0001,
+      "step": 2895
+    },
+    {
+      "epoch": 2.56927510529816,
+      "grad_norm": 0.010401280596852303,
+      "learning_rate": 5.295313523548851e-07,
+      "loss": 0.0,
+      "step": 2896
+    },
+    {
+      "epoch": 2.5701618266459763,
+      "grad_norm": 0.0022809342481195927,
+      "learning_rate": 5.27389153000939e-07,
+      "loss": 0.0,
+      "step": 2897
+    },
+    {
+      "epoch": 2.571048547993793,
+      "grad_norm": 0.005268683657050133,
+      "learning_rate": 5.252510542560268e-07,
+      "loss": 0.0,
+      "step": 2898
+    },
+    {
+      "epoch": 2.5719352693416093,
+      "grad_norm": 0.0015263669192790985,
+      "learning_rate": 5.231170580804146e-07,
+      "loss": 0.0,
+      "step": 2899
+    },
+    {
+      "epoch": 2.572821990689426,
+      "grad_norm": 0.0002925303124357015,
+      "learning_rate": 5.209871664306076e-07,
+      "loss": 0.0,
+      "step": 2900
+    },
+    {
+      "epoch": 2.5737087120372424,
+      "grad_norm": 0.4949420094490051,
+      "learning_rate": 5.188613812593463e-07,
+      "loss": 0.0053,
+      "step": 2901
+    },
+    {
+      "epoch": 2.574595433385059,
+      "grad_norm": 0.005495188757777214,
+      "learning_rate": 5.167397045156075e-07,
+      "loss": 0.0,
+      "step": 2902
+    },
+    {
+      "epoch": 2.5754821547328755,
+      "grad_norm": 0.001007727812975645,
+      "learning_rate": 5.14622138144602e-07,
+      "loss": 0.0,
+      "step": 2903
+    },
+    {
+      "epoch": 2.5763688760806915,
+      "grad_norm": 0.003264603205025196,
+      "learning_rate": 5.125086840877707e-07,
+      "loss": 0.0,
+      "step": 2904
+    },
+    {
+      "epoch": 2.577255597428508,
+      "grad_norm": 0.001805190579034388,
+      "learning_rate": 5.103993442827832e-07,
+      "loss": 0.0,
+      "step": 2905
+    },
+    {
+      "epoch": 2.5781423187763246,
+      "grad_norm": 0.3015930950641632,
+      "learning_rate": 5.082941206635417e-07,
+      "loss": 0.0008,
+      "step": 2906
+    },
+    {
+      "epoch": 2.579029040124141,
+      "grad_norm": 0.14610321819782257,
+      "learning_rate": 5.06193015160169e-07,
+      "loss": 0.0002,
+      "step": 2907
+    },
+    {
+      "epoch": 2.5799157614719572,
+      "grad_norm": 0.003515577409416437,
+      "learning_rate": 5.04096029699016e-07,
+      "loss": 0.0,
+      "step": 2908
+    },
+    {
+      "epoch": 2.5808024828197738,
+      "grad_norm": 0.0014688996598124504,
+      "learning_rate": 5.020031662026526e-07,
+      "loss": 0.0,
+      "step": 2909
+    },
+    {
+      "epoch": 2.5816892041675903,
+      "grad_norm": 0.006348750554025173,
+      "learning_rate": 4.999144265898753e-07,
+      "loss": 0.0,
+      "step": 2910
+    },
+    {
+      "epoch": 2.582575925515407,
+      "grad_norm": 0.6952155232429504,
+      "learning_rate": 4.978298127756937e-07,
+      "loss": 0.001,
+      "step": 2911
+    },
+    {
+      "epoch": 2.5834626468632234,
+      "grad_norm": 0.004262842237949371,
+      "learning_rate": 4.957493266713387e-07,
+      "loss": 0.0,
+      "step": 2912
+    },
+    {
+      "epoch": 2.58434936821104,
+      "grad_norm": 0.0013588215224444866,
+      "learning_rate": 4.936729701842552e-07,
+      "loss": 0.0,
+      "step": 2913
+    },
+    {
+      "epoch": 2.585236089558856,
+      "grad_norm": 0.008707460016012192,
+      "learning_rate": 4.916007452181032e-07,
+      "loss": 0.0,
+      "step": 2914
+    },
+    {
+      "epoch": 2.5861228109066725,
+      "grad_norm": 0.20416337251663208,
+      "learning_rate": 4.895326536727518e-07,
+      "loss": 0.0003,
+      "step": 2915
+    },
+    {
+      "epoch": 2.587009532254489,
+      "grad_norm": 0.05300324037671089,
+      "learning_rate": 4.874686974442839e-07,
+      "loss": 0.0002,
+      "step": 2916
+    },
+    {
+      "epoch": 2.5878962536023056,
+      "grad_norm": 0.8284305334091187,
+      "learning_rate": 4.854088784249889e-07,
+      "loss": 0.0077,
+      "step": 2917
+    },
+    {
+      "epoch": 2.5887829749501217,
+      "grad_norm": 0.00847573857754469,
+      "learning_rate": 4.833531985033657e-07,
+      "loss": 0.0001,
+      "step": 2918
+    },
+    {
+      "epoch": 2.589669696297938,
+      "grad_norm": 0.05608953908085823,
+      "learning_rate": 4.813016595641135e-07,
+      "loss": 0.0001,
+      "step": 2919
+    },
+    {
+      "epoch": 2.5905564176457547,
+      "grad_norm": 0.003506513312458992,
+      "learning_rate": 4.792542634881414e-07,
+      "loss": 0.0,
+      "step": 2920
+    },
+    {
+      "epoch": 2.5914431389935713,
+      "grad_norm": 0.030793076381087303,
+      "learning_rate": 4.772110121525548e-07,
+      "loss": 0.0001,
+      "step": 2921
+    },
+    {
+      "epoch": 2.592329860341388,
+      "grad_norm": 0.1802815943956375,
+      "learning_rate": 4.751719074306604e-07,
+      "loss": 0.001,
+      "step": 2922
+    },
+    {
+      "epoch": 2.5932165816892043,
+      "grad_norm": 0.0012924959883093834,
+      "learning_rate": 4.731369511919653e-07,
+      "loss": 0.0,
+      "step": 2923
+    },
+    {
+      "epoch": 2.594103303037021,
+      "grad_norm": 0.002344125881791115,
+      "learning_rate": 4.7110614530217147e-07,
+      "loss": 0.0,
+      "step": 2924
+    },
+    {
+      "epoch": 2.594990024384837,
+      "grad_norm": 0.018599161878228188,
+      "learning_rate": 4.6907949162317667e-07,
+      "loss": 0.0001,
+      "step": 2925
+    },
+    {
+      "epoch": 2.5958767457326535,
+      "grad_norm": 0.0013417156878858805,
+      "learning_rate": 4.670569920130691e-07,
+      "loss": 0.0,
+      "step": 2926
+    },
+    {
+      "epoch": 2.59676346708047,
+      "grad_norm": 0.29868119955062866,
+      "learning_rate": 4.65038648326131e-07,
+      "loss": 0.0029,
+      "step": 2927
+    },
+    {
+      "epoch": 2.5976501884282865,
+      "grad_norm": 0.030218416824936867,
+      "learning_rate": 4.630244624128349e-07,
+      "loss": 0.0001,
+      "step": 2928
+    },
+    {
+      "epoch": 2.5985369097761026,
+      "grad_norm": 0.2323320060968399,
+      "learning_rate": 4.610144361198399e-07,
+      "loss": 0.0009,
+      "step": 2929
+    },
+    {
+      "epoch": 2.599423631123919,
+      "grad_norm": 0.0015396233648061752,
+      "learning_rate": 4.5900857128998997e-07,
+      "loss": 0.0,
+      "step": 2930
+    },
+    {
+      "epoch": 2.6003103524717357,
+      "grad_norm": 0.3064575493335724,
+      "learning_rate": 4.5700686976231865e-07,
+      "loss": 0.0006,
+      "step": 2931
+    },
+    {
+      "epoch": 2.6011970738195522,
+      "grad_norm": 0.0025372328236699104,
+      "learning_rate": 4.550093333720368e-07,
+      "loss": 0.0,
+      "step": 2932
+    },
+    {
+      "epoch": 2.6020837951673688,
+      "grad_norm": 0.0003127675736322999,
+      "learning_rate": 4.5301596395053993e-07,
+      "loss": 0.0,
+      "step": 2933
+    },
+    {
+      "epoch": 2.6029705165151853,
+      "grad_norm": 0.0011001265374943614,
+      "learning_rate": 4.5102676332540284e-07,
+      "loss": 0.0,
+      "step": 2934
+    },
+    {
+      "epoch": 2.603857237863002,
+      "grad_norm": 0.40902015566825867,
+      "learning_rate": 4.490417333203778e-07,
+      "loss": 0.0028,
+      "step": 2935
+    },
+    {
+      "epoch": 2.604743959210818,
+      "grad_norm": 0.06541017442941666,
+      "learning_rate": 4.470608757553918e-07,
+      "loss": 0.0002,
+      "step": 2936
+    },
+    {
+      "epoch": 2.6056306805586344,
+      "grad_norm": 0.08096663653850555,
+      "learning_rate": 4.4508419244654934e-07,
+      "loss": 0.0004,
+      "step": 2937
+    },
+    {
+      "epoch": 2.606517401906451,
+      "grad_norm": 0.11979025602340698,
+      "learning_rate": 4.431116852061257e-07,
+      "loss": 0.0006,
+      "step": 2938
+    },
+    {
+      "epoch": 2.607404123254267,
+      "grad_norm": 0.00492709269747138,
+      "learning_rate": 4.4114335584256986e-07,
+      "loss": 0.0,
+      "step": 2939
+    },
+    {
+      "epoch": 2.6082908446020836,
+      "grad_norm": 0.004521166905760765,
+      "learning_rate": 4.3917920616049546e-07,
+      "loss": 0.0,
+      "step": 2940
+    },
+    {
+      "epoch": 2.6091775659499,
+      "grad_norm": 0.0010443405481055379,
+      "learning_rate": 4.37219237960691e-07,
+      "loss": 0.0,
+      "step": 2941
+    },
+    {
+      "epoch": 2.6100642872977167,
+      "grad_norm": 0.0015108479419723153,
+      "learning_rate": 4.352634530401051e-07,
+      "loss": 0.0,
+      "step": 2942
+    },
+    {
+      "epoch": 2.610951008645533,
+      "grad_norm": 0.20358669757843018,
+      "learning_rate": 4.3331185319185567e-07,
+      "loss": 0.0006,
+      "step": 2943
+    },
+    {
+      "epoch": 2.6118377299933497,
+      "grad_norm": 0.1645130217075348,
+      "learning_rate": 4.313644402052186e-07,
+      "loss": 0.0011,
+      "step": 2944
+    },
+    {
+      "epoch": 2.6127244513411663,
+      "grad_norm": 0.002132191089913249,
+      "learning_rate": 4.2942121586563745e-07,
+      "loss": 0.0,
+      "step": 2945
+    },
+    {
+      "epoch": 2.6136111726889824,
+      "grad_norm": 0.005285217426717281,
+      "learning_rate": 4.274821819547098e-07,
+      "loss": 0.0,
+      "step": 2946
+    },
+    {
+      "epoch": 2.614497894036799,
+      "grad_norm": 0.0008985325694084167,
+      "learning_rate": 4.255473402501947e-07,
+      "loss": 0.0,
+      "step": 2947
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.0016815599519759417,
+      "learning_rate": 4.23616692526006e-07,
+      "loss": 0.0,
+      "step": 2948
+    },
+    {
+      "epoch": 2.616271336732432,
+      "grad_norm": 0.0033405658323317766,
+      "learning_rate": 4.2169024055221385e-07,
+      "loss": 0.0,
+      "step": 2949
+    },
+    {
+      "epoch": 2.617158058080248,
+      "grad_norm": 0.11762901395559311,
+      "learning_rate": 4.1976798609504155e-07,
+      "loss": 0.0003,
+      "step": 2950
+    },
+    {
+      "epoch": 2.6180447794280646,
+      "grad_norm": 0.0011727182427421212,
+      "learning_rate": 4.1784993091686156e-07,
+      "loss": 0.0,
+      "step": 2951
+    },
+    {
+      "epoch": 2.618931500775881,
+      "grad_norm": 0.16032883524894714,
+      "learning_rate": 4.159360767761989e-07,
+      "loss": 0.0004,
+      "step": 2952
+    },
+    {
+      "epoch": 2.6198182221236976,
+      "grad_norm": 0.0036196603905409575,
+      "learning_rate": 4.1402642542772544e-07,
+      "loss": 0.0,
+      "step": 2953
+    },
+    {
+      "epoch": 2.620704943471514,
+      "grad_norm": 0.0020090683829039335,
+      "learning_rate": 4.1212097862226186e-07,
+      "loss": 0.0,
+      "step": 2954
+    },
+    {
+      "epoch": 2.6215916648193307,
+      "grad_norm": 0.03135655075311661,
+      "learning_rate": 4.1021973810677187e-07,
+      "loss": 0.0001,
+      "step": 2955
+    },
+    {
+      "epoch": 2.6224783861671472,
+      "grad_norm": 0.10065038502216339,
+      "learning_rate": 4.0832270562436436e-07,
+      "loss": 0.0004,
+      "step": 2956
+    },
+    {
+      "epoch": 2.6233651075149633,
+      "grad_norm": 0.23414331674575806,
+      "learning_rate": 4.0642988291428764e-07,
+      "loss": 0.0013,
+      "step": 2957
+    },
+    {
+      "epoch": 2.62425182886278,
+      "grad_norm": 0.039282117038965225,
+      "learning_rate": 4.0454127171193347e-07,
+      "loss": 0.0001,
+      "step": 2958
+    },
+    {
+      "epoch": 2.6251385502105964,
+      "grad_norm": 0.0033170965034514666,
+      "learning_rate": 4.0265687374883014e-07,
+      "loss": 0.0,
+      "step": 2959
+    },
+    {
+      "epoch": 2.626025271558413,
+      "grad_norm": 0.06266362965106964,
+      "learning_rate": 4.0077669075264555e-07,
+      "loss": 0.0003,
+      "step": 2960
+    },
+    {
+      "epoch": 2.626911992906229,
+      "grad_norm": 0.002013074466958642,
+      "learning_rate": 3.9890072444717933e-07,
+      "loss": 0.0,
+      "step": 2961
+    },
+    {
+      "epoch": 2.6277987142540455,
+      "grad_norm": 0.059029195457696915,
+      "learning_rate": 3.970289765523699e-07,
+      "loss": 0.0001,
+      "step": 2962
+    },
+    {
+      "epoch": 2.628685435601862,
+      "grad_norm": 0.005892843008041382,
+      "learning_rate": 3.951614487842842e-07,
+      "loss": 0.0,
+      "step": 2963
+    },
+    {
+      "epoch": 2.6295721569496786,
+      "grad_norm": 0.26164448261260986,
+      "learning_rate": 3.932981428551225e-07,
+      "loss": 0.0069,
+      "step": 2964
+    },
+    {
+      "epoch": 2.630458878297495,
+      "grad_norm": 0.4535243511199951,
+      "learning_rate": 3.914390604732116e-07,
+      "loss": 0.0009,
+      "step": 2965
+    },
+    {
+      "epoch": 2.6313455996453117,
+      "grad_norm": 0.0008455055649392307,
+      "learning_rate": 3.895842033430103e-07,
+      "loss": 0.0,
+      "step": 2966
+    },
+    {
+      "epoch": 2.6322323209931278,
+      "grad_norm": 0.0007907512481324375,
+      "learning_rate": 3.877335731650994e-07,
+      "loss": 0.0,
+      "step": 2967
+    },
+    {
+      "epoch": 2.6331190423409443,
+      "grad_norm": 0.0056696608662605286,
+      "learning_rate": 3.8588717163618673e-07,
+      "loss": 0.0,
+      "step": 2968
+    },
+    {
+      "epoch": 2.634005763688761,
+      "grad_norm": 0.0016626459546387196,
+      "learning_rate": 3.8404500044910197e-07,
+      "loss": 0.0,
+      "step": 2969
+    },
+    {
+      "epoch": 2.6348924850365774,
+      "grad_norm": 0.013582353480160236,
+      "learning_rate": 3.822070612927986e-07,
+      "loss": 0.0,
+      "step": 2970
+    },
+    {
+      "epoch": 2.6357792063843934,
+      "grad_norm": 0.03314371407032013,
+      "learning_rate": 3.803733558523459e-07,
+      "loss": 0.0001,
+      "step": 2971
+    },
+    {
+      "epoch": 2.63666592773221,
+      "grad_norm": 0.02767125330865383,
+      "learning_rate": 3.7854388580893463e-07,
+      "loss": 0.0,
+      "step": 2972
+    },
+    {
+      "epoch": 2.6375526490800265,
+      "grad_norm": 0.09225767105817795,
+      "learning_rate": 3.7671865283987254e-07,
+      "loss": 0.0002,
+      "step": 2973
+    },
+    {
+      "epoch": 2.638439370427843,
+      "grad_norm": 0.005085418000817299,
+      "learning_rate": 3.7489765861858205e-07,
+      "loss": 0.0,
+      "step": 2974
+    },
+    {
+      "epoch": 2.6393260917756596,
+      "grad_norm": 0.0008364695822820067,
+      "learning_rate": 3.7308090481459836e-07,
+      "loss": 0.0,
+      "step": 2975
+    },
+    {
+      "epoch": 2.640212813123476,
+      "grad_norm": 0.005674805957823992,
+      "learning_rate": 3.712683930935701e-07,
+      "loss": 0.0,
+      "step": 2976
+    },
+    {
+      "epoch": 2.6410995344712926,
+      "grad_norm": 0.021331392228603363,
+      "learning_rate": 3.6946012511725806e-07,
+      "loss": 0.0001,
+      "step": 2977
+    },
+    {
+      "epoch": 2.6419862558191087,
+      "grad_norm": 0.030648577958345413,
+      "learning_rate": 3.676561025435277e-07,
+      "loss": 0.0001,
+      "step": 2978
+    },
+    {
+      "epoch": 2.6428729771669253,
+      "grad_norm": 0.006878063082695007,
+      "learning_rate": 3.6585632702635653e-07,
+      "loss": 0.0,
+      "step": 2979
+    },
+    {
+      "epoch": 2.643759698514742,
+      "grad_norm": 0.0028057079762220383,
+      "learning_rate": 3.6406080021582747e-07,
+      "loss": 0.0,
+      "step": 2980
+    },
+    {
+      "epoch": 2.6446464198625583,
+      "grad_norm": 0.12466216087341309,
+      "learning_rate": 3.62269523758127e-07,
+      "loss": 0.0004,
+      "step": 2981
+    },
+    {
+      "epoch": 2.6455331412103744,
+      "grad_norm": 0.016507942229509354,
+      "learning_rate": 3.604824992955436e-07,
+      "loss": 0.0001,
+      "step": 2982
+    },
+    {
+      "epoch": 2.646419862558191,
+      "grad_norm": 0.004891809541732073,
+      "learning_rate": 3.586997284664723e-07,
+      "loss": 0.0,
+      "step": 2983
+    },
+    {
+      "epoch": 2.6473065839060075,
+      "grad_norm": 0.0006478436989709735,
+      "learning_rate": 3.569212129054017e-07,
+      "loss": 0.0,
+      "step": 2984
+    },
+    {
+      "epoch": 2.648193305253824,
+      "grad_norm": 0.009865147061645985,
+      "learning_rate": 3.551469542429248e-07,
+      "loss": 0.0,
+      "step": 2985
+    },
+    {
+      "epoch": 2.6490800266016405,
+      "grad_norm": 0.04954991117119789,
+      "learning_rate": 3.5337695410572694e-07,
+      "loss": 0.0001,
+      "step": 2986
+    },
+    {
+      "epoch": 2.649966747949457,
+      "grad_norm": 0.002256659558042884,
+      "learning_rate": 3.5161121411659395e-07,
+      "loss": 0.0,
+      "step": 2987
+    },
+    {
+      "epoch": 2.650853469297273,
+      "grad_norm": 0.00758238323032856,
+      "learning_rate": 3.4984973589440143e-07,
+      "loss": 0.0,
+      "step": 2988
+    },
+    {
+      "epoch": 2.6517401906450897,
+      "grad_norm": 0.017352283000946045,
+      "learning_rate": 3.480925210541203e-07,
+      "loss": 0.0001,
+      "step": 2989
+    },
+    {
+      "epoch": 2.6526269119929062,
+      "grad_norm": 0.0017505744472146034,
+      "learning_rate": 3.4633957120681294e-07,
+      "loss": 0.0,
+      "step": 2990
+    },
+    {
+      "epoch": 2.6535136333407228,
+      "grad_norm": 0.005053408443927765,
+      "learning_rate": 3.445908879596299e-07,
+      "loss": 0.0,
+      "step": 2991
+    },
+    {
+      "epoch": 2.654400354688539,
+      "grad_norm": 0.011921010911464691,
+      "learning_rate": 3.4284647291580984e-07,
+      "loss": 0.0,
+      "step": 2992
+    },
+    {
+      "epoch": 2.6552870760363554,
+      "grad_norm": 0.0022184166591614485,
+      "learning_rate": 3.411063276746801e-07,
+      "loss": 0.0,
+      "step": 2993
+    },
+    {
+      "epoch": 2.656173797384172,
+      "grad_norm": 0.001728870440274477,
+      "learning_rate": 3.3937045383165226e-07,
+      "loss": 0.0,
+      "step": 2994
+    },
+    {
+      "epoch": 2.6570605187319885,
+      "grad_norm": 0.0017493030754849315,
+      "learning_rate": 3.3763885297822153e-07,
+      "loss": 0.0,
+      "step": 2995
+    },
+    {
+      "epoch": 2.657947240079805,
+      "grad_norm": 0.050390731543302536,
+      "learning_rate": 3.359115267019647e-07,
+      "loss": 0.0002,
+      "step": 2996
+    },
+    {
+      "epoch": 2.6588339614276215,
+      "grad_norm": 0.00843139924108982,
+      "learning_rate": 3.341884765865433e-07,
+      "loss": 0.0,
+      "step": 2997
+    },
+    {
+      "epoch": 2.659720682775438,
+      "grad_norm": 0.1068384125828743,
+      "learning_rate": 3.324697042116931e-07,
+      "loss": 0.0003,
+      "step": 2998
+    },
+    {
+      "epoch": 2.660607404123254,
+      "grad_norm": 0.0007967590354382992,
+      "learning_rate": 3.3075521115323194e-07,
+      "loss": 0.0,
+      "step": 2999
+    },
+    {
+      "epoch": 2.6614941254710707,
+      "grad_norm": 0.028992870822548866,
+      "learning_rate": 3.290449989830519e-07,
+      "loss": 0.0,
+      "step": 3000
+    },
+    {
+      "epoch": 2.662380846818887,
+      "grad_norm": 0.017356855794787407,
+      "learning_rate": 3.2733906926912193e-07,
+      "loss": 0.0001,
+      "step": 3001
+    },
+    {
+      "epoch": 2.6632675681667037,
+      "grad_norm": 0.005700121633708477,
+      "learning_rate": 3.256374235754844e-07,
+      "loss": 0.0,
+      "step": 3002
+    },
+    {
+      "epoch": 2.66415428951452,
+      "grad_norm": 0.0009079916635528207,
+      "learning_rate": 3.239400634622519e-07,
+      "loss": 0.0,
+      "step": 3003
+    },
+    {
+      "epoch": 2.6650410108623364,
+      "grad_norm": 0.22687380015850067,
+      "learning_rate": 3.222469904856107e-07,
+      "loss": 0.001,
+      "step": 3004
+    },
+    {
+      "epoch": 2.665927732210153,
+      "grad_norm": 0.0053674750961363316,
+      "learning_rate": 3.2055820619781466e-07,
+      "loss": 0.0,
+      "step": 3005
+    },
+    {
+      "epoch": 2.6668144535579694,
+      "grad_norm": 0.0032461630180478096,
+      "learning_rate": 3.1887371214718687e-07,
+      "loss": 0.0,
+      "step": 3006
+    },
+    {
+      "epoch": 2.667701174905786,
+      "grad_norm": 0.07514698058366776,
+      "learning_rate": 3.1719350987811537e-07,
+      "loss": 0.0001,
+      "step": 3007
+    },
+    {
+      "epoch": 2.6685878962536025,
+      "grad_norm": 0.13264571130275726,
+      "learning_rate": 3.1551760093105675e-07,
+      "loss": 0.0003,
+      "step": 3008
+    },
+    {
+      "epoch": 2.669474617601419,
+      "grad_norm": 0.0013396762078627944,
+      "learning_rate": 3.1384598684252643e-07,
+      "loss": 0.0,
+      "step": 3009
+    },
+    {
+      "epoch": 2.670361338949235,
+      "grad_norm": 0.00618441728875041,
+      "learning_rate": 3.121786691451062e-07,
+      "loss": 0.0,
+      "step": 3010
+    },
+    {
+      "epoch": 2.6712480602970516,
+      "grad_norm": 0.002804888878017664,
+      "learning_rate": 3.105156493674377e-07,
+      "loss": 0.0,
+      "step": 3011
+    },
+    {
+      "epoch": 2.672134781644868,
+      "grad_norm": 0.010229287669062614,
+      "learning_rate": 3.0885692903422194e-07,
+      "loss": 0.0,
+      "step": 3012
+    },
+    {
+      "epoch": 2.6730215029926847,
+      "grad_norm": 0.024806374683976173,
+      "learning_rate": 3.0720250966621624e-07,
+      "loss": 0.0001,
+      "step": 3013
+    },
+    {
+      "epoch": 2.673908224340501,
+      "grad_norm": 0.01945449411869049,
+      "learning_rate": 3.0555239278023786e-07,
+      "loss": 0.0001,
+      "step": 3014
+    },
+    {
+      "epoch": 2.6747949456883173,
+      "grad_norm": 0.06463182717561722,
+      "learning_rate": 3.039065798891583e-07,
+      "loss": 0.0002,
+      "step": 3015
+    },
+    {
+      "epoch": 2.675681667036134,
+      "grad_norm": 0.0008560666465200484,
+      "learning_rate": 3.0226507250190274e-07,
+      "loss": 0.0,
+      "step": 3016
+    },
+    {
+      "epoch": 2.6765683883839504,
+      "grad_norm": 0.006548132281750441,
+      "learning_rate": 3.0062787212344793e-07,
+      "loss": 0.0,
+      "step": 3017
+    },
+    {
+      "epoch": 2.677455109731767,
+      "grad_norm": 0.013721938244998455,
+      "learning_rate": 2.989949802548231e-07,
+      "loss": 0.0,
+      "step": 3018
+    },
+    {
+      "epoch": 2.6783418310795835,
+      "grad_norm": 0.0003532420960254967,
+      "learning_rate": 2.9736639839310744e-07,
+      "loss": 0.0,
+      "step": 3019
+    },
+    {
+      "epoch": 2.6792285524273995,
+      "grad_norm": 0.001644345000386238,
+      "learning_rate": 2.9574212803142986e-07,
+      "loss": 0.0,
+      "step": 3020
+    },
+    {
+      "epoch": 2.680115273775216,
+      "grad_norm": 0.011774779297411442,
+      "learning_rate": 2.9412217065896085e-07,
+      "loss": 0.0,
+      "step": 3021
+    },
+    {
+      "epoch": 2.6810019951230326,
+      "grad_norm": 0.008171603083610535,
+      "learning_rate": 2.9250652776092513e-07,
+      "loss": 0.0,
+      "step": 3022
+    },
+    {
+      "epoch": 2.681888716470849,
+      "grad_norm": 0.01484227180480957,
+      "learning_rate": 2.9089520081858444e-07,
+      "loss": 0.0001,
+      "step": 3023
+    },
+    {
+      "epoch": 2.6827754378186652,
+      "grad_norm": 0.08469567447900772,
+      "learning_rate": 2.8928819130924656e-07,
+      "loss": 0.0003,
+      "step": 3024
+    },
+    {
+      "epoch": 2.6836621591664818,
+      "grad_norm": 0.017949476838111877,
+      "learning_rate": 2.8768550070626125e-07,
+      "loss": 0.0001,
+      "step": 3025
+    },
+    {
+      "epoch": 2.6845488805142983,
+      "grad_norm": 0.009285989217460155,
+      "learning_rate": 2.8608713047901805e-07,
+      "loss": 0.0,
+      "step": 3026
+    },
+    {
+      "epoch": 2.685435601862115,
+      "grad_norm": 0.0016875972505658865,
+      "learning_rate": 2.8449308209294645e-07,
+      "loss": 0.0,
+      "step": 3027
+    },
+    {
+      "epoch": 2.6863223232099314,
+      "grad_norm": 0.3143994212150574,
+      "learning_rate": 2.829033570095102e-07,
+      "loss": 0.0011,
+      "step": 3028
+    },
+    {
+      "epoch": 2.687209044557748,
+      "grad_norm": 0.001808093162253499,
+      "learning_rate": 2.813179566862134e-07,
+      "loss": 0.0,
+      "step": 3029
+    },
+    {
+      "epoch": 2.6880957659055644,
+      "grad_norm": 0.03166713938117027,
+      "learning_rate": 2.797368825765928e-07,
+      "loss": 0.0001,
+      "step": 3030
+    },
+    {
+      "epoch": 2.6889824872533805,
+      "grad_norm": 0.13500897586345673,
+      "learning_rate": 2.7816013613021995e-07,
+      "loss": 0.0003,
+      "step": 3031
+    },
+    {
+      "epoch": 2.689869208601197,
+      "grad_norm": 0.05272597447037697,
+      "learning_rate": 2.765877187926963e-07,
+      "loss": 0.0001,
+      "step": 3032
+    },
+    {
+      "epoch": 2.6907559299490136,
+      "grad_norm": 0.0012098954757675529,
+      "learning_rate": 2.7501963200565916e-07,
+      "loss": 0.0,
+      "step": 3033
+    },
+    {
+      "epoch": 2.69164265129683,
+      "grad_norm": 0.005726026836782694,
+      "learning_rate": 2.7345587720677026e-07,
+      "loss": 0.0,
+      "step": 3034
+    },
+    {
+      "epoch": 2.692529372644646,
+      "grad_norm": 0.011130993254482746,
+      "learning_rate": 2.7189645582972155e-07,
+      "loss": 0.0,
+      "step": 3035
+    },
+    {
+      "epoch": 2.6934160939924627,
+      "grad_norm": 0.1115150898694992,
+      "learning_rate": 2.703413693042328e-07,
+      "loss": 0.0002,
+      "step": 3036
+    },
+    {
+      "epoch": 2.6943028153402793,
+      "grad_norm": 0.1770944446325302,
+      "learning_rate": 2.6879061905604963e-07,
+      "loss": 0.001,
+      "step": 3037
+    },
+    {
+      "epoch": 2.695189536688096,
+      "grad_norm": 0.0014840234071016312,
+      "learning_rate": 2.672442065069392e-07,
+      "loss": 0.0,
+      "step": 3038
+    },
+    {
+      "epoch": 2.6960762580359123,
+      "grad_norm": 0.002955073956400156,
+      "learning_rate": 2.657021330746956e-07,
+      "loss": 0.0,
+      "step": 3039
+    },
+    {
+      "epoch": 2.696962979383729,
+      "grad_norm": 0.0011210017837584019,
+      "learning_rate": 2.641644001731325e-07,
+      "loss": 0.0,
+      "step": 3040
+    },
+    {
+      "epoch": 2.697849700731545,
+      "grad_norm": 0.0069196210242807865,
+      "learning_rate": 2.6263100921208484e-07,
+      "loss": 0.0,
+      "step": 3041
+    },
+    {
+      "epoch": 2.6987364220793615,
+      "grad_norm": 0.0022021131590008736,
+      "learning_rate": 2.611019615974042e-07,
+      "loss": 0.0,
+      "step": 3042
+    },
+    {
+      "epoch": 2.699623143427178,
+      "grad_norm": 0.03699037432670593,
+      "learning_rate": 2.5957725873096583e-07,
+      "loss": 0.0001,
+      "step": 3043
+    },
+    {
+      "epoch": 2.7005098647749946,
+      "grad_norm": 0.031189635396003723,
+      "learning_rate": 2.5805690201065494e-07,
+      "loss": 0.0,
+      "step": 3044
+    },
+    {
+      "epoch": 2.7013965861228106,
+      "grad_norm": 0.35749873518943787,
+      "learning_rate": 2.565408928303775e-07,
+      "loss": 0.0026,
+      "step": 3045
+    },
+    {
+      "epoch": 2.702283307470627,
+      "grad_norm": 0.013324571773409843,
+      "learning_rate": 2.5502923258004853e-07,
+      "loss": 0.0001,
+      "step": 3046
+    },
+    {
+      "epoch": 2.7031700288184437,
+      "grad_norm": 0.012070167809724808,
+      "learning_rate": 2.535219226456015e-07,
+      "loss": 0.0,
+      "step": 3047
+    },
+    {
+      "epoch": 2.7040567501662602,
+      "grad_norm": 0.004928046837449074,
+      "learning_rate": 2.520189644089766e-07,
+      "loss": 0.0,
+      "step": 3048
+    },
+    {
+      "epoch": 2.7049434715140768,
+      "grad_norm": 0.004704819060862064,
+      "learning_rate": 2.505203592481259e-07,
+      "loss": 0.0,
+      "step": 3049
+    },
+    {
+      "epoch": 2.7058301928618933,
+      "grad_norm": 0.045247048139572144,
+      "learning_rate": 2.49026108537011e-07,
+      "loss": 0.0002,
+      "step": 3050
+    },
+    {
+      "epoch": 2.70671691420971,
+      "grad_norm": 0.022802405059337616,
+      "learning_rate": 2.4753621364560166e-07,
+      "loss": 0.0001,
+      "step": 3051
+    },
+    {
+      "epoch": 2.707603635557526,
+      "grad_norm": 0.02606274001300335,
+      "learning_rate": 2.4605067593987286e-07,
+      "loss": 0.0001,
+      "step": 3052
+    },
+    {
+      "epoch": 2.7084903569053425,
+      "grad_norm": 0.013050327077507973,
+      "learning_rate": 2.445694967818041e-07,
+      "loss": 0.0,
+      "step": 3053
+    },
+    {
+      "epoch": 2.709377078253159,
+      "grad_norm": 0.000353116454789415,
+      "learning_rate": 2.4309267752938126e-07,
+      "loss": 0.0,
+      "step": 3054
+    },
+    {
+      "epoch": 2.7102637996009755,
+      "grad_norm": 0.0003638343187049031,
+      "learning_rate": 2.416202195365913e-07,
+      "loss": 0.0,
+      "step": 3055
+    },
+    {
+      "epoch": 2.7111505209487916,
+      "grad_norm": 0.0013948410050943494,
+      "learning_rate": 2.401521241534227e-07,
+      "loss": 0.0,
+      "step": 3056
+    },
+    {
+      "epoch": 2.712037242296608,
+      "grad_norm": 0.0002752837899606675,
+      "learning_rate": 2.3868839272586485e-07,
+      "loss": 0.0,
+      "step": 3057
+    },
+    {
+      "epoch": 2.7129239636444247,
+      "grad_norm": 0.01548377051949501,
+      "learning_rate": 2.3722902659590653e-07,
+      "loss": 0.0001,
+      "step": 3058
+    },
+    {
+      "epoch": 2.713810684992241,
+      "grad_norm": 0.1762523651123047,
+      "learning_rate": 2.357740271015324e-07,
+      "loss": 0.0004,
+      "step": 3059
+    },
+    {
+      "epoch": 2.7146974063400577,
+      "grad_norm": 0.0014459047233685851,
+      "learning_rate": 2.343233955767249e-07,
+      "loss": 0.0,
+      "step": 3060
+    },
+    {
+      "epoch": 2.7155841276878743,
+      "grad_norm": 0.025118915364146233,
+      "learning_rate": 2.3287713335146177e-07,
+      "loss": 0.0001,
+      "step": 3061
+    },
+    {
+      "epoch": 2.716470849035691,
+      "grad_norm": 0.4808674156665802,
+      "learning_rate": 2.3143524175171628e-07,
+      "loss": 0.0008,
+      "step": 3062
+    },
+    {
+      "epoch": 2.717357570383507,
+      "grad_norm": 0.005553117021918297,
+      "learning_rate": 2.2999772209945037e-07,
+      "loss": 0.0,
+      "step": 3063
+    },
+    {
+      "epoch": 2.7182442917313234,
+      "grad_norm": 0.010888897813856602,
+      "learning_rate": 2.2856457571262314e-07,
+      "loss": 0.0,
+      "step": 3064
+    },
+    {
+      "epoch": 2.71913101307914,
+      "grad_norm": 0.14666052162647247,
+      "learning_rate": 2.271358039051802e-07,
+      "loss": 0.0004,
+      "step": 3065
+    },
+    {
+      "epoch": 2.7200177344269565,
+      "grad_norm": 0.13892030715942383,
+      "learning_rate": 2.2571140798705816e-07,
+      "loss": 0.0003,
+      "step": 3066
+    },
+    {
+      "epoch": 2.7209044557747726,
+      "grad_norm": 0.0004430158587638289,
+      "learning_rate": 2.2429138926417959e-07,
+      "loss": 0.0,
+      "step": 3067
+    },
+    {
+      "epoch": 2.721791177122589,
+      "grad_norm": 0.024874933063983917,
+      "learning_rate": 2.2287574903845866e-07,
+      "loss": 0.0,
+      "step": 3068
+    },
+    {
+      "epoch": 2.7226778984704056,
+      "grad_norm": 0.007251809351146221,
+      "learning_rate": 2.2146448860778934e-07,
+      "loss": 0.0,
+      "step": 3069
+    },
+    {
+      "epoch": 2.723564619818222,
+      "grad_norm": 0.011288401670753956,
+      "learning_rate": 2.200576092660539e-07,
+      "loss": 0.0,
+      "step": 3070
+    },
+    {
+      "epoch": 2.7244513411660387,
+      "grad_norm": 0.02641960419714451,
+      "learning_rate": 2.186551123031172e-07,
+      "loss": 0.0001,
+      "step": 3071
+    },
+    {
+      "epoch": 2.7253380625138552,
+      "grad_norm": 0.021034937351942062,
+      "learning_rate": 2.1725699900482567e-07,
+      "loss": 0.0001,
+      "step": 3072
+    },
+    {
+      "epoch": 2.7262247838616713,
+      "grad_norm": 0.022156720981001854,
+      "learning_rate": 2.1586327065300672e-07,
+      "loss": 0.0,
+      "step": 3073
+    },
+    {
+      "epoch": 2.727111505209488,
+      "grad_norm": 0.028590481728315353,
+      "learning_rate": 2.1447392852546766e-07,
+      "loss": 0.0001,
+      "step": 3074
+    },
+    {
+      "epoch": 2.7279982265573044,
+      "grad_norm": 0.005033856723457575,
+      "learning_rate": 2.130889738959946e-07,
+      "loss": 0.0,
+      "step": 3075
+    },
+    {
+      "epoch": 2.728884947905121,
+      "grad_norm": 0.2219175547361374,
+      "learning_rate": 2.1170840803435066e-07,
+      "loss": 0.0006,
+      "step": 3076
+    },
+    {
+      "epoch": 2.729771669252937,
+      "grad_norm": 0.00024738299543969333,
+      "learning_rate": 2.1033223220627563e-07,
+      "loss": 0.0,
+      "step": 3077
+    },
+    {
+      "epoch": 2.7306583906007535,
+      "grad_norm": 0.06527652591466904,
+      "learning_rate": 2.0896044767348468e-07,
+      "loss": 0.0003,
+      "step": 3078
+    },
+    {
+      "epoch": 2.73154511194857,
+      "grad_norm": 0.0066484929993748665,
+      "learning_rate": 2.075930556936667e-07,
+      "loss": 0.0,
+      "step": 3079
+    },
+    {
+      "epoch": 2.7324318332963866,
+      "grad_norm": 0.012028141878545284,
+      "learning_rate": 2.062300575204823e-07,
+      "loss": 0.0,
+      "step": 3080
+    },
+    {
+      "epoch": 2.733318554644203,
+      "grad_norm": 0.013649419881403446,
+      "learning_rate": 2.048714544035646e-07,
+      "loss": 0.0,
+      "step": 3081
+    },
+    {
+      "epoch": 2.7342052759920197,
+      "grad_norm": 0.01976783014833927,
+      "learning_rate": 2.0351724758851776e-07,
+      "loss": 0.0001,
+      "step": 3082
+    },
+    {
+      "epoch": 2.735091997339836,
+      "grad_norm": 0.07697169482707977,
+      "learning_rate": 2.0216743831691477e-07,
+      "loss": 0.0002,
+      "step": 3083
+    },
+    {
+      "epoch": 2.7359787186876523,
+      "grad_norm": 0.006745744496583939,
+      "learning_rate": 2.0082202782629566e-07,
+      "loss": 0.0,
+      "step": 3084
+    },
+    {
+      "epoch": 2.736865440035469,
+      "grad_norm": 0.0004296990518923849,
+      "learning_rate": 1.9948101735017045e-07,
+      "loss": 0.0,
+      "step": 3085
+    },
+    {
+      "epoch": 2.7377521613832854,
+      "grad_norm": 0.09031447768211365,
+      "learning_rate": 1.9814440811801228e-07,
+      "loss": 0.0004,
+      "step": 3086
+    },
+    {
+      "epoch": 2.738638882731102,
+      "grad_norm": 0.0685887336730957,
+      "learning_rate": 1.968122013552609e-07,
+      "loss": 0.0003,
+      "step": 3087
+    },
+    {
+      "epoch": 2.739525604078918,
+      "grad_norm": 0.0008473261841572821,
+      "learning_rate": 1.954843982833171e-07,
+      "loss": 0.0,
+      "step": 3088
+    },
+    {
+      "epoch": 2.7404123254267345,
+      "grad_norm": 0.37110695242881775,
+      "learning_rate": 1.9416100011954874e-07,
+      "loss": 0.0018,
+      "step": 3089
+    },
+    {
+      "epoch": 2.741299046774551,
+      "grad_norm": 0.0020779715850949287,
+      "learning_rate": 1.9284200807728083e-07,
+      "loss": 0.0,
+      "step": 3090
+    },
+    {
+      "epoch": 2.7421857681223676,
+      "grad_norm": 0.015576466917991638,
+      "learning_rate": 1.9152742336580155e-07,
+      "loss": 0.0001,
+      "step": 3091
+    },
+    {
+      "epoch": 2.743072489470184,
+      "grad_norm": 0.016060447320342064,
+      "learning_rate": 1.902172471903563e-07,
+      "loss": 0.0,
+      "step": 3092
+    },
+    {
+      "epoch": 2.7439592108180006,
+      "grad_norm": 0.002036657417193055,
+      "learning_rate": 1.8891148075215137e-07,
+      "loss": 0.0,
+      "step": 3093
+    },
+    {
+      "epoch": 2.7448459321658167,
+      "grad_norm": 0.0033351199235767126,
+      "learning_rate": 1.8761012524834588e-07,
+      "loss": 0.0,
+      "step": 3094
+    },
+    {
+      "epoch": 2.7457326535136333,
+      "grad_norm": 0.008396641351282597,
+      "learning_rate": 1.8631318187205817e-07,
+      "loss": 0.0,
+      "step": 3095
+    },
+    {
+      "epoch": 2.74661937486145,
+      "grad_norm": 0.02515893056988716,
+      "learning_rate": 1.850206518123615e-07,
+      "loss": 0.0,
+      "step": 3096
+    },
+    {
+      "epoch": 2.7475060962092663,
+      "grad_norm": 0.0037924533244222403,
+      "learning_rate": 1.8373253625428133e-07,
+      "loss": 0.0,
+      "step": 3097
+    },
+    {
+      "epoch": 2.7483928175570824,
+      "grad_norm": 0.0001926338445628062,
+      "learning_rate": 1.8244883637879518e-07,
+      "loss": 0.0,
+      "step": 3098
+    },
+    {
+      "epoch": 2.749279538904899,
+      "grad_norm": 0.014933107420802116,
+      "learning_rate": 1.8116955336283603e-07,
+      "loss": 0.0001,
+      "step": 3099
+    },
+    {
+      "epoch": 2.7501662602527155,
+      "grad_norm": 0.015284646302461624,
+      "learning_rate": 1.7989468837928237e-07,
+      "loss": 0.0001,
+      "step": 3100
+    },
+    {
+      "epoch": 2.751052981600532,
+      "grad_norm": 0.016768403351306915,
+      "learning_rate": 1.7862424259696532e-07,
+      "loss": 0.0001,
+      "step": 3101
+    },
+    {
+      "epoch": 2.7519397029483486,
+      "grad_norm": 0.001939128735102713,
+      "learning_rate": 1.773582171806637e-07,
+      "loss": 0.0,
+      "step": 3102
+    },
+    {
+      "epoch": 2.752826424296165,
+      "grad_norm": 0.0016595263732597232,
+      "learning_rate": 1.7609661329110295e-07,
+      "loss": 0.0,
+      "step": 3103
+    },
+    {
+      "epoch": 2.7537131456439816,
+      "grad_norm": 0.0015289318980649114,
+      "learning_rate": 1.7483943208495558e-07,
+      "loss": 0.0,
+      "step": 3104
+    },
+    {
+      "epoch": 2.7545998669917977,
+      "grad_norm": 0.09125112742185593,
+      "learning_rate": 1.735866747148385e-07,
+      "loss": 0.0004,
+      "step": 3105
+    },
+    {
+      "epoch": 2.7554865883396142,
+      "grad_norm": 0.03696690499782562,
+      "learning_rate": 1.7233834232931345e-07,
+      "loss": 0.0001,
+      "step": 3106
+    },
+    {
+      "epoch": 2.7563733096874308,
+      "grad_norm": 0.0077039701864123344,
+      "learning_rate": 1.7109443607288446e-07,
+      "loss": 0.0,
+      "step": 3107
+    },
+    {
+      "epoch": 2.7572600310352473,
+      "grad_norm": 0.0009759148233570158,
+      "learning_rate": 1.6985495708599863e-07,
+      "loss": 0.0,
+      "step": 3108
+    },
+    {
+      "epoch": 2.7581467523830634,
+      "grad_norm": 0.0027849047910422087,
+      "learning_rate": 1.6861990650504256e-07,
+      "loss": 0.0,
+      "step": 3109
+    },
+    {
+      "epoch": 2.75903347373088,
+      "grad_norm": 0.0005394841427914798,
+      "learning_rate": 1.6738928546234435e-07,
+      "loss": 0.0,
+      "step": 3110
+    },
+    {
+      "epoch": 2.7599201950786965,
+      "grad_norm": 0.027912667021155357,
+      "learning_rate": 1.6616309508617033e-07,
+      "loss": 0.0001,
+      "step": 3111
+    },
+    {
+      "epoch": 2.760806916426513,
+      "grad_norm": 0.034508995711803436,
+      "learning_rate": 1.6494133650072352e-07,
+      "loss": 0.0001,
+      "step": 3112
+    },
+    {
+      "epoch": 2.7616936377743295,
+      "grad_norm": 0.2373371124267578,
+      "learning_rate": 1.6372401082614564e-07,
+      "loss": 0.0012,
+      "step": 3113
+    },
+    {
+      "epoch": 2.762580359122146,
+      "grad_norm": 0.0013953030575066805,
+      "learning_rate": 1.6251111917851391e-07,
+      "loss": 0.0,
+      "step": 3114
+    },
+    {
+      "epoch": 2.7634670804699626,
+      "grad_norm": 0.0005270761903375387,
+      "learning_rate": 1.6130266266983886e-07,
+      "loss": 0.0,
+      "step": 3115
+    },
+    {
+      "epoch": 2.7643538018177787,
+      "grad_norm": 0.0027937826234847307,
+      "learning_rate": 1.6009864240806528e-07,
+      "loss": 0.0,
+      "step": 3116
+    },
+    {
+      "epoch": 2.765240523165595,
+      "grad_norm": 0.03274468332529068,
+      "learning_rate": 1.5889905949707186e-07,
+      "loss": 0.0001,
+      "step": 3117
+    },
+    {
+      "epoch": 2.7661272445134117,
+      "grad_norm": 0.0675559788942337,
+      "learning_rate": 1.5770391503666883e-07,
+      "loss": 0.0001,
+      "step": 3118
+    },
+    {
+      "epoch": 2.7670139658612283,
+      "grad_norm": 0.005557031370699406,
+      "learning_rate": 1.5651321012259525e-07,
+      "loss": 0.0,
+      "step": 3119
+    },
+    {
+      "epoch": 2.7679006872090444,
+      "grad_norm": 0.08115627616643906,
+      "learning_rate": 1.5532694584652175e-07,
+      "loss": 0.0002,
+      "step": 3120
+    },
+    {
+      "epoch": 2.768787408556861,
+      "grad_norm": 0.0014597626868635416,
+      "learning_rate": 1.541451232960467e-07,
+      "loss": 0.0,
+      "step": 3121
+    },
+    {
+      "epoch": 2.7696741299046774,
+      "grad_norm": 0.13055862486362457,
+      "learning_rate": 1.5296774355469778e-07,
+      "loss": 0.0005,
+      "step": 3122
+    },
+    {
+      "epoch": 2.770560851252494,
+      "grad_norm": 0.009585856460034847,
+      "learning_rate": 1.5179480770192602e-07,
+      "loss": 0.0,
+      "step": 3123
+    },
+    {
+      "epoch": 2.7714475726003105,
+      "grad_norm": 0.00751367025077343,
+      "learning_rate": 1.506263168131128e-07,
+      "loss": 0.0,
+      "step": 3124
+    },
+    {
+      "epoch": 2.772334293948127,
+      "grad_norm": 0.03164404258131981,
+      "learning_rate": 1.4946227195956008e-07,
+      "loss": 0.0001,
+      "step": 3125
+    },
+    {
+      "epoch": 2.773221015295943,
+      "grad_norm": 0.25857213139533997,
+      "learning_rate": 1.4830267420849587e-07,
+      "loss": 0.0027,
+      "step": 3126
+    },
+    {
+      "epoch": 2.7741077366437596,
+      "grad_norm": 0.000911539129447192,
+      "learning_rate": 1.4714752462307025e-07,
+      "loss": 0.0,
+      "step": 3127
+    },
+    {
+      "epoch": 2.774994457991576,
+      "grad_norm": 0.006078448612242937,
+      "learning_rate": 1.4599682426235552e-07,
+      "loss": 0.0,
+      "step": 3128
+    },
+    {
+      "epoch": 2.7758811793393927,
+      "grad_norm": 0.0028333021327853203,
+      "learning_rate": 1.4485057418134385e-07,
+      "loss": 0.0,
+      "step": 3129
+    },
+    {
+      "epoch": 2.776767900687209,
+      "grad_norm": 0.007403951603919268,
+      "learning_rate": 1.437087754309485e-07,
+      "loss": 0.0,
+      "step": 3130
+    },
+    {
+      "epoch": 2.7776546220350253,
+      "grad_norm": 0.014319689944386482,
+      "learning_rate": 1.4257142905800048e-07,
+      "loss": 0.0,
+      "step": 3131
+    },
+    {
+      "epoch": 2.778541343382842,
+      "grad_norm": 0.0011654490372166038,
+      "learning_rate": 1.414385361052495e-07,
+      "loss": 0.0,
+      "step": 3132
+    },
+    {
+      "epoch": 2.7794280647306584,
+      "grad_norm": 0.0013293580850586295,
+      "learning_rate": 1.403100976113625e-07,
+      "loss": 0.0,
+      "step": 3133
+    },
+    {
+      "epoch": 2.780314786078475,
+      "grad_norm": 0.002857569605112076,
+      "learning_rate": 1.3918611461092135e-07,
+      "loss": 0.0,
+      "step": 3134
+    },
+    {
+      "epoch": 2.7812015074262915,
+      "grad_norm": 0.5601361989974976,
+      "learning_rate": 1.3806658813442453e-07,
+      "loss": 0.0047,
+      "step": 3135
+    },
+    {
+      "epoch": 2.782088228774108,
+      "grad_norm": 0.3579235076904297,
+      "learning_rate": 1.3695151920828265e-07,
+      "loss": 0.001,
+      "step": 3136
+    },
+    {
+      "epoch": 2.782974950121924,
+      "grad_norm": 0.04795952886343002,
+      "learning_rate": 1.3584090885482183e-07,
+      "loss": 0.0,
+      "step": 3137
+    },
+    {
+      "epoch": 2.7838616714697406,
+      "grad_norm": 0.027433745563030243,
+      "learning_rate": 1.347347580922781e-07,
+      "loss": 0.0001,
+      "step": 3138
+    },
+    {
+      "epoch": 2.784748392817557,
+      "grad_norm": 0.1161801666021347,
+      "learning_rate": 1.3363306793480191e-07,
+      "loss": 0.0006,
+      "step": 3139
+    },
+    {
+      "epoch": 2.7856351141653737,
+      "grad_norm": 0.003158286912366748,
+      "learning_rate": 1.3253583939245028e-07,
+      "loss": 0.0,
+      "step": 3140
+    },
+    {
+      "epoch": 2.7865218355131898,
+      "grad_norm": 0.2493053823709488,
+      "learning_rate": 1.314430734711919e-07,
+      "loss": 0.0043,
+      "step": 3141
+    },
+    {
+      "epoch": 2.7874085568610063,
+      "grad_norm": 0.02230064757168293,
+      "learning_rate": 1.3035477117290472e-07,
+      "loss": 0.0001,
+      "step": 3142
+    },
+    {
+      "epoch": 2.788295278208823,
+      "grad_norm": 0.026541607454419136,
+      "learning_rate": 1.292709334953729e-07,
+      "loss": 0.0001,
+      "step": 3143
+    },
+    {
+      "epoch": 2.7891819995566394,
+      "grad_norm": 0.013466564007103443,
+      "learning_rate": 1.2819156143228706e-07,
+      "loss": 0.0,
+      "step": 3144
+    },
+    {
+      "epoch": 2.790068720904456,
+      "grad_norm": 0.0019223958952352405,
+      "learning_rate": 1.2711665597324563e-07,
+      "loss": 0.0,
+      "step": 3145
+    },
+    {
+      "epoch": 2.7909554422522724,
+      "grad_norm": 0.04089783504605293,
+      "learning_rate": 1.260462181037492e-07,
+      "loss": 0.0001,
+      "step": 3146
+    },
+    {
+      "epoch": 2.7918421636000885,
+      "grad_norm": 0.009505347348749638,
+      "learning_rate": 1.2498024880520543e-07,
+      "loss": 0.0,
+      "step": 3147
+    },
+    {
+      "epoch": 2.792728884947905,
+      "grad_norm": 0.001099666696973145,
+      "learning_rate": 1.239187490549215e-07,
+      "loss": 0.0,
+      "step": 3148
+    },
+    {
+      "epoch": 2.7936156062957216,
+      "grad_norm": 0.0020217658020555973,
+      "learning_rate": 1.2286171982611062e-07,
+      "loss": 0.0,
+      "step": 3149
+    },
+    {
+      "epoch": 2.794502327643538,
+      "grad_norm": 0.008771456778049469,
+      "learning_rate": 1.2180916208788418e-07,
+      "loss": 0.0,
+      "step": 3150
+    },
+    {
+      "epoch": 2.795389048991354,
+      "grad_norm": 0.005936725065112114,
+      "learning_rate": 1.2076107680525529e-07,
+      "loss": 0.0,
+      "step": 3151
+    },
+    {
+      "epoch": 2.7962757703391707,
+      "grad_norm": 0.09092960506677628,
+      "learning_rate": 1.1971746493913693e-07,
+      "loss": 0.0003,
+      "step": 3152
+    },
+    {
+      "epoch": 2.7971624916869873,
+      "grad_norm": 0.2639486491680145,
+      "learning_rate": 1.1867832744633989e-07,
+      "loss": 0.0014,
+      "step": 3153
+    },
+    {
+      "epoch": 2.798049213034804,
+      "grad_norm": 0.0012152879498898983,
+      "learning_rate": 1.1764366527957427e-07,
+      "loss": 0.0,
+      "step": 3154
+    },
+    {
+      "epoch": 2.7989359343826203,
+      "grad_norm": 0.0016898787580430508,
+      "learning_rate": 1.166134793874446e-07,
+      "loss": 0.0,
+      "step": 3155
+    },
+    {
+      "epoch": 2.799822655730437,
+      "grad_norm": 0.000665109371766448,
+      "learning_rate": 1.1558777071445371e-07,
+      "loss": 0.0,
+      "step": 3156
+    },
+    {
+      "epoch": 2.8007093770782534,
+      "grad_norm": 0.16242636740207672,
+      "learning_rate": 1.1456654020099767e-07,
+      "loss": 0.0006,
+      "step": 3157
+    },
+    {
+      "epoch": 2.8015960984260695,
+      "grad_norm": 0.001742783235386014,
+      "learning_rate": 1.135497887833692e-07,
+      "loss": 0.0,
+      "step": 3158
+    },
+    {
+      "epoch": 2.802482819773886,
+      "grad_norm": 0.013286867178976536,
+      "learning_rate": 1.1253751739375207e-07,
+      "loss": 0.0,
+      "step": 3159
+    },
+    {
+      "epoch": 2.8033695411217026,
+      "grad_norm": 0.005383123643696308,
+      "learning_rate": 1.1152972696022447e-07,
+      "loss": 0.0,
+      "step": 3160
+    },
+    {
+      "epoch": 2.804256262469519,
+      "grad_norm": 0.0018168076639994979,
+      "learning_rate": 1.105264184067556e-07,
+      "loss": 0.0,
+      "step": 3161
+    },
+    {
+      "epoch": 2.805142983817335,
+      "grad_norm": 0.001898395363241434,
+      "learning_rate": 1.0952759265320412e-07,
+      "loss": 0.0,
+      "step": 3162
+    },
+    {
+      "epoch": 2.8060297051651517,
+      "grad_norm": 0.03626805543899536,
+      "learning_rate": 1.0853325061532194e-07,
+      "loss": 0.0002,
+      "step": 3163
+    },
+    {
+      "epoch": 2.8069164265129682,
+      "grad_norm": 0.18261322379112244,
+      "learning_rate": 1.0754339320474816e-07,
+      "loss": 0.0011,
+      "step": 3164
+    },
+    {
+      "epoch": 2.8078031478607848,
+      "grad_norm": 0.0031685426365584135,
+      "learning_rate": 1.065580213290085e-07,
+      "loss": 0.0,
+      "step": 3165
+    },
+    {
+      "epoch": 2.8086898692086013,
+      "grad_norm": 0.003848884254693985,
+      "learning_rate": 1.0557713589152086e-07,
+      "loss": 0.0,
+      "step": 3166
+    },
+    {
+      "epoch": 2.809576590556418,
+      "grad_norm": 0.010794474743306637,
+      "learning_rate": 1.0460073779158586e-07,
+      "loss": 0.0,
+      "step": 3167
+    },
+    {
+      "epoch": 2.810463311904234,
+      "grad_norm": 0.003852618858218193,
+      "learning_rate": 1.0362882792439189e-07,
+      "loss": 0.0,
+      "step": 3168
+    },
+    {
+      "epoch": 2.8113500332520505,
+      "grad_norm": 0.006770124193280935,
+      "learning_rate": 1.0266140718101003e-07,
+      "loss": 0.0,
+      "step": 3169
+    },
+    {
+      "epoch": 2.812236754599867,
+      "grad_norm": 0.012775304727256298,
+      "learning_rate": 1.0169847644840026e-07,
+      "loss": 0.0001,
+      "step": 3170
+    },
+    {
+      "epoch": 2.8131234759476835,
+      "grad_norm": 0.06564140319824219,
+      "learning_rate": 1.0074003660940135e-07,
+      "loss": 0.0002,
+      "step": 3171
+    },
+    {
+      "epoch": 2.8140101972954996,
+      "grad_norm": 0.0029760506004095078,
+      "learning_rate": 9.978608854273708e-08,
+      "loss": 0.0,
+      "step": 3172
+    },
+    {
+      "epoch": 2.814896918643316,
+      "grad_norm": 0.002229688921943307,
+      "learning_rate": 9.883663312301228e-08,
+      "loss": 0.0,
+      "step": 3173
+    },
+    {
+      "epoch": 2.8157836399911327,
+      "grad_norm": 0.006304146256297827,
+      "learning_rate": 9.789167122071341e-08,
+      "loss": 0.0,
+      "step": 3174
+    },
+    {
+      "epoch": 2.816670361338949,
+      "grad_norm": 0.0052603622898459435,
+      "learning_rate": 9.695120370220634e-08,
+      "loss": 0.0,
+      "step": 3175
+    },
+    {
+      "epoch": 2.8175570826867657,
+      "grad_norm": 0.0849158987402916,
+      "learning_rate": 9.60152314297369e-08,
+      "loss": 0.0003,
+      "step": 3176
+    },
+    {
+      "epoch": 2.8184438040345823,
+      "grad_norm": 0.088808573782444,
+      "learning_rate": 9.508375526142976e-08,
+      "loss": 0.0005,
+      "step": 3177
+    },
+    {
+      "epoch": 2.819330525382399,
+      "grad_norm": 0.0012514623813331127,
+      "learning_rate": 9.415677605128681e-08,
+      "loss": 0.0,
+      "step": 3178
+    },
+    {
+      "epoch": 2.820217246730215,
+      "grad_norm": 0.005674469750374556,
+      "learning_rate": 9.323429464918766e-08,
+      "loss": 0.0,
+      "step": 3179
+    },
+    {
+      "epoch": 2.8211039680780314,
+      "grad_norm": 0.00882049836218357,
+      "learning_rate": 9.231631190088742e-08,
+      "loss": 0.0,
+      "step": 3180
+    },
+    {
+      "epoch": 2.821990689425848,
+      "grad_norm": 0.0023833143059164286,
+      "learning_rate": 9.140282864801786e-08,
+      "loss": 0.0,
+      "step": 3181
+    },
+    {
+      "epoch": 2.8228774107736645,
+      "grad_norm": 0.07122648507356644,
+      "learning_rate": 9.049384572808407e-08,
+      "loss": 0.0003,
+      "step": 3182
+    },
+    {
+      "epoch": 2.8237641321214806,
+      "grad_norm": 0.010699799284338951,
+      "learning_rate": 8.958936397446605e-08,
+      "loss": 0.0,
+      "step": 3183
+    },
+    {
+      "epoch": 2.824650853469297,
+      "grad_norm": 0.005035745445638895,
+      "learning_rate": 8.868938421641715e-08,
+      "loss": 0.0,
+      "step": 3184
+    },
+    {
+      "epoch": 2.8255375748171137,
+      "grad_norm": 0.0031189401634037495,
+      "learning_rate": 8.77939072790629e-08,
+      "loss": 0.0,
+      "step": 3185
+    },
+    {
+      "epoch": 2.82642429616493,
+      "grad_norm": 0.007225423585623503,
+      "learning_rate": 8.690293398339933e-08,
+      "loss": 0.0,
+      "step": 3186
+    },
+    {
+      "epoch": 2.8273110175127467,
+      "grad_norm": 0.019024573266506195,
+      "learning_rate": 8.601646514629635e-08,
+      "loss": 0.0001,
+      "step": 3187
+    },
+    {
+      "epoch": 2.8281977388605632,
+      "grad_norm": 0.0028227169532328844,
+      "learning_rate": 8.513450158049109e-08,
+      "loss": 0.0,
+      "step": 3188
+    },
+    {
+      "epoch": 2.82908446020838,
+      "grad_norm": 0.04316164553165436,
+      "learning_rate": 8.425704409459168e-08,
+      "loss": 0.0001,
+      "step": 3189
+    },
+    {
+      "epoch": 2.829971181556196,
+      "grad_norm": 0.21143779158592224,
+      "learning_rate": 8.338409349307409e-08,
+      "loss": 0.0012,
+      "step": 3190
+    },
+    {
+      "epoch": 2.8308579029040124,
+      "grad_norm": 0.007404598407447338,
+      "learning_rate": 8.251565057628364e-08,
+      "loss": 0.0,
+      "step": 3191
+    },
+    {
+      "epoch": 2.831744624251829,
+      "grad_norm": 0.03625713661313057,
+      "learning_rate": 8.165171614043066e-08,
+      "loss": 0.0001,
+      "step": 3192
+    },
+    {
+      "epoch": 2.8326313455996455,
+      "grad_norm": 0.0007035653688944876,
+      "learning_rate": 8.079229097759433e-08,
+      "loss": 0.0,
+      "step": 3193
+    },
+    {
+      "epoch": 2.8335180669474616,
+      "grad_norm": 0.04336908459663391,
+      "learning_rate": 7.993737587571825e-08,
+      "loss": 0.0001,
+      "step": 3194
+    },
+    {
+      "epoch": 2.834404788295278,
+      "grad_norm": 0.0009928614599630237,
+      "learning_rate": 7.908697161861212e-08,
+      "loss": 0.0,
+      "step": 3195
+    },
+    {
+      "epoch": 2.8352915096430946,
+      "grad_norm": 0.05311229079961777,
+      "learning_rate": 7.824107898594835e-08,
+      "loss": 0.0001,
+      "step": 3196
+    },
+    {
+      "epoch": 2.836178230990911,
+      "grad_norm": 0.000616305333096534,
+      "learning_rate": 7.739969875326381e-08,
+      "loss": 0.0,
+      "step": 3197
+    },
+    {
+      "epoch": 2.8370649523387277,
+      "grad_norm": 0.0035000985953956842,
+      "learning_rate": 7.656283169195866e-08,
+      "loss": 0.0,
+      "step": 3198
+    },
+    {
+      "epoch": 2.837951673686544,
+      "grad_norm": 0.10979343205690384,
+      "learning_rate": 7.573047856929582e-08,
+      "loss": 0.0002,
+      "step": 3199
+    },
+    {
+      "epoch": 2.8388383950343603,
+      "grad_norm": 0.036622680723667145,
+      "learning_rate": 7.490264014839654e-08,
+      "loss": 0.0001,
+      "step": 3200
+    },
+    {
+      "epoch": 2.839725116382177,
+      "grad_norm": 0.03518826141953468,
+      "learning_rate": 7.407931718824812e-08,
+      "loss": 0.0001,
+      "step": 3201
+    },
+    {
+      "epoch": 2.8406118377299934,
+      "grad_norm": 0.45864933729171753,
+      "learning_rate": 7.326051044369343e-08,
+      "loss": 0.0083,
+      "step": 3202
+    },
+    {
+      "epoch": 2.84149855907781,
+      "grad_norm": 0.0005170101649127901,
+      "learning_rate": 7.244622066543582e-08,
+      "loss": 0.0,
+      "step": 3203
+    },
+    {
+      "epoch": 2.842385280425626,
+      "grad_norm": 0.0011105621233582497,
+      "learning_rate": 7.163644860003927e-08,
+      "loss": 0.0,
+      "step": 3204
+    },
+    {
+      "epoch": 2.8432720017734425,
+      "grad_norm": 0.10616616904735565,
+      "learning_rate": 7.083119498992319e-08,
+      "loss": 0.0007,
+      "step": 3205
+    },
+    {
+      "epoch": 2.844158723121259,
+      "grad_norm": 0.0022827358916401863,
+      "learning_rate": 7.003046057336704e-08,
+      "loss": 0.0,
+      "step": 3206
+    },
+    {
+      "epoch": 2.8450454444690756,
+      "grad_norm": 0.2717021703720093,
+      "learning_rate": 6.923424608450358e-08,
+      "loss": 0.0003,
+      "step": 3207
+    },
+    {
+      "epoch": 2.845932165816892,
+      "grad_norm": 0.4444045424461365,
+      "learning_rate": 6.844255225332552e-08,
+      "loss": 0.0028,
+      "step": 3208
+    },
+    {
+      "epoch": 2.8468188871647087,
+      "grad_norm": 0.09726931899785995,
+      "learning_rate": 6.765537980567727e-08,
+      "loss": 0.0004,
+      "step": 3209
+    },
+    {
+      "epoch": 2.847705608512525,
+      "grad_norm": 0.006602408364415169,
+      "learning_rate": 6.687272946326096e-08,
+      "loss": 0.0,
+      "step": 3210
+    },
+    {
+      "epoch": 2.8485923298603413,
+      "grad_norm": 0.0016501571517437696,
+      "learning_rate": 6.609460194362927e-08,
+      "loss": 0.0,
+      "step": 3211
+    },
+    {
+      "epoch": 2.849479051208158,
+      "grad_norm": 0.012879370711743832,
+      "learning_rate": 6.53209979601932e-08,
+      "loss": 0.0,
+      "step": 3212
+    },
+    {
+      "epoch": 2.8503657725559743,
+      "grad_norm": 0.0010488486150279641,
+      "learning_rate": 6.455191822221097e-08,
+      "loss": 0.0,
+      "step": 3213
+    },
+    {
+      "epoch": 2.851252493903791,
+      "grad_norm": 0.0028409042861312628,
+      "learning_rate": 6.37873634347963e-08,
+      "loss": 0.0,
+      "step": 3214
+    },
+    {
+      "epoch": 2.852139215251607,
+      "grad_norm": 0.012911890633404255,
+      "learning_rate": 6.302733429891406e-08,
+      "loss": 0.0,
+      "step": 3215
+    },
+    {
+      "epoch": 2.8530259365994235,
+      "grad_norm": 0.11188660562038422,
+      "learning_rate": 6.22718315113785e-08,
+      "loss": 0.0003,
+      "step": 3216
+    },
+    {
+      "epoch": 2.85391265794724,
+      "grad_norm": 0.1650157868862152,
+      "learning_rate": 6.1520855764855e-08,
+      "loss": 0.0008,
+      "step": 3217
+    },
+    {
+      "epoch": 2.8547993792950566,
+      "grad_norm": 0.018146656453609467,
+      "learning_rate": 6.077440774785837e-08,
+      "loss": 0.0,
+      "step": 3218
+    },
+    {
+      "epoch": 2.855686100642873,
+      "grad_norm": 0.003937463741749525,
+      "learning_rate": 6.003248814475116e-08,
+      "loss": 0.0,
+      "step": 3219
+    },
+    {
+      "epoch": 2.8565728219906896,
+      "grad_norm": 0.012310229241847992,
+      "learning_rate": 5.929509763574648e-08,
+      "loss": 0.0,
+      "step": 3220
+    },
+    {
+      "epoch": 2.8574595433385057,
+      "grad_norm": 0.00288616050966084,
+      "learning_rate": 5.8562236896902437e-08,
+      "loss": 0.0,
+      "step": 3221
+    },
+    {
+      "epoch": 2.8583462646863222,
+      "grad_norm": 0.0060603683814406395,
+      "learning_rate": 5.7833906600125996e-08,
+      "loss": 0.0,
+      "step": 3222
+    },
+    {
+      "epoch": 2.859232986034139,
+      "grad_norm": 0.03721298649907112,
+      "learning_rate": 5.711010741316969e-08,
+      "loss": 0.0001,
+      "step": 3223
+    },
+    {
+      "epoch": 2.8601197073819553,
+      "grad_norm": 0.0499625988304615,
+      "learning_rate": 5.6390839999631574e-08,
+      "loss": 0.0001,
+      "step": 3224
+    },
+    {
+      "epoch": 2.8610064287297714,
+      "grad_norm": 0.04375501722097397,
+      "learning_rate": 5.567610501895526e-08,
+      "loss": 0.0001,
+      "step": 3225
+    },
+    {
+      "epoch": 2.861893150077588,
+      "grad_norm": 0.002494650427252054,
+      "learning_rate": 5.496590312642991e-08,
+      "loss": 0.0,
+      "step": 3226
+    },
+    {
+      "epoch": 2.8627798714254045,
+      "grad_norm": 0.011626002378761768,
+      "learning_rate": 5.4260234973186335e-08,
+      "loss": 0.0,
+      "step": 3227
+    },
+    {
+      "epoch": 2.863666592773221,
+      "grad_norm": 0.042777977883815765,
+      "learning_rate": 5.3559101206200337e-08,
+      "loss": 0.0001,
+      "step": 3228
+    },
+    {
+      "epoch": 2.8645533141210375,
+      "grad_norm": 0.0015777841908857226,
+      "learning_rate": 5.286250246828994e-08,
+      "loss": 0.0,
+      "step": 3229
+    },
+    {
+      "epoch": 2.865440035468854,
+      "grad_norm": 0.041069697588682175,
+      "learning_rate": 5.217043939811595e-08,
+      "loss": 0.0001,
+      "step": 3230
+    },
+    {
+      "epoch": 2.8663267568166706,
+      "grad_norm": 0.0033635362051427364,
+      "learning_rate": 5.1482912630180234e-08,
+      "loss": 0.0,
+      "step": 3231
+    },
+    {
+      "epoch": 2.8672134781644867,
+      "grad_norm": 0.4834699332714081,
+      "learning_rate": 5.079992279482471e-08,
+      "loss": 0.0004,
+      "step": 3232
+    },
+    {
+      "epoch": 2.868100199512303,
+      "grad_norm": 0.23373693227767944,
+      "learning_rate": 5.012147051823346e-08,
+      "loss": 0.0017,
+      "step": 3233
+    },
+    {
+      "epoch": 2.8689869208601197,
+      "grad_norm": 0.008513184264302254,
+      "learning_rate": 4.9447556422430046e-08,
+      "loss": 0.0,
+      "step": 3234
+    },
+    {
+      "epoch": 2.8698736422079363,
+      "grad_norm": 0.009647476486861706,
+      "learning_rate": 4.877818112527632e-08,
+      "loss": 0.0,
+      "step": 3235
+    },
+    {
+      "epoch": 2.8707603635557524,
+      "grad_norm": 0.017421780154109,
+      "learning_rate": 4.811334524047307e-08,
+      "loss": 0.0,
+      "step": 3236
+    },
+    {
+      "epoch": 2.871647084903569,
+      "grad_norm": 0.013581366278231144,
+      "learning_rate": 4.745304937756101e-08,
+      "loss": 0.0,
+      "step": 3237
+    },
+    {
+      "epoch": 2.8725338062513854,
+      "grad_norm": 0.0025726743042469025,
+      "learning_rate": 4.679729414191536e-08,
+      "loss": 0.0,
+      "step": 3238
+    },
+    {
+      "epoch": 2.873420527599202,
+      "grad_norm": 0.005139803979545832,
+      "learning_rate": 4.614608013475075e-08,
+      "loss": 0.0,
+      "step": 3239
+    },
+    {
+      "epoch": 2.8743072489470185,
+      "grad_norm": 0.004231195896863937,
+      "learning_rate": 4.5499407953117355e-08,
+      "loss": 0.0,
+      "step": 3240
+    },
+    {
+      "epoch": 2.875193970294835,
+      "grad_norm": 0.015432806685566902,
+      "learning_rate": 4.4857278189902046e-08,
+      "loss": 0.0,
+      "step": 3241
+    },
+    {
+      "epoch": 2.8760806916426516,
+      "grad_norm": 0.014841458760201931,
+      "learning_rate": 4.4219691433826094e-08,
+      "loss": 0.0,
+      "step": 3242
+    },
+    {
+      "epoch": 2.8769674129904677,
+      "grad_norm": 0.0013239766703918576,
+      "learning_rate": 4.358664826944636e-08,
+      "loss": 0.0,
+      "step": 3243
+    },
+    {
+      "epoch": 2.877854134338284,
+      "grad_norm": 0.009761502966284752,
+      "learning_rate": 4.295814927715303e-08,
+      "loss": 0.0,
+      "step": 3244
+    },
+    {
+      "epoch": 2.8787408556861007,
+      "grad_norm": 0.003177881008014083,
+      "learning_rate": 4.233419503317182e-08,
+      "loss": 0.0,
+      "step": 3245
+    },
+    {
+      "epoch": 2.8796275770339173,
+      "grad_norm": 0.35972318053245544,
+      "learning_rate": 4.171478610955904e-08,
+      "loss": 0.0009,
+      "step": 3246
+    },
+    {
+      "epoch": 2.8805142983817333,
+      "grad_norm": 0.0046821520663797855,
+      "learning_rate": 4.109992307420707e-08,
+      "loss": 0.0,
+      "step": 3247
+    },
+    {
+      "epoch": 2.88140101972955,
+      "grad_norm": 0.02193521521985531,
+      "learning_rate": 4.048960649083777e-08,
+      "loss": 0.0001,
+      "step": 3248
+    },
+    {
+      "epoch": 2.8822877410773664,
+      "grad_norm": 0.16279035806655884,
+      "learning_rate": 3.9883836919006325e-08,
+      "loss": 0.0009,
+      "step": 3249
+    },
+    {
+      "epoch": 2.883174462425183,
+      "grad_norm": 0.04672611132264137,
+      "learning_rate": 3.928261491409735e-08,
+      "loss": 0.0001,
+      "step": 3250
+    },
+    {
+      "epoch": 2.8840611837729995,
+      "grad_norm": 0.12944093346595764,
+      "learning_rate": 3.868594102732881e-08,
+      "loss": 0.0006,
+      "step": 3251
+    },
+    {
+      "epoch": 2.884947905120816,
+      "grad_norm": 0.4571356773376465,
+      "learning_rate": 3.8093815805745895e-08,
+      "loss": 0.0018,
+      "step": 3252
+    },
+    {
+      "epoch": 2.885834626468632,
+      "grad_norm": 0.21273784339427948,
+      "learning_rate": 3.7506239792225454e-08,
+      "loss": 0.0013,
+      "step": 3253
+    },
+    {
+      "epoch": 2.8867213478164486,
+      "grad_norm": 0.11761754006147385,
+      "learning_rate": 3.6923213525472126e-08,
+      "loss": 0.0004,
+      "step": 3254
+    },
+    {
+      "epoch": 2.887608069164265,
+      "grad_norm": 0.005210244562476873,
+      "learning_rate": 3.63447375400211e-08,
+      "loss": 0.0,
+      "step": 3255
+    },
+    {
+      "epoch": 2.8884947905120817,
+      "grad_norm": 0.025401975959539413,
+      "learning_rate": 3.577081236623425e-08,
+      "loss": 0.0001,
+      "step": 3256
+    },
+    {
+      "epoch": 2.8893815118598978,
+      "grad_norm": 0.0018261591903865337,
+      "learning_rate": 3.5201438530300666e-08,
+      "loss": 0.0,
+      "step": 3257
+    },
+    {
+      "epoch": 2.8902682332077143,
+      "grad_norm": 0.002138376934453845,
+      "learning_rate": 3.463661655423889e-08,
+      "loss": 0.0,
+      "step": 3258
+    },
+    {
+      "epoch": 2.891154954555531,
+      "grad_norm": 0.003766312263906002,
+      "learning_rate": 3.407634695589135e-08,
+      "loss": 0.0,
+      "step": 3259
+    },
+    {
+      "epoch": 2.8920416759033474,
+      "grad_norm": 0.0012409350601956248,
+      "learning_rate": 3.352063024892882e-08,
+      "loss": 0.0,
+      "step": 3260
+    },
+    {
+      "epoch": 2.892928397251164,
+      "grad_norm": 0.7102566361427307,
+      "learning_rate": 3.296946694284764e-08,
+      "loss": 0.0036,
+      "step": 3261
+    },
+    {
+      "epoch": 2.8938151185989804,
+      "grad_norm": 0.1386748105287552,
+      "learning_rate": 3.242285754296859e-08,
+      "loss": 0.0008,
+      "step": 3262
+    },
+    {
+      "epoch": 2.894701839946797,
+      "grad_norm": 0.05686341971158981,
+      "learning_rate": 3.188080255043746e-08,
+      "loss": 0.0001,
+      "step": 3263
+    },
+    {
+      "epoch": 2.895588561294613,
+      "grad_norm": 0.011563396081328392,
+      "learning_rate": 3.1343302462225054e-08,
+      "loss": 0.0,
+      "step": 3264
+    },
+    {
+      "epoch": 2.8964752826424296,
+      "grad_norm": 0.0008911413606256247,
+      "learning_rate": 3.0810357771126064e-08,
+      "loss": 0.0,
+      "step": 3265
+    },
+    {
+      "epoch": 2.897362003990246,
+      "grad_norm": 0.0011407394194975495,
+      "learning_rate": 3.028196896575852e-08,
+      "loss": 0.0,
+      "step": 3266
+    },
+    {
+      "epoch": 2.8982487253380627,
+      "grad_norm": 0.0022614169865846634,
+      "learning_rate": 2.9758136530562143e-08,
+      "loss": 0.0,
+      "step": 3267
+    },
+    {
+      "epoch": 2.8991354466858787,
+      "grad_norm": 0.002719403011724353,
+      "learning_rate": 2.9238860945802193e-08,
+      "loss": 0.0,
+      "step": 3268
+    },
+    {
+      "epoch": 2.9000221680336953,
+      "grad_norm": 0.0018850235501304269,
+      "learning_rate": 2.872414268756285e-08,
+      "loss": 0.0,
+      "step": 3269
+    },
+    {
+      "epoch": 2.900908889381512,
+      "grad_norm": 0.0057585276663303375,
+      "learning_rate": 2.821398222775329e-08,
+      "loss": 0.0,
+      "step": 3270
+    },
+    {
+      "epoch": 2.9017956107293283,
+      "grad_norm": 0.00659179175272584,
+      "learning_rate": 2.7708380034099923e-08,
+      "loss": 0.0,
+      "step": 3271
+    },
+    {
+      "epoch": 2.902682332077145,
+      "grad_norm": 0.0016464820364490151,
+      "learning_rate": 2.720733657015473e-08,
+      "loss": 0.0,
+      "step": 3272
+    },
+    {
+      "epoch": 2.9035690534249614,
+      "grad_norm": 0.0004565789131447673,
+      "learning_rate": 2.6710852295286362e-08,
+      "loss": 0.0,
+      "step": 3273
+    },
+    {
+      "epoch": 2.9044557747727775,
+      "grad_norm": 0.012226813472807407,
+      "learning_rate": 2.6218927664685145e-08,
+      "loss": 0.0,
+      "step": 3274
+    },
+    {
+      "epoch": 2.905342496120594,
+      "grad_norm": 0.00505524966865778,
+      "learning_rate": 2.573156312936087e-08,
+      "loss": 0.0,
+      "step": 3275
+    },
+    {
+      "epoch": 2.9062292174684106,
+      "grad_norm": 0.008225847966969013,
+      "learning_rate": 2.5248759136142222e-08,
+      "loss": 0.0,
+      "step": 3276
+    },
+    {
+      "epoch": 2.907115938816227,
+      "grad_norm": 0.008066046983003616,
+      "learning_rate": 2.4770516127676226e-08,
+      "loss": 0.0,
+      "step": 3277
+    },
+    {
+      "epoch": 2.908002660164043,
+      "grad_norm": 0.037920404225587845,
+      "learning_rate": 2.4296834542429926e-08,
+      "loss": 0.0,
+      "step": 3278
+    },
+    {
+      "epoch": 2.9088893815118597,
+      "grad_norm": 0.01949114352464676,
+      "learning_rate": 2.3827714814686488e-08,
+      "loss": 0.0001,
+      "step": 3279
+    },
+    {
+      "epoch": 2.9097761028596763,
+      "grad_norm": 0.00045723363291472197,
+      "learning_rate": 2.3363157374547417e-08,
+      "loss": 0.0,
+      "step": 3280
+    },
+    {
+      "epoch": 2.910662824207493,
+      "grad_norm": 0.0014682360924780369,
+      "learning_rate": 2.2903162647932022e-08,
+      "loss": 0.0,
+      "step": 3281
+    },
+    {
+      "epoch": 2.9115495455553093,
+      "grad_norm": 0.0035071983002126217,
+      "learning_rate": 2.244773105657516e-08,
+      "loss": 0.0,
+      "step": 3282
+    },
+    {
+      "epoch": 2.912436266903126,
+      "grad_norm": 0.0004839412576984614,
+      "learning_rate": 2.1996863018028945e-08,
+      "loss": 0.0,
+      "step": 3283
+    },
+    {
+      "epoch": 2.9133229882509424,
+      "grad_norm": 0.0006948516820557415,
+      "learning_rate": 2.155055894566105e-08,
+      "loss": 0.0,
+      "step": 3284
+    },
+    {
+      "epoch": 2.9142097095987585,
+      "grad_norm": 0.07006675004959106,
+      "learning_rate": 2.1108819248655266e-08,
+      "loss": 0.0003,
+      "step": 3285
+    },
+    {
+      "epoch": 2.915096430946575,
+      "grad_norm": 0.11140327900648117,
+      "learning_rate": 2.0671644332009854e-08,
+      "loss": 0.0006,
+      "step": 3286
+    },
+    {
+      "epoch": 2.9159831522943915,
+      "grad_norm": 0.0038438565097749233,
+      "learning_rate": 2.0239034596538644e-08,
+      "loss": 0.0,
+      "step": 3287
+    },
+    {
+      "epoch": 2.916869873642208,
+      "grad_norm": 0.0013927739346399903,
+      "learning_rate": 1.9810990438869913e-08,
+      "loss": 0.0,
+      "step": 3288
+    },
+    {
+      "epoch": 2.917756594990024,
+      "grad_norm": 0.004284338094294071,
+      "learning_rate": 1.938751225144586e-08,
+      "loss": 0.0,
+      "step": 3289
+    },
+    {
+      "epoch": 2.9186433163378407,
+      "grad_norm": 0.03040909208357334,
+      "learning_rate": 1.8968600422522577e-08,
+      "loss": 0.0001,
+      "step": 3290
+    },
+    {
+      "epoch": 2.919530037685657,
+      "grad_norm": 0.004622400738298893,
+      "learning_rate": 1.8554255336170056e-08,
+      "loss": 0.0,
+      "step": 3291
+    },
+    {
+      "epoch": 2.9204167590334738,
+      "grad_norm": 0.005725715775042772,
+      "learning_rate": 1.8144477372269988e-08,
+      "loss": 0.0,
+      "step": 3292
+    },
+    {
+      "epoch": 2.9213034803812903,
+      "grad_norm": 0.06149230897426605,
+      "learning_rate": 1.7739266906519058e-08,
+      "loss": 0.0002,
+      "step": 3293
+    },
+    {
+      "epoch": 2.922190201729107,
+      "grad_norm": 0.005780136212706566,
+      "learning_rate": 1.733862431042399e-08,
+      "loss": 0.0,
+      "step": 3294
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.013424360193312168,
+      "learning_rate": 1.6942549951304844e-08,
+      "loss": 0.0,
+      "step": 3295
+    },
+    {
+      "epoch": 2.9239636444247394,
+      "grad_norm": 0.0061098430305719376,
+      "learning_rate": 1.655104419229281e-08,
+      "loss": 0.0,
+      "step": 3296
+    },
+    {
+      "epoch": 2.924850365772556,
+      "grad_norm": 0.0005447586299851537,
+      "learning_rate": 1.6164107392331873e-08,
+      "loss": 0.0,
+      "step": 3297
+    },
+    {
+      "epoch": 2.9257370871203725,
+      "grad_norm": 0.019417855888605118,
+      "learning_rate": 1.5781739906175486e-08,
+      "loss": 0.0,
+      "step": 3298
+    },
+    {
+      "epoch": 2.926623808468189,
+      "grad_norm": 0.0017337505705654621,
+      "learning_rate": 1.5403942084387667e-08,
+      "loss": 0.0,
+      "step": 3299
+    },
+    {
+      "epoch": 2.927510529816005,
+      "grad_norm": 0.006602004170417786,
+      "learning_rate": 1.503071427334468e-08,
+      "loss": 0.0,
+      "step": 3300
+    },
+    {
+      "epoch": 2.9283972511638217,
+      "grad_norm": 0.011092391796410084,
+      "learning_rate": 1.466205681523114e-08,
+      "loss": 0.0001,
+      "step": 3301
+    },
+    {
+      "epoch": 2.929283972511638,
+      "grad_norm": 0.0026441768277436495,
+      "learning_rate": 1.4297970048041676e-08,
+      "loss": 0.0,
+      "step": 3302
+    },
+    {
+      "epoch": 2.9301706938594547,
+      "grad_norm": 0.00047737229033373296,
+      "learning_rate": 1.393845430558205e-08,
+      "loss": 0.0,
+      "step": 3303
+    },
+    {
+      "epoch": 2.9310574152072713,
+      "grad_norm": 0.011745357885956764,
+      "learning_rate": 1.3583509917464155e-08,
+      "loss": 0.0001,
+      "step": 3304
+    },
+    {
+      "epoch": 2.931944136555088,
+      "grad_norm": 0.0005886032013222575,
+      "learning_rate": 1.3233137209112123e-08,
+      "loss": 0.0,
+      "step": 3305
+    },
+    {
+      "epoch": 2.932830857902904,
+      "grad_norm": 0.039524223655462265,
+      "learning_rate": 1.2887336501755666e-08,
+      "loss": 0.0001,
+      "step": 3306
+    },
+    {
+      "epoch": 2.9337175792507204,
+      "grad_norm": 0.008052697405219078,
+      "learning_rate": 1.2546108112435062e-08,
+      "loss": 0.0,
+      "step": 3307
+    },
+    {
+      "epoch": 2.934604300598537,
+      "grad_norm": 0.003623201046139002,
+      "learning_rate": 1.2209452353997841e-08,
+      "loss": 0.0,
+      "step": 3308
+    },
+    {
+      "epoch": 2.9354910219463535,
+      "grad_norm": 0.0007283863378688693,
+      "learning_rate": 1.187736953509766e-08,
+      "loss": 0.0,
+      "step": 3309
+    },
+    {
+      "epoch": 2.9363777432941696,
+      "grad_norm": 0.13052868843078613,
+      "learning_rate": 1.154985996019764e-08,
+      "loss": 0.0003,
+      "step": 3310
+    },
+    {
+      "epoch": 2.937264464641986,
+      "grad_norm": 0.002719426527619362,
+      "learning_rate": 1.122692392956759e-08,
+      "loss": 0.0,
+      "step": 3311
+    },
+    {
+      "epoch": 2.9381511859898026,
+      "grad_norm": 0.10330086201429367,
+      "learning_rate": 1.0908561739283452e-08,
+      "loss": 0.0004,
+      "step": 3312
+    },
+    {
+      "epoch": 2.939037907337619,
+      "grad_norm": 0.009346824139356613,
+      "learning_rate": 1.059477368122841e-08,
+      "loss": 0.0,
+      "step": 3313
+    },
+    {
+      "epoch": 2.9399246286854357,
+      "grad_norm": 0.004005473107099533,
+      "learning_rate": 1.028556004309178e-08,
+      "loss": 0.0,
+      "step": 3314
+    },
+    {
+      "epoch": 2.9408113500332522,
+      "grad_norm": 0.0023341896012425423,
+      "learning_rate": 9.980921108368457e-09,
+      "loss": 0.0,
+      "step": 3315
+    },
+    {
+      "epoch": 2.9416980713810688,
+      "grad_norm": 0.0034011153038591146,
+      "learning_rate": 9.68085715636058e-09,
+      "loss": 0.0,
+      "step": 3316
+    },
+    {
+      "epoch": 2.942584792728885,
+      "grad_norm": 0.0011600746074691415,
+      "learning_rate": 9.385368462173083e-09,
+      "loss": 0.0,
+      "step": 3317
+    },
+    {
+      "epoch": 2.9434715140767014,
+      "grad_norm": 0.0015895559918135405,
+      "learning_rate": 9.094455296719817e-09,
+      "loss": 0.0,
+      "step": 3318
+    },
+    {
+      "epoch": 2.944358235424518,
+      "grad_norm": 0.014281976036727428,
+      "learning_rate": 8.80811792671632e-09,
+      "loss": 0.0,
+      "step": 3319
+    },
+    {
+      "epoch": 2.9452449567723344,
+      "grad_norm": 0.2445225864648819,
+      "learning_rate": 8.526356614684816e-09,
+      "loss": 0.001,
+      "step": 3320
+    },
+    {
+      "epoch": 2.9461316781201505,
+      "grad_norm": 0.011881627142429352,
+      "learning_rate": 8.249171618952002e-09,
+      "loss": 0.0,
+      "step": 3321
+    },
+    {
+      "epoch": 2.947018399467967,
+      "grad_norm": 0.1268923282623291,
+      "learning_rate": 7.976563193647924e-09,
+      "loss": 0.0003,
+      "step": 3322
+    },
+    {
+      "epoch": 2.9479051208157836,
+      "grad_norm": 0.013145245611667633,
+      "learning_rate": 7.708531588707103e-09,
+      "loss": 0.0,
+      "step": 3323
+    },
+    {
+      "epoch": 2.9487918421636,
+      "grad_norm": 0.0036690891720354557,
+      "learning_rate": 7.445077049868521e-09,
+      "loss": 0.0,
+      "step": 3324
+    },
+    {
+      "epoch": 2.9496785635114167,
+      "grad_norm": 0.0030290221329778433,
+      "learning_rate": 7.186199818673967e-09,
+      "loss": 0.0,
+      "step": 3325
+    },
+    {
+      "epoch": 2.950565284859233,
+      "grad_norm": 0.006416503805667162,
+      "learning_rate": 6.931900132469694e-09,
+      "loss": 0.0,
+      "step": 3326
+    },
+    {
+      "epoch": 2.9514520062070493,
+      "grad_norm": 0.059920359402894974,
+      "learning_rate": 6.682178224403646e-09,
+      "loss": 0.0003,
+      "step": 3327
+    },
+    {
+      "epoch": 2.952338727554866,
+      "grad_norm": 0.042692240327596664,
+      "learning_rate": 6.4370343234282375e-09,
+      "loss": 0.0001,
+      "step": 3328
+    },
+    {
+      "epoch": 2.9532254489026823,
+      "grad_norm": 0.01230137050151825,
+      "learning_rate": 6.1964686542975716e-09,
+      "loss": 0.0,
+      "step": 3329
+    },
+    {
+      "epoch": 2.954112170250499,
+      "grad_norm": 0.0041125863790512085,
+      "learning_rate": 5.9604814375685546e-09,
+      "loss": 0.0,
+      "step": 3330
+    },
+    {
+      "epoch": 2.954998891598315,
+      "grad_norm": 0.01572389528155327,
+      "learning_rate": 5.72907288960034e-09,
+      "loss": 0.0,
+      "step": 3331
+    },
+    {
+      "epoch": 2.9558856129461315,
+      "grad_norm": 0.0016322958981618285,
+      "learning_rate": 5.502243222555437e-09,
+      "loss": 0.0,
+      "step": 3332
+    },
+    {
+      "epoch": 2.956772334293948,
+      "grad_norm": 0.30885788798332214,
+      "learning_rate": 5.279992644396381e-09,
+      "loss": 0.0061,
+      "step": 3333
+    },
+    {
+      "epoch": 2.9576590556417646,
+      "grad_norm": 0.0038180218543857336,
+      "learning_rate": 5.0623213588885114e-09,
+      "loss": 0.0,
+      "step": 3334
+    },
+    {
+      "epoch": 2.958545776989581,
+      "grad_norm": 0.01314638089388609,
+      "learning_rate": 4.849229565599412e-09,
+      "loss": 0.0001,
+      "step": 3335
+    },
+    {
+      "epoch": 2.9594324983373976,
+      "grad_norm": 0.005611849948763847,
+      "learning_rate": 4.640717459896693e-09,
+      "loss": 0.0,
+      "step": 3336
+    },
+    {
+      "epoch": 2.960319219685214,
+      "grad_norm": 0.006579660344868898,
+      "learning_rate": 4.436785232950214e-09,
+      "loss": 0.0,
+      "step": 3337
+    },
+    {
+      "epoch": 2.9612059410330303,
+      "grad_norm": 0.00035080299130640924,
+      "learning_rate": 4.237433071729857e-09,
+      "loss": 0.0,
+      "step": 3338
+    },
+    {
+      "epoch": 2.962092662380847,
+      "grad_norm": 0.0008623858448117971,
+      "learning_rate": 4.042661159007755e-09,
+      "loss": 0.0,
+      "step": 3339
+    },
+    {
+      "epoch": 2.9629793837286633,
+      "grad_norm": 0.1388547718524933,
+      "learning_rate": 3.852469673355508e-09,
+      "loss": 0.0007,
+      "step": 3340
+    },
+    {
+      "epoch": 2.96386610507648,
+      "grad_norm": 0.06763661652803421,
+      "learning_rate": 3.666858789145855e-09,
+      "loss": 0.0003,
+      "step": 3341
+    },
+    {
+      "epoch": 2.964752826424296,
+      "grad_norm": 0.0008432602626271546,
+      "learning_rate": 3.4858286765515614e-09,
+      "loss": 0.0,
+      "step": 3342
+    },
+    {
+      "epoch": 2.9656395477721125,
+      "grad_norm": 0.004892279859632254,
+      "learning_rate": 3.309379501546528e-09,
+      "loss": 0.0,
+      "step": 3343
+    },
+    {
+      "epoch": 2.966526269119929,
+      "grad_norm": 0.003920522518455982,
+      "learning_rate": 3.1375114259035723e-09,
+      "loss": 0.0,
+      "step": 3344
+    },
+    {
+      "epoch": 2.9674129904677455,
+      "grad_norm": 0.005874779541045427,
+      "learning_rate": 2.970224607196093e-09,
+      "loss": 0.0,
+      "step": 3345
+    },
+    {
+      "epoch": 2.968299711815562,
+      "grad_norm": 0.05062325671315193,
+      "learning_rate": 2.8075191987969597e-09,
+      "loss": 0.0002,
+      "step": 3346
+    },
+    {
+      "epoch": 2.9691864331633786,
+      "grad_norm": 0.10245689749717712,
+      "learning_rate": 2.6493953498790692e-09,
+      "loss": 0.0002,
+      "step": 3347
+    },
+    {
+      "epoch": 2.9700731545111947,
+      "grad_norm": 0.01170055940747261,
+      "learning_rate": 2.4958532054142338e-09,
+      "loss": 0.0,
+      "step": 3348
+    },
+    {
+      "epoch": 2.9709598758590112,
+      "grad_norm": 0.001403836184181273,
+      "learning_rate": 2.346892906174847e-09,
+      "loss": 0.0,
+      "step": 3349
+    },
+    {
+      "epoch": 2.9718465972068278,
+      "grad_norm": 0.06057799234986305,
+      "learning_rate": 2.2025145887305533e-09,
+      "loss": 0.0001,
+      "step": 3350
+    },
+    {
+      "epoch": 2.9727333185546443,
+      "grad_norm": 0.04310649633407593,
+      "learning_rate": 2.062718385451579e-09,
+      "loss": 0.0001,
+      "step": 3351
+    },
+    {
+      "epoch": 2.9736200399024604,
+      "grad_norm": 0.3466622531414032,
+      "learning_rate": 1.9275044245076203e-09,
+      "loss": 0.0006,
+      "step": 3352
+    },
+    {
+      "epoch": 2.974506761250277,
+      "grad_norm": 0.0031994169112294912,
+      "learning_rate": 1.7968728298650705e-09,
+      "loss": 0.0,
+      "step": 3353
+    },
+    {
+      "epoch": 2.9753934825980934,
+      "grad_norm": 1.4261260032653809,
+      "learning_rate": 1.6708237212920142e-09,
+      "loss": 0.0008,
+      "step": 3354
+    },
+    {
+      "epoch": 2.97628020394591,
+      "grad_norm": 0.0007013491122052073,
+      "learning_rate": 1.5493572143521207e-09,
+      "loss": 0.0,
+      "step": 3355
+    },
+    {
+      "epoch": 2.9771669252937265,
+      "grad_norm": 0.1456507444381714,
+      "learning_rate": 1.4324734204101964e-09,
+      "loss": 0.0007,
+      "step": 3356
+    },
+    {
+      "epoch": 2.978053646641543,
+      "grad_norm": 0.006528434809297323,
+      "learning_rate": 1.3201724466277432e-09,
+      "loss": 0.0,
+      "step": 3357
+    },
+    {
+      "epoch": 2.9789403679893596,
+      "grad_norm": 0.00108535576146096,
+      "learning_rate": 1.2124543959662894e-09,
+      "loss": 0.0,
+      "step": 3358
+    },
+    {
+      "epoch": 2.9798270893371757,
+      "grad_norm": 0.015093772672116756,
+      "learning_rate": 1.1093193671835035e-09,
+      "loss": 0.0001,
+      "step": 3359
+    },
+    {
+      "epoch": 2.980713810684992,
+      "grad_norm": 0.009435775689780712,
+      "learning_rate": 1.0107674548365255e-09,
+      "loss": 0.0,
+      "step": 3360
+    },
+    {
+      "epoch": 2.9816005320328087,
+      "grad_norm": 0.001818368211388588,
+      "learning_rate": 9.167987492808561e-10,
+      "loss": 0.0,
+      "step": 3361
+    },
+    {
+      "epoch": 2.9824872533806253,
+      "grad_norm": 0.7512038946151733,
+      "learning_rate": 8.274133366698023e-10,
+      "loss": 0.0029,
+      "step": 3362
+    },
+    {
+      "epoch": 2.9833739747284413,
+      "grad_norm": 0.33148589730262756,
+      "learning_rate": 7.426112989533663e-10,
+      "loss": 0.0026,
+      "step": 3363
+    },
+    {
+      "epoch": 2.984260696076258,
+      "grad_norm": 0.08316826820373535,
+      "learning_rate": 6.623927138804665e-10,
+      "loss": 0.0004,
+      "step": 3364
+    },
+    {
+      "epoch": 2.9851474174240744,
+      "grad_norm": 0.023770712316036224,
+      "learning_rate": 5.867576549983822e-10,
+      "loss": 0.0001,
+      "step": 3365
+    },
+    {
+      "epoch": 2.986034138771891,
+      "grad_norm": 0.006840167101472616,
+      "learning_rate": 5.157061916505335e-10,
+      "loss": 0.0,
+      "step": 3366
+    },
+    {
+      "epoch": 2.9869208601197075,
+      "grad_norm": 0.005075496155768633,
+      "learning_rate": 4.492383889792562e-10,
+      "loss": 0.0,
+      "step": 3367
+    },
+    {
+      "epoch": 2.987807581467524,
+      "grad_norm": 0.03411711007356644,
+      "learning_rate": 3.873543079241371e-10,
+      "loss": 0.0001,
+      "step": 3368
+    },
+    {
+      "epoch": 2.9886943028153405,
+      "grad_norm": 0.38265281915664673,
+      "learning_rate": 3.300540052214585e-10,
+      "loss": 0.0014,
+      "step": 3369
+    },
+    {
+      "epoch": 2.9895810241631566,
+      "grad_norm": 0.0326530784368515,
+      "learning_rate": 2.773375334064188e-10,
+      "loss": 0.0001,
+      "step": 3370
+    },
+    {
+      "epoch": 2.990467745510973,
+      "grad_norm": 0.002753943670541048,
+      "learning_rate": 2.2920494081091204e-10,
+      "loss": 0.0,
+      "step": 3371
+    },
+    {
+      "epoch": 2.9913544668587897,
+      "grad_norm": 0.07752973586320877,
+      "learning_rate": 1.8565627156352794e-10,
+      "loss": 0.0003,
+      "step": 3372
+    },
+    {
+      "epoch": 2.9922411882066062,
+      "grad_norm": 0.0007558520883321762,
+      "learning_rate": 1.466915655912171e-10,
+      "loss": 0.0,
+      "step": 3373
+    },
+    {
+      "epoch": 2.9931279095544223,
+      "grad_norm": 0.052483346313238144,
+      "learning_rate": 1.123108586176258e-10,
+      "loss": 0.0002,
+      "step": 3374
+    },
+    {
+      "epoch": 2.994014630902239,
+      "grad_norm": 0.0038302731700241566,
+      "learning_rate": 8.251418216420615e-11,
+      "loss": 0.0,
+      "step": 3375
+    },
+    {
+      "epoch": 2.9949013522500554,
+      "grad_norm": 0.0008110279450193048,
+      "learning_rate": 5.730156354966099e-11,
+      "loss": 0.0,
+      "step": 3376
+    },
+    {
+      "epoch": 2.995788073597872,
+      "grad_norm": 0.00048128835624083877,
+      "learning_rate": 3.66730258888337e-11,
+      "loss": 0.0,
+      "step": 3377
+    },
+    {
+      "epoch": 2.9966747949456884,
+      "grad_norm": 0.004053472075611353,
+      "learning_rate": 2.0628588094928627e-11,
+      "loss": 0.0,
+      "step": 3378
+    },
+    {
+      "epoch": 2.997561516293505,
+      "grad_norm": 0.014079282991588116,
+      "learning_rate": 9.168264877845723e-12,
+      "loss": 0.0001,
+      "step": 3379
+    },
+    {
+      "epoch": 2.998448237641321,
+      "grad_norm": 0.0015199949266389012,
+      "learning_rate": 2.292066744735699e-12,
+      "loss": 0.0,
+      "step": 3380
+    },
+    {
+      "epoch": 2.9993349589891376,
+      "grad_norm": 0.001595963491126895,
+      "learning_rate": 0.0,
+      "loss": 0.0,
+      "step": 3381
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3381,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 564,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.102510967617749e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}