diff --git "a/checkpoint-6102/trainer_state.json" "b/checkpoint-6102/trainer_state.json"
deleted file mode 100644--- "a/checkpoint-6102/trainer_state.json"
+++ /dev/null
@@ -1,42803 +0,0 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 2.9964976958525344,
-  "eval_steps": 1017,
-  "global_step": 6102,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 0.0004918839153959665,
-      "grad_norm": 3.2944885419116114,
-      "learning_rate": 2.0000000000000002e-07,
-      "loss": 1.3754,
-      "step": 1
-    },
-    {
-      "epoch": 0.0004918839153959665,
-      "eval_loss": 1.7429497241973877,
-      "eval_runtime": 6757.8165,
-      "eval_samples_per_second": 4.217,
-      "eval_steps_per_second": 2.109,
-      "step": 1
-    },
-    {
-      "epoch": 0.000983767830791933,
-      "grad_norm": 4.387903167670078,
-      "learning_rate": 4.0000000000000003e-07,
-      "loss": 1.4888,
-      "step": 2
-    },
-    {
-      "epoch": 0.0014756517461878996,
-      "grad_norm": 3.702486400269614,
-      "learning_rate": 6.000000000000001e-07,
-      "loss": 1.42,
-      "step": 3
-    },
-    {
-      "epoch": 0.001967535661583866,
-      "grad_norm": 3.6873356554967187,
-      "learning_rate": 8.000000000000001e-07,
-      "loss": 1.4333,
-      "step": 4
-    },
-    {
-      "epoch": 0.002459419576979833,
-      "grad_norm": 3.4722776683329206,
-      "learning_rate": 1.0000000000000002e-06,
-      "loss": 1.4012,
-      "step": 5
-    },
-    {
-      "epoch": 0.002951303492375799,
-      "grad_norm": 3.501303887226817,
-      "learning_rate": 1.2000000000000002e-06,
-      "loss": 1.4415,
-      "step": 6
-    },
-    {
-      "epoch": 0.003443187407771766,
-      "grad_norm": 3.7234243820409483,
-      "learning_rate": 1.4000000000000001e-06,
-      "loss": 1.393,
-      "step": 7
-    },
-    {
-      "epoch": 0.003935071323167732,
-      "grad_norm": 4.291923849986159,
-      "learning_rate": 1.6000000000000001e-06,
-      "loss": 1.4528,
-      "step": 8
-    },
-    {
-      "epoch": 0.004426955238563699,
-      "grad_norm": 3.551363047162373,
-      "learning_rate": 1.8e-06,
-      "loss": 1.5339,
-      "step": 9
-    },
-    {
-      "epoch": 0.004918839153959666,
-      "grad_norm": 3.2095887045357925,
-      "learning_rate": 2.0000000000000003e-06,
-      "loss": 1.4037,
-      "step": 10
-    },
-    {
-      "epoch": 0.0054107230693556324,
-      "grad_norm": 3.2136633682112525,
-      "learning_rate": 2.2e-06,
-      "loss": 1.4643,
-      "step": 11
-    },
-    {
-      "epoch": 0.005902606984751598,
-      "grad_norm": 3.2755879900685083,
-      "learning_rate": 2.4000000000000003e-06,
-      "loss": 1.4317,
-      "step": 12
-    },
-    {
-      "epoch": 0.006394490900147565,
-      "grad_norm": 3.0262191829479654,
-      "learning_rate": 2.6e-06,
-      "loss": 1.385,
-      "step": 13
-    },
-    {
-      "epoch": 0.006886374815543532,
-      "grad_norm": 2.829759375766714,
-      "learning_rate": 2.8000000000000003e-06,
-      "loss": 1.4019,
-      "step": 14
-    },
-    {
-      "epoch": 0.0073782587309394985,
-      "grad_norm": 3.3734116571047603,
-      "learning_rate": 3e-06,
-      "loss": 1.5036,
-      "step": 15
-    },
-    {
-      "epoch": 0.007870142646335464,
-      "grad_norm": 2.2223043949951697,
-      "learning_rate": 3.2000000000000003e-06,
-      "loss": 1.4211,
-      "step": 16
-    },
-    {
-      "epoch": 0.008362026561731432,
-      "grad_norm": 2.150664036408096,
-      "learning_rate": 3.4000000000000005e-06,
-      "loss": 1.421,
-      "step": 17
-    },
-    {
-      "epoch": 0.008853910477127398,
-      "grad_norm": 2.167189487369783,
-      "learning_rate": 3.6e-06,
-      "loss": 1.412,
-      "step": 18
-    },
-    {
-      "epoch": 0.009345794392523364,
-      "grad_norm": 1.9248200012962078,
-      "learning_rate": 3.8e-06,
-      "loss": 1.3272,
-      "step": 19
-    },
-    {
-      "epoch": 0.009837678307919331,
-      "grad_norm": 1.4357536891109497,
-      "learning_rate": 4.000000000000001e-06,
-      "loss": 1.2681,
-      "step": 20
-    },
-    {
-      "epoch": 0.010329562223315297,
-      "grad_norm": 1.4074010741692329,
-      "learning_rate": 4.2000000000000004e-06,
-      "loss": 1.3999,
-      "step": 21
-    },
-    {
-      "epoch": 0.010821446138711265,
-      "grad_norm": 1.3417126612322257,
-      "learning_rate": 4.4e-06,
-      "loss": 1.3597,
-      "step": 22
-    },
-    {
-      "epoch": 0.01131333005410723,
-      "grad_norm": 1.2085318394846978,
-      "learning_rate": 4.6e-06,
-      "loss": 1.2399,
-      "step": 23
-    },
-    {
-      "epoch": 0.011805213969503197,
-      "grad_norm": 1.2314668941371152,
-      "learning_rate": 4.800000000000001e-06,
-      "loss": 1.2098,
-      "step": 24
-    },
-    {
-      "epoch": 0.012297097884899164,
-      "grad_norm": 1.1065431950366,
-      "learning_rate": 5e-06,
-      "loss": 1.1893,
-      "step": 25
-    },
-    {
-      "epoch": 0.01278898180029513,
-      "grad_norm": 1.2923599312074627,
-      "learning_rate": 5.2e-06,
-      "loss": 1.4414,
-      "step": 26
-    },
-    {
-      "epoch": 0.013280865715691098,
-      "grad_norm": 1.4380339902364112,
-      "learning_rate": 5.4e-06,
-      "loss": 1.3331,
-      "step": 27
-    },
-    {
-      "epoch": 0.013772749631087064,
-      "grad_norm": 1.0684373704524008,
-      "learning_rate": 5.600000000000001e-06,
-      "loss": 1.2566,
-      "step": 28
-    },
-    {
-      "epoch": 0.01426463354648303,
-      "grad_norm": 1.9974303492749381,
-      "learning_rate": 5.8e-06,
-      "loss": 1.392,
-      "step": 29
-    },
-    {
-      "epoch": 0.014756517461878997,
-      "grad_norm": 1.0635238293595115,
-      "learning_rate": 6e-06,
-      "loss": 1.2207,
-      "step": 30
-    },
-    {
-      "epoch": 0.015248401377274963,
-      "grad_norm": 1.144610170563909,
-      "learning_rate": 6.2e-06,
-      "loss": 1.1776,
-      "step": 31
-    },
-    {
-      "epoch": 0.01574028529267093,
-      "grad_norm": 1.0215236404064907,
-      "learning_rate": 6.4000000000000006e-06,
-      "loss": 1.2454,
-      "step": 32
-    },
-    {
-      "epoch": 0.016232169208066895,
-      "grad_norm": 1.0978098723676712,
-      "learning_rate": 6.6e-06,
-      "loss": 1.3075,
-      "step": 33
-    },
-    {
-      "epoch": 0.016724053123462864,
-      "grad_norm": 1.011060190230393,
-      "learning_rate": 6.800000000000001e-06,
-      "loss": 1.3027,
-      "step": 34
-    },
-    {
-      "epoch": 0.01721593703885883,
-      "grad_norm": 0.8804453134798843,
-      "learning_rate": 7.000000000000001e-06,
-      "loss": 1.2454,
-      "step": 35
-    },
-    {
-      "epoch": 0.017707820954254796,
-      "grad_norm": 1.0044730642582949,
-      "learning_rate": 7.2e-06,
-      "loss": 1.1784,
-      "step": 36
-    },
-    {
-      "epoch": 0.01819970486965076,
-      "grad_norm": 1.0003621918042445,
-      "learning_rate": 7.4e-06,
-      "loss": 1.1357,
-      "step": 37
-    },
-    {
-      "epoch": 0.018691588785046728,
-      "grad_norm": 0.902655380789519,
-      "learning_rate": 7.6e-06,
-      "loss": 1.2277,
-      "step": 38
-    },
-    {
-      "epoch": 0.019183472700442697,
-      "grad_norm": 0.8620586528524807,
-      "learning_rate": 7.8e-06,
-      "loss": 1.1915,
-      "step": 39
-    },
-    {
-      "epoch": 0.019675356615838663,
-      "grad_norm": 0.8137353744254079,
-      "learning_rate": 8.000000000000001e-06,
-      "loss": 1.2105,
-      "step": 40
-    },
-    {
-      "epoch": 0.02016724053123463,
-      "grad_norm": 0.8498537846473997,
-      "learning_rate": 8.200000000000001e-06,
-      "loss": 1.2529,
-      "step": 41
-    },
-    {
-      "epoch": 0.020659124446630595,
-      "grad_norm": 0.7727069062986878,
-      "learning_rate": 8.400000000000001e-06,
-      "loss": 1.1524,
-      "step": 42
-    },
-    {
-      "epoch": 0.02115100836202656,
-      "grad_norm": 0.7866519624230008,
-      "learning_rate": 8.599999999999999e-06,
-      "loss": 1.1283,
-      "step": 43
-    },
-    {
-      "epoch": 0.02164289227742253,
-      "grad_norm": 0.8400864244937326,
-      "learning_rate": 8.8e-06,
-      "loss": 1.1706,
-      "step": 44
-    },
-    {
-      "epoch": 0.022134776192818496,
-      "grad_norm": 0.7547645574164925,
-      "learning_rate": 9e-06,
-      "loss": 1.1232,
-      "step": 45
-    },
-    {
-      "epoch": 0.02262666010821446,
-      "grad_norm": 0.8158199982108677,
-      "learning_rate": 9.2e-06,
-      "loss": 1.1204,
-      "step": 46
-    },
-    {
-      "epoch": 0.023118544023610427,
-      "grad_norm": 1.3611423960474989,
-      "learning_rate": 9.4e-06,
-      "loss": 1.2444,
-      "step": 47
-    },
-    {
-      "epoch": 0.023610427939006393,
-      "grad_norm": 2.665278252165078,
-      "learning_rate": 9.600000000000001e-06,
-      "loss": 1.245,
-      "step": 48
-    },
-    {
-      "epoch": 0.024102311854402363,
-      "grad_norm": 0.7473276580820656,
-      "learning_rate": 9.800000000000001e-06,
-      "loss": 1.1831,
-      "step": 49
-    },
-    {
-      "epoch": 0.02459419576979833,
-      "grad_norm": 0.867805055812748,
-      "learning_rate": 1e-05,
-      "loss": 1.1472,
-      "step": 50
-    },
-    {
-      "epoch": 0.025086079685194294,
-      "grad_norm": 0.7651704029177023,
-      "learning_rate": 1.02e-05,
-      "loss": 1.1587,
-      "step": 51
-    },
-    {
-      "epoch": 0.02557796360059026,
-      "grad_norm": 0.8363948308926066,
-      "learning_rate": 1.04e-05,
-      "loss": 1.0964,
-      "step": 52
-    },
-    {
-      "epoch": 0.026069847515986226,
-      "grad_norm": 0.7505930186457537,
-      "learning_rate": 1.06e-05,
-      "loss": 1.078,
-      "step": 53
-    },
-    {
-      "epoch": 0.026561731431382195,
-      "grad_norm": 0.8102793549862889,
-      "learning_rate": 1.08e-05,
-      "loss": 1.2748,
-      "step": 54
-    },
-    {
-      "epoch": 0.02705361534677816,
-      "grad_norm": 0.7249072559891252,
-      "learning_rate": 1.1000000000000001e-05,
-      "loss": 1.0854,
-      "step": 55
-    },
-    {
-      "epoch": 0.027545499262174127,
-      "grad_norm": 3.1671466786532454,
-      "learning_rate": 1.1200000000000001e-05,
-      "loss": 1.124,
-      "step": 56
-    },
-    {
-      "epoch": 0.028037383177570093,
-      "grad_norm": 0.824497189137833,
-      "learning_rate": 1.1400000000000001e-05,
-      "loss": 1.1202,
-      "step": 57
-    },
-    {
-      "epoch": 0.02852926709296606,
-      "grad_norm": 0.7272267118712353,
-      "learning_rate": 1.16e-05,
-      "loss": 1.1422,
-      "step": 58
-    },
-    {
-      "epoch": 0.029021151008362025,
-      "grad_norm": 0.6765223066810455,
-      "learning_rate": 1.18e-05,
-      "loss": 1.2085,
-      "step": 59
-    },
-    {
-      "epoch": 0.029513034923757994,
-      "grad_norm": 0.8119323091496039,
-      "learning_rate": 1.2e-05,
-      "loss": 1.1647,
-      "step": 60
-    },
-    {
-      "epoch": 0.03000491883915396,
-      "grad_norm": 0.808311334107385,
-      "learning_rate": 1.22e-05,
-      "loss": 1.1931,
-      "step": 61
-    },
-    {
-      "epoch": 0.030496802754549926,
-      "grad_norm": 0.7333493427730208,
-      "learning_rate": 1.24e-05,
-      "loss": 1.1925,
-      "step": 62
-    },
-    {
-      "epoch": 0.030988686669945892,
-      "grad_norm": 0.6744329214605398,
-      "learning_rate": 1.2600000000000001e-05,
-      "loss": 1.0442,
-      "step": 63
-    },
-    {
-      "epoch": 0.03148057058534186,
-      "grad_norm": 0.6926905933941976,
-      "learning_rate": 1.2800000000000001e-05,
-      "loss": 1.0809,
-      "step": 64
-    },
-    {
-      "epoch": 0.031972454500737824,
-      "grad_norm": 1.3668458130847467,
-      "learning_rate": 1.3000000000000001e-05,
-      "loss": 1.0877,
-      "step": 65
-    },
-    {
-      "epoch": 0.03246433841613379,
-      "grad_norm": 0.8289258812132557,
-      "learning_rate": 1.32e-05,
-      "loss": 1.1428,
-      "step": 66
-    },
-    {
-      "epoch": 0.03295622233152976,
-      "grad_norm": 0.6980913641316498,
-      "learning_rate": 1.3400000000000002e-05,
-      "loss": 1.0939,
-      "step": 67
-    },
-    {
-      "epoch": 0.03344810624692573,
-      "grad_norm": 0.6979371882691208,
-      "learning_rate": 1.3600000000000002e-05,
-      "loss": 1.0171,
-      "step": 68
-    },
-    {
-      "epoch": 0.033939990162321694,
-      "grad_norm": 0.7772969332213725,
-      "learning_rate": 1.3800000000000002e-05,
-      "loss": 1.193,
-      "step": 69
-    },
-    {
-      "epoch": 0.03443187407771766,
-      "grad_norm": 0.7028342002384361,
-      "learning_rate": 1.4000000000000001e-05,
-      "loss": 1.1459,
-      "step": 70
-    },
-    {
-      "epoch": 0.034923757993113626,
-      "grad_norm": 0.6968410058706372,
-      "learning_rate": 1.42e-05,
-      "loss": 1.0484,
-      "step": 71
-    },
-    {
-      "epoch": 0.03541564190850959,
-      "grad_norm": 0.695112229906698,
-      "learning_rate": 1.44e-05,
-      "loss": 1.0834,
-      "step": 72
-    },
-    {
-      "epoch": 0.03590752582390556,
-      "grad_norm": 0.6971728236831181,
-      "learning_rate": 1.4599999999999999e-05,
-      "loss": 0.9687,
-      "step": 73
-    },
-    {
-      "epoch": 0.03639940973930152,
-      "grad_norm": 0.7529709591379297,
-      "learning_rate": 1.48e-05,
-      "loss": 1.0345,
-      "step": 74
-    },
-    {
-      "epoch": 0.03689129365469749,
-      "grad_norm": 0.7219911603689171,
-      "learning_rate": 1.5e-05,
-      "loss": 1.127,
-      "step": 75
-    },
-    {
-      "epoch": 0.037383177570093455,
-      "grad_norm": 2.3000076928278923,
-      "learning_rate": 1.52e-05,
-      "loss": 1.1482,
-      "step": 76
-    },
-    {
-      "epoch": 0.03787506148548942,
-      "grad_norm": 0.7809471483200394,
-      "learning_rate": 1.54e-05,
-      "loss": 1.1334,
-      "step": 77
-    },
-    {
-      "epoch": 0.038366945400885394,
-      "grad_norm": 0.6682890681656087,
-      "learning_rate": 1.56e-05,
-      "loss": 1.0889,
-      "step": 78
-    },
-    {
-      "epoch": 0.03885882931628136,
-      "grad_norm": 0.6544825794684948,
-      "learning_rate": 1.58e-05,
-      "loss": 1.0849,
-      "step": 79
-    },
-    {
-      "epoch": 0.039350713231677326,
-      "grad_norm": 0.7193471748761666,
-      "learning_rate": 1.6000000000000003e-05,
-      "loss": 1.0591,
-      "step": 80
-    },
-    {
-      "epoch": 0.03984259714707329,
-      "grad_norm": 0.7047661426725917,
-      "learning_rate": 1.62e-05,
-      "loss": 1.1813,
-      "step": 81
-    },
-    {
-      "epoch": 0.04033448106246926,
-      "grad_norm": 0.7908701188113396,
-      "learning_rate": 1.6400000000000002e-05,
-      "loss": 1.1497,
-      "step": 82
-    },
-    {
-      "epoch": 0.04082636497786522,
-      "grad_norm": 0.7708018644658138,
-      "learning_rate": 1.66e-05,
-      "loss": 1.1542,
-      "step": 83
-    },
-    {
-      "epoch": 0.04131824889326119,
-      "grad_norm": 0.7903847098228288,
-      "learning_rate": 1.6800000000000002e-05,
-      "loss": 1.0924,
-      "step": 84
-    },
-    {
-      "epoch": 0.041810132808657155,
-      "grad_norm": 0.7175005138594569,
-      "learning_rate": 1.7000000000000003e-05,
-      "loss": 1.0804,
-      "step": 85
-    },
-    {
-      "epoch": 0.04230201672405312,
-      "grad_norm": 0.6586142525496985,
-      "learning_rate": 1.7199999999999998e-05,
-      "loss": 1.1146,
-      "step": 86
-    },
-    {
-      "epoch": 0.04279390063944909,
-      "grad_norm": 0.653452996418673,
-      "learning_rate": 1.74e-05,
-      "loss": 1.0572,
-      "step": 87
-    },
-    {
-      "epoch": 0.04328578455484506,
-      "grad_norm": 0.6977335177281804,
-      "learning_rate": 1.76e-05,
-      "loss": 1.0954,
-      "step": 88
-    },
-    {
-      "epoch": 0.043777668470241025,
-      "grad_norm": 0.7965568824498416,
-      "learning_rate": 1.78e-05,
-      "loss": 1.1306,
-      "step": 89
-    },
-    {
-      "epoch": 0.04426955238563699,
-      "grad_norm": 0.7702049866863492,
-      "learning_rate": 1.8e-05,
-      "loss": 1.0624,
-      "step": 90
-    },
-    {
-      "epoch": 0.04476143630103296,
-      "grad_norm": 0.833380055984766,
-      "learning_rate": 1.8200000000000002e-05,
-      "loss": 1.1674,
-      "step": 91
-    },
-    {
-      "epoch": 0.04525332021642892,
-      "grad_norm": 0.7292333702665459,
-      "learning_rate": 1.84e-05,
-      "loss": 1.119,
-      "step": 92
-    },
-    {
-      "epoch": 0.04574520413182489,
-      "grad_norm": 1.085524626858829,
-      "learning_rate": 1.86e-05,
-      "loss": 1.1564,
-      "step": 93
-    },
-    {
-      "epoch": 0.046237088047220855,
-      "grad_norm": 0.7419165242760628,
-      "learning_rate": 1.88e-05,
-      "loss": 1.0865,
-      "step": 94
-    },
-    {
-      "epoch": 0.04672897196261682,
-      "grad_norm": 0.7521136841453322,
-      "learning_rate": 1.9e-05,
-      "loss": 1.0778,
-      "step": 95
-    },
-    {
-      "epoch": 0.04722085587801279,
-      "grad_norm": 0.8250024974505459,
-      "learning_rate": 1.9200000000000003e-05,
-      "loss": 1.0457,
-      "step": 96
-    },
-    {
-      "epoch": 0.04771273979340875,
-      "grad_norm": 0.6965644981293239,
-      "learning_rate": 1.94e-05,
-      "loss": 1.1203,
-      "step": 97
-    },
-    {
-      "epoch": 0.048204623708804725,
-      "grad_norm": 0.7725116537341091,
-      "learning_rate": 1.9600000000000002e-05,
-      "loss": 1.1314,
-      "step": 98
-    },
-    {
-      "epoch": 0.04869650762420069,
-      "grad_norm": 0.8954655800907969,
-      "learning_rate": 1.9800000000000004e-05,
-      "loss": 1.1697,
-      "step": 99
-    },
-    {
-      "epoch": 0.04918839153959666,
-      "grad_norm": 0.7099715953824428,
-      "learning_rate": 2e-05,
-      "loss": 1.0653,
-      "step": 100
-    },
-    {
-      "epoch": 0.04968027545499262,
-      "grad_norm": 0.6979198082373752,
-      "learning_rate": 2.0200000000000003e-05,
-      "loss": 1.0092,
-      "step": 101
-    },
-    {
-      "epoch": 0.05017215937038859,
-      "grad_norm": 0.7542658397206478,
-      "learning_rate": 2.04e-05,
-      "loss": 0.9632,
-      "step": 102
-    },
-    {
-      "epoch": 0.050664043285784555,
-      "grad_norm": 0.6828660555164732,
-      "learning_rate": 2.06e-05,
-      "loss": 1.0614,
-      "step": 103
-    },
-    {
-      "epoch": 0.05115592720118052,
-      "grad_norm": 0.6721539069619723,
-      "learning_rate": 2.08e-05,
-      "loss": 1.0285,
-      "step": 104
-    },
-    {
-      "epoch": 0.051647811116576486,
-      "grad_norm": 0.7159114501125087,
-      "learning_rate": 2.1e-05,
-      "loss": 1.0312,
-      "step": 105
-    },
-    {
-      "epoch": 0.05213969503197245,
-      "grad_norm": 0.6545850965584485,
-      "learning_rate": 2.12e-05,
-      "loss": 1.0698,
-      "step": 106
-    },
-    {
-      "epoch": 0.05263157894736842,
-      "grad_norm": 0.658675131654431,
-      "learning_rate": 2.1400000000000002e-05,
-      "loss": 1.1183,
-      "step": 107
-    },
-    {
-      "epoch": 0.05312346286276439,
-      "grad_norm": 0.6765638401749785,
-      "learning_rate": 2.16e-05,
-      "loss": 1.0923,
-      "step": 108
-    },
-    {
-      "epoch": 0.05361534677816036,
-      "grad_norm": 0.6774305162394292,
-      "learning_rate": 2.18e-05,
-      "loss": 1.0696,
-      "step": 109
-    },
-    {
-      "epoch": 0.05410723069355632,
-      "grad_norm": 0.6941450033816315,
-      "learning_rate": 2.2000000000000003e-05,
-      "loss": 1.0278,
-      "step": 110
-    },
-    {
-      "epoch": 0.05459911460895229,
-      "grad_norm": 0.7298702161046116,
-      "learning_rate": 2.22e-05,
-      "loss": 0.9942,
-      "step": 111
-    },
-    {
-      "epoch": 0.055090998524348254,
-      "grad_norm": 0.78359048411605,
-      "learning_rate": 2.2400000000000002e-05,
-      "loss": 1.0722,
-      "step": 112
-    },
-    {
-      "epoch": 0.05558288243974422,
-      "grad_norm": 0.7234222510303335,
-      "learning_rate": 2.26e-05,
-      "loss": 1.0851,
-      "step": 113
-    },
-    {
-      "epoch": 0.056074766355140186,
-      "grad_norm": 0.7517380728841061,
-      "learning_rate": 2.2800000000000002e-05,
-      "loss": 0.9929,
-      "step": 114
-    },
-    {
-      "epoch": 0.05656665027053615,
-      "grad_norm": 0.688888491301893,
-      "learning_rate": 2.3000000000000003e-05,
-      "loss": 1.0505,
-      "step": 115
-    },
-    {
-      "epoch": 0.05705853418593212,
-      "grad_norm": 0.815491863848212,
-      "learning_rate": 2.32e-05,
-      "loss": 1.1339,
-      "step": 116
-    },
-    {
-      "epoch": 0.057550418101328084,
-      "grad_norm": 0.75531131045591,
-      "learning_rate": 2.3400000000000003e-05,
-      "loss": 1.0716,
-      "step": 117
-    },
-    {
-      "epoch": 0.05804230201672405,
-      "grad_norm": 0.7625830133131772,
-      "learning_rate": 2.36e-05,
-      "loss": 1.0524,
-      "step": 118
-    },
-    {
-      "epoch": 0.05853418593212002,
-      "grad_norm": 0.7566379286598796,
-      "learning_rate": 2.38e-05,
-      "loss": 1.0334,
-      "step": 119
-    },
-    {
-      "epoch": 0.05902606984751599,
-      "grad_norm": 0.7028004568598516,
-      "learning_rate": 2.4e-05,
-      "loss": 1.0436,
-      "step": 120
-    },
-    {
-      "epoch": 0.059517953762911954,
-      "grad_norm": 0.7070567970425724,
-      "learning_rate": 2.4200000000000002e-05,
-      "loss": 1.0651,
-      "step": 121
-    },
-    {
-      "epoch": 0.06000983767830792,
-      "grad_norm": 0.7098752497716972,
-      "learning_rate": 2.44e-05,
-      "loss": 1.0092,
-      "step": 122
-    },
-    {
-      "epoch": 0.060501721593703886,
-      "grad_norm": 0.6923985079217133,
-      "learning_rate": 2.46e-05,
-      "loss": 0.9777,
-      "step": 123
-    },
-    {
-      "epoch": 0.06099360550909985,
-      "grad_norm": 0.736397162563532,
-      "learning_rate": 2.48e-05,
-      "loss": 1.0494,
-      "step": 124
-    },
-    {
-      "epoch": 0.06148548942449582,
-      "grad_norm": 0.6717044407705316,
-      "learning_rate": 2.5e-05,
-      "loss": 1.0805,
-      "step": 125
-    },
-    {
-      "epoch": 0.061977373339891784,
-      "grad_norm": 0.6918045884854129,
-      "learning_rate": 2.5200000000000003e-05,
-      "loss": 1.0736,
-      "step": 126
-    },
-    {
-      "epoch": 0.06246925725528775,
-      "grad_norm": 0.7451366287208874,
-      "learning_rate": 2.54e-05,
-      "loss": 1.1339,
-      "step": 127
-    },
-    {
-      "epoch": 0.06296114117068372,
-      "grad_norm": 0.7280019983557976,
-      "learning_rate": 2.5600000000000002e-05,
-      "loss": 1.0258,
-      "step": 128
-    },
-    {
-      "epoch": 0.06345302508607968,
-      "grad_norm": 0.8510806423225098,
-      "learning_rate": 2.58e-05,
-      "loss": 0.9742,
-      "step": 129
-    },
-    {
-      "epoch": 0.06394490900147565,
-      "grad_norm": 0.6722285390124173,
-      "learning_rate": 2.6000000000000002e-05,
-      "loss": 1.091,
-      "step": 130
-    },
-    {
-      "epoch": 0.06443679291687161,
-      "grad_norm": 0.7882995585151922,
-      "learning_rate": 2.6200000000000003e-05,
-      "loss": 1.1108,
-      "step": 131
-    },
-    {
-      "epoch": 0.06492867683226758,
-      "grad_norm": 0.7124573821253081,
-      "learning_rate": 2.64e-05,
-      "loss": 1.094,
-      "step": 132
-    },
-    {
-      "epoch": 0.06542056074766354,
-      "grad_norm": 0.7173359737653685,
-      "learning_rate": 2.6600000000000003e-05,
-      "loss": 0.9962,
-      "step": 133
-    },
-    {
-      "epoch": 0.06591244466305952,
-      "grad_norm": 0.7834358909342277,
-      "learning_rate": 2.6800000000000004e-05,
-      "loss": 1.1099,
-      "step": 134
-    },
-    {
-      "epoch": 0.06640432857845549,
-      "grad_norm": 0.7008793922142209,
-      "learning_rate": 2.7000000000000002e-05,
-      "loss": 1.0077,
-      "step": 135
-    },
-    {
-      "epoch": 0.06689621249385146,
-      "grad_norm": 0.7634356340968927,
-      "learning_rate": 2.7200000000000004e-05,
-      "loss": 1.1082,
-      "step": 136
-    },
-    {
-      "epoch": 0.06738809640924742,
-      "grad_norm": 0.7378999561528924,
-      "learning_rate": 2.7400000000000002e-05,
-      "loss": 1.1178,
-      "step": 137
-    },
-    {
-      "epoch": 0.06787998032464339,
-      "grad_norm": 0.700386136828009,
-      "learning_rate": 2.7600000000000003e-05,
-      "loss": 1.1726,
-      "step": 138
-    },
-    {
-      "epoch": 0.06837186424003935,
-      "grad_norm": 0.68094249806194,
-      "learning_rate": 2.7800000000000005e-05,
-      "loss": 1.0505,
-      "step": 139
-    },
-    {
-      "epoch": 0.06886374815543532,
-      "grad_norm": 0.6844926094012175,
-      "learning_rate": 2.8000000000000003e-05,
-      "loss": 1.0492,
-      "step": 140
-    },
-    {
-      "epoch": 0.06935563207083129,
-      "grad_norm": 0.6903824887607274,
-      "learning_rate": 2.8199999999999998e-05,
-      "loss": 1.0364,
-      "step": 141
-    },
-    {
-      "epoch": 0.06984751598622725,
-      "grad_norm": 0.7116078298920068,
-      "learning_rate": 2.84e-05,
-      "loss": 1.1036,
-      "step": 142
-    },
-    {
-      "epoch": 0.07033939990162322,
-      "grad_norm": 0.6756583405629968,
-      "learning_rate": 2.86e-05,
-      "loss": 0.9928,
-      "step": 143
-    },
-    {
-      "epoch": 0.07083128381701918,
-      "grad_norm": 0.7394965661508411,
-      "learning_rate": 2.88e-05,
-      "loss": 1.102,
-      "step": 144
-    },
-    {
-      "epoch": 0.07132316773241515,
-      "grad_norm": 0.7541310822808502,
-      "learning_rate": 2.9e-05,
-      "loss": 1.0452,
-      "step": 145
-    },
-    {
-      "epoch": 0.07181505164781112,
-      "grad_norm": 0.7239753447013636,
-      "learning_rate": 2.9199999999999998e-05,
-      "loss": 1.0378,
-      "step": 146
-    },
-    {
-      "epoch": 0.07230693556320708,
-      "grad_norm": 0.6864664784282436,
-      "learning_rate": 2.94e-05,
-      "loss": 1.0373,
-      "step": 147
-    },
-    {
-      "epoch": 0.07279881947860305,
-      "grad_norm": 0.6457006672999801,
-      "learning_rate": 2.96e-05,
-      "loss": 1.0271,
-      "step": 148
-    },
-    {
-      "epoch": 0.07329070339399901,
-      "grad_norm": 0.7029187156226748,
-      "learning_rate": 2.98e-05,
-      "loss": 1.1029,
-      "step": 149
-    },
-    {
-      "epoch": 0.07378258730939498,
-      "grad_norm": 0.6923306808409229,
-      "learning_rate": 3e-05,
-      "loss": 1.0088,
-      "step": 150
-    },
-    {
-      "epoch": 0.07427447122479094,
-      "grad_norm": 0.6863830617495799,
-      "learning_rate": 3.02e-05,
-      "loss": 1.0504,
-      "step": 151
-    },
-    {
-      "epoch": 0.07476635514018691,
-      "grad_norm": 0.6593531065686705,
-      "learning_rate": 3.04e-05,
-      "loss": 1.0617,
-      "step": 152
-    },
-    {
-      "epoch": 0.07525823905558288,
-      "grad_norm": 0.7881517346051917,
-      "learning_rate": 3.06e-05,
-      "loss": 1.0784,
-      "step": 153
-    },
-    {
-      "epoch": 0.07575012297097884,
-      "grad_norm": 0.6915901349067198,
-      "learning_rate": 3.08e-05,
-      "loss": 1.0124,
-      "step": 154
-    },
-    {
-      "epoch": 0.07624200688637482,
-      "grad_norm": 0.7266366648143068,
-      "learning_rate": 3.1e-05,
-      "loss": 1.1224,
-      "step": 155
-    },
-    {
-      "epoch": 0.07673389080177079,
-      "grad_norm": 0.7162444071348569,
-      "learning_rate": 3.12e-05,
-      "loss": 1.149,
-      "step": 156
-    },
-    {
-      "epoch": 0.07722577471716675,
-      "grad_norm": 0.7237776602931134,
-      "learning_rate": 3.1400000000000004e-05,
-      "loss": 1.0799,
-      "step": 157
-    },
-    {
-      "epoch": 0.07771765863256272,
-      "grad_norm": 0.8258091547112008,
-      "learning_rate": 3.16e-05,
-      "loss": 1.0523,
-      "step": 158
-    },
-    {
-      "epoch": 0.07820954254795869,
-      "grad_norm": 0.6852296673700097,
-      "learning_rate": 3.18e-05,
-      "loss": 1.0213,
-      "step": 159
-    },
-    {
-      "epoch": 0.07870142646335465,
-      "grad_norm": 0.6350465704546835,
-      "learning_rate": 3.2000000000000005e-05,
-      "loss": 0.9624,
-      "step": 160
-    },
-    {
-      "epoch": 0.07919331037875062,
-      "grad_norm": 1.4962584316561658,
-      "learning_rate": 3.2200000000000003e-05,
-      "loss": 1.0782,
-      "step": 161
-    },
-    {
-      "epoch": 0.07968519429414658,
-      "grad_norm": 0.7353035875932009,
-      "learning_rate": 3.24e-05,
-      "loss": 0.9777,
-      "step": 162
-    },
-    {
-      "epoch": 0.08017707820954255,
-      "grad_norm": 0.8083481878994074,
-      "learning_rate": 3.26e-05,
-      "loss": 1.0283,
-      "step": 163
-    },
-    {
-      "epoch": 0.08066896212493851,
-      "grad_norm": 0.7376876744622659,
-      "learning_rate": 3.2800000000000004e-05,
-      "loss": 1.0049,
-      "step": 164
-    },
-    {
-      "epoch": 0.08116084604033448,
-      "grad_norm": 0.7684385423867578,
-      "learning_rate": 3.3e-05,
-      "loss": 1.0919,
-      "step": 165
-    },
-    {
-      "epoch": 0.08165272995573045,
-      "grad_norm": 0.6828848133442723,
-      "learning_rate": 3.32e-05,
-      "loss": 1.0262,
-      "step": 166
-    },
-    {
-      "epoch": 0.08214461387112641,
-      "grad_norm": 0.6435240309407694,
-      "learning_rate": 3.3400000000000005e-05,
-      "loss": 1.0324,
-      "step": 167
-    },
-    {
-      "epoch": 0.08263649778652238,
-      "grad_norm": 0.720314282911722,
-      "learning_rate": 3.3600000000000004e-05,
-      "loss": 1.0566,
-      "step": 168
-    },
-    {
-      "epoch": 0.08312838170191834,
-      "grad_norm": 0.7727153801819339,
-      "learning_rate": 3.38e-05,
-      "loss": 0.9888,
-      "step": 169
-    },
-    {
-      "epoch": 0.08362026561731431,
-      "grad_norm": 0.6505013326914232,
-      "learning_rate": 3.4000000000000007e-05,
-      "loss": 1.0101,
-      "step": 170
-    },
-    {
-      "epoch": 0.08411214953271028,
-      "grad_norm": 0.6728550885628142,
-      "learning_rate": 3.4200000000000005e-05,
-      "loss": 1.0934,
-      "step": 171
-    },
-    {
-      "epoch": 0.08460403344810624,
-      "grad_norm": 0.6665239884376515,
-      "learning_rate": 3.4399999999999996e-05,
-      "loss": 1.0947,
-      "step": 172
-    },
-    {
-      "epoch": 0.08509591736350221,
-      "grad_norm": 0.6959342104513762,
-      "learning_rate": 3.46e-05,
-      "loss": 1.0673,
-      "step": 173
-    },
-    {
-      "epoch": 0.08558780127889817,
-      "grad_norm": 0.7429585913136677,
-      "learning_rate": 3.48e-05,
-      "loss": 0.9795,
-      "step": 174
-    },
-    {
-      "epoch": 0.08607968519429415,
-      "grad_norm": 0.6739954355198908,
-      "learning_rate": 3.5e-05,
-      "loss": 1.0087,
-      "step": 175
-    },
-    {
-      "epoch": 0.08657156910969012,
-      "grad_norm": 0.7196630795576944,
-      "learning_rate": 3.52e-05,
-      "loss": 1.042,
-      "step": 176
-    },
-    {
-      "epoch": 0.08706345302508608,
-      "grad_norm": 0.674245145598711,
-      "learning_rate": 3.54e-05,
-      "loss": 0.9691,
-      "step": 177
-    },
-    {
-      "epoch": 0.08755533694048205,
-      "grad_norm": 0.6237692105197545,
-      "learning_rate": 3.56e-05,
-      "loss": 1.0078,
-      "step": 178
-    },
-    {
-      "epoch": 0.08804722085587802,
-      "grad_norm": 0.7338926542494251,
-      "learning_rate": 3.58e-05,
-      "loss": 1.0325,
-      "step": 179
-    },
-    {
-      "epoch": 0.08853910477127398,
-      "grad_norm": 0.6797934842299513,
-      "learning_rate": 3.6e-05,
-      "loss": 1.0204,
-      "step": 180
-    },
-    {
-      "epoch": 0.08903098868666995,
-      "grad_norm": 0.7299173452740791,
-      "learning_rate": 3.62e-05,
-      "loss": 1.047,
-      "step": 181
-    },
-    {
-      "epoch": 0.08952287260206591,
-      "grad_norm": 0.7045963197748994,
-      "learning_rate": 3.6400000000000004e-05,
-      "loss": 1.0916,
-      "step": 182
-    },
-    {
-      "epoch": 0.09001475651746188,
-      "grad_norm": 0.7436237348213588,
-      "learning_rate": 3.66e-05,
-      "loss": 1.0694,
-      "step": 183
-    },
-    {
-      "epoch": 0.09050664043285785,
-      "grad_norm": 0.7114595640434511,
-      "learning_rate": 3.68e-05,
-      "loss": 1.0079,
-      "step": 184
-    },
-    {
-      "epoch": 0.09099852434825381,
-      "grad_norm": 0.6846091209296655,
-      "learning_rate": 3.7e-05,
-      "loss": 0.9537,
-      "step": 185
-    },
-    {
-      "epoch": 0.09149040826364978,
-      "grad_norm": 0.685900598420304,
-      "learning_rate": 3.72e-05,
-      "loss": 1.0499,
-      "step": 186
-    },
-    {
-      "epoch": 0.09198229217904574,
-      "grad_norm": 0.6801716444252568,
-      "learning_rate": 3.74e-05,
-      "loss": 1.0011,
-      "step": 187
-    },
-    {
-      "epoch": 0.09247417609444171,
-      "grad_norm": 0.6959834732767167,
-      "learning_rate": 3.76e-05,
-      "loss": 1.0543,
-      "step": 188
-    },
-    {
-      "epoch": 0.09296606000983768,
-      "grad_norm": 0.6526864408010312,
-      "learning_rate": 3.7800000000000004e-05,
-      "loss": 1.0366,
-      "step": 189
-    },
-    {
-      "epoch": 0.09345794392523364,
-      "grad_norm": 0.6982655777808178,
-      "learning_rate": 3.8e-05,
-      "loss": 1.0264,
-      "step": 190
-    },
-    {
-      "epoch": 0.09394982784062961,
-      "grad_norm": 0.6967368126765116,
-      "learning_rate": 3.82e-05,
-      "loss": 1.005,
-      "step": 191
-    },
-    {
-      "epoch": 0.09444171175602557,
-      "grad_norm": 0.6571475095270259,
-      "learning_rate": 3.8400000000000005e-05,
-      "loss": 0.9507,
-      "step": 192
-    },
-    {
-      "epoch": 0.09493359567142154,
-      "grad_norm": 0.7893579623123786,
-      "learning_rate": 3.86e-05,
-      "loss": 1.0194,
-      "step": 193
-    },
-    {
-      "epoch": 0.0954254795868175,
-      "grad_norm": 0.6747872226437224,
-      "learning_rate": 3.88e-05,
-      "loss": 0.9702,
-      "step": 194
-    },
-    {
-      "epoch": 0.09591736350221347,
-      "grad_norm": 0.6951769070722532,
-      "learning_rate": 3.9000000000000006e-05,
-      "loss": 1.0154,
-      "step": 195
-    },
-    {
-      "epoch": 0.09640924741760945,
-      "grad_norm": 0.6729600273811861,
-      "learning_rate": 3.9200000000000004e-05,
-      "loss": 0.9737,
-      "step": 196
-    },
-    {
-      "epoch": 0.09690113133300542,
-      "grad_norm": 0.646289442356639,
-      "learning_rate": 3.94e-05,
-      "loss": 0.9964,
-      "step": 197
-    },
-    {
-      "epoch": 0.09739301524840138,
-      "grad_norm": 0.7111022014767776,
-      "learning_rate": 3.960000000000001e-05,
-      "loss": 0.9115,
-      "step": 198
-    },
-    {
-      "epoch": 0.09788489916379735,
-      "grad_norm": 0.7852036200596197,
-      "learning_rate": 3.9800000000000005e-05,
-      "loss": 1.05,
-      "step": 199
-    },
-    {
-      "epoch": 0.09837678307919331,
-      "grad_norm": 0.7088405268610213,
-      "learning_rate": 4e-05,
-      "loss": 1.0435,
-      "step": 200
-    },
-    {
-      "epoch": 0.09886866699458928,
-      "grad_norm": 0.7777103623163572,
-      "learning_rate": 4.02e-05,
-      "loss": 1.0491,
-      "step": 201
-    },
-    {
-      "epoch": 0.09936055090998525,
-      "grad_norm": 0.6579744250873005,
-      "learning_rate": 4.0400000000000006e-05,
-      "loss": 1.0442,
-      "step": 202
-    },
-    {
-      "epoch": 0.09985243482538121,
-      "grad_norm": 0.6543156891174083,
-      "learning_rate": 4.0600000000000004e-05,
-      "loss": 0.975,
-      "step": 203
-    },
-    {
-      "epoch": 0.10034431874077718,
-      "grad_norm": 0.7020232902689696,
-      "learning_rate": 4.08e-05,
-      "loss": 1.0331,
-      "step": 204
-    },
-    {
-      "epoch": 0.10083620265617314,
-      "grad_norm": 0.6556250690855042,
-      "learning_rate": 4.1e-05,
-      "loss": 0.9965,
-      "step": 205
-    },
-    {
-      "epoch": 0.10132808657156911,
-      "grad_norm": 0.6496179732912809,
-      "learning_rate": 4.12e-05,
-      "loss": 0.9488,
-      "step": 206
-    },
-    {
-      "epoch": 0.10181997048696508,
-      "grad_norm": 0.6590304062714621,
-      "learning_rate": 4.14e-05,
-      "loss": 1.01,
-      "step": 207
-    },
-    {
-      "epoch": 0.10231185440236104,
-      "grad_norm": 0.6984098889798658,
-      "learning_rate": 4.16e-05,
-      "loss": 1.0673,
-      "step": 208
-    },
-    {
-      "epoch": 0.102803738317757,
-      "grad_norm": 0.6751148224917966,
-      "learning_rate": 4.18e-05,
-      "loss": 1.0013,
-      "step": 209
-    },
-    {
-      "epoch": 0.10329562223315297,
-      "grad_norm": 0.6661081975498573,
-      "learning_rate": 4.2e-05,
-      "loss": 1.0373,
-      "step": 210
-    },
-    {
-      "epoch": 0.10378750614854894,
-      "grad_norm": 0.7460179270770505,
-      "learning_rate": 4.22e-05,
-      "loss": 1.0583,
-      "step": 211
-    },
-    {
-      "epoch": 0.1042793900639449,
-      "grad_norm": 0.6823917332904713,
-      "learning_rate": 4.24e-05,
-      "loss": 0.9444,
-      "step": 212
-    },
-    {
-      "epoch": 0.10477127397934087,
-      "grad_norm": 0.6308296487006928,
-      "learning_rate": 4.26e-05,
-      "loss": 0.9662,
-      "step": 213
-    },
-    {
-      "epoch": 0.10526315789473684,
-      "grad_norm": 0.6405975064168019,
-      "learning_rate": 4.2800000000000004e-05,
-      "loss": 0.964,
-      "step": 214
-    },
-    {
-      "epoch": 0.1057550418101328,
-      "grad_norm": 0.6332265604039056,
-      "learning_rate": 4.3e-05,
-      "loss": 1.0117,
-      "step": 215
-    },
-    {
-      "epoch": 0.10624692572552878,
-      "grad_norm": 0.6234050429175113,
-      "learning_rate": 4.32e-05,
-      "loss": 0.8902,
-      "step": 216
-    },
-    {
-      "epoch": 0.10673880964092475,
-      "grad_norm": 0.6873878864046749,
-      "learning_rate": 4.3400000000000005e-05,
-      "loss": 1.0024,
-      "step": 217
-    },
-    {
-      "epoch": 0.10723069355632071,
-      "grad_norm": 0.6497392131207163,
-      "learning_rate": 4.36e-05,
-      "loss": 0.9846,
-      "step": 218
-    },
-    {
-      "epoch": 0.10772257747171668,
-      "grad_norm": 0.6821587564436689,
-      "learning_rate": 4.38e-05,
-      "loss": 0.9138,
-      "step": 219
-    },
-    {
-      "epoch": 0.10821446138711265,
-      "grad_norm": 0.7071003711939765,
-      "learning_rate": 4.4000000000000006e-05,
-      "loss": 1.0503,
-      "step": 220
-    },
-    {
-      "epoch": 0.10870634530250861,
-      "grad_norm": 0.6535807129084836,
-      "learning_rate": 4.4200000000000004e-05,
-      "loss": 1.0347,
-      "step": 221
-    },
-    {
-      "epoch": 0.10919822921790458,
-      "grad_norm": 0.6675568978536691,
-      "learning_rate": 4.44e-05,
-      "loss": 1.0242,
-      "step": 222
-    },
-    {
-      "epoch": 0.10969011313330054,
-      "grad_norm": 0.6433035642719022,
-      "learning_rate": 4.46e-05,
-      "loss": 0.9678,
-      "step": 223
-    },
-    {
-      "epoch": 0.11018199704869651,
-      "grad_norm": 0.734775071588386,
-      "learning_rate": 4.4800000000000005e-05,
-      "loss": 0.9407,
-      "step": 224
-    },
-    {
-      "epoch": 0.11067388096409247,
-      "grad_norm": 0.6463459120325686,
-      "learning_rate": 4.5e-05,
-      "loss": 0.9922,
-      "step": 225
-    },
-    {
-      "epoch": 0.11116576487948844,
-      "grad_norm": 0.699785016750569,
-      "learning_rate": 4.52e-05,
-      "loss": 0.9674,
-      "step": 226
-    },
-    {
-      "epoch": 0.1116576487948844,
-      "grad_norm": 0.6187364073010676,
-      "learning_rate": 4.5400000000000006e-05,
-      "loss": 0.9845,
-      "step": 227
-    },
-    {
-      "epoch": 0.11214953271028037,
-      "grad_norm": 0.6506296005006268,
-      "learning_rate": 4.5600000000000004e-05,
-      "loss": 1.0033,
-      "step": 228
-    },
-    {
-      "epoch": 0.11264141662567634,
-      "grad_norm": 0.6769159452757093,
-      "learning_rate": 4.58e-05,
-      "loss": 1.0231,
-      "step": 229
-    },
-    {
-      "epoch": 0.1131333005410723,
-      "grad_norm": 0.6581875106445415,
-      "learning_rate": 4.600000000000001e-05,
-      "loss": 0.9595,
-      "step": 230
-    },
-    {
-      "epoch": 0.11362518445646827,
-      "grad_norm": 0.7094660326901466,
-      "learning_rate": 4.6200000000000005e-05,
-      "loss": 0.9939,
-      "step": 231
-    },
-    {
-      "epoch": 0.11411706837186424,
-      "grad_norm": 0.6298366474174654,
-      "learning_rate": 4.64e-05,
-      "loss": 0.9612,
-      "step": 232
-    },
-    {
-      "epoch": 0.1146089522872602,
-      "grad_norm": 0.6700127579688815,
-      "learning_rate": 4.660000000000001e-05,
-      "loss": 1.0655,
-      "step": 233
-    },
-    {
-      "epoch": 0.11510083620265617,
-      "grad_norm": 0.6908235779413228,
-      "learning_rate": 4.6800000000000006e-05,
-      "loss": 1.0331,
-      "step": 234
-    },
-    {
-      "epoch": 0.11559272011805213,
-      "grad_norm": 0.7278533721042479,
-      "learning_rate": 4.7e-05,
-      "loss": 1.0268,
-      "step": 235
-    },
-    {
-      "epoch": 0.1160846040334481,
-      "grad_norm": 0.6561769300017106,
-      "learning_rate": 4.72e-05,
-      "loss": 0.9314,
-      "step": 236
-    },
-    {
-      "epoch": 0.11657648794884408,
-      "grad_norm": 0.6838436557643636,
-      "learning_rate": 4.74e-05,
-      "loss": 1.0146,
-      "step": 237
-    },
-    {
-      "epoch": 0.11706837186424005,
-      "grad_norm": 0.7077889514774452,
-      "learning_rate": 4.76e-05,
-      "loss": 1.0245,
-      "step": 238
-    },
-    {
-      "epoch": 0.11756025577963601,
-      "grad_norm": 0.7519909427913024,
-      "learning_rate": 4.78e-05,
-      "loss": 1.0359,
-      "step": 239
-    },
-    {
-      "epoch": 0.11805213969503198,
-      "grad_norm": 0.672648659379634,
-      "learning_rate": 4.8e-05,
-      "loss": 1.0815,
-      "step": 240
-    },
-    {
-      "epoch": 0.11854402361042794,
-      "grad_norm": 0.6810172690834393,
-      "learning_rate": 4.82e-05,
-      "loss": 1.0053,
-      "step": 241
-    },
-    {
-      "epoch": 0.11903590752582391,
-      "grad_norm": 0.6708663972642332,
-      "learning_rate": 4.8400000000000004e-05,
-      "loss": 1.0072,
-      "step": 242
-    },
-    {
-      "epoch": 0.11952779144121987,
-      "grad_norm": 0.7168031253779362,
-      "learning_rate": 4.86e-05,
-      "loss": 0.9631,
-      "step": 243
-    },
-    {
-      "epoch": 0.12001967535661584,
-      "grad_norm": 0.6583600524485893,
-      "learning_rate": 4.88e-05,
-      "loss": 1.0597,
-      "step": 244
-    },
-    {
-      "epoch": 0.1205115592720118,
-      "grad_norm": 0.6753110485508403,
-      "learning_rate": 4.9e-05,
-      "loss": 1.0366,
-      "step": 245
-    },
-    {
-      "epoch": 0.12100344318740777,
-      "grad_norm": 0.7026537939048871,
-      "learning_rate": 4.92e-05,
-      "loss": 1.0412,
-      "step": 246
-    },
-    {
-      "epoch": 0.12149532710280374,
-      "grad_norm": 0.6840738596363672,
-      "learning_rate": 4.94e-05,
-      "loss": 0.8955,
-      "step": 247
-    },
-    {
-      "epoch": 0.1219872110181997,
-      "grad_norm": 0.6787167522875908,
-      "learning_rate": 4.96e-05,
-      "loss": 1.0285,
-      "step": 248
-    },
-    {
-      "epoch": 0.12247909493359567,
-      "grad_norm": 0.6447409625317669,
-      "learning_rate": 4.9800000000000004e-05,
-      "loss": 1.0053,
-      "step": 249
-    },
-    {
-      "epoch": 0.12297097884899164,
-      "grad_norm": 0.6540923109167576,
-      "learning_rate": 5e-05,
-      "loss": 0.9068,
-      "step": 250
-    },
-    {
-      "epoch": 0.1234628627643876,
-      "grad_norm": 0.6565019553654861,
-      "learning_rate": 5.02e-05,
-      "loss": 1.0011,
-      "step": 251
-    },
-    {
-      "epoch": 0.12395474667978357,
-      "grad_norm": 1.909933657652513,
-      "learning_rate": 5.0400000000000005e-05,
-      "loss": 1.0145,
-      "step": 252
-    },
-    {
-      "epoch": 0.12444663059517953,
-      "grad_norm": 0.66278666736166,
-      "learning_rate": 5.0600000000000003e-05,
-      "loss": 0.9843,
-      "step": 253
-    },
-    {
-      "epoch": 0.1249385145105755,
-      "grad_norm": 0.6836835743581384,
-      "learning_rate": 5.08e-05,
-      "loss": 0.9417,
-      "step": 254
-    },
-    {
-      "epoch": 0.12543039842597148,
-      "grad_norm": 0.6863239949032678,
-      "learning_rate": 5.1000000000000006e-05,
-      "loss": 0.933,
-      "step": 255
-    },
-    {
-      "epoch": 0.12592228234136743,
-      "grad_norm": 0.6904796237804453,
-      "learning_rate": 5.1200000000000004e-05,
-      "loss": 0.9765,
-      "step": 256
-    },
-    {
-      "epoch": 0.1264141662567634,
-      "grad_norm": 0.685186929661333,
-      "learning_rate": 5.14e-05,
-      "loss": 0.9448,
-      "step": 257
-    },
-    {
-      "epoch": 0.12690605017215936,
-      "grad_norm": 0.6414016224391275,
-      "learning_rate": 5.16e-05,
-      "loss": 1.0642,
-      "step": 258
-    },
-    {
-      "epoch": 0.12739793408755534,
-      "grad_norm": 0.6389222696585697,
-      "learning_rate": 5.1800000000000005e-05,
-      "loss": 0.9642,
-      "step": 259
-    },
-    {
-      "epoch": 0.1278898180029513,
-      "grad_norm": 0.6741485892556393,
-      "learning_rate": 5.2000000000000004e-05,
-      "loss": 1.0645,
-      "step": 260
-    },
-    {
-      "epoch": 0.12838170191834727,
-      "grad_norm": 0.6229141794339108,
-      "learning_rate": 5.22e-05,
-      "loss": 1.013,
-      "step": 261
-    },
-    {
-      "epoch": 0.12887358583374323,
-      "grad_norm": 0.6907834356049275,
-      "learning_rate": 5.2400000000000007e-05,
-      "loss": 0.9936,
-      "step": 262
-    },
-    {
-      "epoch": 0.1293654697491392,
-      "grad_norm": 0.7038791661126186,
-      "learning_rate": 5.2600000000000005e-05,
-      "loss": 1.0136,
-      "step": 263
-    },
-    {
-      "epoch": 0.12985735366453516,
-      "grad_norm": 0.640725198404905,
-      "learning_rate": 5.28e-05,
-      "loss": 0.9825,
-      "step": 264
-    },
-    {
-      "epoch": 0.13034923757993114,
-      "grad_norm": 0.5842738998671183,
-      "learning_rate": 5.300000000000001e-05,
-      "loss": 0.8905,
-      "step": 265
-    },
-    {
-      "epoch": 0.1308411214953271,
-      "grad_norm": 0.7302625337561353,
-      "learning_rate": 5.3200000000000006e-05,
-      "loss": 0.9707,
-      "step": 266
-    },
-    {
-      "epoch": 0.13133300541072307,
-      "grad_norm": 0.640591046196343,
-      "learning_rate": 5.3400000000000004e-05,
-      "loss": 0.9923,
-      "step": 267
-    },
-    {
-      "epoch": 0.13182488932611905,
-      "grad_norm": 0.6401230104389162,
-      "learning_rate": 5.360000000000001e-05,
-      "loss": 1.0509,
-      "step": 268
-    },
-    {
-      "epoch": 0.132316773241515,
-      "grad_norm": 0.6168517912433212,
-      "learning_rate": 5.380000000000001e-05,
-      "loss": 1.0207,
-      "step": 269
-    },
-    {
-      "epoch": 0.13280865715691098,
-      "grad_norm": 0.6477707857028956,
-      "learning_rate": 5.4000000000000005e-05,
-      "loss": 1.0227,
-      "step": 270
-    },
-    {
-      "epoch": 0.13330054107230693,
-      "grad_norm": 0.6832697192287229,
-      "learning_rate": 5.420000000000001e-05,
-      "loss": 1.0356,
-      "step": 271
-    },
-    {
-      "epoch": 0.1337924249877029,
-      "grad_norm": 0.6243425458645113,
-      "learning_rate": 5.440000000000001e-05,
-      "loss": 0.9723,
-      "step": 272
-    },
-    {
-      "epoch": 0.13428430890309886,
-      "grad_norm": 0.6659761821502914,
-      "learning_rate": 5.4600000000000006e-05,
-      "loss": 0.9364,
-      "step": 273
-    },
-    {
-      "epoch": 0.13477619281849484,
-      "grad_norm": 0.655424441774203,
-      "learning_rate": 5.4800000000000004e-05,
-      "loss": 0.9311,
-      "step": 274
-    },
-    {
-      "epoch": 0.1352680767338908,
-      "grad_norm": 0.6658295218882189,
-      "learning_rate": 5.500000000000001e-05,
-      "loss": 0.9409,
-      "step": 275
-    },
-    {
-      "epoch": 0.13575996064928678,
-      "grad_norm": 4.3427010379394915,
-      "learning_rate": 5.520000000000001e-05,
-      "loss": 1.1593,
-      "step": 276
-    },
-    {
-      "epoch": 0.13625184456468273,
-      "grad_norm": 0.651728430891512,
-      "learning_rate": 5.5400000000000005e-05,
-      "loss": 0.9168,
-      "step": 277
-    },
-    {
-      "epoch": 0.1367437284800787,
-      "grad_norm": 0.6611257694815104,
-      "learning_rate": 5.560000000000001e-05,
-      "loss": 0.9477,
-      "step": 278
-    },
-    {
-      "epoch": 0.13723561239547466,
-      "grad_norm": 0.677143186101974,
-      "learning_rate": 5.580000000000001e-05,
-      "loss": 0.9808,
-      "step": 279
-    },
-    {
-      "epoch": 0.13772749631087064,
-      "grad_norm": 0.6757312948008557,
-      "learning_rate": 5.6000000000000006e-05,
-      "loss": 1.0211,
-      "step": 280
-    },
-    {
-      "epoch": 0.1382193802262666,
-      "grad_norm": 0.6204458845897312,
-      "learning_rate": 5.620000000000001e-05,
-      "loss": 0.9787,
-      "step": 281
-    },
-    {
-      "epoch": 0.13871126414166257,
-      "grad_norm": 0.6301827758158104,
-      "learning_rate": 5.6399999999999995e-05,
-      "loss": 0.9815,
-      "step": 282
-    },
-    {
-      "epoch": 0.13920314805705852,
-      "grad_norm": 0.8061051866889022,
-      "learning_rate": 5.66e-05,
-      "loss": 1.0097,
-      "step": 283
-    },
-    {
-      "epoch": 0.1396950319724545,
-      "grad_norm": 0.6337588057022924,
-      "learning_rate": 5.68e-05,
-      "loss": 0.9373,
-      "step": 284
-    },
-    {
-      "epoch": 0.14018691588785046,
-      "grad_norm": 0.6629923742300748,
-      "learning_rate": 5.6999999999999996e-05,
-      "loss": 1.105,
-      "step": 285
-    },
-    {
-      "epoch": 0.14067879980324643,
-      "grad_norm": 0.6202099379314775,
-      "learning_rate": 5.72e-05,
-      "loss": 1.0256,
-      "step": 286
-    },
-    {
-      "epoch": 0.1411706837186424,
-      "grad_norm": 0.6364234732947897,
-      "learning_rate": 5.74e-05,
-      "loss": 0.996,
-      "step": 287
-    },
-    {
-      "epoch": 0.14166256763403837,
-      "grad_norm": 0.6345000851230562,
-      "learning_rate": 5.76e-05,
-      "loss": 0.9256,
-      "step": 288
-    },
-    {
-      "epoch": 0.14215445154943435,
-      "grad_norm": 0.6356287411447527,
-      "learning_rate": 5.7799999999999995e-05,
-      "loss": 0.9884,
-      "step": 289
-    },
-    {
-      "epoch": 0.1426463354648303,
-      "grad_norm": 0.7056133469436392,
-      "learning_rate": 5.8e-05,
-      "loss": 0.981,
-      "step": 290
-    },
-    {
-      "epoch": 0.14313821938022628,
-      "grad_norm": 0.6850636267710177,
-      "learning_rate": 5.82e-05,
-      "loss": 0.9667,
-      "step": 291
-    },
-    {
-      "epoch": 0.14363010329562223,
-      "grad_norm": 0.6282739333968476,
-      "learning_rate": 5.8399999999999997e-05,
-      "loss": 0.9947,
-      "step": 292
-    },
-    {
-      "epoch": 0.1441219872110182,
-      "grad_norm": 0.6418596479252349,
-      "learning_rate": 5.86e-05,
-      "loss": 1.0141,
-      "step": 293
-    },
-    {
-      "epoch": 0.14461387112641416,
-      "grad_norm": 0.6595528178588359,
-      "learning_rate": 5.88e-05,
-      "loss": 1.0151,
-      "step": 294
-    },
-    {
-      "epoch": 0.14510575504181014,
-      "grad_norm": 0.6920906659420407,
-      "learning_rate": 5.9e-05,
-      "loss": 0.9947,
-      "step": 295
-    },
-    {
-      "epoch": 0.1455976389572061,
-      "grad_norm": 0.6518087338599203,
-      "learning_rate": 5.92e-05,
-      "loss": 0.9352,
-      "step": 296
-    },
-    {
-      "epoch": 0.14608952287260207,
-      "grad_norm": 0.611941517521001,
-      "learning_rate": 5.94e-05,
-      "loss": 0.9139,
-      "step": 297
-    },
-    {
-      "epoch": 0.14658140678799803,
-      "grad_norm": 0.6447816591904986,
-      "learning_rate": 5.96e-05,
-      "loss": 0.9703,
-      "step": 298
-    },
-    {
-      "epoch": 0.147073290703394,
-      "grad_norm": 0.6306005109194481,
-      "learning_rate": 5.9800000000000003e-05,
-      "loss": 0.9748,
-      "step": 299
-    },
-    {
-      "epoch": 0.14756517461878996,
-      "grad_norm": 0.6632237582326308,
-      "learning_rate": 6e-05,
-      "loss": 1.0416,
-      "step": 300
-    },
-    {
-      "epoch": 0.14805705853418594,
-      "grad_norm": 0.6517410541130536,
-      "learning_rate": 6.02e-05,
-      "loss": 0.9308,
-      "step": 301
-    },
-    {
-      "epoch": 0.1485489424495819,
-      "grad_norm": 0.7062565571754538,
-      "learning_rate": 6.04e-05,
-      "loss": 1.0257,
-      "step": 302
-    },
-    {
-      "epoch": 0.14904082636497787,
-      "grad_norm": 0.6459974349047483,
-      "learning_rate": 6.06e-05,
-      "loss": 0.9867,
-      "step": 303
-    },
-    {
-      "epoch": 0.14953271028037382,
-      "grad_norm": 0.6671134833116954,
-      "learning_rate": 6.08e-05,
-      "loss": 1.0088,
-      "step": 304
-    },
-    {
-      "epoch": 0.1500245941957698,
-      "grad_norm": 0.6682947966042925,
-      "learning_rate": 6.1e-05,
-      "loss": 1.0053,
-      "step": 305
-    },
-    {
-      "epoch": 0.15051647811116575,
-      "grad_norm": 0.6132659312827933,
-      "learning_rate": 6.12e-05,
-      "loss": 0.9845,
-      "step": 306
-    },
-    {
-      "epoch": 0.15100836202656173,
-      "grad_norm": 0.7433304321908412,
-      "learning_rate": 6.14e-05,
-      "loss": 1.0108,
-      "step": 307
-    },
-    {
-      "epoch": 0.15150024594195768,
-      "grad_norm": 0.7092648473648441,
-      "learning_rate": 6.16e-05,
-      "loss": 0.9951,
-      "step": 308
-    },
-    {
-      "epoch": 0.15199212985735366,
-      "grad_norm": 0.6355870188527534,
-      "learning_rate": 6.18e-05,
-      "loss": 0.9423,
-      "step": 309
-    },
-    {
-      "epoch": 0.15248401377274964,
-      "grad_norm": 0.6242503654633647,
-      "learning_rate": 6.2e-05,
-      "loss": 0.9129,
-      "step": 310
-    },
-    {
-      "epoch": 0.1529758976881456,
-      "grad_norm": 0.6109391351151449,
-      "learning_rate": 6.220000000000001e-05,
-      "loss": 0.9067,
-      "step": 311
-    },
-    {
-      "epoch": 0.15346778160354158,
-      "grad_norm": 0.6300248986974131,
-      "learning_rate": 6.24e-05,
-      "loss": 0.9957,
-      "step": 312
-    },
-    {
-      "epoch": 0.15395966551893753,
-      "grad_norm": 0.6165606343456851,
-      "learning_rate": 6.26e-05,
-      "loss": 1.0315,
-      "step": 313
-    },
-    {
-      "epoch": 0.1544515494343335,
-      "grad_norm": 0.616948344272319,
-      "learning_rate": 6.280000000000001e-05,
-      "loss": 0.964,
-      "step": 314
-    },
-    {
-      "epoch": 0.15494343334972946,
-      "grad_norm": 0.6465167406806364,
-      "learning_rate": 6.3e-05,
-      "loss": 1.0329,
-      "step": 315
-    },
-    {
-      "epoch": 0.15543531726512544,
-      "grad_norm": 0.6153339495121846,
-      "learning_rate": 6.32e-05,
-      "loss": 1.035,
-      "step": 316
-    },
-    {
-      "epoch": 0.1559272011805214,
-      "grad_norm": 0.6170904248104213,
-      "learning_rate": 6.340000000000001e-05,
-      "loss": 0.97,
-      "step": 317
-    },
-    {
-      "epoch": 0.15641908509591737,
-      "grad_norm": 0.696445964079362,
-      "learning_rate": 6.36e-05,
-      "loss": 1.0645,
-      "step": 318
-    },
-    {
-      "epoch": 0.15691096901131332,
-      "grad_norm": 0.5936689219966211,
-      "learning_rate": 6.38e-05,
-      "loss": 0.9516,
-      "step": 319
-    },
-    {
-      "epoch": 0.1574028529267093,
-      "grad_norm": 0.6766475677939298,
-      "learning_rate": 6.400000000000001e-05,
-      "loss": 0.9874,
-      "step": 320
-    },
-    {
-      "epoch": 0.15789473684210525,
-      "grad_norm": 0.6199318696925801,
-      "learning_rate": 6.42e-05,
-      "loss": 0.9286,
-      "step": 321
-    },
-    {
-      "epoch": 0.15838662075750123,
-      "grad_norm": 0.6643394593362091,
-      "learning_rate": 6.440000000000001e-05,
-      "loss": 1.0774,
-      "step": 322
-    },
-    {
-      "epoch": 0.1588785046728972,
-      "grad_norm": 0.5898421821171212,
-      "learning_rate": 6.460000000000001e-05,
-      "loss": 0.8951,
-      "step": 323
-    },
-    {
-      "epoch": 0.15937038858829317,
-      "grad_norm": 0.6146746666462884,
-      "learning_rate": 6.48e-05,
-      "loss": 0.9303,
-      "step": 324
-    },
-    {
-      "epoch": 0.15986227250368912,
-      "grad_norm": 0.6423463369567147,
-      "learning_rate": 6.500000000000001e-05,
-      "loss": 1.0393,
-      "step": 325
-    },
-    {
-      "epoch": 0.1603541564190851,
-      "grad_norm": 0.6315059932248648,
-      "learning_rate": 6.52e-05,
-      "loss": 1.0133,
-      "step": 326
-    },
-    {
-      "epoch": 0.16084604033448105,
-      "grad_norm": 0.6815156367432513,
-      "learning_rate": 6.54e-05,
-      "loss": 0.9572,
-      "step": 327
-    },
-    {
-      "epoch": 0.16133792424987703,
-      "grad_norm": 0.6519666208797611,
-      "learning_rate": 6.560000000000001e-05,
-      "loss": 0.9082,
-      "step": 328
-    },
-    {
-      "epoch": 0.161829808165273,
-      "grad_norm": 0.62523957910431,
-      "learning_rate": 6.58e-05,
-      "loss": 0.9654,
-      "step": 329
-    },
-    {
-      "epoch": 0.16232169208066896,
-      "grad_norm": 0.7309020196058403,
-      "learning_rate": 6.6e-05,
-      "loss": 0.9375,
-      "step": 330
-    },
-    {
-      "epoch": 0.16281357599606494,
-      "grad_norm": 0.6695844264781273,
-      "learning_rate": 6.620000000000001e-05,
-      "loss": 0.9489,
-      "step": 331
-    },
-    {
-      "epoch": 0.1633054599114609,
-      "grad_norm": 0.6494479688502096,
-      "learning_rate": 6.64e-05,
-      "loss": 0.9377,
-      "step": 332
-    },
-    {
-      "epoch": 0.16379734382685687,
-      "grad_norm": 0.6336750522480793,
-      "learning_rate": 6.66e-05,
-      "loss": 0.9992,
-      "step": 333
-    },
-    {
-      "epoch": 0.16428922774225282,
-      "grad_norm": 0.6693397113005544,
-      "learning_rate": 6.680000000000001e-05,
-      "loss": 0.9959,
-      "step": 334
-    },
-    {
-      "epoch": 0.1647811116576488,
-      "grad_norm": 0.6002755640350812,
-      "learning_rate": 6.7e-05,
-      "loss": 0.9596,
-      "step": 335
-    },
-    {
-      "epoch": 0.16527299557304476,
-      "grad_norm": 0.6559694357848178,
-      "learning_rate": 6.720000000000001e-05,
-      "loss": 1.0273,
-      "step": 336
-    },
-    {
-      "epoch": 0.16576487948844074,
-      "grad_norm": 0.6545341522027837,
-      "learning_rate": 6.740000000000001e-05,
-      "loss": 1.0007,
-      "step": 337
-    },
-    {
-      "epoch": 0.1662567634038367,
-      "grad_norm": 0.640746586799343,
-      "learning_rate": 6.76e-05,
-      "loss": 1.0115,
-      "step": 338
-    },
-    {
-      "epoch": 0.16674864731923267,
-      "grad_norm": 0.6186436669731425,
-      "learning_rate": 6.780000000000001e-05,
-      "loss": 1.0891,
-      "step": 339
-    },
-    {
-      "epoch": 0.16724053123462862,
-      "grad_norm": 0.6119028812273772,
-      "learning_rate": 6.800000000000001e-05,
-      "loss": 0.9275,
-      "step": 340
-    },
-    {
-      "epoch": 0.1677324151500246,
-      "grad_norm": 0.6579939475287446,
-      "learning_rate": 6.82e-05,
-      "loss": 1.0235,
-      "step": 341
-    },
-    {
-      "epoch": 0.16822429906542055,
-      "grad_norm": 0.6383679027580579,
-      "learning_rate": 6.840000000000001e-05,
-      "loss": 1.0284,
-      "step": 342
-    },
-    {
-      "epoch": 0.16871618298081653,
-      "grad_norm": 0.6507778291909052,
-      "learning_rate": 6.860000000000001e-05,
-      "loss": 1.0164,
-      "step": 343
-    },
-    {
-      "epoch": 0.16920806689621248,
-      "grad_norm": 0.6559899499233537,
-      "learning_rate": 6.879999999999999e-05,
-      "loss": 1.0207,
-      "step": 344
-    },
-    {
-      "epoch": 0.16969995081160846,
-      "grad_norm": 0.6307205703320318,
-      "learning_rate": 6.9e-05,
-      "loss": 0.9824,
-      "step": 345
-    },
-    {
-      "epoch": 0.17019183472700442,
-      "grad_norm": 0.6484572895058113,
-      "learning_rate": 6.92e-05,
-      "loss": 0.9476,
-      "step": 346
-    },
-    {
-      "epoch": 0.1706837186424004,
-      "grad_norm": 0.5938424560381667,
-      "learning_rate": 6.939999999999999e-05,
-      "loss": 0.9291,
-      "step": 347
-    },
-    {
-      "epoch": 0.17117560255779635,
-      "grad_norm": 0.6077052340881578,
-      "learning_rate": 6.96e-05,
-      "loss": 0.9314,
-      "step": 348
-    },
-    {
-      "epoch": 0.17166748647319233,
-      "grad_norm": 0.5793665866913059,
-      "learning_rate": 6.98e-05,
-      "loss": 0.8881,
-      "step": 349
-    },
-    {
-      "epoch": 0.1721593703885883,
-      "grad_norm": 0.6217622467644931,
-      "learning_rate": 7e-05,
-      "loss": 0.9426,
-      "step": 350
-    },
-    {
-      "epoch": 0.17265125430398426,
-      "grad_norm": 0.6492966649268521,
-      "learning_rate": 7.02e-05,
-      "loss": 0.9888,
-      "step": 351
-    },
-    {
-      "epoch": 0.17314313821938024,
-      "grad_norm": 0.638633022430336,
-      "learning_rate": 7.04e-05,
-      "loss": 1.0077,
-      "step": 352
-    },
-    {
-      "epoch": 0.1736350221347762,
-      "grad_norm": 0.6228845062610631,
-      "learning_rate": 7.06e-05,
-      "loss": 0.9699,
-      "step": 353
-    },
-    {
-      "epoch": 0.17412690605017217,
-      "grad_norm": 0.5702229019173722,
-      "learning_rate": 7.08e-05,
-      "loss": 0.9429,
-      "step": 354
-    },
-    {
-      "epoch": 0.17461878996556812,
-      "grad_norm": 0.5924801567771402,
-      "learning_rate": 7.1e-05,
-      "loss": 0.9659,
-      "step": 355
-    },
-    {
-      "epoch": 0.1751106738809641,
-      "grad_norm": 0.6232365364163425,
-      "learning_rate": 7.12e-05,
-      "loss": 0.993,
-      "step": 356
-    },
-    {
-      "epoch": 0.17560255779636005,
-      "grad_norm": 0.6183797922468482,
-      "learning_rate": 7.14e-05,
-      "loss": 0.9918,
-      "step": 357
-    },
-    {
-      "epoch": 0.17609444171175603,
-      "grad_norm": 0.8735174977501221,
-      "learning_rate": 7.16e-05,
-      "loss": 1.1078,
-      "step": 358
-    },
-    {
-      "epoch": 0.17658632562715199,
-      "grad_norm": 0.6011381747794056,
-      "learning_rate": 7.18e-05,
-      "loss": 0.9236,
-      "step": 359
-    },
-    {
-      "epoch": 0.17707820954254797,
-      "grad_norm": 0.5965327186295076,
-      "learning_rate": 7.2e-05,
-      "loss": 0.9969,
-      "step": 360
-    },
-    {
-      "epoch": 0.17757009345794392,
-      "grad_norm": 0.6331964093992425,
-      "learning_rate": 7.22e-05,
-      "loss": 0.9924,
-      "step": 361
-    },
-    {
-      "epoch": 0.1780619773733399,
-      "grad_norm": 0.6412859155950007,
-      "learning_rate": 7.24e-05,
-      "loss": 0.9406,
-      "step": 362
-    },
-    {
-      "epoch": 0.17855386128873585,
-      "grad_norm": 0.638583426526138,
-      "learning_rate": 7.26e-05,
-      "loss": 1.0027,
-      "step": 363
-    },
-    {
-      "epoch": 0.17904574520413183,
-      "grad_norm": 0.5709523375451793,
-      "learning_rate": 7.280000000000001e-05,
-      "loss": 0.9009,
-      "step": 364
-    },
-    {
-      "epoch": 0.17953762911952778,
-      "grad_norm": 0.6474681443412542,
-      "learning_rate": 7.3e-05,
-      "loss": 1.0186,
-      "step": 365
-    },
-    {
-      "epoch": 0.18002951303492376,
-      "grad_norm": 0.7193510038246659,
-      "learning_rate": 7.32e-05,
-      "loss": 0.9623,
-      "step": 366
-    },
-    {
-      "epoch": 0.1805213969503197,
-      "grad_norm": 0.6479023597226308,
-      "learning_rate": 7.340000000000001e-05,
-      "loss": 1.0473,
-      "step": 367
-    },
-    {
-      "epoch": 0.1810132808657157,
-      "grad_norm": 0.6714577198688532,
-      "learning_rate": 7.36e-05,
-      "loss": 0.9741,
-      "step": 368
-    },
-    {
-      "epoch": 0.18150516478111164,
-      "grad_norm": 0.6626593218023097,
-      "learning_rate": 7.38e-05,
-      "loss": 1.0352,
-      "step": 369
-    },
-    {
-      "epoch": 0.18199704869650762,
-      "grad_norm": 0.6860491249523826,
-      "learning_rate": 7.4e-05,
-      "loss": 0.9544,
-      "step": 370
-    },
-    {
-      "epoch": 0.1824889326119036,
-      "grad_norm": 0.6577681496370567,
-      "learning_rate": 7.42e-05,
-      "loss": 1.0932,
-      "step": 371
-    },
-    {
-      "epoch": 0.18298081652729956,
-      "grad_norm": 0.5896122652070747,
-      "learning_rate": 7.44e-05,
-      "loss": 1.0346,
-      "step": 372
-    },
-    {
-      "epoch": 0.18347270044269554,
-      "grad_norm": 0.6350856778845727,
-      "learning_rate": 7.46e-05,
-      "loss": 1.0566,
-      "step": 373
-    },
-    {
-      "epoch": 0.1839645843580915,
-      "grad_norm": 0.6310622245330231,
-      "learning_rate": 7.48e-05,
-      "loss": 0.9427,
-      "step": 374
-    },
-    {
-      "epoch": 0.18445646827348747,
-      "grad_norm": 0.591393808769729,
-      "learning_rate": 7.500000000000001e-05,
-      "loss": 0.9367,
-      "step": 375
-    },
-    {
-      "epoch": 0.18494835218888342,
-      "grad_norm": 0.5983444228918225,
-      "learning_rate": 7.52e-05,
-      "loss": 0.8887,
-      "step": 376
-    },
-    {
-      "epoch": 0.1854402361042794,
-      "grad_norm": 0.6300800008512292,
-      "learning_rate": 7.54e-05,
-      "loss": 1.0046,
-      "step": 377
-    },
-    {
-      "epoch": 0.18593212001967535,
-      "grad_norm": 0.6248792310246031,
-      "learning_rate": 7.560000000000001e-05,
-      "loss": 0.9309,
-      "step": 378
-    },
-    {
-      "epoch": 0.18642400393507133,
-      "grad_norm": 0.6224165965967048,
-      "learning_rate": 7.58e-05,
-      "loss": 0.9787,
-      "step": 379
-    },
-    {
-      "epoch": 0.18691588785046728,
-      "grad_norm": 0.7023975725077084,
-      "learning_rate": 7.6e-05,
-      "loss": 1.0844,
-      "step": 380
-    },
-    {
-      "epoch": 0.18740777176586326,
-      "grad_norm": 0.7741352574238073,
-      "learning_rate": 7.620000000000001e-05,
-      "loss": 1.0394,
-      "step": 381
-    },
-    {
-      "epoch": 0.18789965568125921,
-      "grad_norm": 0.6098810052004804,
-      "learning_rate": 7.64e-05,
-      "loss": 0.9726,
-      "step": 382
-    },
-    {
-      "epoch": 0.1883915395966552,
-      "grad_norm": 0.5818813489381275,
-      "learning_rate": 7.66e-05,
-      "loss": 0.9709,
-      "step": 383
-    },
-    {
-      "epoch": 0.18888342351205115,
-      "grad_norm": 0.6226189074036584,
-      "learning_rate": 7.680000000000001e-05,
-      "loss": 0.9544,
-      "step": 384
-    },
-    {
-      "epoch": 0.18937530742744713,
-      "grad_norm": 0.6356790232097046,
-      "learning_rate": 7.7e-05,
-      "loss": 0.9817,
-      "step": 385
-    },
-    {
-      "epoch": 0.18986719134284308,
-      "grad_norm": 0.671564537895718,
-      "learning_rate": 7.72e-05,
-      "loss": 0.9338,
-      "step": 386
-    },
-    {
-      "epoch": 0.19035907525823906,
-      "grad_norm": 0.664535078510209,
-      "learning_rate": 7.740000000000001e-05,
-      "loss": 0.9018,
-      "step": 387
-    },
-    {
-      "epoch": 0.190850959173635,
-      "grad_norm": 0.6394011220372047,
-      "learning_rate": 7.76e-05,
-      "loss": 1.0591,
-      "step": 388
-    },
-    {
-      "epoch": 0.191342843089031,
-      "grad_norm": 0.5892373379666562,
-      "learning_rate": 7.780000000000001e-05,
-      "loss": 0.9627,
-      "step": 389
-    },
-    {
-      "epoch": 0.19183472700442694,
-      "grad_norm": 0.6377552139327307,
-      "learning_rate": 7.800000000000001e-05,
-      "loss": 1.0187,
-      "step": 390
-    },
-    {
-      "epoch": 0.19232661091982292,
-      "grad_norm": 0.6050739601789275,
-      "learning_rate": 7.82e-05,
-      "loss": 0.9378,
-      "step": 391
-    },
-    {
-      "epoch": 0.1928184948352189,
-      "grad_norm": 0.6674680863375297,
-      "learning_rate": 7.840000000000001e-05,
-      "loss": 1.0179,
-      "step": 392
-    },
-    {
-      "epoch": 0.19331037875061485,
-      "grad_norm": 0.6328515579503435,
-      "learning_rate": 7.860000000000001e-05,
-      "loss": 1.0017,
-      "step": 393
-    },
-    {
-      "epoch": 0.19380226266601083,
-      "grad_norm": 0.6908769012404306,
-      "learning_rate": 7.88e-05,
-      "loss": 1.0762,
-      "step": 394
-    },
-    {
-      "epoch": 0.19429414658140678,
-      "grad_norm": 0.6482324781379857,
-      "learning_rate": 7.900000000000001e-05,
-      "loss": 0.9735,
-      "step": 395
-    },
-    {
-      "epoch": 0.19478603049680276,
-      "grad_norm": 0.6018402507831343,
-      "learning_rate": 7.920000000000001e-05,
-      "loss": 0.9507,
-      "step": 396
-    },
-    {
-      "epoch": 0.19527791441219872,
-      "grad_norm": 0.6019084942068132,
-      "learning_rate": 7.94e-05,
-      "loss": 0.9597,
-      "step": 397
-    },
-    {
-      "epoch": 0.1957697983275947,
-      "grad_norm": 0.6818325977268064,
-      "learning_rate": 7.960000000000001e-05,
-      "loss": 0.8991,
-      "step": 398
-    },
-    {
-      "epoch": 0.19626168224299065,
-      "grad_norm": 0.5914021323075259,
-      "learning_rate": 7.98e-05,
-      "loss": 0.9103,
-      "step": 399
-    },
-    {
-      "epoch": 0.19675356615838663,
-      "grad_norm": 0.6391699256528783,
-      "learning_rate": 8e-05,
-      "loss": 0.9233,
-      "step": 400
-    },
-    {
-      "epoch": 0.19724545007378258,
-      "grad_norm": 0.6336439557662955,
-      "learning_rate": 8.020000000000001e-05,
-      "loss": 0.9096,
-      "step": 401
-    },
-    {
-      "epoch": 0.19773733398917856,
-      "grad_norm": 0.6266381712881892,
-      "learning_rate": 8.04e-05,
-      "loss": 0.997,
-      "step": 402
-    },
-    {
-      "epoch": 0.1982292179045745,
-      "grad_norm": 0.600600584008272,
-      "learning_rate": 8.060000000000001e-05,
-      "loss": 0.992,
-      "step": 403
-    },
-    {
-      "epoch": 0.1987211018199705,
-      "grad_norm": 0.6517734815494437,
-      "learning_rate": 8.080000000000001e-05,
-      "loss": 0.9817,
-      "step": 404
-    },
-    {
-      "epoch": 0.19921298573536644,
-      "grad_norm": 0.6293416217886094,
-      "learning_rate": 8.1e-05,
-      "loss": 1.0526,
-      "step": 405
-    },
-    {
-      "epoch": 0.19970486965076242,
-      "grad_norm": 0.6465733209160941,
-      "learning_rate": 8.120000000000001e-05,
-      "loss": 0.976,
-      "step": 406
-    },
-    {
-      "epoch": 0.20019675356615838,
-      "grad_norm": 0.6485332141795432,
-      "learning_rate": 8.14e-05,
-      "loss": 1.0016,
-      "step": 407
-    },
-    {
-      "epoch": 0.20068863748155436,
-      "grad_norm": 0.5718975648508007,
-      "learning_rate": 8.16e-05,
-      "loss": 0.9666,
-      "step": 408
-    },
-    {
-      "epoch": 0.2011805213969503,
-      "grad_norm": 0.6283887150004627,
-      "learning_rate": 8.18e-05,
-      "loss": 0.9837,
-      "step": 409
-    },
-    {
-      "epoch": 0.2016724053123463,
-      "grad_norm": 0.6345487037460759,
-      "learning_rate": 8.2e-05,
-      "loss": 0.981,
-      "step": 410
-    },
-    {
-      "epoch": 0.20216428922774227,
-      "grad_norm": 0.6248364172555066,
-      "learning_rate": 8.22e-05,
-      "loss": 0.9581,
-      "step": 411
-    },
-    {
-      "epoch": 0.20265617314313822,
-      "grad_norm": 0.6102819030010944,
-      "learning_rate": 8.24e-05,
-      "loss": 0.9207,
-      "step": 412
-    },
-    {
-      "epoch": 0.2031480570585342,
-      "grad_norm": 0.9706049727359666,
-      "learning_rate": 8.26e-05,
-      "loss": 0.9608,
-      "step": 413
-    },
-    {
-      "epoch": 0.20363994097393015,
-      "grad_norm": 0.6044328622763054,
-      "learning_rate": 8.28e-05,
-      "loss": 0.9308,
-      "step": 414
-    },
-    {
-      "epoch": 0.20413182488932613,
-      "grad_norm": 0.639778068725871,
-      "learning_rate": 8.3e-05,
-      "loss": 1.0425,
-      "step": 415
-    },
-    {
-      "epoch": 0.20462370880472208,
-      "grad_norm": 0.6167918577777772,
-      "learning_rate": 8.32e-05,
-      "loss": 0.9816,
-      "step": 416
-    },
-    {
-      "epoch": 0.20511559272011806,
-      "grad_norm": 0.5999991449288878,
-      "learning_rate": 8.34e-05,
-      "loss": 0.8474,
-      "step": 417
-    },
-    {
-      "epoch": 0.205607476635514,
-      "grad_norm": 0.6076168181208658,
-      "learning_rate": 8.36e-05,
-      "loss": 0.9624,
-      "step": 418
-    },
-    {
-      "epoch": 0.20609936055091,
-      "grad_norm": 0.6165342411399956,
-      "learning_rate": 8.38e-05,
-      "loss": 0.9529,
-      "step": 419
-    },
-    {
-      "epoch": 0.20659124446630595,
-      "grad_norm": 0.6132122494477279,
-      "learning_rate": 8.4e-05,
-      "loss": 0.989,
-      "step": 420
-    },
-    {
-      "epoch": 0.20708312838170193,
-      "grad_norm": 0.6027449919133129,
-      "learning_rate": 8.42e-05,
-      "loss": 0.9066,
-      "step": 421
-    },
-    {
-      "epoch": 0.20757501229709788,
-      "grad_norm": 0.5928239968088229,
-      "learning_rate": 8.44e-05,
-      "loss": 1.0044,
-      "step": 422
-    },
-    {
-      "epoch": 0.20806689621249386,
-      "grad_norm": 0.5869899950440051,
-      "learning_rate": 8.46e-05,
-      "loss": 0.9863,
-      "step": 423
-    },
-    {
-      "epoch": 0.2085587801278898,
-      "grad_norm": 0.5766767880122016,
-      "learning_rate": 8.48e-05,
-      "loss": 1.0377,
-      "step": 424
-    },
-    {
-      "epoch": 0.2090506640432858,
-      "grad_norm": 0.6264523722554275,
-      "learning_rate": 8.5e-05,
-      "loss": 0.947,
-      "step": 425
-    },
-    {
-      "epoch": 0.20954254795868174,
-      "grad_norm": 0.6045702512587927,
-      "learning_rate": 8.52e-05,
-      "loss": 0.9296,
-      "step": 426
-    },
-    {
-      "epoch": 0.21003443187407772,
-      "grad_norm": 0.5860636391406551,
-      "learning_rate": 8.54e-05,
-      "loss": 0.9624,
-      "step": 427
-    },
-    {
-      "epoch": 0.21052631578947367,
-      "grad_norm": 0.8151675635611291,
-      "learning_rate": 8.560000000000001e-05,
-      "loss": 0.997,
-      "step": 428
-    },
-    {
-      "epoch": 0.21101819970486965,
-      "grad_norm": 0.6030314739242377,
-      "learning_rate": 8.58e-05,
-      "loss": 0.9598,
-      "step": 429
-    },
-    {
-      "epoch": 0.2115100836202656,
-      "grad_norm": 0.5792174344697232,
-      "learning_rate": 8.6e-05,
-      "loss": 1.0389,
-      "step": 430
-    },
-    {
-      "epoch": 0.21200196753566158,
-      "grad_norm": 0.5684645313201525,
-      "learning_rate": 8.620000000000001e-05,
-      "loss": 0.8997,
-      "step": 431
-    },
-    {
-      "epoch": 0.21249385145105756,
-      "grad_norm": 0.5876143641689083,
-      "learning_rate": 8.64e-05,
-      "loss": 1.0035,
-      "step": 432
-    },
-    {
-      "epoch": 0.21298573536645352,
-      "grad_norm": 0.5872667882912961,
-      "learning_rate": 8.66e-05,
-      "loss": 0.9598,
-      "step": 433
-    },
-    {
-      "epoch": 0.2134776192818495,
-      "grad_norm": 0.637253451970368,
-      "learning_rate": 8.680000000000001e-05,
-      "loss": 0.9934,
-      "step": 434
-    },
-    {
-      "epoch": 0.21396950319724545,
-      "grad_norm": 0.5481379443674183,
-      "learning_rate": 8.7e-05,
-      "loss": 0.953,
-      "step": 435
-    },
-    {
-      "epoch": 0.21446138711264143,
-      "grad_norm": 0.5971331846018932,
-      "learning_rate": 8.72e-05,
-      "loss": 0.9386,
-      "step": 436
-    },
-    {
-      "epoch": 0.21495327102803738,
-      "grad_norm": 0.6307097366754557,
-      "learning_rate": 8.740000000000001e-05,
-      "loss": 0.9256,
-      "step": 437
-    },
-    {
-      "epoch": 0.21544515494343336,
-      "grad_norm": 0.5788073659558398,
-      "learning_rate": 8.76e-05,
-      "loss": 0.8838,
-      "step": 438
-    },
-    {
-      "epoch": 0.2159370388588293,
-      "grad_norm": 0.601923427178012,
-      "learning_rate": 8.78e-05,
-      "loss": 0.9056,
-      "step": 439
-    },
-    {
-      "epoch": 0.2164289227742253,
-      "grad_norm": 0.6370454166547089,
-      "learning_rate": 8.800000000000001e-05,
-      "loss": 1.0072,
-      "step": 440
-    },
-    {
-      "epoch": 0.21692080668962124,
-      "grad_norm": 0.6189606659633952,
-      "learning_rate": 8.82e-05,
-      "loss": 0.9057,
-      "step": 441
-    },
-    {
-      "epoch": 0.21741269060501722,
-      "grad_norm": 0.5936340681748756,
-      "learning_rate": 8.840000000000001e-05,
-      "loss": 0.9471,
-      "step": 442
-    },
-    {
-      "epoch": 0.21790457452041317,
-      "grad_norm": 0.7206104620301254,
-      "learning_rate": 8.86e-05,
-      "loss": 1.0461,
-      "step": 443
-    },
-    {
-      "epoch": 0.21839645843580915,
-      "grad_norm": 0.6075997639143946,
-      "learning_rate": 8.88e-05,
-      "loss": 0.9781,
-      "step": 444
-    },
-    {
-      "epoch": 0.2188883423512051,
-      "grad_norm": 0.6248001399556955,
-      "learning_rate": 8.900000000000001e-05,
-      "loss": 0.9315,
-      "step": 445
-    },
-    {
-      "epoch": 0.21938022626660109,
-      "grad_norm": 0.5813944293621506,
-      "learning_rate": 8.92e-05,
-      "loss": 0.9133,
-      "step": 446
-    },
-    {
-      "epoch": 0.21987211018199704,
-      "grad_norm": 0.636632987875858,
-      "learning_rate": 8.94e-05,
-      "loss": 0.9739,
-      "step": 447
-    },
-    {
-      "epoch": 0.22036399409739302,
-      "grad_norm": 0.590767131415477,
-      "learning_rate": 8.960000000000001e-05,
-      "loss": 0.8692,
-      "step": 448
-    },
-    {
-      "epoch": 0.22085587801278897,
-      "grad_norm": 0.617883557634732,
-      "learning_rate": 8.98e-05,
-      "loss": 0.961,
-      "step": 449
-    },
-    {
-      "epoch": 0.22134776192818495,
-      "grad_norm": 0.5876025309993035,
-      "learning_rate": 9e-05,
-      "loss": 0.8839,
-      "step": 450
-    },
-    {
-      "epoch": 0.2218396458435809,
-      "grad_norm": 0.567536847812842,
-      "learning_rate": 9.020000000000001e-05,
-      "loss": 0.9234,
-      "step": 451
-    },
-    {
-      "epoch": 0.22233152975897688,
-      "grad_norm": 0.6249958799726273,
-      "learning_rate": 9.04e-05,
-      "loss": 0.9916,
-      "step": 452
-    },
-    {
-      "epoch": 0.22282341367437286,
-      "grad_norm": 0.5903837871692948,
-      "learning_rate": 9.06e-05,
-      "loss": 0.9584,
-      "step": 453
-    },
-    {
-      "epoch": 0.2233152975897688,
-      "grad_norm": 0.5975533024049463,
-      "learning_rate": 9.080000000000001e-05,
-      "loss": 0.9771,
-      "step": 454
-    },
-    {
-      "epoch": 0.2238071815051648,
-      "grad_norm": 0.5772160499556087,
-      "learning_rate": 9.1e-05,
-      "loss": 1.0129,
-      "step": 455
-    },
-    {
-      "epoch": 0.22429906542056074,
-      "grad_norm": 0.6639280119847331,
-      "learning_rate": 9.120000000000001e-05,
-      "loss": 1.068,
-      "step": 456
-    },
-    {
-      "epoch": 0.22479094933595672,
-      "grad_norm": 0.5642570348460942,
-      "learning_rate": 9.140000000000001e-05,
-      "loss": 0.9018,
-      "step": 457
-    },
-    {
-      "epoch": 0.22528283325135268,
-      "grad_norm": 0.6103778845508339,
-      "learning_rate": 9.16e-05,
-      "loss": 0.932,
-      "step": 458
-    },
-    {
-      "epoch": 0.22577471716674866,
-      "grad_norm": 0.6456272560340363,
-      "learning_rate": 9.180000000000001e-05,
-      "loss": 1.0029,
-      "step": 459
-    },
-    {
-      "epoch": 0.2262666010821446,
-      "grad_norm": 0.5961434170382809,
-      "learning_rate": 9.200000000000001e-05,
-      "loss": 0.9617,
-      "step": 460
-    },
-    {
-      "epoch": 0.2267584849975406,
-      "grad_norm": 0.6314598969589529,
-      "learning_rate": 9.22e-05,
-      "loss": 0.9515,
-      "step": 461
-    },
-    {
-      "epoch": 0.22725036891293654,
-      "grad_norm": 0.5602422225660095,
-      "learning_rate": 9.240000000000001e-05,
-      "loss": 0.9644,
-      "step": 462
-    },
-    {
-      "epoch": 0.22774225282833252,
-      "grad_norm": 0.5977278837906834,
-      "learning_rate": 9.260000000000001e-05,
-      "loss": 0.8666,
-      "step": 463
-    },
-    {
-      "epoch": 0.22823413674372847,
-      "grad_norm": 0.5733762249981974,
-      "learning_rate": 9.28e-05,
-      "loss": 0.9176,
-      "step": 464
-    },
-    {
-      "epoch": 0.22872602065912445,
-      "grad_norm": 0.6065979413259693,
-      "learning_rate": 9.300000000000001e-05,
-      "loss": 0.9304,
-      "step": 465
-    },
-    {
-      "epoch": 0.2292179045745204,
-      "grad_norm": 0.6682669750658685,
-      "learning_rate": 9.320000000000002e-05,
-      "loss": 1.0568,
-      "step": 466
-    },
-    {
-      "epoch": 0.22970978848991638,
-      "grad_norm": 0.6114510975349654,
-      "learning_rate": 9.340000000000001e-05,
-      "loss": 0.9325,
-      "step": 467
-    },
-    {
-      "epoch": 0.23020167240531234,
-      "grad_norm": 0.9915938410046113,
-      "learning_rate": 9.360000000000001e-05,
-      "loss": 0.9733,
-      "step": 468
-    },
-    {
-      "epoch": 0.23069355632070832,
-      "grad_norm": 0.5858992558347075,
-      "learning_rate": 9.38e-05,
-      "loss": 0.9279,
-      "step": 469
-    },
-    {
-      "epoch": 0.23118544023610427,
-      "grad_norm": 0.5521035516591665,
-      "learning_rate": 9.4e-05,
-      "loss": 0.9738,
-      "step": 470
-    },
-    {
-      "epoch": 0.23167732415150025,
-      "grad_norm": 0.5754273126337525,
-      "learning_rate": 9.42e-05,
-      "loss": 0.9224,
-      "step": 471
-    },
-    {
-      "epoch": 0.2321692080668962,
-      "grad_norm": 0.6288424898995382,
-      "learning_rate": 9.44e-05,
-      "loss": 0.934,
-      "step": 472
-    },
-    {
-      "epoch": 0.23266109198229218,
-      "grad_norm": 0.5930926280595387,
-      "learning_rate": 9.46e-05,
-      "loss": 0.9048,
-      "step": 473
-    },
-    {
-      "epoch": 0.23315297589768816,
-      "grad_norm": 1.0573852034903584,
-      "learning_rate": 9.48e-05,
-      "loss": 0.9652,
-      "step": 474
-    },
-    {
-      "epoch": 0.2336448598130841,
-      "grad_norm": 0.7957217544718668,
-      "learning_rate": 9.5e-05,
-      "loss": 1.0241,
-      "step": 475
-    },
-    {
-      "epoch": 0.2341367437284801,
-      "grad_norm": 0.6637946045342709,
-      "learning_rate": 9.52e-05,
-      "loss": 1.0212,
-      "step": 476
-    },
-    {
-      "epoch": 0.23462862764387604,
-      "grad_norm": 0.6091148671336437,
-      "learning_rate": 9.54e-05,
-      "loss": 1.0706,
-      "step": 477
-    },
-    {
-      "epoch": 0.23512051155927202,
-      "grad_norm": 0.5928870673422334,
-      "learning_rate": 9.56e-05,
-      "loss": 1.024,
-      "step": 478
-    },
-    {
-      "epoch": 0.23561239547466797,
-      "grad_norm": 0.6328201225197264,
-      "learning_rate": 9.58e-05,
-      "loss": 1.004,
-      "step": 479
-    },
-    {
-      "epoch": 0.23610427939006395,
-      "grad_norm": 1.3909007846120647,
-      "learning_rate": 9.6e-05,
-      "loss": 1.0172,
-      "step": 480
-    },
-    {
-      "epoch": 0.2365961633054599,
-      "grad_norm": 0.6446533249348028,
-      "learning_rate": 9.620000000000001e-05,
-      "loss": 0.9927,
-      "step": 481
-    },
-    {
-      "epoch": 0.23708804722085589,
-      "grad_norm": 0.7692567910839619,
-      "learning_rate": 9.64e-05,
-      "loss": 0.9427,
-      "step": 482
-    },
-    {
-      "epoch": 0.23757993113625184,
-      "grad_norm": 0.6093984042747421,
-      "learning_rate": 9.66e-05,
-      "loss": 0.9684,
-      "step": 483
-    },
-    {
-      "epoch": 0.23807181505164782,
-      "grad_norm": 1.0005629998863774,
-      "learning_rate": 9.680000000000001e-05,
-      "loss": 0.9163,
-      "step": 484
-    },
-    {
-      "epoch": 0.23856369896704377,
-      "grad_norm": 2.1010922008411193,
-      "learning_rate": 9.7e-05,
-      "loss": 1.0023,
-      "step": 485
-    },
-    {
-      "epoch": 0.23905558288243975,
-      "grad_norm": 0.6795740877841306,
-      "learning_rate": 9.72e-05,
-      "loss": 0.9061,
-      "step": 486
-    },
-    {
-      "epoch": 0.2395474667978357,
-      "grad_norm": 1.3821862397613611,
-      "learning_rate": 9.74e-05,
-      "loss": 1.0303,
-      "step": 487
-    },
-    {
-      "epoch": 0.24003935071323168,
-      "grad_norm": 0.5980079275840284,
-      "learning_rate": 9.76e-05,
-      "loss": 0.8701,
-      "step": 488
-    },
-    {
-      "epoch": 0.24053123462862763,
-      "grad_norm": 0.6332421610937169,
-      "learning_rate": 9.78e-05,
-      "loss": 0.9758,
-      "step": 489
-    },
-    {
-      "epoch": 0.2410231185440236,
-      "grad_norm": 0.5999201682022608,
-      "learning_rate": 9.8e-05,
-      "loss": 0.9127,
-      "step": 490
-    },
-    {
-      "epoch": 0.24151500245941956,
-      "grad_norm": 1.2712794830931637,
-      "learning_rate": 9.82e-05,
-      "loss": 1.062,
-      "step": 491
-    },
-    {
-      "epoch": 0.24200688637481554,
-      "grad_norm": 0.6504036380794753,
-      "learning_rate": 9.84e-05,
-      "loss": 0.9864,
-      "step": 492
-    },
-    {
-      "epoch": 0.24249877029021152,
-      "grad_norm": 0.5651148499618845,
-      "learning_rate": 9.86e-05,
-      "loss": 0.9324,
-      "step": 493
-    },
-    {
-      "epoch": 0.24299065420560748,
-      "grad_norm": 0.5793954144082308,
-      "learning_rate": 9.88e-05,
-      "loss": 1.0368,
-      "step": 494
-    },
-    {
-      "epoch": 0.24348253812100346,
-      "grad_norm": 0.6124134428526112,
-      "learning_rate": 9.900000000000001e-05,
-      "loss": 0.9498,
-      "step": 495
-    },
-    {
-      "epoch": 0.2439744220363994,
-      "grad_norm": 0.5744591301399047,
-      "learning_rate": 9.92e-05,
-      "loss": 0.9139,
-      "step": 496
-    },
-    {
-      "epoch": 0.2444663059517954,
-      "grad_norm": 0.7292901522520773,
-      "learning_rate": 9.94e-05,
-      "loss": 0.96,
-      "step": 497
-    },
-    {
-      "epoch": 0.24495818986719134,
-      "grad_norm": 0.6198598397505529,
-      "learning_rate": 9.960000000000001e-05,
-      "loss": 0.9709,
-      "step": 498
-    },
-    {
-      "epoch": 0.24545007378258732,
-      "grad_norm": 0.7475634835580965,
-      "learning_rate": 9.98e-05,
-      "loss": 0.9326,
-      "step": 499
-    },
-    {
-      "epoch": 0.24594195769798327,
-      "grad_norm": 0.576130778925058,
-      "learning_rate": 0.0001,
-      "loss": 0.9667,
-      "step": 500
-    },
-    {
-      "epoch": 0.24643384161337925,
-      "grad_norm": 1.6126086957373478,
-      "learning_rate": 0.00010020000000000001,
-      "loss": 0.9386,
-      "step": 501
-    },
-    {
-      "epoch": 0.2469257255287752,
-      "grad_norm": 0.6396943451122885,
-      "learning_rate": 0.0001004,
-      "loss": 0.9154,
-      "step": 502
-    },
-    {
-      "epoch": 0.24741760944417118,
-      "grad_norm": 0.593136779240249,
-      "learning_rate": 0.0001006,
-      "loss": 0.9406,
-      "step": 503
-    },
-    {
-      "epoch": 0.24790949335956713,
-      "grad_norm": 0.5942229056137495,
-      "learning_rate": 0.00010080000000000001,
-      "loss": 1.0724,
-      "step": 504
-    },
-    {
-      "epoch": 0.24840137727496311,
-      "grad_norm": 0.5619219133034947,
-      "learning_rate": 0.000101,
-      "loss": 1.0055,
-      "step": 505
-    },
-    {
-      "epoch": 0.24889326119035907,
-      "grad_norm": 0.6672551418770766,
-      "learning_rate": 0.00010120000000000001,
-      "loss": 1.0772,
-      "step": 506
-    },
-    {
-      "epoch": 0.24938514510575505,
-      "grad_norm": 0.5996092644056796,
-      "learning_rate": 0.00010140000000000001,
-      "loss": 0.9409,
-      "step": 507
-    },
-    {
-      "epoch": 0.249877029021151,
-      "grad_norm": 0.5923281147260792,
-      "learning_rate": 0.0001016,
-      "loss": 0.9666,
-      "step": 508
-    },
-    {
-      "epoch": 0.25036891293654695,
-      "grad_norm": 0.636100143332497,
-      "learning_rate": 0.00010180000000000001,
-      "loss": 0.9769,
-      "step": 509
-    },
-    {
-      "epoch": 0.25086079685194296,
-      "grad_norm": 0.6200372518917829,
-      "learning_rate": 0.00010200000000000001,
-      "loss": 1.0639,
-      "step": 510
-    },
-    {
-      "epoch": 0.2513526807673389,
-      "grad_norm": 0.6502389809828354,
-      "learning_rate": 0.0001022,
-      "loss": 0.8989,
-      "step": 511
-    },
-    {
-      "epoch": 0.25184456468273486,
-      "grad_norm": 0.6335856938392092,
-      "learning_rate": 0.00010240000000000001,
-      "loss": 1.0302,
-      "step": 512
-    },
-    {
-      "epoch": 0.2523364485981308,
-      "grad_norm": 0.6054997347827139,
-      "learning_rate": 0.00010260000000000001,
-      "loss": 0.9907,
-      "step": 513
-    },
-    {
-      "epoch": 0.2528283325135268,
-      "grad_norm": 0.570723575147577,
-      "learning_rate": 0.0001028,
-      "loss": 0.9767,
-      "step": 514
-    },
-    {
-      "epoch": 0.2533202164289228,
-      "grad_norm": 0.6220939076031705,
-      "learning_rate": 0.00010300000000000001,
-      "loss": 0.9529,
-      "step": 515
-    },
-    {
-      "epoch": 0.2538121003443187,
-      "grad_norm": 0.6155466649132726,
-      "learning_rate": 0.0001032,
-      "loss": 1.0067,
-      "step": 516
-    },
-    {
-      "epoch": 0.25430398425971473,
-      "grad_norm": 0.8726025671843395,
-      "learning_rate": 0.0001034,
-      "loss": 0.9961,
-      "step": 517
-    },
-    {
-      "epoch": 0.2547958681751107,
-      "grad_norm": 0.6620076482195109,
-      "learning_rate": 0.00010360000000000001,
-      "loss": 0.9331,
-      "step": 518
-    },
-    {
-      "epoch": 0.25528775209050664,
-      "grad_norm": 1.2595302428131354,
-      "learning_rate": 0.0001038,
-      "loss": 0.8981,
-      "step": 519
-    },
-    {
-      "epoch": 0.2557796360059026,
-      "grad_norm": 0.6458264834672185,
-      "learning_rate": 0.00010400000000000001,
-      "loss": 0.9958,
-      "step": 520
-    },
-    {
-      "epoch": 0.2562715199212986,
-      "grad_norm": 0.7635091611089441,
-      "learning_rate": 0.00010420000000000001,
-      "loss": 0.9467,
-      "step": 521
-    },
-    {
-      "epoch": 0.25676340383669455,
-      "grad_norm": 0.714982120373078,
-      "learning_rate": 0.0001044,
-      "loss": 0.9966,
-      "step": 522
-    },
-    {
-      "epoch": 0.2572552877520905,
-      "grad_norm": 0.6425717367298969,
-      "learning_rate": 0.00010460000000000001,
-      "loss": 0.913,
-      "step": 523
-    },
-    {
-      "epoch": 0.25774717166748645,
-      "grad_norm": 0.5721649715746685,
-      "learning_rate": 0.00010480000000000001,
-      "loss": 1.0209,
-      "step": 524
-    },
-    {
-      "epoch": 0.25823905558288246,
-      "grad_norm": 0.7066309254383983,
-      "learning_rate": 0.000105,
-      "loss": 1.0178,
-      "step": 525
-    },
-    {
-      "epoch": 0.2587309394982784,
-      "grad_norm": 0.602685793418877,
-      "learning_rate": 0.00010520000000000001,
-      "loss": 0.9892,
-      "step": 526
-    },
-    {
-      "epoch": 0.25922282341367436,
-      "grad_norm": 0.6250830114089143,
-      "learning_rate": 0.00010540000000000001,
-      "loss": 1.038,
-      "step": 527
-    },
-    {
-      "epoch": 0.2597147073290703,
-      "grad_norm": 0.5996294671164073,
-      "learning_rate": 0.0001056,
-      "loss": 0.9697,
-      "step": 528
-    },
-    {
-      "epoch": 0.2602065912444663,
-      "grad_norm": 0.5670041707916343,
-      "learning_rate": 0.00010580000000000001,
-      "loss": 0.8589,
-      "step": 529
-    },
-    {
-      "epoch": 0.2606984751598623,
-      "grad_norm": 0.5992300105808054,
-      "learning_rate": 0.00010600000000000002,
-      "loss": 0.9784,
-      "step": 530
-    },
-    {
-      "epoch": 0.2611903590752582,
-      "grad_norm": 0.5734369344300806,
-      "learning_rate": 0.0001062,
-      "loss": 0.9789,
-      "step": 531
-    },
-    {
-      "epoch": 0.2616822429906542,
-      "grad_norm": 0.6202809724939895,
-      "learning_rate": 0.00010640000000000001,
-      "loss": 0.9404,
-      "step": 532
-    },
-    {
-      "epoch": 0.2621741269060502,
-      "grad_norm": 0.5708965613496635,
-      "learning_rate": 0.00010660000000000002,
-      "loss": 0.9989,
-      "step": 533
-    },
-    {
-      "epoch": 0.26266601082144614,
-      "grad_norm": 0.6050720664565258,
-      "learning_rate": 0.00010680000000000001,
-      "loss": 1.0119,
-      "step": 534
-    },
-    {
-      "epoch": 0.2631578947368421,
-      "grad_norm": 0.6503025204290864,
-      "learning_rate": 0.00010700000000000001,
-      "loss": 0.8838,
-      "step": 535
-    },
-    {
-      "epoch": 0.2636497786522381,
-      "grad_norm": 0.6003814831427035,
-      "learning_rate": 0.00010720000000000002,
-      "loss": 0.9797,
-      "step": 536
-    },
-    {
-      "epoch": 0.26414166256763405,
-      "grad_norm": 0.5700604678362049,
-      "learning_rate": 0.00010740000000000001,
-      "loss": 0.9529,
-      "step": 537
-    },
-    {
-      "epoch": 0.26463354648303,
-      "grad_norm": 0.6286797156268352,
-      "learning_rate": 0.00010760000000000001,
-      "loss": 1.0507,
-      "step": 538
-    },
-    {
-      "epoch": 0.26512543039842595,
-      "grad_norm": 0.5648484304652618,
-      "learning_rate": 0.00010780000000000002,
-      "loss": 0.9681,
-      "step": 539
-    },
-    {
-      "epoch": 0.26561731431382196,
-      "grad_norm": 0.6284114762154561,
-      "learning_rate": 0.00010800000000000001,
-      "loss": 0.9653,
-      "step": 540
-    },
-    {
-      "epoch": 0.2661091982292179,
-      "grad_norm": 0.5569323701062922,
-      "learning_rate": 0.00010820000000000001,
-      "loss": 0.9292,
-      "step": 541
-    },
-    {
-      "epoch": 0.26660108214461387,
-      "grad_norm": 0.5887770254355695,
-      "learning_rate": 0.00010840000000000002,
-      "loss": 0.9844,
-      "step": 542
-    },
-    {
-      "epoch": 0.2670929660600098,
-      "grad_norm": 0.6248044039199642,
-      "learning_rate": 0.00010860000000000001,
-      "loss": 0.8658,
-      "step": 543
-    },
-    {
-      "epoch": 0.2675848499754058,
-      "grad_norm": 0.662119932700776,
-      "learning_rate": 0.00010880000000000002,
-      "loss": 1.0137,
-      "step": 544
-    },
-    {
-      "epoch": 0.2680767338908018,
-      "grad_norm": 1.262688816523999,
-      "learning_rate": 0.000109,
-      "loss": 0.8925,
-      "step": 545
-    },
-    {
-      "epoch": 0.26856861780619773,
-      "grad_norm": 0.667479850878831,
-      "learning_rate": 0.00010920000000000001,
-      "loss": 1.0127,
-      "step": 546
-    },
-    {
-      "epoch": 0.2690605017215937,
-      "grad_norm": 0.5988917410592548,
-      "learning_rate": 0.00010940000000000002,
-      "loss": 0.9813,
-      "step": 547
-    },
-    {
-      "epoch": 0.2695523856369897,
-      "grad_norm": 0.646787588104879,
-      "learning_rate": 0.00010960000000000001,
-      "loss": 0.968,
-      "step": 548
-    },
-    {
-      "epoch": 0.27004426955238564,
-      "grad_norm": 0.5934709955522586,
-      "learning_rate": 0.00010980000000000001,
-      "loss": 0.9106,
-      "step": 549
-    },
-    {
-      "epoch": 0.2705361534677816,
-      "grad_norm": 0.6890356002678032,
-      "learning_rate": 0.00011000000000000002,
-      "loss": 0.896,
-      "step": 550
-    },
-    {
-      "epoch": 0.27102803738317754,
-      "grad_norm": 0.635952440140494,
-      "learning_rate": 0.00011020000000000001,
-      "loss": 1.0253,
-      "step": 551
-    },
-    {
-      "epoch": 0.27151992129857355,
-      "grad_norm": 0.5480143941867203,
-      "learning_rate": 0.00011040000000000001,
-      "loss": 0.9053,
-      "step": 552
-    },
-    {
-      "epoch": 0.2720118052139695,
-      "grad_norm": 0.5458926973694216,
-      "learning_rate": 0.00011060000000000002,
-      "loss": 0.9718,
-      "step": 553
-    },
-    {
-      "epoch": 0.27250368912936546,
-      "grad_norm": 0.5891788945755541,
-      "learning_rate": 0.00011080000000000001,
-      "loss": 0.937,
-      "step": 554
-    },
-    {
-      "epoch": 0.27299557304476146,
-      "grad_norm": 0.6604709314184619,
-      "learning_rate": 0.00011100000000000001,
-      "loss": 0.941,
-      "step": 555
-    },
-    {
-      "epoch": 0.2734874569601574,
-      "grad_norm": 0.6087029346118886,
-      "learning_rate": 0.00011120000000000002,
-      "loss": 1.0306,
-      "step": 556
-    },
-    {
-      "epoch": 0.27397934087555337,
-      "grad_norm": 0.5801967623798765,
-      "learning_rate": 0.00011140000000000001,
-      "loss": 0.9579,
-      "step": 557
-    },
-    {
-      "epoch": 0.2744712247909493,
-      "grad_norm": 1.7686976341139045,
-      "learning_rate": 0.00011160000000000002,
-      "loss": 0.9981,
-      "step": 558
-    },
-    {
-      "epoch": 0.2749631087063453,
-      "grad_norm": 0.606144888287221,
-      "learning_rate": 0.00011180000000000002,
-      "loss": 0.9892,
-      "step": 559
-    },
-    {
-      "epoch": 0.2754549926217413,
-      "grad_norm": 0.5641112522030708,
-      "learning_rate": 0.00011200000000000001,
-      "loss": 0.9197,
-      "step": 560
-    },
-    {
-      "epoch": 0.27594687653713723,
-      "grad_norm": 0.5348434591364023,
-      "learning_rate": 0.00011220000000000002,
-      "loss": 0.8833,
-      "step": 561
-    },
-    {
-      "epoch": 0.2764387604525332,
-      "grad_norm": 0.5621195777422913,
-      "learning_rate": 0.00011240000000000002,
-      "loss": 0.9392,
-      "step": 562
-    },
-    {
-      "epoch": 0.2769306443679292,
-      "grad_norm": 0.6183412922022422,
-      "learning_rate": 0.0001126,
-      "loss": 0.9989,
-      "step": 563
-    },
-    {
-      "epoch": 0.27742252828332514,
-      "grad_norm": 1.566745653338309,
-      "learning_rate": 0.00011279999999999999,
-      "loss": 0.9692,
-      "step": 564
-    },
-    {
-      "epoch": 0.2779144121987211,
-      "grad_norm": 0.5884266700151978,
-      "learning_rate": 0.000113,
-      "loss": 0.996,
-      "step": 565
-    },
-    {
-      "epoch": 0.27840629611411705,
-      "grad_norm": 0.5941883832990468,
-      "learning_rate": 0.0001132,
-      "loss": 0.8867,
-      "step": 566
-    },
-    {
-      "epoch": 0.27889818002951305,
-      "grad_norm": 0.5663337462783874,
-      "learning_rate": 0.00011339999999999999,
-      "loss": 0.9749,
-      "step": 567
-    },
-    {
-      "epoch": 0.279390063944909,
-      "grad_norm": 0.7083122369348663,
-      "learning_rate": 0.0001136,
-      "loss": 0.983,
-      "step": 568
-    },
-    {
-      "epoch": 0.27988194786030496,
-      "grad_norm": 0.6082393826155303,
-      "learning_rate": 0.0001138,
-      "loss": 0.996,
-      "step": 569
-    },
-    {
-      "epoch": 0.2803738317757009,
-      "grad_norm": 0.5778035016558593,
-      "learning_rate": 0.00011399999999999999,
-      "loss": 0.9854,
-      "step": 570
-    },
-    {
-      "epoch": 0.2808657156910969,
-      "grad_norm": 0.5871357641220516,
-      "learning_rate": 0.0001142,
-      "loss": 0.9928,
-      "step": 571
-    },
-    {
-      "epoch": 0.28135759960649287,
-      "grad_norm": 0.5950779514570566,
-      "learning_rate": 0.0001144,
-      "loss": 0.9732,
-      "step": 572
-    },
-    {
-      "epoch": 0.2818494835218888,
-      "grad_norm": 0.6282290640040025,
-      "learning_rate": 0.0001146,
-      "loss": 0.9407,
-      "step": 573
-    },
-    {
-      "epoch": 0.2823413674372848,
-      "grad_norm": 0.6091611582888863,
-      "learning_rate": 0.0001148,
-      "loss": 0.9945,
-      "step": 574
-    },
-    {
-      "epoch": 0.2828332513526808,
-      "grad_norm": 0.5445033036319282,
-      "learning_rate": 0.00011499999999999999,
-      "loss": 0.9237,
-      "step": 575
-    },
-    {
-      "epoch": 0.28332513526807673,
-      "grad_norm": 0.6028655740862312,
-      "learning_rate": 0.0001152,
-      "loss": 1.0068,
-      "step": 576
-    },
-    {
-      "epoch": 0.2838170191834727,
-      "grad_norm": 0.6450170796685454,
-      "learning_rate": 0.0001154,
-      "loss": 0.9833,
-      "step": 577
-    },
-    {
-      "epoch": 0.2843089030988687,
-      "grad_norm": 0.5774276129265211,
-      "learning_rate": 0.00011559999999999999,
-      "loss": 0.9708,
-      "step": 578
-    },
-    {
-      "epoch": 0.28480078701426464,
-      "grad_norm": 0.5252295207085985,
-      "learning_rate": 0.0001158,
-      "loss": 0.9146,
-      "step": 579
-    },
-    {
-      "epoch": 0.2852926709296606,
-      "grad_norm": 0.5846219170257404,
-      "learning_rate": 0.000116,
-      "loss": 0.9593,
-      "step": 580
-    },
-    {
-      "epoch": 0.28578455484505655,
-      "grad_norm": 0.6406247034473914,
-      "learning_rate": 0.00011619999999999999,
-      "loss": 0.9634,
-      "step": 581
-    },
-    {
-      "epoch": 0.28627643876045256,
-      "grad_norm": 0.5407092545690616,
-      "learning_rate": 0.0001164,
-      "loss": 0.9844,
-      "step": 582
-    },
-    {
-      "epoch": 0.2867683226758485,
-      "grad_norm": 0.5405736655392542,
-      "learning_rate": 0.0001166,
-      "loss": 0.885,
-      "step": 583
-    },
-    {
-      "epoch": 0.28726020659124446,
-      "grad_norm": 0.5635996908391186,
-      "learning_rate": 0.00011679999999999999,
-      "loss": 0.9904,
-      "step": 584
-    },
-    {
-      "epoch": 0.2877520905066404,
-      "grad_norm": 0.5386855826499043,
-      "learning_rate": 0.000117,
-      "loss": 0.9462,
-      "step": 585
-    },
-    {
-      "epoch": 0.2882439744220364,
-      "grad_norm": 0.5625326433463138,
-      "learning_rate": 0.0001172,
-      "loss": 0.9852,
-      "step": 586
-    },
-    {
-      "epoch": 0.28873585833743237,
-      "grad_norm": 0.5754031622934517,
-      "learning_rate": 0.0001174,
-      "loss": 0.9888,
-      "step": 587
-    },
-    {
-      "epoch": 0.2892277422528283,
-      "grad_norm": 0.5602175216543137,
-      "learning_rate": 0.0001176,
-      "loss": 1.0083,
-      "step": 588
-    },
-    {
-      "epoch": 0.2897196261682243,
-      "grad_norm": 0.6661953671814983,
-      "learning_rate": 0.0001178,
-      "loss": 0.9044,
-      "step": 589
-    },
-    {
-      "epoch": 0.2902115100836203,
-      "grad_norm": 0.5607320927918982,
-      "learning_rate": 0.000118,
-      "loss": 1.0375,
-      "step": 590
-    },
-    {
-      "epoch": 0.29070339399901624,
-      "grad_norm": 0.6091870680391808,
-      "learning_rate": 0.0001182,
-      "loss": 0.969,
-      "step": 591
-    },
-    {
-      "epoch": 0.2911952779144122,
-      "grad_norm": 0.6194195534546384,
-      "learning_rate": 0.0001184,
-      "loss": 0.9679,
-      "step": 592
-    },
-    {
-      "epoch": 0.29168716182980814,
-      "grad_norm": 0.5379078512348062,
-      "learning_rate": 0.0001186,
-      "loss": 0.9136,
-      "step": 593
-    },
-    {
-      "epoch": 0.29217904574520415,
-      "grad_norm": 0.5881165054322028,
-      "learning_rate": 0.0001188,
-      "loss": 0.8879,
-      "step": 594
-    },
-    {
-      "epoch": 0.2926709296606001,
-      "grad_norm": 0.5693384812772315,
-      "learning_rate": 0.000119,
-      "loss": 0.9097,
-      "step": 595
-    },
-    {
-      "epoch": 0.29316281357599605,
-      "grad_norm": 0.6220722328965506,
-      "learning_rate": 0.0001192,
-      "loss": 1.0315,
-      "step": 596
-    },
-    {
-      "epoch": 0.29365469749139206,
-      "grad_norm": 0.5885920602717346,
-      "learning_rate": 0.0001194,
-      "loss": 1.0039,
-      "step": 597
-    },
-    {
-      "epoch": 0.294146581406788,
-      "grad_norm": 0.5342119882180294,
-      "learning_rate": 0.00011960000000000001,
-      "loss": 0.9946,
-      "step": 598
-    },
-    {
-      "epoch": 0.29463846532218396,
-      "grad_norm": 0.5578159619400491,
-      "learning_rate": 0.0001198,
-      "loss": 0.975,
-      "step": 599
-    },
-    {
-      "epoch": 0.2951303492375799,
-      "grad_norm": 0.5673528225237494,
-      "learning_rate": 0.00012,
-      "loss": 0.8774,
-      "step": 600
-    },
-    {
-      "epoch": 0.2956222331529759,
-      "grad_norm": 0.6363250119030278,
-      "learning_rate": 0.00012020000000000001,
-      "loss": 0.9999,
-      "step": 601
-    },
-    {
-      "epoch": 0.2961141170683719,
-      "grad_norm": 0.5701824144888316,
-      "learning_rate": 0.0001204,
-      "loss": 1.0054,
-      "step": 602
-    },
-    {
-      "epoch": 0.2966060009837678,
-      "grad_norm": 0.5925396106953534,
-      "learning_rate": 0.0001206,
-      "loss": 0.9801,
-      "step": 603
-    },
-    {
-      "epoch": 0.2970978848991638,
-      "grad_norm": 0.8132860321261537,
-      "learning_rate": 0.0001208,
-      "loss": 0.9329,
-      "step": 604
-    },
-    {
-      "epoch": 0.2975897688145598,
-      "grad_norm": 0.5979623554943858,
-      "learning_rate": 0.000121,
-      "loss": 0.9376,
-      "step": 605
-    },
-    {
-      "epoch": 0.29808165272995574,
-      "grad_norm": 0.5537701303285555,
-      "learning_rate": 0.0001212,
-      "loss": 0.8985,
-      "step": 606
-    },
-    {
-      "epoch": 0.2985735366453517,
-      "grad_norm": 1.1934420587099575,
-      "learning_rate": 0.0001214,
-      "loss": 1.0415,
-      "step": 607
-    },
-    {
-      "epoch": 0.29906542056074764,
-      "grad_norm": 0.6390295391691274,
-      "learning_rate": 0.0001216,
-      "loss": 1.0092,
-      "step": 608
-    },
-    {
-      "epoch": 0.29955730447614365,
-      "grad_norm": 0.5632358051191003,
-      "learning_rate": 0.0001218,
-      "loss": 0.9522,
-      "step": 609
-    },
-    {
-      "epoch": 0.3000491883915396,
-      "grad_norm": 0.5846962460493994,
-      "learning_rate": 0.000122,
-      "loss": 0.9943,
-      "step": 610
-    },
-    {
-      "epoch": 0.30054107230693555,
-      "grad_norm": 0.720362221624888,
-      "learning_rate": 0.00012220000000000002,
-      "loss": 1.0228,
-      "step": 611
-    },
-    {
-      "epoch": 0.3010329562223315,
-      "grad_norm": 0.6473326791386469,
-      "learning_rate": 0.0001224,
-      "loss": 0.9099,
-      "step": 612
-    },
-    {
-      "epoch": 0.3015248401377275,
-      "grad_norm": 0.6463604247118291,
-      "learning_rate": 0.0001226,
-      "loss": 0.9361,
-      "step": 613
-    },
-    {
-      "epoch": 0.30201672405312346,
-      "grad_norm": 0.6032454455578362,
-      "learning_rate": 0.0001228,
-      "loss": 0.9172,
-      "step": 614
-    },
-    {
-      "epoch": 0.3025086079685194,
-      "grad_norm": 0.5975245447925146,
-      "learning_rate": 0.000123,
-      "loss": 0.9364,
-      "step": 615
-    },
-    {
-      "epoch": 0.30300049188391537,
-      "grad_norm": 0.6286030008050292,
-      "learning_rate": 0.0001232,
-      "loss": 0.9243,
-      "step": 616
-    },
-    {
-      "epoch": 0.3034923757993114,
-      "grad_norm": 0.7795705212149114,
-      "learning_rate": 0.00012340000000000002,
-      "loss": 0.8647,
-      "step": 617
-    },
-    {
-      "epoch": 0.30398425971470733,
-      "grad_norm": 0.6254254168046154,
-      "learning_rate": 0.0001236,
-      "loss": 0.9356,
-      "step": 618
-    },
-    {
-      "epoch": 0.3044761436301033,
-      "grad_norm": 0.5790348057840939,
-      "learning_rate": 0.0001238,
-      "loss": 0.9544,
-      "step": 619
-    },
-    {
-      "epoch": 0.3049680275454993,
-      "grad_norm": 0.587364870786678,
-      "learning_rate": 0.000124,
-      "loss": 0.9516,
-      "step": 620
-    },
-    {
-      "epoch": 0.30545991146089524,
-      "grad_norm": 0.5954110385983872,
-      "learning_rate": 0.0001242,
-      "loss": 1.0545,
-      "step": 621
-    },
-    {
-      "epoch": 0.3059517953762912,
-      "grad_norm": 0.5733986106469642,
-      "learning_rate": 0.00012440000000000002,
-      "loss": 0.9221,
-      "step": 622
-    },
-    {
-      "epoch": 0.30644367929168714,
-      "grad_norm": 0.6133244232392819,
-      "learning_rate": 0.0001246,
-      "loss": 0.9729,
-      "step": 623
-    },
-    {
-      "epoch": 0.30693556320708315,
-      "grad_norm": 0.6156606416669301,
-      "learning_rate": 0.0001248,
-      "loss": 1.077,
-      "step": 624
-    },
-    {
-      "epoch": 0.3074274471224791,
-      "grad_norm": 0.7156742684026377,
-      "learning_rate": 0.000125,
-      "loss": 0.9544,
-      "step": 625
-    },
-    {
-      "epoch": 0.30791933103787505,
-      "grad_norm": 0.5891986448979375,
-      "learning_rate": 0.0001252,
-      "loss": 0.9714,
-      "step": 626
-    },
-    {
-      "epoch": 0.308411214953271,
-      "grad_norm": 0.6154447166478924,
-      "learning_rate": 0.0001254,
-      "loss": 1.0105,
-      "step": 627
-    },
-    {
-      "epoch": 0.308903098868667,
-      "grad_norm": 0.5712629776576733,
-      "learning_rate": 0.00012560000000000002,
-      "loss": 0.9743,
-      "step": 628
-    },
-    {
-      "epoch": 0.30939498278406297,
-      "grad_norm": 0.603402460788961,
-      "learning_rate": 0.0001258,
-      "loss": 0.9682,
-      "step": 629
-    },
-    {
-      "epoch": 0.3098868666994589,
-      "grad_norm": 0.7361243548316179,
-      "learning_rate": 0.000126,
-      "loss": 0.9361,
-      "step": 630
-    },
-    {
-      "epoch": 0.31037875061485487,
-      "grad_norm": 0.5982206894600505,
-      "learning_rate": 0.0001262,
-      "loss": 0.9961,
-      "step": 631
-    },
-    {
-      "epoch": 0.3108706345302509,
-      "grad_norm": 0.5686996284709666,
-      "learning_rate": 0.0001264,
-      "loss": 1.0046,
-      "step": 632
-    },
-    {
-      "epoch": 0.31136251844564683,
-      "grad_norm": 0.6957390022948259,
-      "learning_rate": 0.00012660000000000001,
-      "loss": 1.1283,
-      "step": 633
-    },
-    {
-      "epoch": 0.3118544023610428,
-      "grad_norm": 0.5659975015734552,
-      "learning_rate": 0.00012680000000000002,
-      "loss": 1.0171,
-      "step": 634
-    },
-    {
-      "epoch": 0.31234628627643873,
-      "grad_norm": 0.6356091656285305,
-      "learning_rate": 0.000127,
-      "loss": 1.0067,
-      "step": 635
-    },
-    {
-      "epoch": 0.31283817019183474,
-      "grad_norm": 0.5765472435552592,
-      "learning_rate": 0.0001272,
-      "loss": 0.9492,
-      "step": 636
-    },
-    {
-      "epoch": 0.3133300541072307,
-      "grad_norm": 0.593052903331575,
-      "learning_rate": 0.0001274,
-      "loss": 0.9961,
-      "step": 637
-    },
-    {
-      "epoch": 0.31382193802262665,
-      "grad_norm": 0.5737673808109037,
-      "learning_rate": 0.0001276,
-      "loss": 0.8553,
-      "step": 638
-    },
-    {
-      "epoch": 0.31431382193802265,
-      "grad_norm": 0.583876297288899,
-      "learning_rate": 0.00012780000000000002,
-      "loss": 0.9115,
-      "step": 639
-    },
-    {
-      "epoch": 0.3148057058534186,
-      "grad_norm": 0.5507556107242984,
-      "learning_rate": 0.00012800000000000002,
-      "loss": 0.9111,
-      "step": 640
-    },
-    {
-      "epoch": 0.31529758976881456,
-      "grad_norm": 0.5657350366364802,
-      "learning_rate": 0.0001282,
-      "loss": 0.947,
-      "step": 641
-    },
-    {
-      "epoch": 0.3157894736842105,
-      "grad_norm": 0.5809533663037738,
-      "learning_rate": 0.0001284,
-      "loss": 0.9645,
-      "step": 642
-    },
-    {
-      "epoch": 0.3162813575996065,
-      "grad_norm": 0.5346005995438732,
-      "learning_rate": 0.0001286,
-      "loss": 0.916,
-      "step": 643
-    },
-    {
-      "epoch": 0.31677324151500247,
-      "grad_norm": 0.6486732450104193,
-      "learning_rate": 0.00012880000000000001,
-      "loss": 0.9752,
-      "step": 644
-    },
-    {
-      "epoch": 0.3172651254303984,
-      "grad_norm": 0.6276238792109974,
-      "learning_rate": 0.00012900000000000002,
-      "loss": 0.959,
-      "step": 645
-    },
-    {
-      "epoch": 0.3177570093457944,
-      "grad_norm": 0.5786657577934453,
-      "learning_rate": 0.00012920000000000002,
-      "loss": 0.9983,
-      "step": 646
-    },
-    {
-      "epoch": 0.3182488932611904,
-      "grad_norm": 0.600133902239411,
-      "learning_rate": 0.0001294,
-      "loss": 1.1194,
-      "step": 647
-    },
-    {
-      "epoch": 0.31874077717658633,
-      "grad_norm": 0.5756013538106385,
-      "learning_rate": 0.0001296,
-      "loss": 0.9643,
-      "step": 648
-    },
-    {
-      "epoch": 0.3192326610919823,
-      "grad_norm": 0.5724358955757081,
-      "learning_rate": 0.0001298,
-      "loss": 0.9263,
-      "step": 649
-    },
-    {
-      "epoch": 0.31972454500737824,
-      "grad_norm": 0.5865977260259286,
-      "learning_rate": 0.00013000000000000002,
-      "loss": 1.0196,
-      "step": 650
-    },
-    {
-      "epoch": 0.32021642892277424,
-      "grad_norm": 0.5509247698582813,
-      "learning_rate": 0.00013020000000000002,
-      "loss": 0.9469,
-      "step": 651
-    },
-    {
-      "epoch": 0.3207083128381702,
-      "grad_norm": 0.5375325028727678,
-      "learning_rate": 0.0001304,
-      "loss": 0.9322,
-      "step": 652
-    },
-    {
-      "epoch": 0.32120019675356615,
-      "grad_norm": 0.549825036834768,
-      "learning_rate": 0.0001306,
-      "loss": 0.9052,
-      "step": 653
-    },
-    {
-      "epoch": 0.3216920806689621,
-      "grad_norm": 0.7103188525342564,
-      "learning_rate": 0.0001308,
-      "loss": 1.0165,
-      "step": 654
-    },
-    {
-      "epoch": 0.3221839645843581,
-      "grad_norm": 0.5398051571072059,
-      "learning_rate": 0.000131,
-      "loss": 0.9455,
-      "step": 655
-    },
-    {
-      "epoch": 0.32267584849975406,
-      "grad_norm": 0.5815471566735355,
-      "learning_rate": 0.00013120000000000002,
-      "loss": 0.9447,
-      "step": 656
-    },
-    {
-      "epoch": 0.32316773241515,
-      "grad_norm": 0.5417798807684112,
-      "learning_rate": 0.00013140000000000002,
-      "loss": 0.9406,
-      "step": 657
-    },
-    {
-      "epoch": 0.323659616330546,
-      "grad_norm": 0.5301739122946627,
-      "learning_rate": 0.0001316,
-      "loss": 0.9318,
-      "step": 658
-    },
-    {
-      "epoch": 0.32415150024594197,
-      "grad_norm": 0.5534740199402641,
-      "learning_rate": 0.0001318,
-      "loss": 0.964,
-      "step": 659
-    },
-    {
-      "epoch": 0.3246433841613379,
-      "grad_norm": 0.5654722268298933,
-      "learning_rate": 0.000132,
-      "loss": 0.9994,
-      "step": 660
-    },
-    {
-      "epoch": 0.3251352680767339,
-      "grad_norm": 0.5967544855156218,
-      "learning_rate": 0.00013220000000000001,
-      "loss": 1.0159,
-      "step": 661
-    },
-    {
-      "epoch": 0.3256271519921299,
-      "grad_norm": 0.6466246670710497,
-      "learning_rate": 0.00013240000000000002,
-      "loss": 0.9938,
-      "step": 662
-    },
-    {
-      "epoch": 0.32611903590752583,
-      "grad_norm": 0.5539070788172011,
-      "learning_rate": 0.00013260000000000002,
-      "loss": 0.9236,
-      "step": 663
-    },
-    {
-      "epoch": 0.3266109198229218,
-      "grad_norm": 0.5357911496242144,
-      "learning_rate": 0.0001328,
-      "loss": 0.8679,
-      "step": 664
-    },
-    {
-      "epoch": 0.32710280373831774,
-      "grad_norm": 0.5757556362805178,
-      "learning_rate": 0.000133,
-      "loss": 0.9713,
-      "step": 665
-    },
-    {
-      "epoch": 0.32759468765371375,
-      "grad_norm": 1.595516807778794,
-      "learning_rate": 0.0001332,
-      "loss": 1.0152,
-      "step": 666
-    },
-    {
-      "epoch": 0.3280865715691097,
-      "grad_norm": 0.8092171184687358,
-      "learning_rate": 0.00013340000000000002,
-      "loss": 0.9439,
-      "step": 667
-    },
-    {
-      "epoch": 0.32857845548450565,
-      "grad_norm": 0.5547824246192261,
-      "learning_rate": 0.00013360000000000002,
-      "loss": 0.9708,
-      "step": 668
-    },
-    {
-      "epoch": 0.3290703393999016,
-      "grad_norm": 0.5667719349306606,
-      "learning_rate": 0.00013380000000000003,
-      "loss": 0.9921,
-      "step": 669
-    },
-    {
-      "epoch": 0.3295622233152976,
-      "grad_norm": 0.5602335299747019,
-      "learning_rate": 0.000134,
-      "loss": 0.9328,
-      "step": 670
-    },
-    {
-      "epoch": 0.33005410723069356,
-      "grad_norm": 1.542235688252758,
-      "learning_rate": 0.0001342,
-      "loss": 1.0415,
-      "step": 671
-    },
-    {
-      "epoch": 0.3305459911460895,
-      "grad_norm": 0.528859524564198,
-      "learning_rate": 0.00013440000000000001,
-      "loss": 0.9635,
-      "step": 672
-    },
-    {
-      "epoch": 0.33103787506148546,
-      "grad_norm": 0.570686563030655,
-      "learning_rate": 0.00013460000000000002,
-      "loss": 1.0653,
-      "step": 673
-    },
-    {
-      "epoch": 0.3315297589768815,
-      "grad_norm": 0.5205478000916222,
-      "learning_rate": 0.00013480000000000002,
-      "loss": 0.9937,
-      "step": 674
-    },
-    {
-      "epoch": 0.3320216428922774,
-      "grad_norm": 0.5968896498283343,
-      "learning_rate": 0.00013500000000000003,
-      "loss": 0.9731,
-      "step": 675
-    },
-    {
-      "epoch": 0.3325135268076734,
-      "grad_norm": 0.5933439978690628,
-      "learning_rate": 0.0001352,
-      "loss": 0.9518,
-      "step": 676
-    },
-    {
-      "epoch": 0.33300541072306933,
-      "grad_norm": 0.6183079211299322,
-      "learning_rate": 0.0001354,
-      "loss": 0.9279,
-      "step": 677
-    },
-    {
-      "epoch": 0.33349729463846534,
-      "grad_norm": 0.5681382320807932,
-      "learning_rate": 0.00013560000000000002,
-      "loss": 0.9453,
-      "step": 678
-    },
-    {
-      "epoch": 0.3339891785538613,
-      "grad_norm": 0.580113713966502,
-      "learning_rate": 0.00013580000000000002,
-      "loss": 0.9855,
-      "step": 679
-    },
-    {
-      "epoch": 0.33448106246925724,
-      "grad_norm": 0.6179511418209821,
-      "learning_rate": 0.00013600000000000003,
-      "loss": 1.0135,
-      "step": 680
-    },
-    {
-      "epoch": 0.33497294638465325,
-      "grad_norm": 0.6647931623300862,
-      "learning_rate": 0.0001362,
-      "loss": 0.8864,
-      "step": 681
-    },
-    {
-      "epoch": 0.3354648303000492,
-      "grad_norm": 0.5474827309408606,
-      "learning_rate": 0.0001364,
-      "loss": 0.9699,
-      "step": 682
-    },
-    {
-      "epoch": 0.33595671421544515,
-      "grad_norm": 0.71983895648374,
-      "learning_rate": 0.0001366,
-      "loss": 1.0405,
-      "step": 683
-    },
-    {
-      "epoch": 0.3364485981308411,
-      "grad_norm": 0.5996720854516058,
-      "learning_rate": 0.00013680000000000002,
-      "loss": 0.9509,
-      "step": 684
-    },
-    {
-      "epoch": 0.3369404820462371,
-      "grad_norm": 0.5638459937753646,
-      "learning_rate": 0.00013700000000000002,
-      "loss": 0.9376,
-      "step": 685
-    },
-    {
-      "epoch": 0.33743236596163306,
-      "grad_norm": 0.6071855878622703,
-      "learning_rate": 0.00013720000000000003,
-      "loss": 1.0412,
-      "step": 686
-    },
-    {
-      "epoch": 0.337924249877029,
-      "grad_norm": 0.6709708960381234,
-      "learning_rate": 0.0001374,
-      "loss": 0.9346,
-      "step": 687
-    },
-    {
-      "epoch": 0.33841613379242497,
-      "grad_norm": 0.5882971338580069,
-      "learning_rate": 0.00013759999999999998,
-      "loss": 1.0748,
-      "step": 688
-    },
-    {
-      "epoch": 0.338908017707821,
-      "grad_norm": 0.6895579608375398,
-      "learning_rate": 0.0001378,
-      "loss": 0.9703,
-      "step": 689
-    },
-    {
-      "epoch": 0.3393999016232169,
-      "grad_norm": 0.5549688261016558,
-      "learning_rate": 0.000138,
-      "loss": 0.8929,
-      "step": 690
-    },
-    {
-      "epoch": 0.3398917855386129,
-      "grad_norm": 0.5963572160505828,
-      "learning_rate": 0.0001382,
-      "loss": 1.0304,
-      "step": 691
-    },
-    {
-      "epoch": 0.34038366945400883,
-      "grad_norm": 0.6198648727259731,
-      "learning_rate": 0.0001384,
-      "loss": 0.9831,
-      "step": 692
-    },
-    {
-      "epoch": 0.34087555336940484,
-      "grad_norm": 0.5719964600673073,
-      "learning_rate": 0.0001386,
-      "loss": 0.9806,
-      "step": 693
-    },
-    {
-      "epoch": 0.3413674372848008,
-      "grad_norm": 0.5731736669696705,
-      "learning_rate": 0.00013879999999999999,
-      "loss": 0.8825,
-      "step": 694
-    },
-    {
-      "epoch": 0.34185932120019674,
-      "grad_norm": 0.5419254295863805,
-      "learning_rate": 0.000139,
-      "loss": 0.8876,
-      "step": 695
-    },
-    {
-      "epoch": 0.3423512051155927,
-      "grad_norm": 0.5480778481079349,
-      "learning_rate": 0.0001392,
-      "loss": 0.8689,
-      "step": 696
-    },
-    {
-      "epoch": 0.3428430890309887,
-      "grad_norm": 0.6061965287289426,
-      "learning_rate": 0.0001394,
-      "loss": 0.9779,
-      "step": 697
-    },
-    {
-      "epoch": 0.34333497294638465,
-      "grad_norm": 0.56418799698156,
-      "learning_rate": 0.0001396,
-      "loss": 0.9965,
-      "step": 698
-    },
-    {
-      "epoch": 0.3438268568617806,
-      "grad_norm": 0.5474847123144074,
-      "learning_rate": 0.0001398,
-      "loss": 0.9504,
-      "step": 699
-    },
-    {
-      "epoch": 0.3443187407771766,
-      "grad_norm": 0.6020447593211263,
-      "learning_rate": 0.00014,
-      "loss": 0.9916,
-      "step": 700
-    },
-    {
-      "epoch": 0.34481062469257256,
-      "grad_norm": 0.5716767111427228,
-      "learning_rate": 0.0001402,
-      "loss": 0.9785,
-      "step": 701
-    },
-    {
-      "epoch": 0.3453025086079685,
-      "grad_norm": 0.53738356089577,
-      "learning_rate": 0.0001404,
-      "loss": 0.913,
-      "step": 702
-    },
-    {
-      "epoch": 0.34579439252336447,
-      "grad_norm": 0.5533695050238833,
-      "learning_rate": 0.0001406,
-      "loss": 0.8827,
-      "step": 703
-    },
-    {
-      "epoch": 0.3462862764387605,
-      "grad_norm": 0.521861791445525,
-      "learning_rate": 0.0001408,
-      "loss": 0.963,
-      "step": 704
-    },
-    {
-      "epoch": 0.34677816035415643,
-      "grad_norm": 0.5467343292585995,
-      "learning_rate": 0.000141,
-      "loss": 0.9585,
-      "step": 705
-    },
-    {
-      "epoch": 0.3472700442695524,
-      "grad_norm": 0.6613807586137288,
-      "learning_rate": 0.0001412,
-      "loss": 0.9471,
-      "step": 706
-    },
-    {
-      "epoch": 0.34776192818494833,
-      "grad_norm": 0.5387654674317737,
-      "learning_rate": 0.0001414,
-      "loss": 0.894,
-      "step": 707
-    },
-    {
-      "epoch": 0.34825381210034434,
-      "grad_norm": 0.6187900254092165,
-      "learning_rate": 0.0001416,
-      "loss": 1.0506,
-      "step": 708
-    },
-    {
-      "epoch": 0.3487456960157403,
-      "grad_norm": 0.5842576555732705,
-      "learning_rate": 0.0001418,
-      "loss": 0.972,
-      "step": 709
-    },
-    {
-      "epoch": 0.34923757993113624,
-      "grad_norm": 0.5471657682259135,
-      "learning_rate": 0.000142,
-      "loss": 0.9363,
-      "step": 710
-    },
-    {
-      "epoch": 0.3497294638465322,
-      "grad_norm": 0.6713128120037045,
-      "learning_rate": 0.0001422,
-      "loss": 1.0076,
-      "step": 711
-    },
-    {
-      "epoch": 0.3502213477619282,
-      "grad_norm": 0.562939754525997,
-      "learning_rate": 0.0001424,
-      "loss": 0.994,
-      "step": 712
-    },
-    {
-      "epoch": 0.35071323167732416,
-      "grad_norm": 0.5427219860424152,
-      "learning_rate": 0.0001426,
-      "loss": 0.9664,
-      "step": 713
-    },
-    {
-      "epoch": 0.3512051155927201,
-      "grad_norm": 0.5336172312438766,
-      "learning_rate": 0.0001428,
-      "loss": 0.8873,
-      "step": 714
-    },
-    {
-      "epoch": 0.35169699950811606,
-      "grad_norm": 0.5759279718130945,
-      "learning_rate": 0.000143,
-      "loss": 0.994,
-      "step": 715
-    },
-    {
-      "epoch": 0.35218888342351207,
-      "grad_norm": 0.6210830795173621,
-      "learning_rate": 0.0001432,
-      "loss": 1.0437,
-      "step": 716
-    },
-    {
-      "epoch": 0.352680767338908,
-      "grad_norm": 0.5279473344404623,
-      "learning_rate": 0.0001434,
-      "loss": 0.8987,
-      "step": 717
-    },
-    {
-      "epoch": 0.35317265125430397,
-      "grad_norm": 0.575646122945982,
-      "learning_rate": 0.0001436,
-      "loss": 0.9883,
-      "step": 718
-    },
-    {
-      "epoch": 0.3536645351697,
-      "grad_norm": 0.5866741707982973,
-      "learning_rate": 0.0001438,
-      "loss": 0.9654,
-      "step": 719
-    },
-    {
-      "epoch": 0.35415641908509593,
-      "grad_norm": 0.8804767483817335,
-      "learning_rate": 0.000144,
-      "loss": 1.0005,
-      "step": 720
-    },
-    {
-      "epoch": 0.3546483030004919,
-      "grad_norm": 0.49721230503684766,
-      "learning_rate": 0.0001442,
-      "loss": 0.8645,
-      "step": 721
-    },
-    {
-      "epoch": 0.35514018691588783,
-      "grad_norm": 0.6135127465838539,
-      "learning_rate": 0.0001444,
-      "loss": 0.9552,
-      "step": 722
-    },
-    {
-      "epoch": 0.35563207083128384,
-      "grad_norm": 0.6006179727170639,
-      "learning_rate": 0.0001446,
-      "loss": 0.9883,
-      "step": 723
-    },
-    {
-      "epoch": 0.3561239547466798,
-      "grad_norm": 0.5791762925281315,
-      "learning_rate": 0.0001448,
-      "loss": 1.0127,
-      "step": 724
-    },
-    {
-      "epoch": 0.35661583866207575,
-      "grad_norm": 0.6211190579227377,
-      "learning_rate": 0.000145,
-      "loss": 0.9896,
-      "step": 725
-    },
-    {
-      "epoch": 0.3571077225774717,
-      "grad_norm": 0.5927954254139796,
-      "learning_rate": 0.0001452,
-      "loss": 0.9654,
-      "step": 726
-    },
-    {
-      "epoch": 0.3575996064928677,
-      "grad_norm": 0.5627595411888491,
-      "learning_rate": 0.0001454,
-      "loss": 0.9103,
-      "step": 727
-    },
-    {
-      "epoch": 0.35809149040826366,
-      "grad_norm": 0.7494298053751162,
-      "learning_rate": 0.00014560000000000002,
-      "loss": 1.0005,
-      "step": 728
-    },
-    {
-      "epoch": 0.3585833743236596,
-      "grad_norm": 0.536368405130208,
-      "learning_rate": 0.0001458,
-      "loss": 0.8954,
-      "step": 729
-    },
-    {
-      "epoch": 0.35907525823905556,
-      "grad_norm": 0.5323406282397959,
-      "learning_rate": 0.000146,
-      "loss": 0.9857,
-      "step": 730
-    },
-    {
-      "epoch": 0.35956714215445157,
-      "grad_norm": 0.5674356330933522,
-      "learning_rate": 0.0001462,
-      "loss": 0.927,
-      "step": 731
-    },
-    {
-      "epoch": 0.3600590260698475,
-      "grad_norm": 0.6209798732214817,
-      "learning_rate": 0.0001464,
-      "loss": 1.0156,
-      "step": 732
-    },
-    {
-      "epoch": 0.3605509099852435,
-      "grad_norm": 0.5546862564497353,
-      "learning_rate": 0.0001466,
-      "loss": 0.8843,
-      "step": 733
-    },
-    {
-      "epoch": 0.3610427939006394,
-      "grad_norm": 0.533114061617224,
-      "learning_rate": 0.00014680000000000002,
-      "loss": 0.8848,
-      "step": 734
-    },
-    {
-      "epoch": 0.36153467781603543,
-      "grad_norm": 0.6350819782389399,
-      "learning_rate": 0.000147,
-      "loss": 0.9055,
-      "step": 735
-    },
-    {
-      "epoch": 0.3620265617314314,
-      "grad_norm": 0.5678788134404464,
-      "learning_rate": 0.0001472,
-      "loss": 0.9804,
-      "step": 736
-    },
-    {
-      "epoch": 0.36251844564682734,
-      "grad_norm": 0.5760551838489592,
-      "learning_rate": 0.0001474,
-      "loss": 0.9949,
-      "step": 737
-    },
-    {
-      "epoch": 0.3630103295622233,
-      "grad_norm": 2.2923759269385813,
-      "learning_rate": 0.0001476,
-      "loss": 0.9776,
-      "step": 738
-    },
-    {
-      "epoch": 0.3635022134776193,
-      "grad_norm": 0.632221799207917,
-      "learning_rate": 0.00014780000000000001,
-      "loss": 1.0221,
-      "step": 739
-    },
-    {
-      "epoch": 0.36399409739301525,
-      "grad_norm": 0.6136556819957433,
-      "learning_rate": 0.000148,
-      "loss": 1.0076,
-      "step": 740
-    },
-    {
-      "epoch": 0.3644859813084112,
-      "grad_norm": 0.5539084756184474,
-      "learning_rate": 0.0001482,
-      "loss": 0.9471,
-      "step": 741
-    },
-    {
-      "epoch": 0.3649778652238072,
-      "grad_norm": 0.600388803994492,
-      "learning_rate": 0.0001484,
-      "loss": 1.0066,
-      "step": 742
-    },
-    {
-      "epoch": 0.36546974913920316,
-      "grad_norm": 0.5904080160729892,
-      "learning_rate": 0.0001486,
-      "loss": 0.9935,
-      "step": 743
-    },
-    {
-      "epoch": 0.3659616330545991,
-      "grad_norm": 0.6062535085595192,
-      "learning_rate": 0.0001488,
-      "loss": 0.9043,
-      "step": 744
-    },
-    {
-      "epoch": 0.36645351696999506,
-      "grad_norm": 0.581266851539047,
-      "learning_rate": 0.00014900000000000002,
-      "loss": 0.9436,
-      "step": 745
-    },
-    {
-      "epoch": 0.36694540088539107,
-      "grad_norm": 0.5531548464262217,
-      "learning_rate": 0.0001492,
-      "loss": 0.9334,
-      "step": 746
-    },
-    {
-      "epoch": 0.367437284800787,
-      "grad_norm": 0.8283535851545999,
-      "learning_rate": 0.0001494,
-      "loss": 0.9869,
-      "step": 747
-    },
-    {
-      "epoch": 0.367929168716183,
-      "grad_norm": 0.5998592619281596,
-      "learning_rate": 0.0001496,
-      "loss": 1.0054,
-      "step": 748
-    },
-    {
-      "epoch": 0.3684210526315789,
-      "grad_norm": 0.5254187775306348,
-      "learning_rate": 0.0001498,
-      "loss": 0.8801,
-      "step": 749
-    },
-    {
-      "epoch": 0.36891293654697493,
-      "grad_norm": 0.5840307586056854,
-      "learning_rate": 0.00015000000000000001,
-      "loss": 1.0775,
-      "step": 750
-    },
-    {
-      "epoch": 0.3694048204623709,
-      "grad_norm": 0.5548845893528461,
-      "learning_rate": 0.00015020000000000002,
-      "loss": 0.8954,
-      "step": 751
-    },
-    {
-      "epoch": 0.36989670437776684,
-      "grad_norm": 0.4779657950486935,
-      "learning_rate": 0.0001504,
-      "loss": 0.8264,
-      "step": 752
-    },
-    {
-      "epoch": 0.3703885882931628,
-      "grad_norm": 0.5680652462758233,
-      "learning_rate": 0.0001506,
-      "loss": 1.0293,
-      "step": 753
-    },
-    {
-      "epoch": 0.3708804722085588,
-      "grad_norm": 0.5506113131006278,
-      "learning_rate": 0.0001508,
-      "loss": 0.9605,
-      "step": 754
-    },
-    {
-      "epoch": 0.37137235612395475,
-      "grad_norm": 0.5241419248314438,
-      "learning_rate": 0.000151,
-      "loss": 0.95,
-      "step": 755
-    },
-    {
-      "epoch": 0.3718642400393507,
-      "grad_norm": 0.5548976086666146,
-      "learning_rate": 0.00015120000000000002,
-      "loss": 0.9271,
-      "step": 756
-    },
-    {
-      "epoch": 0.37235612395474665,
-      "grad_norm": 0.545441963837688,
-      "learning_rate": 0.00015140000000000002,
-      "loss": 0.9555,
-      "step": 757
-    },
-    {
-      "epoch": 0.37284800787014266,
-      "grad_norm": 0.51534507000048,
-      "learning_rate": 0.0001516,
-      "loss": 0.8994,
-      "step": 758
-    },
-    {
-      "epoch": 0.3733398917855386,
-      "grad_norm": 0.5373985678785738,
-      "learning_rate": 0.0001518,
-      "loss": 1.0127,
-      "step": 759
-    },
-    {
-      "epoch": 0.37383177570093457,
-      "grad_norm": 0.5746685531454623,
-      "learning_rate": 0.000152,
-      "loss": 1.0257,
-      "step": 760
-    },
-    {
-      "epoch": 0.3743236596163306,
-      "grad_norm": 0.503979910793055,
-      "learning_rate": 0.0001522,
-      "loss": 0.9056,
-      "step": 761
-    },
-    {
-      "epoch": 0.3748155435317265,
-      "grad_norm": 0.5913604557559807,
-      "learning_rate": 0.00015240000000000002,
-      "loss": 1.0301,
-      "step": 762
-    },
-    {
-      "epoch": 0.3753074274471225,
-      "grad_norm": 0.5398864774426727,
-      "learning_rate": 0.00015260000000000002,
-      "loss": 0.9607,
-      "step": 763
-    },
-    {
-      "epoch": 0.37579931136251843,
-      "grad_norm": 0.5495286862756644,
-      "learning_rate": 0.0001528,
-      "loss": 0.9597,
-      "step": 764
-    },
-    {
-      "epoch": 0.37629119527791444,
-      "grad_norm": 0.5907199039930229,
-      "learning_rate": 0.000153,
-      "loss": 0.8879,
-      "step": 765
-    },
-    {
-      "epoch": 0.3767830791933104,
-      "grad_norm": 0.6058728144562922,
-      "learning_rate": 0.0001532,
-      "loss": 1.0295,
-      "step": 766
-    },
-    {
-      "epoch": 0.37727496310870634,
-      "grad_norm": 0.5535252317067916,
-      "learning_rate": 0.00015340000000000002,
-      "loss": 0.9721,
-      "step": 767
-    },
-    {
-      "epoch": 0.3777668470241023,
-      "grad_norm": 0.5689401640933246,
-      "learning_rate": 0.00015360000000000002,
-      "loss": 0.9659,
-      "step": 768
-    },
-    {
-      "epoch": 0.3782587309394983,
-      "grad_norm": 0.5492311772960732,
-      "learning_rate": 0.0001538,
-      "loss": 1.0209,
-      "step": 769
-    },
-    {
-      "epoch": 0.37875061485489425,
-      "grad_norm": 0.547940922474521,
-      "learning_rate": 0.000154,
-      "loss": 0.989,
-      "step": 770
-    },
-    {
-      "epoch": 0.3792424987702902,
-      "grad_norm": 0.50370578360955,
-      "learning_rate": 0.0001542,
-      "loss": 0.9145,
-      "step": 771
-    },
-    {
-      "epoch": 0.37973438268568616,
-      "grad_norm": 0.5633163359060936,
-      "learning_rate": 0.0001544,
-      "loss": 0.9606,
-      "step": 772
-    },
-    {
-      "epoch": 0.38022626660108216,
-      "grad_norm": 0.5672102348479715,
-      "learning_rate": 0.00015460000000000002,
-      "loss": 1.0797,
-      "step": 773
-    },
-    {
-      "epoch": 0.3807181505164781,
-      "grad_norm": 0.48875960371628935,
-      "learning_rate": 0.00015480000000000002,
-      "loss": 0.8897,
-      "step": 774
-    },
-    {
-      "epoch": 0.38121003443187407,
-      "grad_norm": 0.5183070695492531,
-      "learning_rate": 0.000155,
-      "loss": 0.9114,
-      "step": 775
-    },
-    {
-      "epoch": 0.38170191834727,
-      "grad_norm": 0.5260847222140161,
-      "learning_rate": 0.0001552,
-      "loss": 1.0111,
-      "step": 776
-    },
-    {
-      "epoch": 0.382193802262666,
-      "grad_norm": 0.4873770357123154,
-      "learning_rate": 0.0001554,
-      "loss": 0.8981,
-      "step": 777
-    },
-    {
-      "epoch": 0.382685686178062,
-      "grad_norm": 0.533184635208334,
-      "learning_rate": 0.00015560000000000001,
-      "loss": 0.912,
-      "step": 778
-    },
-    {
-      "epoch": 0.38317757009345793,
-      "grad_norm": 0.5726377304478081,
-      "learning_rate": 0.00015580000000000002,
-      "loss": 1.0016,
-      "step": 779
-    },
-    {
-      "epoch": 0.3836694540088539,
-      "grad_norm": 0.5699433809365081,
-      "learning_rate": 0.00015600000000000002,
-      "loss": 1.0063,
-      "step": 780
-    },
-    {
-      "epoch": 0.3841613379242499,
-      "grad_norm": 0.52631644676918,
-      "learning_rate": 0.0001562,
-      "loss": 0.9563,
-      "step": 781
-    },
-    {
-      "epoch": 0.38465322183964584,
-      "grad_norm": 0.5634430508864812,
-      "learning_rate": 0.0001564,
-      "loss": 1.0055,
-      "step": 782
-    },
-    {
-      "epoch": 0.3851451057550418,
-      "grad_norm": 0.5221500631664022,
-      "learning_rate": 0.0001566,
-      "loss": 1.0167,
-      "step": 783
-    },
-    {
-      "epoch": 0.3856369896704378,
-      "grad_norm": 0.5654864366000731,
-      "learning_rate": 0.00015680000000000002,
-      "loss": 0.9569,
-      "step": 784
-    },
-    {
-      "epoch": 0.38612887358583375,
-      "grad_norm": 0.5180728219205175,
-      "learning_rate": 0.00015700000000000002,
-      "loss": 0.9176,
-      "step": 785
-    },
-    {
-      "epoch": 0.3866207575012297,
-      "grad_norm": 0.538064697836738,
-      "learning_rate": 0.00015720000000000003,
-      "loss": 0.9627,
-      "step": 786
-    },
-    {
-      "epoch": 0.38711264141662566,
-      "grad_norm": 0.5519014505895645,
-      "learning_rate": 0.0001574,
-      "loss": 0.986,
-      "step": 787
-    },
-    {
-      "epoch": 0.38760452533202167,
-      "grad_norm": 0.5921326184710332,
-      "learning_rate": 0.0001576,
-      "loss": 1.1092,
-      "step": 788
-    },
-    {
-      "epoch": 0.3880964092474176,
-      "grad_norm": 0.5365549041365112,
-      "learning_rate": 0.00015780000000000001,
-      "loss": 0.9762,
-      "step": 789
-    },
-    {
-      "epoch": 0.38858829316281357,
-      "grad_norm": 0.5285861559667572,
-      "learning_rate": 0.00015800000000000002,
-      "loss": 1.0393,
-      "step": 790
-    },
-    {
-      "epoch": 0.3890801770782095,
-      "grad_norm": 0.5353608332868288,
-      "learning_rate": 0.00015820000000000002,
-      "loss": 0.9721,
-      "step": 791
-    },
-    {
-      "epoch": 0.38957206099360553,
-      "grad_norm": 0.5676796635103992,
-      "learning_rate": 0.00015840000000000003,
-      "loss": 0.9707,
-      "step": 792
-    },
-    {
-      "epoch": 0.3900639449090015,
-      "grad_norm": 0.5324974277172961,
-      "learning_rate": 0.0001586,
-      "loss": 1.0181,
-      "step": 793
-    },
-    {
-      "epoch": 0.39055582882439743,
-      "grad_norm": 0.5444144065736729,
-      "learning_rate": 0.0001588,
-      "loss": 1.0322,
-      "step": 794
-    },
-    {
-      "epoch": 0.3910477127397934,
-      "grad_norm": 0.5216984893352097,
-      "learning_rate": 0.00015900000000000002,
-      "loss": 0.9468,
-      "step": 795
-    },
-    {
-      "epoch": 0.3915395966551894,
-      "grad_norm": 0.5118829818178605,
-      "learning_rate": 0.00015920000000000002,
-      "loss": 0.9302,
-      "step": 796
-    },
-    {
-      "epoch": 0.39203148057058534,
-      "grad_norm": 0.5560684653179324,
-      "learning_rate": 0.00015940000000000003,
-      "loss": 1.0283,
-      "step": 797
-    },
-    {
-      "epoch": 0.3925233644859813,
-      "grad_norm": 0.5137964259683437,
-      "learning_rate": 0.0001596,
-      "loss": 0.9219,
-      "step": 798
-    },
-    {
-      "epoch": 0.39301524840137725,
-      "grad_norm": 0.5972357390162277,
-      "learning_rate": 0.0001598,
-      "loss": 0.8816,
-      "step": 799
-    },
-    {
-      "epoch": 0.39350713231677326,
-      "grad_norm": 0.5731293092568074,
-      "learning_rate": 0.00016,
-      "loss": 1.0413,
-      "step": 800
-    },
-    {
-      "epoch": 0.3939990162321692,
-      "grad_norm": 0.5832125955169386,
-      "learning_rate": 0.00016020000000000002,
-      "loss": 0.9656,
-      "step": 801
-    },
-    {
-      "epoch": 0.39449090014756516,
-      "grad_norm": 0.5703505863315125,
-      "learning_rate": 0.00016040000000000002,
-      "loss": 0.9717,
-      "step": 802
-    },
-    {
-      "epoch": 0.39498278406296117,
-      "grad_norm": 0.5233933183593751,
-      "learning_rate": 0.00016060000000000003,
-      "loss": 0.986,
-      "step": 803
-    },
-    {
-      "epoch": 0.3954746679783571,
-      "grad_norm": 0.5239916997866086,
-      "learning_rate": 0.0001608,
-      "loss": 0.9461,
-      "step": 804
-    },
-    {
-      "epoch": 0.39596655189375307,
-      "grad_norm": 0.5361896576371097,
-      "learning_rate": 0.000161,
-      "loss": 0.931,
-      "step": 805
-    },
-    {
-      "epoch": 0.396458435809149,
-      "grad_norm": 1.84624704067933,
-      "learning_rate": 0.00016120000000000002,
-      "loss": 1.07,
-      "step": 806
-    },
-    {
-      "epoch": 0.39695031972454503,
-      "grad_norm": 0.6194161425941437,
-      "learning_rate": 0.00016140000000000002,
-      "loss": 0.9299,
-      "step": 807
-    },
-    {
-      "epoch": 0.397442203639941,
-      "grad_norm": 0.5598157158310759,
-      "learning_rate": 0.00016160000000000002,
-      "loss": 0.8952,
-      "step": 808
-    },
-    {
-      "epoch": 0.39793408755533693,
-      "grad_norm": 0.5147892749426699,
-      "learning_rate": 0.00016180000000000003,
-      "loss": 0.9648,
-      "step": 809
-    },
-    {
-      "epoch": 0.3984259714707329,
-      "grad_norm": 0.53381884733348,
-      "learning_rate": 0.000162,
-      "loss": 0.965,
-      "step": 810
-    },
-    {
-      "epoch": 0.3989178553861289,
-      "grad_norm": 0.5228813426312591,
-      "learning_rate": 0.0001622,
-      "loss": 0.9231,
-      "step": 811
-    },
-    {
-      "epoch": 0.39940973930152485,
-      "grad_norm": 0.5837534109522405,
-      "learning_rate": 0.00016240000000000002,
-      "loss": 0.9328,
-      "step": 812
-    },
-    {
-      "epoch": 0.3999016232169208,
-      "grad_norm": 0.5275571830323403,
-      "learning_rate": 0.0001626,
-      "loss": 0.8928,
-      "step": 813
-    },
-    {
-      "epoch": 0.40039350713231675,
-      "grad_norm": 0.5876104356865259,
-      "learning_rate": 0.0001628,
-      "loss": 1.0121,
-      "step": 814
-    },
-    {
-      "epoch": 0.40088539104771276,
-      "grad_norm": 0.6382169634620648,
-      "learning_rate": 0.000163,
-      "loss": 0.9843,
-      "step": 815
-    },
-    {
-      "epoch": 0.4013772749631087,
-      "grad_norm": 0.5801866766826601,
-      "learning_rate": 0.0001632,
-      "loss": 1.0471,
-      "step": 816
-    },
-    {
-      "epoch": 0.40186915887850466,
-      "grad_norm": 0.5510639872096187,
-      "learning_rate": 0.0001634,
-      "loss": 1.0638,
-      "step": 817
-    },
-    {
-      "epoch": 0.4023610427939006,
-      "grad_norm": 0.5140283673856436,
-      "learning_rate": 0.0001636,
-      "loss": 0.9217,
-      "step": 818
-    },
-    {
-      "epoch": 0.4028529267092966,
-      "grad_norm": 0.7924911609540576,
-      "learning_rate": 0.0001638,
-      "loss": 0.9755,
-      "step": 819
-    },
-    {
-      "epoch": 0.4033448106246926,
-      "grad_norm": 0.5331762791413481,
-      "learning_rate": 0.000164,
-      "loss": 0.9539,
-      "step": 820
-    },
-    {
-      "epoch": 0.4038366945400885,
-      "grad_norm": 0.5966286424829558,
-      "learning_rate": 0.0001642,
-      "loss": 0.8722,
-      "step": 821
-    },
-    {
-      "epoch": 0.40432857845548453,
-      "grad_norm": 1.309276772968705,
-      "learning_rate": 0.0001644,
-      "loss": 1.0592,
-      "step": 822
-    },
-    {
-      "epoch": 0.4048204623708805,
-      "grad_norm": 0.5380584739269623,
-      "learning_rate": 0.0001646,
-      "loss": 0.9616,
-      "step": 823
-    },
-    {
-      "epoch": 0.40531234628627644,
-      "grad_norm": 0.5424035960763122,
-      "learning_rate": 0.0001648,
-      "loss": 0.9627,
-      "step": 824
-    },
-    {
-      "epoch": 0.4058042302016724,
-      "grad_norm": 0.5538177464111131,
-      "learning_rate": 0.000165,
-      "loss": 1.0005,
-      "step": 825
-    },
-    {
-      "epoch": 0.4062961141170684,
-      "grad_norm": 0.6408546066646377,
-      "learning_rate": 0.0001652,
-      "loss": 0.8947,
-      "step": 826
-    },
-    {
-      "epoch": 0.40678799803246435,
-      "grad_norm": 0.6640867349404851,
-      "learning_rate": 0.0001654,
-      "loss": 1.0078,
-      "step": 827
-    },
-    {
-      "epoch": 0.4072798819478603,
-      "grad_norm": 0.6337291990645043,
-      "learning_rate": 0.0001656,
-      "loss": 1.0795,
-      "step": 828
-    },
-    {
-      "epoch": 0.40777176586325625,
-      "grad_norm": 0.5994443818623946,
-      "learning_rate": 0.0001658,
-      "loss": 1.0258,
-      "step": 829
-    },
-    {
-      "epoch": 0.40826364977865226,
-      "grad_norm": 0.6256523434556119,
-      "learning_rate": 0.000166,
-      "loss": 1.0448,
-      "step": 830
-    },
-    {
-      "epoch": 0.4087555336940482,
-      "grad_norm": 0.5442887937180465,
-      "learning_rate": 0.0001662,
-      "loss": 0.9358,
-      "step": 831
-    },
-    {
-      "epoch": 0.40924741760944416,
-      "grad_norm": 0.8138607813124575,
-      "learning_rate": 0.0001664,
-      "loss": 0.9839,
-      "step": 832
-    },
-    {
-      "epoch": 0.4097393015248401,
-      "grad_norm": 0.547143371707003,
-      "learning_rate": 0.0001666,
-      "loss": 0.9262,
-      "step": 833
-    },
-    {
-      "epoch": 0.4102311854402361,
-      "grad_norm": 0.5750895604971006,
-      "learning_rate": 0.0001668,
-      "loss": 1.0468,
-      "step": 834
-    },
-    {
-      "epoch": 0.4107230693556321,
-      "grad_norm": 0.5904281252379038,
-      "learning_rate": 0.000167,
-      "loss": 0.963,
-      "step": 835
-    },
-    {
-      "epoch": 0.411214953271028,
-      "grad_norm": 0.5016551397659511,
-      "learning_rate": 0.0001672,
-      "loss": 0.9286,
-      "step": 836
-    },
-    {
-      "epoch": 0.411706837186424,
-      "grad_norm": 0.5351125384196013,
-      "learning_rate": 0.0001674,
-      "loss": 0.9714,
-      "step": 837
-    },
-    {
-      "epoch": 0.41219872110182,
-      "grad_norm": 0.5382761575591485,
-      "learning_rate": 0.0001676,
-      "loss": 0.9532,
-      "step": 838
-    },
-    {
-      "epoch": 0.41269060501721594,
-      "grad_norm": 1.0701240247269224,
-      "learning_rate": 0.0001678,
-      "loss": 1.0374,
-      "step": 839
-    },
-    {
-      "epoch": 0.4131824889326119,
-      "grad_norm": 0.4991462766725644,
-      "learning_rate": 0.000168,
-      "loss": 0.8981,
-      "step": 840
-    },
-    {
-      "epoch": 0.41367437284800784,
-      "grad_norm": 0.6255019505856168,
-      "learning_rate": 0.0001682,
-      "loss": 0.9986,
-      "step": 841
-    },
-    {
-      "epoch": 0.41416625676340385,
-      "grad_norm": 0.5226316612893023,
-      "learning_rate": 0.0001684,
-      "loss": 0.8961,
-      "step": 842
-    },
-    {
-      "epoch": 0.4146581406787998,
-      "grad_norm": 0.5686473192385973,
-      "learning_rate": 0.0001686,
-      "loss": 0.9205,
-      "step": 843
-    },
-    {
-      "epoch": 0.41515002459419575,
-      "grad_norm": 0.5761042202049704,
-      "learning_rate": 0.0001688,
-      "loss": 1.008,
-      "step": 844
-    },
-    {
-      "epoch": 0.41564190850959176,
-      "grad_norm": 0.5630637985316533,
-      "learning_rate": 0.00016900000000000002,
-      "loss": 0.9981,
-      "step": 845
-    },
-    {
-      "epoch": 0.4161337924249877,
-      "grad_norm": 0.6032313914065442,
-      "learning_rate": 0.0001692,
-      "loss": 1.0199,
-      "step": 846
-    },
-    {
-      "epoch": 0.41662567634038367,
-      "grad_norm": 0.5261433904762187,
-      "learning_rate": 0.0001694,
-      "loss": 0.9452,
-      "step": 847
-    },
-    {
-      "epoch": 0.4171175602557796,
-      "grad_norm": 0.5330558170318862,
-      "learning_rate": 0.0001696,
-      "loss": 0.9222,
-      "step": 848
-    },
-    {
-      "epoch": 0.4176094441711756,
-      "grad_norm": 0.5732281296425276,
-      "learning_rate": 0.0001698,
-      "loss": 0.9501,
-      "step": 849
-    },
-    {
-      "epoch": 0.4181013280865716,
-      "grad_norm": 0.5491208816730045,
-      "learning_rate": 0.00017,
-      "loss": 0.9096,
-      "step": 850
-    },
-    {
-      "epoch": 0.41859321200196753,
-      "grad_norm": 0.5597683458893487,
-      "learning_rate": 0.00017020000000000002,
-      "loss": 1.0032,
-      "step": 851
-    },
-    {
-      "epoch": 0.4190850959173635,
-      "grad_norm": 0.5622591927744834,
-      "learning_rate": 0.0001704,
-      "loss": 0.9857,
-      "step": 852
-    },
-    {
-      "epoch": 0.4195769798327595,
-      "grad_norm": 0.5885794587196331,
-      "learning_rate": 0.0001706,
-      "loss": 1.0659,
-      "step": 853
-    },
-    {
-      "epoch": 0.42006886374815544,
-      "grad_norm": 0.5514053204871268,
-      "learning_rate": 0.0001708,
-      "loss": 0.9635,
-      "step": 854
-    },
-    {
-      "epoch": 0.4205607476635514,
-      "grad_norm": 0.5518416797926007,
-      "learning_rate": 0.000171,
-      "loss": 0.954,
-      "step": 855
-    },
-    {
-      "epoch": 0.42105263157894735,
-      "grad_norm": 0.6023882068182663,
-      "learning_rate": 0.00017120000000000001,
-      "loss": 0.9739,
-      "step": 856
-    },
-    {
-      "epoch": 0.42154451549434335,
-      "grad_norm": 0.5500016689819863,
-      "learning_rate": 0.0001714,
-      "loss": 0.9494,
-      "step": 857
-    },
-    {
-      "epoch": 0.4220363994097393,
-      "grad_norm": 0.5649553119783989,
-      "learning_rate": 0.0001716,
-      "loss": 0.9781,
-      "step": 858
-    },
-    {
-      "epoch": 0.42252828332513526,
-      "grad_norm": 0.5459792426412101,
-      "learning_rate": 0.0001718,
-      "loss": 0.941,
-      "step": 859
-    },
-    {
-      "epoch": 0.4230201672405312,
-      "grad_norm": 0.5652491524840089,
-      "learning_rate": 0.000172,
-      "loss": 0.9732,
-      "step": 860
-    },
-    {
-      "epoch": 0.4235120511559272,
-      "grad_norm": 0.5270656850538266,
-      "learning_rate": 0.0001722,
-      "loss": 0.9065,
-      "step": 861
-    },
-    {
-      "epoch": 0.42400393507132317,
-      "grad_norm": 0.5769523725720945,
-      "learning_rate": 0.00017240000000000002,
-      "loss": 1.0541,
-      "step": 862
-    },
-    {
-      "epoch": 0.4244958189867191,
-      "grad_norm": 0.7200570869145366,
-      "learning_rate": 0.0001726,
-      "loss": 0.9736,
-      "step": 863
-    },
-    {
-      "epoch": 0.4249877029021151,
-      "grad_norm": 0.5161154721973787,
-      "learning_rate": 0.0001728,
-      "loss": 1.0092,
-      "step": 864
-    },
-    {
-      "epoch": 0.4254795868175111,
-      "grad_norm": 0.5296402663634671,
-      "learning_rate": 0.000173,
-      "loss": 0.9159,
-      "step": 865
-    },
-    {
-      "epoch": 0.42597147073290703,
-      "grad_norm": 0.5562736193802396,
-      "learning_rate": 0.0001732,
-      "loss": 0.8968,
-      "step": 866
-    },
-    {
-      "epoch": 0.426463354648303,
-      "grad_norm": 0.5270375029235659,
-      "learning_rate": 0.0001734,
-      "loss": 0.9101,
-      "step": 867
-    },
-    {
-      "epoch": 0.426955238563699,
-      "grad_norm": 0.5405629707161682,
-      "learning_rate": 0.00017360000000000002,
-      "loss": 1.0218,
-      "step": 868
-    },
-    {
-      "epoch": 0.42744712247909494,
-      "grad_norm": 0.5370615414631368,
-      "learning_rate": 0.0001738,
-      "loss": 0.9493,
-      "step": 869
-    },
-    {
-      "epoch": 0.4279390063944909,
-      "grad_norm": 0.5836169487693725,
-      "learning_rate": 0.000174,
-      "loss": 1.0389,
-      "step": 870
-    },
-    {
-      "epoch": 0.42843089030988685,
-      "grad_norm": 0.5692881217128932,
-      "learning_rate": 0.0001742,
-      "loss": 1.0029,
-      "step": 871
-    },
-    {
-      "epoch": 0.42892277422528285,
-      "grad_norm": 0.5162749207602118,
-      "learning_rate": 0.0001744,
-      "loss": 0.9913,
-      "step": 872
-    },
-    {
-      "epoch": 0.4294146581406788,
-      "grad_norm": 0.5179704644692854,
-      "learning_rate": 0.00017460000000000002,
-      "loss": 0.9416,
-      "step": 873
-    },
-    {
-      "epoch": 0.42990654205607476,
-      "grad_norm": 0.5413434863970977,
-      "learning_rate": 0.00017480000000000002,
-      "loss": 0.9138,
-      "step": 874
-    },
-    {
-      "epoch": 0.4303984259714707,
-      "grad_norm": 0.5278059859029571,
-      "learning_rate": 0.000175,
-      "loss": 0.9468,
-      "step": 875
-    },
-    {
-      "epoch": 0.4308903098868667,
-      "grad_norm": 0.5119656198059847,
-      "learning_rate": 0.0001752,
-      "loss": 1.0543,
-      "step": 876
-    },
-    {
-      "epoch": 0.43138219380226267,
-      "grad_norm": 0.5377509485318009,
-      "learning_rate": 0.0001754,
-      "loss": 1.0107,
-      "step": 877
-    },
-    {
-      "epoch": 0.4318740777176586,
-      "grad_norm": 0.5820007835814583,
-      "learning_rate": 0.0001756,
-      "loss": 0.9246,
-      "step": 878
-    },
-    {
-      "epoch": 0.4323659616330546,
-      "grad_norm": 0.5343999036018037,
-      "learning_rate": 0.00017580000000000002,
-      "loss": 0.9566,
-      "step": 879
-    },
-    {
-      "epoch": 0.4328578455484506,
-      "grad_norm": 0.5062821082498787,
-      "learning_rate": 0.00017600000000000002,
-      "loss": 0.8971,
-      "step": 880
-    },
-    {
-      "epoch": 0.43334972946384653,
-      "grad_norm": 0.5503322884070123,
-      "learning_rate": 0.0001762,
-      "loss": 1.1004,
-      "step": 881
-    },
-    {
-      "epoch": 0.4338416133792425,
-      "grad_norm": 0.5212493269418079,
-      "learning_rate": 0.0001764,
-      "loss": 0.9533,
-      "step": 882
-    },
-    {
-      "epoch": 0.43433349729463844,
-      "grad_norm": 0.5388116816718315,
-      "learning_rate": 0.0001766,
-      "loss": 0.9838,
-      "step": 883
-    },
-    {
-      "epoch": 0.43482538121003445,
-      "grad_norm": 0.5735442764615539,
-      "learning_rate": 0.00017680000000000001,
-      "loss": 1.0515,
-      "step": 884
-    },
-    {
-      "epoch": 0.4353172651254304,
-      "grad_norm": 0.5155939464902994,
-      "learning_rate": 0.00017700000000000002,
-      "loss": 0.9761,
-      "step": 885
-    },
-    {
-      "epoch": 0.43580914904082635,
-      "grad_norm": 0.5067177828353471,
-      "learning_rate": 0.0001772,
-      "loss": 0.9218,
-      "step": 886
-    },
-    {
-      "epoch": 0.43630103295622236,
-      "grad_norm": 0.516882852883826,
-      "learning_rate": 0.0001774,
-      "loss": 0.9928,
-      "step": 887
-    },
-    {
-      "epoch": 0.4367929168716183,
-      "grad_norm": 0.541258076851383,
-      "learning_rate": 0.0001776,
-      "loss": 1.0355,
-      "step": 888
-    },
-    {
-      "epoch": 0.43728480078701426,
-      "grad_norm": 0.5145423495342223,
-      "learning_rate": 0.0001778,
-      "loss": 0.9292,
-      "step": 889
-    },
-    {
-      "epoch": 0.4377766847024102,
-      "grad_norm": 0.5142489348461433,
-      "learning_rate": 0.00017800000000000002,
-      "loss": 0.9434,
-      "step": 890
-    },
-    {
-      "epoch": 0.4382685686178062,
-      "grad_norm": 0.5150661220780518,
-      "learning_rate": 0.00017820000000000002,
-      "loss": 0.8795,
-      "step": 891
-    },
-    {
-      "epoch": 0.43876045253320217,
-      "grad_norm": 0.5739910608494103,
-      "learning_rate": 0.0001784,
-      "loss": 1.0217,
-      "step": 892
-    },
-    {
-      "epoch": 0.4392523364485981,
-      "grad_norm": 0.5210480855295413,
-      "learning_rate": 0.0001786,
-      "loss": 0.9,
-      "step": 893
-    },
-    {
-      "epoch": 0.4397442203639941,
-      "grad_norm": 0.5226029458504111,
-      "learning_rate": 0.0001788,
-      "loss": 0.9757,
-      "step": 894
-    },
-    {
-      "epoch": 0.4402361042793901,
-      "grad_norm": 0.5390926276584241,
-      "learning_rate": 0.00017900000000000001,
-      "loss": 0.9254,
-      "step": 895
-    },
-    {
-      "epoch": 0.44072798819478604,
-      "grad_norm": 0.6096776794834838,
-      "learning_rate": 0.00017920000000000002,
-      "loss": 1.0365,
-      "step": 896
-    },
-    {
-      "epoch": 0.441219872110182,
-      "grad_norm": 0.9843822292629351,
-      "learning_rate": 0.00017940000000000002,
-      "loss": 0.9873,
-      "step": 897
-    },
-    {
-      "epoch": 0.44171175602557794,
-      "grad_norm": 0.5597177313333231,
-      "learning_rate": 0.0001796,
-      "loss": 1.0479,
-      "step": 898
-    },
-    {
-      "epoch": 0.44220363994097395,
-      "grad_norm": 0.5127121453283586,
-      "learning_rate": 0.0001798,
-      "loss": 0.9685,
-      "step": 899
-    },
-    {
-      "epoch": 0.4426955238563699,
-      "grad_norm": 0.5030124116175764,
-      "learning_rate": 0.00018,
-      "loss": 0.8948,
-      "step": 900
-    },
-    {
-      "epoch": 0.44318740777176585,
-      "grad_norm": 0.5293459541657133,
-      "learning_rate": 0.00018020000000000002,
-      "loss": 0.9261,
-      "step": 901
-    },
-    {
-      "epoch": 0.4436792916871618,
-      "grad_norm": 0.5054173854921744,
-      "learning_rate": 0.00018040000000000002,
-      "loss": 1.0131,
-      "step": 902
-    },
-    {
-      "epoch": 0.4441711756025578,
-      "grad_norm": 0.5217743352033825,
-      "learning_rate": 0.00018060000000000003,
-      "loss": 0.9877,
-      "step": 903
-    },
-    {
-      "epoch": 0.44466305951795376,
-      "grad_norm": 0.501600899425512,
-      "learning_rate": 0.0001808,
-      "loss": 0.82,
-      "step": 904
-    },
-    {
-      "epoch": 0.4451549434333497,
-      "grad_norm": 0.525501434091119,
-      "learning_rate": 0.000181,
-      "loss": 0.9267,
-      "step": 905
-    },
-    {
-      "epoch": 0.4456468273487457,
-      "grad_norm": 0.6120434569760427,
-      "learning_rate": 0.0001812,
-      "loss": 1.1088,
-      "step": 906
-    },
-    {
-      "epoch": 0.4461387112641417,
-      "grad_norm": 0.493055662564238,
-      "learning_rate": 0.00018140000000000002,
-      "loss": 0.9278,
-      "step": 907
-    },
-    {
-      "epoch": 0.4466305951795376,
-      "grad_norm": 0.4836970336324819,
-      "learning_rate": 0.00018160000000000002,
-      "loss": 0.8893,
-      "step": 908
-    },
-    {
-      "epoch": 0.4471224790949336,
-      "grad_norm": 0.5465285130057251,
-      "learning_rate": 0.00018180000000000003,
-      "loss": 0.9715,
-      "step": 909
-    },
-    {
-      "epoch": 0.4476143630103296,
-      "grad_norm": 0.5609985643623988,
-      "learning_rate": 0.000182,
-      "loss": 0.9634,
-      "step": 910
-    },
-    {
-      "epoch": 0.44810624692572554,
-      "grad_norm": 0.5492145145037122,
-      "learning_rate": 0.0001822,
-      "loss": 0.98,
-      "step": 911
-    },
-    {
-      "epoch": 0.4485981308411215,
-      "grad_norm": 0.5143204844427829,
-      "learning_rate": 0.00018240000000000002,
-      "loss": 0.9637,
-      "step": 912
-    },
-    {
-      "epoch": 0.44909001475651744,
-      "grad_norm": 0.50718221347543,
-      "learning_rate": 0.00018260000000000002,
-      "loss": 0.9899,
-      "step": 913
-    },
-    {
-      "epoch": 0.44958189867191345,
-      "grad_norm": 0.4881790127376402,
-      "learning_rate": 0.00018280000000000003,
-      "loss": 1.0146,
-      "step": 914
-    },
-    {
-      "epoch": 0.4500737825873094,
-      "grad_norm": 0.5273502353074515,
-      "learning_rate": 0.000183,
-      "loss": 0.9837,
-      "step": 915
-    },
-    {
-      "epoch": 0.45056566650270535,
-      "grad_norm": 1.4151392968827463,
-      "learning_rate": 0.0001832,
-      "loss": 1.0776,
-      "step": 916
-    },
-    {
-      "epoch": 0.4510575504181013,
-      "grad_norm": 0.5161037772931225,
-      "learning_rate": 0.0001834,
-      "loss": 0.9127,
-      "step": 917
-    },
-    {
-      "epoch": 0.4515494343334973,
-      "grad_norm": 0.553790492621958,
-      "learning_rate": 0.00018360000000000002,
-      "loss": 0.982,
-      "step": 918
-    },
-    {
-      "epoch": 0.45204131824889326,
-      "grad_norm": 0.8272049234808955,
-      "learning_rate": 0.00018380000000000002,
-      "loss": 0.9753,
-      "step": 919
-    },
-    {
-      "epoch": 0.4525332021642892,
-      "grad_norm": 0.5646572338940326,
-      "learning_rate": 0.00018400000000000003,
-      "loss": 1.0917,
-      "step": 920
-    },
-    {
-      "epoch": 0.45302508607968517,
-      "grad_norm": 1.1220925918626086,
-      "learning_rate": 0.0001842,
-      "loss": 0.9684,
-      "step": 921
-    },
-    {
-      "epoch": 0.4535169699950812,
-      "grad_norm": 0.5288359765330555,
-      "learning_rate": 0.0001844,
-      "loss": 0.964,
-      "step": 922
-    },
-    {
-      "epoch": 0.45400885391047713,
-      "grad_norm": 0.5561258676542358,
-      "learning_rate": 0.00018460000000000001,
-      "loss": 1.0354,
-      "step": 923
-    },
-    {
-      "epoch": 0.4545007378258731,
-      "grad_norm": 0.6450660996673966,
-      "learning_rate": 0.00018480000000000002,
-      "loss": 0.9265,
-      "step": 924
-    },
-    {
-      "epoch": 0.4549926217412691,
-      "grad_norm": 0.6319923842774048,
-      "learning_rate": 0.00018500000000000002,
-      "loss": 0.969,
-      "step": 925
-    },
-    {
-      "epoch": 0.45548450565666504,
-      "grad_norm": 0.5424189796922686,
-      "learning_rate": 0.00018520000000000003,
-      "loss": 0.9816,
-      "step": 926
-    },
-    {
-      "epoch": 0.455976389572061,
-      "grad_norm": 0.5781436446478081,
-      "learning_rate": 0.0001854,
-      "loss": 0.9374,
-      "step": 927
-    },
-    {
-      "epoch": 0.45646827348745694,
-      "grad_norm": 0.5578660553072525,
-      "learning_rate": 0.0001856,
-      "loss": 0.9408,
-      "step": 928
-    },
-    {
-      "epoch": 0.45696015740285295,
-      "grad_norm": 0.5143529191819146,
-      "learning_rate": 0.00018580000000000002,
-      "loss": 0.9493,
-      "step": 929
-    },
-    {
-      "epoch": 0.4574520413182489,
-      "grad_norm": 0.5566530266352897,
-      "learning_rate": 0.00018600000000000002,
-      "loss": 0.9306,
-      "step": 930
-    },
-    {
-      "epoch": 0.45794392523364486,
-      "grad_norm": 0.5693834575300597,
-      "learning_rate": 0.00018620000000000003,
-      "loss": 0.9072,
-      "step": 931
-    },
-    {
-      "epoch": 0.4584358091490408,
-      "grad_norm": 0.5883519629457379,
-      "learning_rate": 0.00018640000000000003,
-      "loss": 0.878,
-      "step": 932
-    },
-    {
-      "epoch": 0.4589276930644368,
-      "grad_norm": 0.5931664627910979,
-      "learning_rate": 0.0001866,
-      "loss": 1.0526,
-      "step": 933
-    },
-    {
-      "epoch": 0.45941957697983277,
-      "grad_norm": 0.5474015587714414,
-      "learning_rate": 0.00018680000000000001,
-      "loss": 0.9676,
-      "step": 934
-    },
-    {
-      "epoch": 0.4599114608952287,
-      "grad_norm": 0.5177988921047931,
-      "learning_rate": 0.00018700000000000002,
-      "loss": 0.9288,
-      "step": 935
-    },
-    {
-      "epoch": 0.46040334481062467,
-      "grad_norm": 0.545548390053368,
-      "learning_rate": 0.00018720000000000002,
-      "loss": 1.0001,
-      "step": 936
-    },
-    {
-      "epoch": 0.4608952287260207,
-      "grad_norm": 0.519325393867732,
-      "learning_rate": 0.00018740000000000003,
-      "loss": 0.9482,
-      "step": 937
-    },
-    {
-      "epoch": 0.46138711264141663,
-      "grad_norm": 0.5456394285516738,
-      "learning_rate": 0.0001876,
-      "loss": 0.9549,
-      "step": 938
-    },
-    {
-      "epoch": 0.4618789965568126,
-      "grad_norm": 0.5610463252748996,
-      "learning_rate": 0.0001878,
-      "loss": 0.9887,
-      "step": 939
-    },
-    {
-      "epoch": 0.46237088047220853,
-      "grad_norm": 0.5088615249359095,
-      "learning_rate": 0.000188,
-      "loss": 0.9997,
-      "step": 940
-    },
-    {
-      "epoch": 0.46286276438760454,
-      "grad_norm": 0.6403892090296743,
-      "learning_rate": 0.0001882,
-      "loss": 0.9992,
-      "step": 941
-    },
-    {
-      "epoch": 0.4633546483030005,
-      "grad_norm": 0.5459758884645787,
-      "learning_rate": 0.0001884,
-      "loss": 0.9136,
-      "step": 942
-    },
-    {
-      "epoch": 0.46384653221839645,
-      "grad_norm": 0.5264417005084878,
-      "learning_rate": 0.0001886,
-      "loss": 1.0359,
-      "step": 943
-    },
-    {
-      "epoch": 0.4643384161337924,
-      "grad_norm": 0.5452103362565185,
-      "learning_rate": 0.0001888,
-      "loss": 1.0104,
-      "step": 944
-    },
-    {
-      "epoch": 0.4648303000491884,
-      "grad_norm": 0.49040713618929604,
-      "learning_rate": 0.00018899999999999999,
-      "loss": 1.0123,
-      "step": 945
-    },
-    {
-      "epoch": 0.46532218396458436,
-      "grad_norm": 0.5444843316477975,
-      "learning_rate": 0.0001892,
-      "loss": 0.9674,
-      "step": 946
-    },
-    {
-      "epoch": 0.4658140678799803,
-      "grad_norm": 0.5154842838561677,
-      "learning_rate": 0.0001894,
-      "loss": 0.954,
-      "step": 947
-    },
-    {
-      "epoch": 0.4663059517953763,
-      "grad_norm": 0.5311303516764446,
-      "learning_rate": 0.0001896,
-      "loss": 0.9436,
-      "step": 948
-    },
-    {
-      "epoch": 0.46679783571077227,
-      "grad_norm": 0.6332795107793114,
-      "learning_rate": 0.0001898,
-      "loss": 0.9715,
-      "step": 949
-    },
-    {
-      "epoch": 0.4672897196261682,
-      "grad_norm": 0.5310883218087131,
-      "learning_rate": 0.00019,
-      "loss": 0.9733,
-      "step": 950
-    },
-    {
-      "epoch": 0.4677816035415642,
-      "grad_norm": 0.5585087599292401,
-      "learning_rate": 0.0001902,
-      "loss": 0.9766,
-      "step": 951
-    },
-    {
-      "epoch": 0.4682734874569602,
-      "grad_norm": 0.5379772982249258,
-      "learning_rate": 0.0001904,
-      "loss": 0.9725,
-      "step": 952
-    },
-    {
-      "epoch": 0.46876537137235613,
-      "grad_norm": 1.0927315848471355,
-      "learning_rate": 0.0001906,
-      "loss": 1.0,
-      "step": 953
-    },
-    {
-      "epoch": 0.4692572552877521,
-      "grad_norm": 0.522722072916259,
-      "learning_rate": 0.0001908,
-      "loss": 0.9734,
-      "step": 954
-    },
-    {
-      "epoch": 0.46974913920314804,
-      "grad_norm": 0.5722683426740471,
-      "learning_rate": 0.000191,
-      "loss": 0.9747,
-      "step": 955
-    },
-    {
-      "epoch": 0.47024102311854404,
-      "grad_norm": 0.6054667509782405,
-      "learning_rate": 0.0001912,
-      "loss": 0.9534,
-      "step": 956
-    },
-    {
-      "epoch": 0.47073290703394,
-      "grad_norm": 0.5028530415420397,
-      "learning_rate": 0.0001914,
-      "loss": 0.9583,
-      "step": 957
-    },
-    {
-      "epoch": 0.47122479094933595,
-      "grad_norm": 0.4567247279556181,
-      "learning_rate": 0.0001916,
-      "loss": 0.9365,
-      "step": 958
-    },
-    {
-      "epoch": 0.4717166748647319,
-      "grad_norm": 0.5415736824580925,
-      "learning_rate": 0.0001918,
-      "loss": 1.0211,
-      "step": 959
-    },
-    {
-      "epoch": 0.4722085587801279,
-      "grad_norm": 0.4893423926343375,
-      "learning_rate": 0.000192,
-      "loss": 0.9563,
-      "step": 960
-    },
-    {
-      "epoch": 0.47270044269552386,
-      "grad_norm": 0.5508130258707056,
-      "learning_rate": 0.0001922,
-      "loss": 1.008,
-      "step": 961
-    },
-    {
-      "epoch": 0.4731923266109198,
-      "grad_norm": 0.5786410786759495,
-      "learning_rate": 0.00019240000000000001,
-      "loss": 1.0173,
-      "step": 962
-    },
-    {
-      "epoch": 0.47368421052631576,
-      "grad_norm": 0.5611499337597293,
-      "learning_rate": 0.0001926,
-      "loss": 0.9662,
-      "step": 963
-    },
-    {
-      "epoch": 0.47417609444171177,
-      "grad_norm": 0.5198201756826512,
-      "learning_rate": 0.0001928,
-      "loss": 0.9753,
-      "step": 964
-    },
-    {
-      "epoch": 0.4746679783571077,
-      "grad_norm": 0.5852887651233413,
-      "learning_rate": 0.000193,
-      "loss": 0.9793,
-      "step": 965
-    },
-    {
-      "epoch": 0.4751598622725037,
-      "grad_norm": 0.6813951087616439,
-      "learning_rate": 0.0001932,
-      "loss": 0.9821,
-      "step": 966
-    },
-    {
-      "epoch": 0.4756517461878997,
-      "grad_norm": 0.5327706348908753,
-      "learning_rate": 0.0001934,
-      "loss": 0.8946,
-      "step": 967
-    },
-    {
-      "epoch": 0.47614363010329563,
-      "grad_norm": 0.5161668091121259,
-      "learning_rate": 0.00019360000000000002,
-      "loss": 0.9447,
-      "step": 968
-    },
-    {
-      "epoch": 0.4766355140186916,
-      "grad_norm": 0.5417441809512521,
-      "learning_rate": 0.0001938,
-      "loss": 0.9962,
-      "step": 969
-    },
-    {
-      "epoch": 0.47712739793408754,
-      "grad_norm": 0.5813128131414306,
-      "learning_rate": 0.000194,
-      "loss": 1.0569,
-      "step": 970
-    },
-    {
-      "epoch": 0.47761928184948355,
-      "grad_norm": 0.5322364446816521,
-      "learning_rate": 0.0001942,
-      "loss": 0.9452,
-      "step": 971
-    },
-    {
-      "epoch": 0.4781111657648795,
-      "grad_norm": 0.5523793272808772,
-      "learning_rate": 0.0001944,
-      "loss": 0.9407,
-      "step": 972
-    },
-    {
-      "epoch": 0.47860304968027545,
-      "grad_norm": 0.5598749774485714,
-      "learning_rate": 0.00019460000000000001,
-      "loss": 1.0484,
-      "step": 973
-    },
-    {
-      "epoch": 0.4790949335956714,
-      "grad_norm": 0.5043454954607122,
-      "learning_rate": 0.0001948,
-      "loss": 0.8841,
-      "step": 974
-    },
-    {
-      "epoch": 0.4795868175110674,
-      "grad_norm": 0.5271800626889586,
-      "learning_rate": 0.000195,
-      "loss": 1.0506,
-      "step": 975
-    },
-    {
-      "epoch": 0.48007870142646336,
-      "grad_norm": 0.5863380204930064,
-      "learning_rate": 0.0001952,
-      "loss": 1.0006,
-      "step": 976
-    },
-    {
-      "epoch": 0.4805705853418593,
-      "grad_norm": 0.5863752827782015,
-      "learning_rate": 0.0001954,
-      "loss": 1.0227,
-      "step": 977
-    },
-    {
-      "epoch": 0.48106246925725527,
-      "grad_norm": 0.5420702202322971,
-      "learning_rate": 0.0001956,
-      "loss": 1.0879,
-      "step": 978
-    },
-    {
-      "epoch": 0.4815543531726513,
-      "grad_norm": 0.46710007809882115,
-      "learning_rate": 0.00019580000000000002,
-      "loss": 0.9074,
-      "step": 979
-    },
-    {
-      "epoch": 0.4820462370880472,
-      "grad_norm": 0.4979039222094224,
-      "learning_rate": 0.000196,
-      "loss": 0.9127,
-      "step": 980
-    },
-    {
-      "epoch": 0.4825381210034432,
-      "grad_norm": 0.5270304948270585,
-      "learning_rate": 0.0001962,
-      "loss": 0.9595,
-      "step": 981
-    },
-    {
-      "epoch": 0.48303000491883913,
-      "grad_norm": 0.5074612848390806,
-      "learning_rate": 0.0001964,
-      "loss": 1.0176,
-      "step": 982
-    },
-    {
-      "epoch": 0.48352188883423514,
-      "grad_norm": 0.5303100630890872,
-      "learning_rate": 0.0001966,
-      "loss": 0.89,
-      "step": 983
-    },
-    {
-      "epoch": 0.4840137727496311,
-      "grad_norm": 0.47382999672479026,
-      "learning_rate": 0.0001968,
-      "loss": 0.9367,
-      "step": 984
-    },
-    {
-      "epoch": 0.48450565666502704,
-      "grad_norm": 0.5519787323789412,
-      "learning_rate": 0.00019700000000000002,
-      "loss": 1.0038,
-      "step": 985
-    },
-    {
-      "epoch": 0.48499754058042305,
-      "grad_norm": 0.4897366081420991,
-      "learning_rate": 0.0001972,
-      "loss": 0.9696,
-      "step": 986
-    },
-    {
-      "epoch": 0.485489424495819,
-      "grad_norm": 0.49158245345946805,
-      "learning_rate": 0.0001974,
-      "loss": 0.952,
-      "step": 987
-    },
-    {
-      "epoch": 0.48598130841121495,
-      "grad_norm": 0.5105032086419014,
-      "learning_rate": 0.0001976,
-      "loss": 0.9443,
-      "step": 988
-    },
-    {
-      "epoch": 0.4864731923266109,
-      "grad_norm": 0.4699398564080473,
-      "learning_rate": 0.0001978,
-      "loss": 0.9536,
-      "step": 989
-    },
-    {
-      "epoch": 0.4869650762420069,
-      "grad_norm": 0.5173446919886702,
-      "learning_rate": 0.00019800000000000002,
-      "loss": 0.8655,
-      "step": 990
-    },
-    {
-      "epoch": 0.48745696015740286,
-      "grad_norm": 0.5004505477278638,
-      "learning_rate": 0.00019820000000000002,
-      "loss": 0.9124,
-      "step": 991
-    },
-    {
-      "epoch": 0.4879488440727988,
-      "grad_norm": 0.48220413607150886,
-      "learning_rate": 0.0001984,
-      "loss": 0.9546,
-      "step": 992
-    },
-    {
-      "epoch": 0.48844072798819477,
-      "grad_norm": 0.4819792774684789,
-      "learning_rate": 0.0001986,
-      "loss": 0.8738,
-      "step": 993
-    },
-    {
-      "epoch": 0.4889326119035908,
-      "grad_norm": 0.5164599873717493,
-      "learning_rate": 0.0001988,
-      "loss": 0.9594,
-      "step": 994
-    },
-    {
-      "epoch": 0.4894244958189867,
-      "grad_norm": 0.5672787019265206,
-      "learning_rate": 0.000199,
-      "loss": 0.9997,
-      "step": 995
-    },
-    {
-      "epoch": 0.4899163797343827,
-      "grad_norm": 0.48163904907656424,
-      "learning_rate": 0.00019920000000000002,
-      "loss": 0.9249,
-      "step": 996
-    },
-    {
-      "epoch": 0.49040826364977863,
-      "grad_norm": 0.4992747022764088,
-      "learning_rate": 0.00019940000000000002,
-      "loss": 0.9101,
-      "step": 997
-    },
-    {
-      "epoch": 0.49090014756517464,
-      "grad_norm": 0.505152562500964,
-      "learning_rate": 0.0001996,
-      "loss": 0.98,
-      "step": 998
-    },
-    {
-      "epoch": 0.4913920314805706,
-      "grad_norm": 0.49220963895087283,
-      "learning_rate": 0.0001998,
-      "loss": 1.0135,
-      "step": 999
-    },
-    {
-      "epoch": 0.49188391539596654,
-      "grad_norm": 0.499785327862804,
-      "learning_rate": 0.0002,
-      "loss": 0.9809,
-      "step": 1000
-    },
-    {
-      "epoch": 0.4923757993113625,
-      "grad_norm": 0.5155171431912795,
-      "learning_rate": 0.00019999998104216116,
-      "loss": 1.0481,
-      "step": 1001
-    },
-    {
-      "epoch": 0.4928676832267585,
-      "grad_norm": 0.4980545715907866,
-      "learning_rate": 0.00019999992416865176,
-      "loss": 0.8527,
-      "step": 1002
-    },
-    {
-      "epoch": 0.49335956714215445,
-      "grad_norm": 0.47724253947130507,
-      "learning_rate": 0.0001999998293794934,
-      "loss": 0.9135,
-      "step": 1003
-    },
-    {
-      "epoch": 0.4938514510575504,
-      "grad_norm": 0.5720455995618264,
-      "learning_rate": 0.00019999969667472203,
-      "loss": 0.9878,
-      "step": 1004
-    },
-    {
-      "epoch": 0.49434333497294636,
-      "grad_norm": 0.5371333285470227,
-      "learning_rate": 0.00019999952605438795,
-      "loss": 1.0797,
-      "step": 1005
-    },
-    {
-      "epoch": 0.49483521888834237,
-      "grad_norm": 0.4819222664190339,
-      "learning_rate": 0.00019999931751855582,
-      "loss": 0.9576,
-      "step": 1006
-    },
-    {
-      "epoch": 0.4953271028037383,
-      "grad_norm": 0.4823918029759922,
-      "learning_rate": 0.00019999907106730479,
-      "loss": 0.9503,
-      "step": 1007
-    },
-    {
-      "epoch": 0.49581898671913427,
-      "grad_norm": 0.5089751526246875,
-      "learning_rate": 0.0001999987867007282,
-      "loss": 0.9793,
-      "step": 1008
-    },
-    {
-      "epoch": 0.4963108706345303,
-      "grad_norm": 0.5511405448092508,
-      "learning_rate": 0.00019999846441893398,
-      "loss": 1.1435,
-      "step": 1009
-    },
-    {
-      "epoch": 0.49680275454992623,
-      "grad_norm": 0.5157899143785711,
-      "learning_rate": 0.00019999810422204422,
-      "loss": 1.017,
-      "step": 1010
-    },
-    {
-      "epoch": 0.4972946384653222,
-      "grad_norm": 0.5026713002369468,
-      "learning_rate": 0.00019999770611019556,
-      "loss": 1.0423,
-      "step": 1011
-    },
-    {
-      "epoch": 0.49778652238071813,
-      "grad_norm": 0.4821734284298085,
-      "learning_rate": 0.0001999972700835389,
-      "loss": 0.9666,
-      "step": 1012
-    },
-    {
-      "epoch": 0.49827840629611414,
-      "grad_norm": 0.5215944398758797,
-      "learning_rate": 0.00019999679614223962,
-      "loss": 0.9825,
-      "step": 1013
-    },
-    {
-      "epoch": 0.4987702902115101,
-      "grad_norm": 0.5304865325619055,
-      "learning_rate": 0.00019999628428647736,
-      "loss": 0.995,
-      "step": 1014
-    },
-    {
-      "epoch": 0.49926217412690604,
-      "grad_norm": 0.5204191713213968,
-      "learning_rate": 0.00019999573451644627,
-      "loss": 0.9588,
-      "step": 1015
-    },
-    {
-      "epoch": 0.499754058042302,
-      "grad_norm": 0.5174085296709169,
-      "learning_rate": 0.0001999951468323547,
-      "loss": 0.9716,
-      "step": 1016
-    },
-    {
-      "epoch": 0.500245941957698,
-      "grad_norm": 0.4938326578972806,
-      "learning_rate": 0.00019999452123442554,
-      "loss": 1.0,
-      "step": 1017
-    },
-    {
-      "epoch": 0.500245941957698,
-      "eval_loss": 0.8863905668258667,
-      "eval_runtime": 6679.255,
-      "eval_samples_per_second": 4.266,
-      "eval_steps_per_second": 2.133,
-      "step": 1017
-    },
-    {
-      "epoch": 0.5007378258730939,
-      "grad_norm": 0.49214519001737467,
-      "learning_rate": 0.00019999385772289597,
-      "loss": 0.9059,
-      "step": 1018
-    },
-    {
-      "epoch": 0.5012297097884899,
-      "grad_norm": 0.5509060070684917,
-      "learning_rate": 0.00019999315629801759,
-      "loss": 0.9234,
-      "step": 1019
-    },
-    {
-      "epoch": 0.5017215937038859,
-      "grad_norm": 0.5360077984580891,
-      "learning_rate": 0.00019999241696005632,
-      "loss": 1.0058,
-      "step": 1020
-    },
-    {
-      "epoch": 0.5022134776192818,
-      "grad_norm": 0.6575060966897581,
-      "learning_rate": 0.00019999163970929246,
-      "loss": 0.9754,
-      "step": 1021
-    },
-    {
-      "epoch": 0.5027053615346778,
-      "grad_norm": 0.4964465882401881,
-      "learning_rate": 0.00019999082454602078,
-      "loss": 0.8975,
-      "step": 1022
-    },
-    {
-      "epoch": 0.5031972454500738,
-      "grad_norm": 0.48294639824987906,
-      "learning_rate": 0.0001999899714705503,
-      "loss": 0.9591,
-      "step": 1023
-    },
-    {
-      "epoch": 0.5036891293654697,
-      "grad_norm": 0.5251594738406836,
-      "learning_rate": 0.0001999890804832045,
-      "loss": 0.9166,
-      "step": 1024
-    },
-    {
-      "epoch": 0.5041810132808657,
-      "grad_norm": 0.5536997769465125,
-      "learning_rate": 0.00019998815158432117,
-      "loss": 1.0339,
-      "step": 1025
-    },
-    {
-      "epoch": 0.5046728971962616,
-      "grad_norm": 0.5365066362511384,
-      "learning_rate": 0.00019998718477425256,
-      "loss": 0.9394,
-      "step": 1026
-    },
-    {
-      "epoch": 0.5051647811116576,
-      "grad_norm": 0.5220932228818115,
-      "learning_rate": 0.0001999861800533652,
-      "loss": 0.9513,
-      "step": 1027
-    },
-    {
-      "epoch": 0.5056566650270536,
-      "grad_norm": 0.48675880030109264,
-      "learning_rate": 0.00019998513742204005,
-      "loss": 0.9856,
-      "step": 1028
-    },
-    {
-      "epoch": 0.5061485489424495,
-      "grad_norm": 0.5340826835480033,
-      "learning_rate": 0.0001999840568806724,
-      "loss": 0.9793,
-      "step": 1029
-    },
-    {
-      "epoch": 0.5066404328578455,
-      "grad_norm": 0.5435428108318967,
-      "learning_rate": 0.000199982938429672,
-      "loss": 0.971,
-      "step": 1030
-    },
-    {
-      "epoch": 0.5071323167732416,
-      "grad_norm": 0.49665018666396576,
-      "learning_rate": 0.0001999817820694629,
-      "loss": 1.0116,
-      "step": 1031
-    },
-    {
-      "epoch": 0.5076242006886375,
-      "grad_norm": 0.5290700436374691,
-      "learning_rate": 0.00019998058780048352,
-      "loss": 1.0281,
-      "step": 1032
-    },
-    {
-      "epoch": 0.5081160846040335,
-      "grad_norm": 0.46656726365676354,
-      "learning_rate": 0.0001999793556231867,
-      "loss": 0.9635,
-      "step": 1033
-    },
-    {
-      "epoch": 0.5086079685194295,
-      "grad_norm": 0.5230829519946276,
-      "learning_rate": 0.0001999780855380396,
-      "loss": 1.0314,
-      "step": 1034
-    },
-    {
-      "epoch": 0.5090998524348254,
-      "grad_norm": 0.4974803748014261,
-      "learning_rate": 0.00019997677754552383,
-      "loss": 0.9608,
-      "step": 1035
-    },
-    {
-      "epoch": 0.5095917363502214,
-      "grad_norm": 0.48499834789667756,
-      "learning_rate": 0.00019997543164613525,
-      "loss": 0.9624,
-      "step": 1036
-    },
-    {
-      "epoch": 0.5100836202656173,
-      "grad_norm": 0.5263013642858179,
-      "learning_rate": 0.00019997404784038425,
-      "loss": 1.0824,
-      "step": 1037
-    },
-    {
-      "epoch": 0.5105755041810133,
-      "grad_norm": 0.48615387278539696,
-      "learning_rate": 0.00019997262612879543,
-      "loss": 1.0257,
-      "step": 1038
-    },
-    {
-      "epoch": 0.5110673880964093,
-      "grad_norm": 0.4918583874695705,
-      "learning_rate": 0.00019997116651190793,
-      "loss": 0.9189,
-      "step": 1039
-    },
-    {
-      "epoch": 0.5115592720118052,
-      "grad_norm": 0.5158751454023438,
-      "learning_rate": 0.0001999696689902751,
-      "loss": 0.9687,
-      "step": 1040
-    },
-    {
-      "epoch": 0.5120511559272012,
-      "grad_norm": 0.5091459685499787,
-      "learning_rate": 0.00019996813356446477,
-      "loss": 0.9564,
-      "step": 1041
-    },
-    {
-      "epoch": 0.5125430398425972,
-      "grad_norm": 0.5316267225900095,
-      "learning_rate": 0.00019996656023505907,
-      "loss": 0.9381,
-      "step": 1042
-    },
-    {
-      "epoch": 0.5130349237579931,
-      "grad_norm": 0.5552813111100937,
-      "learning_rate": 0.0001999649490026546,
-      "loss": 0.9379,
-      "step": 1043
-    },
-    {
-      "epoch": 0.5135268076733891,
-      "grad_norm": 0.5229496871549363,
-      "learning_rate": 0.0001999632998678622,
-      "loss": 0.9411,
-      "step": 1044
-    },
-    {
-      "epoch": 0.514018691588785,
-      "grad_norm": 0.5479240671596419,
-      "learning_rate": 0.0001999616128313072,
-      "loss": 0.9772,
-      "step": 1045
-    },
-    {
-      "epoch": 0.514510575504181,
-      "grad_norm": 0.5028138794727268,
-      "learning_rate": 0.00019995988789362924,
-      "loss": 0.9436,
-      "step": 1046
-    },
-    {
-      "epoch": 0.515002459419577,
-      "grad_norm": 0.5192800745783225,
-      "learning_rate": 0.00019995812505548235,
-      "loss": 1.0119,
-      "step": 1047
-    },
-    {
-      "epoch": 0.5154943433349729,
-      "grad_norm": 0.5052205027465987,
-      "learning_rate": 0.0001999563243175349,
-      "loss": 0.965,
-      "step": 1048
-    },
-    {
-      "epoch": 0.5159862272503689,
-      "grad_norm": 0.5076807151419906,
-      "learning_rate": 0.00019995448568046968,
-      "loss": 1.0358,
-      "step": 1049
-    },
-    {
-      "epoch": 0.5164781111657649,
-      "grad_norm": 0.5583453877277026,
-      "learning_rate": 0.0001999526091449838,
-      "loss": 0.9359,
-      "step": 1050
-    },
-    {
-      "epoch": 0.5169699950811608,
-      "grad_norm": 0.4883208945615239,
-      "learning_rate": 0.00019995069471178873,
-      "loss": 0.8411,
-      "step": 1051
-    },
-    {
-      "epoch": 0.5174618789965568,
-      "grad_norm": 0.48140272826016023,
-      "learning_rate": 0.0001999487423816104,
-      "loss": 0.9294,
-      "step": 1052
-    },
-    {
-      "epoch": 0.5179537629119528,
-      "grad_norm": 0.49741183449374904,
-      "learning_rate": 0.00019994675215518904,
-      "loss": 1.0319,
-      "step": 1053
-    },
-    {
-      "epoch": 0.5184456468273487,
-      "grad_norm": 0.5049671413313106,
-      "learning_rate": 0.00019994472403327924,
-      "loss": 0.9564,
-      "step": 1054
-    },
-    {
-      "epoch": 0.5189375307427447,
-      "grad_norm": 0.5014596762915716,
-      "learning_rate": 0.00019994265801664995,
-      "loss": 1.0328,
-      "step": 1055
-    },
-    {
-      "epoch": 0.5194294146581406,
-      "grad_norm": 0.5439278599357487,
-      "learning_rate": 0.0001999405541060846,
-      "loss": 0.9626,
-      "step": 1056
-    },
-    {
-      "epoch": 0.5199212985735366,
-      "grad_norm": 0.516105235953036,
-      "learning_rate": 0.0001999384123023808,
-      "loss": 0.9725,
-      "step": 1057
-    },
-    {
-      "epoch": 0.5204131824889326,
-      "grad_norm": 0.5134502517325076,
-      "learning_rate": 0.0001999362326063507,
-      "loss": 0.9918,
-      "step": 1058
-    },
-    {
-      "epoch": 0.5209050664043285,
-      "grad_norm": 1.130623670607306,
-      "learning_rate": 0.0001999340150188207,
-      "loss": 0.9511,
-      "step": 1059
-    },
-    {
-      "epoch": 0.5213969503197246,
-      "grad_norm": 0.571277939096185,
-      "learning_rate": 0.00019993175954063162,
-      "loss": 1.0004,
-      "step": 1060
-    },
-    {
-      "epoch": 0.5218888342351206,
-      "grad_norm": 0.5638754527966803,
-      "learning_rate": 0.0001999294661726387,
-      "loss": 1.0458,
-      "step": 1061
-    },
-    {
-      "epoch": 0.5223807181505165,
-      "grad_norm": 0.4915903459752852,
-      "learning_rate": 0.00019992713491571141,
-      "loss": 0.9286,
-      "step": 1062
-    },
-    {
-      "epoch": 0.5228726020659125,
-      "grad_norm": 0.526938834777462,
-      "learning_rate": 0.00019992476577073372,
-      "loss": 0.9886,
-      "step": 1063
-    },
-    {
-      "epoch": 0.5233644859813084,
-      "grad_norm": 0.4815199240735547,
-      "learning_rate": 0.00019992235873860387,
-      "loss": 0.9171,
-      "step": 1064
-    },
-    {
-      "epoch": 0.5238563698967044,
-      "grad_norm": 0.5525309180666556,
-      "learning_rate": 0.0001999199138202345,
-      "loss": 0.9285,
-      "step": 1065
-    },
-    {
-      "epoch": 0.5243482538121004,
-      "grad_norm": 0.4876592580799642,
-      "learning_rate": 0.0001999174310165526,
-      "loss": 0.9306,
-      "step": 1066
-    },
-    {
-      "epoch": 0.5248401377274963,
-      "grad_norm": 0.4956962542956257,
-      "learning_rate": 0.00019991491032849963,
-      "loss": 0.952,
-      "step": 1067
-    },
-    {
-      "epoch": 0.5253320216428923,
-      "grad_norm": 0.5124942889923908,
-      "learning_rate": 0.00019991235175703125,
-      "loss": 0.9061,
-      "step": 1068
-    },
-    {
-      "epoch": 0.5258239055582883,
-      "grad_norm": 0.5240364330047325,
-      "learning_rate": 0.0001999097553031176,
-      "loss": 0.9607,
-      "step": 1069
-    },
-    {
-      "epoch": 0.5263157894736842,
-      "grad_norm": 2.4267417582396087,
-      "learning_rate": 0.0001999071209677431,
-      "loss": 1.0853,
-      "step": 1070
-    },
-    {
-      "epoch": 0.5268076733890802,
-      "grad_norm": 0.6139702116859586,
-      "learning_rate": 0.00019990444875190658,
-      "loss": 1.0365,
-      "step": 1071
-    },
-    {
-      "epoch": 0.5272995573044762,
-      "grad_norm": 0.519507374423165,
-      "learning_rate": 0.00019990173865662124,
-      "loss": 0.9761,
-      "step": 1072
-    },
-    {
-      "epoch": 0.5277914412198721,
-      "grad_norm": 0.5584447614688508,
-      "learning_rate": 0.00019989899068291467,
-      "loss": 0.9908,
-      "step": 1073
-    },
-    {
-      "epoch": 0.5282833251352681,
-      "grad_norm": 0.6773970980327138,
-      "learning_rate": 0.00019989620483182874,
-      "loss": 1.0332,
-      "step": 1074
-    },
-    {
-      "epoch": 0.528775209050664,
-      "grad_norm": 0.5557178461152342,
-      "learning_rate": 0.00019989338110441973,
-      "loss": 0.9724,
-      "step": 1075
-    },
-    {
-      "epoch": 0.52926709296606,
-      "grad_norm": 0.5603461502619792,
-      "learning_rate": 0.00019989051950175827,
-      "loss": 0.9609,
-      "step": 1076
-    },
-    {
-      "epoch": 0.529758976881456,
-      "grad_norm": 0.6223679108164312,
-      "learning_rate": 0.00019988762002492937,
-      "loss": 1.0384,
-      "step": 1077
-    },
-    {
-      "epoch": 0.5302508607968519,
-      "grad_norm": 0.5221436427607583,
-      "learning_rate": 0.0001998846826750324,
-      "loss": 0.9731,
-      "step": 1078
-    },
-    {
-      "epoch": 0.5307427447122479,
-      "grad_norm": 0.5006124844249952,
-      "learning_rate": 0.00019988170745318103,
-      "loss": 1.029,
-      "step": 1079
-    },
-    {
-      "epoch": 0.5312346286276439,
-      "grad_norm": 0.6447036478746674,
-      "learning_rate": 0.00019987869436050338,
-      "loss": 0.9187,
-      "step": 1080
-    },
-    {
-      "epoch": 0.5317265125430398,
-      "grad_norm": 0.4770607221759239,
-      "learning_rate": 0.00019987564339814183,
-      "loss": 0.894,
-      "step": 1081
-    },
-    {
-      "epoch": 0.5322183964584358,
-      "grad_norm": 0.5530214993105562,
-      "learning_rate": 0.00019987255456725326,
-      "loss": 0.966,
-      "step": 1082
-    },
-    {
-      "epoch": 0.5327102803738317,
-      "grad_norm": 0.49400308451337255,
-      "learning_rate": 0.00019986942786900876,
-      "loss": 0.9563,
-      "step": 1083
-    },
-    {
-      "epoch": 0.5332021642892277,
-      "grad_norm": 0.5066660160933474,
-      "learning_rate": 0.00019986626330459383,
-      "loss": 0.9655,
-      "step": 1084
-    },
-    {
-      "epoch": 0.5336940482046237,
-      "grad_norm": 0.48185408793585677,
-      "learning_rate": 0.00019986306087520838,
-      "loss": 0.8772,
-      "step": 1085
-    },
-    {
-      "epoch": 0.5341859321200196,
-      "grad_norm": 0.5570005012138782,
-      "learning_rate": 0.0001998598205820666,
-      "loss": 1.0364,
-      "step": 1086
-    },
-    {
-      "epoch": 0.5346778160354156,
-      "grad_norm": 0.4649427073411664,
-      "learning_rate": 0.00019985654242639709,
-      "loss": 0.9675,
-      "step": 1087
-    },
-    {
-      "epoch": 0.5351696999508117,
-      "grad_norm": 0.5067231805794385,
-      "learning_rate": 0.00019985322640944276,
-      "loss": 0.9838,
-      "step": 1088
-    },
-    {
-      "epoch": 0.5356615838662075,
-      "grad_norm": 0.5434555214676685,
-      "learning_rate": 0.0001998498725324609,
-      "loss": 0.9817,
-      "step": 1089
-    },
-    {
-      "epoch": 0.5361534677816036,
-      "grad_norm": 0.5006330847052317,
-      "learning_rate": 0.0001998464807967232,
-      "loss": 0.9275,
-      "step": 1090
-    },
-    {
-      "epoch": 0.5366453516969996,
-      "grad_norm": 0.48403912095323887,
-      "learning_rate": 0.0001998430512035156,
-      "loss": 0.98,
-      "step": 1091
-    },
-    {
-      "epoch": 0.5371372356123955,
-      "grad_norm": 0.8596836891088817,
-      "learning_rate": 0.0001998395837541385,
-      "loss": 0.9859,
-      "step": 1092
-    },
-    {
-      "epoch": 0.5376291195277915,
-      "grad_norm": 0.5041580903732553,
-      "learning_rate": 0.00019983607844990662,
-      "loss": 1.0673,
-      "step": 1093
-    },
-    {
-      "epoch": 0.5381210034431874,
-      "grad_norm": 0.5221800187321489,
-      "learning_rate": 0.00019983253529214892,
-      "loss": 0.9914,
-      "step": 1094
-    },
-    {
-      "epoch": 0.5386128873585834,
-      "grad_norm": 0.5192987066849946,
-      "learning_rate": 0.00019982895428220893,
-      "loss": 0.9408,
-      "step": 1095
-    },
-    {
-      "epoch": 0.5391047712739794,
-      "grad_norm": 0.8540398302865382,
-      "learning_rate": 0.00019982533542144438,
-      "loss": 0.9425,
-      "step": 1096
-    },
-    {
-      "epoch": 0.5395966551893753,
-      "grad_norm": 0.5089958604203334,
-      "learning_rate": 0.0001998216787112273,
-      "loss": 0.9503,
-      "step": 1097
-    },
-    {
-      "epoch": 0.5400885391047713,
-      "grad_norm": 0.5214487492607818,
-      "learning_rate": 0.0001998179841529443,
-      "loss": 0.9329,
-      "step": 1098
-    },
-    {
-      "epoch": 0.5405804230201673,
-      "grad_norm": 0.5145236945942359,
-      "learning_rate": 0.00019981425174799607,
-      "loss": 0.976,
-      "step": 1099
-    },
-    {
-      "epoch": 0.5410723069355632,
-      "grad_norm": 0.5933597180220942,
-      "learning_rate": 0.00019981048149779785,
-      "loss": 0.9122,
-      "step": 1100
-    },
-    {
-      "epoch": 0.5415641908509592,
-      "grad_norm": 1.0519914958129526,
-      "learning_rate": 0.00019980667340377915,
-      "loss": 1.0156,
-      "step": 1101
-    },
-    {
-      "epoch": 0.5420560747663551,
-      "grad_norm": 0.5179856400801137,
-      "learning_rate": 0.00019980282746738385,
-      "loss": 0.9248,
-      "step": 1102
-    },
-    {
-      "epoch": 0.5425479586817511,
-      "grad_norm": 0.5259129704713565,
-      "learning_rate": 0.00019979894369007007,
-      "loss": 0.9591,
-      "step": 1103
-    },
-    {
-      "epoch": 0.5430398425971471,
-      "grad_norm": 0.523228801819585,
-      "learning_rate": 0.00019979502207331047,
-      "loss": 1.023,
-      "step": 1104
-    },
-    {
-      "epoch": 0.543531726512543,
-      "grad_norm": 1.138365192252825,
-      "learning_rate": 0.00019979106261859192,
-      "loss": 0.9991,
-      "step": 1105
-    },
-    {
-      "epoch": 0.544023610427939,
-      "grad_norm": 0.509377639787038,
-      "learning_rate": 0.0001997870653274157,
-      "loss": 0.9631,
-      "step": 1106
-    },
-    {
-      "epoch": 0.544515494343335,
-      "grad_norm": 0.5211249290231439,
-      "learning_rate": 0.00019978303020129734,
-      "loss": 0.9886,
-      "step": 1107
-    },
-    {
-      "epoch": 0.5450073782587309,
-      "grad_norm": 0.5462673297624403,
-      "learning_rate": 0.00019977895724176685,
-      "loss": 1.1203,
-      "step": 1108
-    },
-    {
-      "epoch": 0.5454992621741269,
-      "grad_norm": 0.5584142311810938,
-      "learning_rate": 0.0001997748464503685,
-      "loss": 0.9207,
-      "step": 1109
-    },
-    {
-      "epoch": 0.5459911460895229,
-      "grad_norm": 0.543254237701427,
-      "learning_rate": 0.00019977069782866092,
-      "loss": 1.0187,
-      "step": 1110
-    },
-    {
-      "epoch": 0.5464830300049188,
-      "grad_norm": 0.7324250838584815,
-      "learning_rate": 0.00019976651137821713,
-      "loss": 1.044,
-      "step": 1111
-    },
-    {
-      "epoch": 0.5469749139203148,
-      "grad_norm": 0.5505484318466812,
-      "learning_rate": 0.0001997622871006244,
-      "loss": 0.9891,
-      "step": 1112
-    },
-    {
-      "epoch": 0.5474667978357107,
-      "grad_norm": 0.6030395790239121,
-      "learning_rate": 0.00019975802499748438,
-      "loss": 0.9314,
-      "step": 1113
-    },
-    {
-      "epoch": 0.5479586817511067,
-      "grad_norm": 0.5111429687043084,
-      "learning_rate": 0.00019975372507041313,
-      "loss": 0.9423,
-      "step": 1114
-    },
-    {
-      "epoch": 0.5484505656665027,
-      "grad_norm": 0.5204916826717528,
-      "learning_rate": 0.00019974938732104095,
-      "loss": 0.9904,
-      "step": 1115
-    },
-    {
-      "epoch": 0.5489424495818986,
-      "grad_norm": 0.54202739853523,
-      "learning_rate": 0.00019974501175101258,
-      "loss": 1.0219,
-      "step": 1116
-    },
-    {
-      "epoch": 0.5494343334972946,
-      "grad_norm": 0.5908784365319103,
-      "learning_rate": 0.00019974059836198697,
-      "loss": 0.9277,
-      "step": 1117
-    },
-    {
-      "epoch": 0.5499262174126907,
-      "grad_norm": 0.5817756149253414,
-      "learning_rate": 0.00019973614715563758,
-      "loss": 0.9432,
-      "step": 1118
-    },
-    {
-      "epoch": 0.5504181013280866,
-      "grad_norm": 3.5849082606997693,
-      "learning_rate": 0.00019973165813365204,
-      "loss": 0.9221,
-      "step": 1119
-    },
-    {
-      "epoch": 0.5509099852434826,
-      "grad_norm": 0.5720745065681223,
-      "learning_rate": 0.00019972713129773242,
-      "loss": 1.0715,
-      "step": 1120
-    },
-    {
-      "epoch": 0.5514018691588785,
-      "grad_norm": 0.5050276723287347,
-      "learning_rate": 0.00019972256664959514,
-      "loss": 0.9377,
-      "step": 1121
-    },
-    {
-      "epoch": 0.5518937530742745,
-      "grad_norm": 0.5225068931361693,
-      "learning_rate": 0.00019971796419097082,
-      "loss": 1.008,
-      "step": 1122
-    },
-    {
-      "epoch": 0.5523856369896705,
-      "grad_norm": 0.523775827439192,
-      "learning_rate": 0.0001997133239236046,
-      "loss": 0.9946,
-      "step": 1123
-    },
-    {
-      "epoch": 0.5528775209050664,
-      "grad_norm": 0.48384120477333176,
-      "learning_rate": 0.00019970864584925582,
-      "loss": 0.9144,
-      "step": 1124
-    },
-    {
-      "epoch": 0.5533694048204624,
-      "grad_norm": 0.5114787640705357,
-      "learning_rate": 0.00019970392996969825,
-      "loss": 1.0152,
-      "step": 1125
-    },
-    {
-      "epoch": 0.5538612887358584,
-      "grad_norm": 0.6021346846711332,
-      "learning_rate": 0.0001996991762867199,
-      "loss": 0.9429,
-      "step": 1126
-    },
-    {
-      "epoch": 0.5543531726512543,
-      "grad_norm": 0.5523618315595479,
-      "learning_rate": 0.00019969438480212318,
-      "loss": 1.0008,
-      "step": 1127
-    },
-    {
-      "epoch": 0.5548450565666503,
-      "grad_norm": 0.5380313924950736,
-      "learning_rate": 0.00019968955551772483,
-      "loss": 1.0406,
-      "step": 1128
-    },
-    {
-      "epoch": 0.5553369404820462,
-      "grad_norm": 0.48910840408788725,
-      "learning_rate": 0.00019968468843535592,
-      "loss": 0.9843,
-      "step": 1129
-    },
-    {
-      "epoch": 0.5558288243974422,
-      "grad_norm": 0.5355562704481304,
-      "learning_rate": 0.00019967978355686175,
-      "loss": 0.9486,
-      "step": 1130
-    },
-    {
-      "epoch": 0.5563207083128382,
-      "grad_norm": 1.016066263257823,
-      "learning_rate": 0.00019967484088410212,
-      "loss": 1.0011,
-      "step": 1131
-    },
-    {
-      "epoch": 0.5568125922282341,
-      "grad_norm": 0.5723587103276893,
-      "learning_rate": 0.00019966986041895107,
-      "loss": 1.0672,
-      "step": 1132
-    },
-    {
-      "epoch": 0.5573044761436301,
-      "grad_norm": 0.5703152223843464,
-      "learning_rate": 0.00019966484216329695,
-      "loss": 1.051,
-      "step": 1133
-    },
-    {
-      "epoch": 0.5577963600590261,
-      "grad_norm": 2.282358305729226,
-      "learning_rate": 0.00019965978611904248,
-      "loss": 1.0211,
-      "step": 1134
-    },
-    {
-      "epoch": 0.558288243974422,
-      "grad_norm": 0.5525840883553625,
-      "learning_rate": 0.00019965469228810467,
-      "loss": 0.962,
-      "step": 1135
-    },
-    {
-      "epoch": 0.558780127889818,
-      "grad_norm": 0.5215527967589574,
-      "learning_rate": 0.00019964956067241492,
-      "loss": 0.9639,
-      "step": 1136
-    },
-    {
-      "epoch": 0.559272011805214,
-      "grad_norm": 0.7241043949117516,
-      "learning_rate": 0.0001996443912739189,
-      "loss": 1.0431,
-      "step": 1137
-    },
-    {
-      "epoch": 0.5597638957206099,
-      "grad_norm": 0.6687373612723888,
-      "learning_rate": 0.0001996391840945766,
-      "loss": 0.9453,
-      "step": 1138
-    },
-    {
-      "epoch": 0.5602557796360059,
-      "grad_norm": 0.5545674299287187,
-      "learning_rate": 0.0001996339391363624,
-      "loss": 1.0497,
-      "step": 1139
-    },
-    {
-      "epoch": 0.5607476635514018,
-      "grad_norm": 0.5352437956107986,
-      "learning_rate": 0.00019962865640126495,
-      "loss": 0.9849,
-      "step": 1140
-    },
-    {
-      "epoch": 0.5612395474667978,
-      "grad_norm": 0.6055837858580955,
-      "learning_rate": 0.0001996233358912872,
-      "loss": 0.9791,
-      "step": 1141
-    },
-    {
-      "epoch": 0.5617314313821938,
-      "grad_norm": 0.4846389948189242,
-      "learning_rate": 0.0001996179776084465,
-      "loss": 0.9185,
-      "step": 1142
-    },
-    {
-      "epoch": 0.5622233152975897,
-      "grad_norm": 0.5139802802931714,
-      "learning_rate": 0.00019961258155477447,
-      "loss": 0.966,
-      "step": 1143
-    },
-    {
-      "epoch": 0.5627151992129857,
-      "grad_norm": 0.6870938004187139,
-      "learning_rate": 0.00019960714773231702,
-      "loss": 0.9575,
-      "step": 1144
-    },
-    {
-      "epoch": 0.5632070831283817,
-      "grad_norm": 0.5113594529829487,
-      "learning_rate": 0.00019960167614313447,
-      "loss": 0.9578,
-      "step": 1145
-    },
-    {
-      "epoch": 0.5636989670437776,
-      "grad_norm": 3.9031183834515217,
-      "learning_rate": 0.0001995961667893014,
-      "loss": 1.0727,
-      "step": 1146
-    },
-    {
-      "epoch": 0.5641908509591737,
-      "grad_norm": 0.5574019687648437,
-      "learning_rate": 0.00019959061967290672,
-      "loss": 0.8689,
-      "step": 1147
-    },
-    {
-      "epoch": 0.5646827348745695,
-      "grad_norm": 0.5221563977062531,
-      "learning_rate": 0.0001995850347960536,
-      "loss": 1.0548,
-      "step": 1148
-    },
-    {
-      "epoch": 0.5651746187899656,
-      "grad_norm": 0.5267773017129473,
-      "learning_rate": 0.00019957941216085966,
-      "loss": 0.9983,
-      "step": 1149
-    },
-    {
-      "epoch": 0.5656665027053616,
-      "grad_norm": 0.526975611698099,
-      "learning_rate": 0.00019957375176945675,
-      "loss": 1.0295,
-      "step": 1150
-    },
-    {
-      "epoch": 0.5661583866207575,
-      "grad_norm": 0.577297577131239,
-      "learning_rate": 0.000199568053623991,
-      "loss": 0.9194,
-      "step": 1151
-    },
-    {
-      "epoch": 0.5666502705361535,
-      "grad_norm": 0.502403890036141,
-      "learning_rate": 0.00019956231772662292,
-      "loss": 0.8761,
-      "step": 1152
-    },
-    {
-      "epoch": 0.5671421544515495,
-      "grad_norm": 0.8534168950724117,
-      "learning_rate": 0.00019955654407952735,
-      "loss": 0.9752,
-      "step": 1153
-    },
-    {
-      "epoch": 0.5676340383669454,
-      "grad_norm": 0.4922632287741393,
-      "learning_rate": 0.00019955073268489336,
-      "loss": 1.0172,
-      "step": 1154
-    },
-    {
-      "epoch": 0.5681259222823414,
-      "grad_norm": 0.5336179384604482,
-      "learning_rate": 0.0001995448835449244,
-      "loss": 1.0188,
-      "step": 1155
-    },
-    {
-      "epoch": 0.5686178061977374,
-      "grad_norm": 0.4844042264946027,
-      "learning_rate": 0.00019953899666183824,
-      "loss": 0.9233,
-      "step": 1156
-    },
-    {
-      "epoch": 0.5691096901131333,
-      "grad_norm": 0.507882364764521,
-      "learning_rate": 0.00019953307203786688,
-      "loss": 0.957,
-      "step": 1157
-    },
-    {
-      "epoch": 0.5696015740285293,
-      "grad_norm": 0.591541215870336,
-      "learning_rate": 0.0001995271096752567,
-      "loss": 0.9356,
-      "step": 1158
-    },
-    {
-      "epoch": 0.5700934579439252,
-      "grad_norm": 0.6186382260669341,
-      "learning_rate": 0.0001995211095762684,
-      "loss": 0.9369,
-      "step": 1159
-    },
-    {
-      "epoch": 0.5705853418593212,
-      "grad_norm": 0.5616378717608467,
-      "learning_rate": 0.0001995150717431769,
-      "loss": 0.9946,
-      "step": 1160
-    },
-    {
-      "epoch": 0.5710772257747172,
-      "grad_norm": 0.5799201922520058,
-      "learning_rate": 0.00019950899617827153,
-      "loss": 1.0119,
-      "step": 1161
-    },
-    {
-      "epoch": 0.5715691096901131,
-      "grad_norm": 0.5253536377945148,
-      "learning_rate": 0.00019950288288385587,
-      "loss": 1.0241,
-      "step": 1162
-    },
-    {
-      "epoch": 0.5720609936055091,
-      "grad_norm": 0.5309270499178786,
-      "learning_rate": 0.00019949673186224782,
-      "loss": 1.0884,
-      "step": 1163
-    },
-    {
-      "epoch": 0.5725528775209051,
-      "grad_norm": 0.47403768845019634,
-      "learning_rate": 0.00019949054311577957,
-      "loss": 0.8939,
-      "step": 1164
-    },
-    {
-      "epoch": 0.573044761436301,
-      "grad_norm": 0.47354505756863635,
-      "learning_rate": 0.00019948431664679764,
-      "loss": 0.9801,
-      "step": 1165
-    },
-    {
-      "epoch": 0.573536645351697,
-      "grad_norm": 1.58911521402257,
-      "learning_rate": 0.0001994780524576628,
-      "loss": 0.9705,
-      "step": 1166
-    },
-    {
-      "epoch": 0.5740285292670929,
-      "grad_norm": 0.7408555909645079,
-      "learning_rate": 0.00019947175055075021,
-      "loss": 1.0306,
-      "step": 1167
-    },
-    {
-      "epoch": 0.5745204131824889,
-      "grad_norm": 0.5432257419818766,
-      "learning_rate": 0.0001994654109284493,
-      "loss": 0.976,
-      "step": 1168
-    },
-    {
-      "epoch": 0.5750122970978849,
-      "grad_norm": 0.49371878758607346,
-      "learning_rate": 0.00019945903359316368,
-      "loss": 0.913,
-      "step": 1169
-    },
-    {
-      "epoch": 0.5755041810132808,
-      "grad_norm": 0.5461157822068061,
-      "learning_rate": 0.0001994526185473115,
-      "loss": 1.0219,
-      "step": 1170
-    },
-    {
-      "epoch": 0.5759960649286768,
-      "grad_norm": 0.5938609107667081,
-      "learning_rate": 0.0001994461657933249,
-      "loss": 0.9624,
-      "step": 1171
-    },
-    {
-      "epoch": 0.5764879488440728,
-      "grad_norm": 2.2098702053471238,
-      "learning_rate": 0.00019943967533365061,
-      "loss": 1.1616,
-      "step": 1172
-    },
-    {
-      "epoch": 0.5769798327594687,
-      "grad_norm": 0.5867682418120316,
-      "learning_rate": 0.00019943314717074952,
-      "loss": 0.9791,
-      "step": 1173
-    },
-    {
-      "epoch": 0.5774717166748647,
-      "grad_norm": 0.6149364776768556,
-      "learning_rate": 0.0001994265813070968,
-      "loss": 1.045,
-      "step": 1174
-    },
-    {
-      "epoch": 0.5779636005902608,
-      "grad_norm": 0.580305060422531,
-      "learning_rate": 0.00019941997774518193,
-      "loss": 1.0056,
-      "step": 1175
-    },
-    {
-      "epoch": 0.5784554845056566,
-      "grad_norm": 0.576886147710524,
-      "learning_rate": 0.0001994133364875087,
-      "loss": 0.9277,
-      "step": 1176
-    },
-    {
-      "epoch": 0.5789473684210527,
-      "grad_norm": 0.5761376880421364,
-      "learning_rate": 0.00019940665753659522,
-      "loss": 1.007,
-      "step": 1177
-    },
-    {
-      "epoch": 0.5794392523364486,
-      "grad_norm": 0.538725198753228,
-      "learning_rate": 0.00019939994089497384,
-      "loss": 0.9512,
-      "step": 1178
-    },
-    {
-      "epoch": 0.5799311362518446,
-      "grad_norm": 0.5906770112529988,
-      "learning_rate": 0.00019939318656519119,
-      "loss": 1.0418,
-      "step": 1179
-    },
-    {
-      "epoch": 0.5804230201672406,
-      "grad_norm": 0.5235094859182425,
-      "learning_rate": 0.00019938639454980826,
-      "loss": 0.9882,
-      "step": 1180
-    },
-    {
-      "epoch": 0.5809149040826365,
-      "grad_norm": 0.5289012994896675,
-      "learning_rate": 0.0001993795648514003,
-      "loss": 0.9857,
-      "step": 1181
-    },
-    {
-      "epoch": 0.5814067879980325,
-      "grad_norm": 0.5185599758845685,
-      "learning_rate": 0.0001993726974725568,
-      "loss": 0.9397,
-      "step": 1182
-    },
-    {
-      "epoch": 0.5818986719134285,
-      "grad_norm": 0.4900753841084743,
-      "learning_rate": 0.0001993657924158816,
-      "loss": 0.9016,
-      "step": 1183
-    },
-    {
-      "epoch": 0.5823905558288244,
-      "grad_norm": 0.5322886410713817,
-      "learning_rate": 0.00019935884968399277,
-      "loss": 0.9766,
-      "step": 1184
-    },
-    {
-      "epoch": 0.5828824397442204,
-      "grad_norm": 0.49780528534311047,
-      "learning_rate": 0.0001993518692795227,
-      "loss": 1.0136,
-      "step": 1185
-    },
-    {
-      "epoch": 0.5833743236596163,
-      "grad_norm": 0.5118283118364332,
-      "learning_rate": 0.0001993448512051181,
-      "loss": 1.0426,
-      "step": 1186
-    },
-    {
-      "epoch": 0.5838662075750123,
-      "grad_norm": 0.5038756348443018,
-      "learning_rate": 0.00019933779546343983,
-      "loss": 1.017,
-      "step": 1187
-    },
-    {
-      "epoch": 0.5843580914904083,
-      "grad_norm": 0.5377683615665799,
-      "learning_rate": 0.00019933070205716328,
-      "loss": 0.9851,
-      "step": 1188
-    },
-    {
-      "epoch": 0.5848499754058042,
-      "grad_norm": 0.4875958919404516,
-      "learning_rate": 0.00019932357098897775,
-      "loss": 0.9638,
-      "step": 1189
-    },
-    {
-      "epoch": 0.5853418593212002,
-      "grad_norm": 0.6352966387789497,
-      "learning_rate": 0.0001993164022615872,
-      "loss": 0.9676,
-      "step": 1190
-    },
-    {
-      "epoch": 0.5858337432365962,
-      "grad_norm": 0.5477111727893049,
-      "learning_rate": 0.00019930919587770967,
-      "loss": 1.035,
-      "step": 1191
-    },
-    {
-      "epoch": 0.5863256271519921,
-      "grad_norm": 0.49050963486189864,
-      "learning_rate": 0.00019930195184007747,
-      "loss": 0.9813,
-      "step": 1192
-    },
-    {
-      "epoch": 0.5868175110673881,
-      "grad_norm": 1.6622331873903418,
-      "learning_rate": 0.00019929467015143723,
-      "loss": 1.0632,
-      "step": 1193
-    },
-    {
-      "epoch": 0.5873093949827841,
-      "grad_norm": 0.5400732997976715,
-      "learning_rate": 0.00019928735081454986,
-      "loss": 1.0002,
-      "step": 1194
-    },
-    {
-      "epoch": 0.58780127889818,
-      "grad_norm": 0.5174335529269121,
-      "learning_rate": 0.00019927999383219055,
-      "loss": 1.0093,
-      "step": 1195
-    },
-    {
-      "epoch": 0.588293162813576,
-      "grad_norm": 0.5367307503099527,
-      "learning_rate": 0.00019927259920714873,
-      "loss": 0.9744,
-      "step": 1196
-    },
-    {
-      "epoch": 0.5887850467289719,
-      "grad_norm": 0.510879953323215,
-      "learning_rate": 0.00019926516694222817,
-      "loss": 0.9288,
-      "step": 1197
-    },
-    {
-      "epoch": 0.5892769306443679,
-      "grad_norm": 0.5591430284311459,
-      "learning_rate": 0.0001992576970402468,
-      "loss": 1.0922,
-      "step": 1198
-    },
-    {
-      "epoch": 0.5897688145597639,
-      "grad_norm": 0.68786396492244,
-      "learning_rate": 0.00019925018950403688,
-      "loss": 1.0722,
-      "step": 1199
-    },
-    {
-      "epoch": 0.5902606984751598,
-      "grad_norm": 0.5197894423013034,
-      "learning_rate": 0.000199242644336445,
-      "loss": 0.9052,
-      "step": 1200
-    },
-    {
-      "epoch": 0.5907525823905558,
-      "grad_norm": 0.540254752280019,
-      "learning_rate": 0.00019923506154033197,
-      "loss": 0.9469,
-      "step": 1201
-    },
-    {
-      "epoch": 0.5912444663059518,
-      "grad_norm": 0.5093910387920063,
-      "learning_rate": 0.00019922744111857278,
-      "loss": 1.0185,
-      "step": 1202
-    },
-    {
-      "epoch": 0.5917363502213477,
-      "grad_norm": 0.46809841121528273,
-      "learning_rate": 0.00019921978307405684,
-      "loss": 0.9627,
-      "step": 1203
-    },
-    {
-      "epoch": 0.5922282341367437,
-      "grad_norm": 0.504913090098908,
-      "learning_rate": 0.00019921208740968769,
-      "loss": 1.0126,
-      "step": 1204
-    },
-    {
-      "epoch": 0.5927201180521396,
-      "grad_norm": 0.8418977179759569,
-      "learning_rate": 0.00019920435412838322,
-      "loss": 0.9788,
-      "step": 1205
-    },
-    {
-      "epoch": 0.5932120019675357,
-      "grad_norm": 0.5038224844723608,
-      "learning_rate": 0.00019919658323307559,
-      "loss": 0.9563,
-      "step": 1206
-    },
-    {
-      "epoch": 0.5937038858829317,
-      "grad_norm": 0.5071611037513739,
-      "learning_rate": 0.0001991887747267111,
-      "loss": 0.9306,
-      "step": 1207
-    },
-    {
-      "epoch": 0.5941957697983276,
-      "grad_norm": 0.4857967851656961,
-      "learning_rate": 0.0001991809286122505,
-      "loss": 0.8986,
-      "step": 1208
-    },
-    {
-      "epoch": 0.5946876537137236,
-      "grad_norm": 0.47897108328724575,
-      "learning_rate": 0.0001991730448926686,
-      "loss": 0.9278,
-      "step": 1209
-    },
-    {
-      "epoch": 0.5951795376291196,
-      "grad_norm": 0.4990242027411677,
-      "learning_rate": 0.00019916512357095467,
-      "loss": 0.9378,
-      "step": 1210
-    },
-    {
-      "epoch": 0.5956714215445155,
-      "grad_norm": 0.505568450647096,
-      "learning_rate": 0.00019915716465011208,
-      "loss": 1.0102,
-      "step": 1211
-    },
-    {
-      "epoch": 0.5961633054599115,
-      "grad_norm": 0.4829731223656389,
-      "learning_rate": 0.00019914916813315844,
-      "loss": 0.9408,
-      "step": 1212
-    },
-    {
-      "epoch": 0.5966551893753075,
-      "grad_norm": 0.4668168249073365,
-      "learning_rate": 0.00019914113402312583,
-      "loss": 0.9245,
-      "step": 1213
-    },
-    {
-      "epoch": 0.5971470732907034,
-      "grad_norm": 0.5323856425343063,
-      "learning_rate": 0.0001991330623230603,
-      "loss": 0.8927,
-      "step": 1214
-    },
-    {
-      "epoch": 0.5976389572060994,
-      "grad_norm": 0.5509541021295224,
-      "learning_rate": 0.00019912495303602237,
-      "loss": 1.0505,
-      "step": 1215
-    },
-    {
-      "epoch": 0.5981308411214953,
-      "grad_norm": 0.488808472489773,
-      "learning_rate": 0.00019911680616508672,
-      "loss": 0.9372,
-      "step": 1216
-    },
-    {
-      "epoch": 0.5986227250368913,
-      "grad_norm": 0.5384313294484143,
-      "learning_rate": 0.00019910862171334227,
-      "loss": 0.9781,
-      "step": 1217
-    },
-    {
-      "epoch": 0.5991146089522873,
-      "grad_norm": 0.5379481451615976,
-      "learning_rate": 0.00019910039968389223,
-      "loss": 0.9996,
-      "step": 1218
-    },
-    {
-      "epoch": 0.5996064928676832,
-      "grad_norm": 0.4961902093761553,
-      "learning_rate": 0.00019909214007985402,
-      "loss": 0.9365,
-      "step": 1219
-    },
-    {
-      "epoch": 0.6000983767830792,
-      "grad_norm": 0.5302134253201395,
-      "learning_rate": 0.00019908384290435934,
-      "loss": 0.9992,
-      "step": 1220
-    },
-    {
-      "epoch": 0.6005902606984752,
-      "grad_norm": 0.5672680562736032,
-      "learning_rate": 0.00019907550816055408,
-      "loss": 1.0397,
-      "step": 1221
-    },
-    {
-      "epoch": 0.6010821446138711,
-      "grad_norm": 0.46181095225446894,
-      "learning_rate": 0.00019906713585159848,
-      "loss": 0.9246,
-      "step": 1222
-    },
-    {
-      "epoch": 0.6015740285292671,
-      "grad_norm": 0.5811155767195028,
-      "learning_rate": 0.00019905872598066692,
-      "loss": 1.0781,
-      "step": 1223
-    },
-    {
-      "epoch": 0.602065912444663,
-      "grad_norm": 0.49787784398474455,
-      "learning_rate": 0.00019905027855094808,
-      "loss": 0.9936,
-      "step": 1224
-    },
-    {
-      "epoch": 0.602557796360059,
-      "grad_norm": 0.5149302314602863,
-      "learning_rate": 0.00019904179356564482,
-      "loss": 0.9727,
-      "step": 1225
-    },
-    {
-      "epoch": 0.603049680275455,
-      "grad_norm": 0.5116850293447144,
-      "learning_rate": 0.00019903327102797433,
-      "loss": 0.9836,
-      "step": 1226
-    },
-    {
-      "epoch": 0.6035415641908509,
-      "grad_norm": 0.485077962795156,
-      "learning_rate": 0.00019902471094116795,
-      "loss": 0.9013,
-      "step": 1227
-    },
-    {
-      "epoch": 0.6040334481062469,
-      "grad_norm": 0.5385744118882313,
-      "learning_rate": 0.00019901611330847132,
-      "loss": 0.9741,
-      "step": 1228
-    },
-    {
-      "epoch": 0.6045253320216429,
-      "grad_norm": 0.4745263928640794,
-      "learning_rate": 0.00019900747813314426,
-      "loss": 0.9313,
-      "step": 1229
-    },
-    {
-      "epoch": 0.6050172159370388,
-      "grad_norm": 0.5045603418455769,
-      "learning_rate": 0.0001989988054184609,
-      "loss": 0.9431,
-      "step": 1230
-    },
-    {
-      "epoch": 0.6055090998524348,
-      "grad_norm": 0.5058346863207296,
-      "learning_rate": 0.00019899009516770953,
-      "loss": 1.0214,
-      "step": 1231
-    },
-    {
-      "epoch": 0.6060009837678307,
-      "grad_norm": 0.47596987066450536,
-      "learning_rate": 0.00019898134738419268,
-      "loss": 0.8772,
-      "step": 1232
-    },
-    {
-      "epoch": 0.6064928676832267,
-      "grad_norm": 0.4769610363097206,
-      "learning_rate": 0.00019897256207122719,
-      "loss": 1.0398,
-      "step": 1233
-    },
-    {
-      "epoch": 0.6069847515986228,
-      "grad_norm": 0.5993985205285812,
-      "learning_rate": 0.000198963739232144,
-      "loss": 0.9669,
-      "step": 1234
-    },
-    {
-      "epoch": 0.6074766355140186,
-      "grad_norm": 0.5040906972160556,
-      "learning_rate": 0.00019895487887028841,
-      "loss": 0.9493,
-      "step": 1235
-    },
-    {
-      "epoch": 0.6079685194294147,
-      "grad_norm": 0.4917920275499752,
-      "learning_rate": 0.00019894598098901988,
-      "loss": 0.9296,
-      "step": 1236
-    },
-    {
-      "epoch": 0.6084604033448107,
-      "grad_norm": 0.43862876354585245,
-      "learning_rate": 0.00019893704559171202,
-      "loss": 0.8733,
-      "step": 1237
-    },
-    {
-      "epoch": 0.6089522872602066,
-      "grad_norm": 0.5183451409292822,
-      "learning_rate": 0.00019892807268175283,
-      "loss": 1.0264,
-      "step": 1238
-    },
-    {
-      "epoch": 0.6094441711756026,
-      "grad_norm": 0.8865642362813331,
-      "learning_rate": 0.0001989190622625445,
-      "loss": 1.0417,
-      "step": 1239
-    },
-    {
-      "epoch": 0.6099360550909986,
-      "grad_norm": 0.5153707302595621,
-      "learning_rate": 0.00019891001433750325,
-      "loss": 0.97,
-      "step": 1240
-    },
-    {
-      "epoch": 0.6104279390063945,
-      "grad_norm": 0.5403104228621635,
-      "learning_rate": 0.00019890092891005975,
-      "loss": 1.0097,
-      "step": 1241
-    },
-    {
-      "epoch": 0.6109198229217905,
-      "grad_norm": 0.5487171784369699,
-      "learning_rate": 0.00019889180598365878,
-      "loss": 0.8768,
-      "step": 1242
-    },
-    {
-      "epoch": 0.6114117068371864,
-      "grad_norm": 0.5709273233038086,
-      "learning_rate": 0.00019888264556175935,
-      "loss": 0.9997,
-      "step": 1243
-    },
-    {
-      "epoch": 0.6119035907525824,
-      "grad_norm": 0.5536470447507988,
-      "learning_rate": 0.00019887344764783475,
-      "loss": 1.053,
-      "step": 1244
-    },
-    {
-      "epoch": 0.6123954746679784,
-      "grad_norm": 0.49447844921601386,
-      "learning_rate": 0.00019886421224537239,
-      "loss": 0.9231,
-      "step": 1245
-    },
-    {
-      "epoch": 0.6128873585833743,
-      "grad_norm": 0.47170407712589835,
-      "learning_rate": 0.0001988549393578739,
-      "loss": 0.8496,
-      "step": 1246
-    },
-    {
-      "epoch": 0.6133792424987703,
-      "grad_norm": 0.5020565154047474,
-      "learning_rate": 0.0001988456289888552,
-      "loss": 0.9665,
-      "step": 1247
-    },
-    {
-      "epoch": 0.6138711264141663,
-      "grad_norm": 0.5190702352179734,
-      "learning_rate": 0.0001988362811418464,
-      "loss": 1.01,
-      "step": 1248
-    },
-    {
-      "epoch": 0.6143630103295622,
-      "grad_norm": 0.4880703088966864,
-      "learning_rate": 0.00019882689582039175,
-      "loss": 0.904,
-      "step": 1249
-    },
-    {
-      "epoch": 0.6148548942449582,
-      "grad_norm": 0.47574493833336534,
-      "learning_rate": 0.0001988174730280498,
-      "loss": 0.9665,
-      "step": 1250
-    },
-    {
-      "epoch": 0.6153467781603541,
-      "grad_norm": 0.5148282636836543,
-      "learning_rate": 0.00019880801276839325,
-      "loss": 0.9884,
-      "step": 1251
-    },
-    {
-      "epoch": 0.6158386620757501,
-      "grad_norm": 0.5235160138996259,
-      "learning_rate": 0.000198798515045009,
-      "loss": 0.9887,
-      "step": 1252
-    },
-    {
-      "epoch": 0.6163305459911461,
-      "grad_norm": 0.5085056626140083,
-      "learning_rate": 0.00019878897986149824,
-      "loss": 0.9317,
-      "step": 1253
-    },
-    {
-      "epoch": 0.616822429906542,
-      "grad_norm": 0.5445136987605921,
-      "learning_rate": 0.0001987794072214762,
-      "loss": 1.0149,
-      "step": 1254
-    },
-    {
-      "epoch": 0.617314313821938,
-      "grad_norm": 0.4930655828797307,
-      "learning_rate": 0.0001987697971285725,
-      "loss": 0.9999,
-      "step": 1255
-    },
-    {
-      "epoch": 0.617806197737334,
-      "grad_norm": 0.465599374379101,
-      "learning_rate": 0.0001987601495864308,
-      "loss": 0.8981,
-      "step": 1256
-    },
-    {
-      "epoch": 0.6182980816527299,
-      "grad_norm": 0.5238286908374731,
-      "learning_rate": 0.0001987504645987091,
-      "loss": 0.9931,
-      "step": 1257
-    },
-    {
-      "epoch": 0.6187899655681259,
-      "grad_norm": 0.4479033955221958,
-      "learning_rate": 0.0001987407421690795,
-      "loss": 0.9034,
-      "step": 1258
-    },
-    {
-      "epoch": 0.6192818494835219,
-      "grad_norm": 0.46701498124585433,
-      "learning_rate": 0.0001987309823012283,
-      "loss": 0.9245,
-      "step": 1259
-    },
-    {
-      "epoch": 0.6197737333989178,
-      "grad_norm": 0.48555597397308403,
-      "learning_rate": 0.0001987211849988561,
-      "loss": 0.939,
-      "step": 1260
-    },
-    {
-      "epoch": 0.6202656173143138,
-      "grad_norm": 0.5381445047831256,
-      "learning_rate": 0.00019871135026567748,
-      "loss": 0.9292,
-      "step": 1261
-    },
-    {
-      "epoch": 0.6207575012297097,
-      "grad_norm": 0.5216798455491977,
-      "learning_rate": 0.00019870147810542148,
-      "loss": 0.9461,
-      "step": 1262
-    },
-    {
-      "epoch": 0.6212493851451057,
-      "grad_norm": 0.5029059724053576,
-      "learning_rate": 0.00019869156852183112,
-      "loss": 0.9747,
-      "step": 1263
-    },
-    {
-      "epoch": 0.6217412690605018,
-      "grad_norm": 0.47676542086643253,
-      "learning_rate": 0.00019868162151866371,
-      "loss": 0.9085,
-      "step": 1264
-    },
-    {
-      "epoch": 0.6222331529758977,
-      "grad_norm": 0.4891345162549392,
-      "learning_rate": 0.0001986716370996907,
-      "loss": 1.0671,
-      "step": 1265
-    },
-    {
-      "epoch": 0.6227250368912937,
-      "grad_norm": 0.6370750809827045,
-      "learning_rate": 0.0001986616152686978,
-      "loss": 0.9901,
-      "step": 1266
-    },
-    {
-      "epoch": 0.6232169208066897,
-      "grad_norm": 0.45380153940893,
-      "learning_rate": 0.0001986515560294848,
-      "loss": 0.9482,
-      "step": 1267
-    },
-    {
-      "epoch": 0.6237088047220856,
-      "grad_norm": 0.49341269333462745,
-      "learning_rate": 0.00019864145938586574,
-      "loss": 0.9426,
-      "step": 1268
-    },
-    {
-      "epoch": 0.6242006886374816,
-      "grad_norm": 0.5444189677286736,
-      "learning_rate": 0.00019863132534166886,
-      "loss": 1.0082,
-      "step": 1269
-    },
-    {
-      "epoch": 0.6246925725528775,
-      "grad_norm": 0.4401484698050124,
-      "learning_rate": 0.00019862115390073654,
-      "loss": 0.9203,
-      "step": 1270
-    },
-    {
-      "epoch": 0.6251844564682735,
-      "grad_norm": 0.559815902272095,
-      "learning_rate": 0.00019861094506692535,
-      "loss": 0.9326,
-      "step": 1271
-    },
-    {
-      "epoch": 0.6256763403836695,
-      "grad_norm": 0.5123525668852091,
-      "learning_rate": 0.00019860069884410604,
-      "loss": 0.9681,
-      "step": 1272
-    },
-    {
-      "epoch": 0.6261682242990654,
-      "grad_norm": 0.5021160581899634,
-      "learning_rate": 0.0001985904152361635,
-      "loss": 1.0304,
-      "step": 1273
-    },
-    {
-      "epoch": 0.6266601082144614,
-      "grad_norm": 0.6376400508102466,
-      "learning_rate": 0.00019858009424699686,
-      "loss": 0.9213,
-      "step": 1274
-    },
-    {
-      "epoch": 0.6271519921298574,
-      "grad_norm": 0.4733845157988784,
-      "learning_rate": 0.00019856973588051942,
-      "loss": 0.9236,
-      "step": 1275
-    },
-    {
-      "epoch": 0.6276438760452533,
-      "grad_norm": 0.49203738543908704,
-      "learning_rate": 0.00019855934014065857,
-      "loss": 1.0325,
-      "step": 1276
-    },
-    {
-      "epoch": 0.6281357599606493,
-      "grad_norm": 0.46907799363323144,
-      "learning_rate": 0.00019854890703135597,
-      "loss": 1.0026,
-      "step": 1277
-    },
-    {
-      "epoch": 0.6286276438760453,
-      "grad_norm": 0.45569539638688455,
-      "learning_rate": 0.00019853843655656737,
-      "loss": 0.9547,
-      "step": 1278
-    },
-    {
-      "epoch": 0.6291195277914412,
-      "grad_norm": 0.49872856225546003,
-      "learning_rate": 0.00019852792872026274,
-      "loss": 0.9291,
-      "step": 1279
-    },
-    {
-      "epoch": 0.6296114117068372,
-      "grad_norm": 0.5584940782919813,
-      "learning_rate": 0.0001985173835264262,
-      "loss": 0.947,
-      "step": 1280
-    },
-    {
-      "epoch": 0.6301032956222331,
-      "grad_norm": 0.5770565277492296,
-      "learning_rate": 0.00019850680097905602,
-      "loss": 0.9817,
-      "step": 1281
-    },
-    {
-      "epoch": 0.6305951795376291,
-      "grad_norm": 0.48890068341395715,
-      "learning_rate": 0.00019849618108216466,
-      "loss": 0.9959,
-      "step": 1282
-    },
-    {
-      "epoch": 0.6310870634530251,
-      "grad_norm": 0.49117862102377047,
-      "learning_rate": 0.00019848552383977872,
-      "loss": 0.9741,
-      "step": 1283
-    },
-    {
-      "epoch": 0.631578947368421,
-      "grad_norm": 0.5035023302634901,
-      "learning_rate": 0.00019847482925593895,
-      "loss": 0.9934,
-      "step": 1284
-    },
-    {
-      "epoch": 0.632070831283817,
-      "grad_norm": 0.4952634613565297,
-      "learning_rate": 0.0001984640973347003,
-      "loss": 0.8869,
-      "step": 1285
-    },
-    {
-      "epoch": 0.632562715199213,
-      "grad_norm": 0.49315288560193077,
-      "learning_rate": 0.00019845332808013182,
-      "loss": 0.9259,
-      "step": 1286
-    },
-    {
-      "epoch": 0.6330545991146089,
-      "grad_norm": 0.47740099661206553,
-      "learning_rate": 0.00019844252149631676,
-      "loss": 0.9996,
-      "step": 1287
-    },
-    {
-      "epoch": 0.6335464830300049,
-      "grad_norm": 0.46708786273246256,
-      "learning_rate": 0.00019843167758735254,
-      "loss": 0.9772,
-      "step": 1288
-    },
-    {
-      "epoch": 0.6340383669454008,
-      "grad_norm": 0.4705108678393681,
-      "learning_rate": 0.00019842079635735066,
-      "loss": 0.9669,
-      "step": 1289
-    },
-    {
-      "epoch": 0.6345302508607968,
-      "grad_norm": 0.5466072337028248,
-      "learning_rate": 0.00019840987781043684,
-      "loss": 0.9398,
-      "step": 1290
-    },
-    {
-      "epoch": 0.6350221347761928,
-      "grad_norm": 0.4874484074629659,
-      "learning_rate": 0.0001983989219507509,
-      "loss": 1.0452,
-      "step": 1291
-    },
-    {
-      "epoch": 0.6355140186915887,
-      "grad_norm": 0.46611852462326536,
-      "learning_rate": 0.0001983879287824468,
-      "loss": 0.8998,
-      "step": 1292
-    },
-    {
-      "epoch": 0.6360059026069848,
-      "grad_norm": 0.5081887752375128,
-      "learning_rate": 0.00019837689830969274,
-      "loss": 1.0055,
-      "step": 1293
-    },
-    {
-      "epoch": 0.6364977865223808,
-      "grad_norm": 0.4677023400470637,
-      "learning_rate": 0.00019836583053667095,
-      "loss": 0.9816,
-      "step": 1294
-    },
-    {
-      "epoch": 0.6369896704377767,
-      "grad_norm": 0.47492532520981673,
-      "learning_rate": 0.0001983547254675779,
-      "loss": 0.9368,
-      "step": 1295
-    },
-    {
-      "epoch": 0.6374815543531727,
-      "grad_norm": 0.49346995310388697,
-      "learning_rate": 0.0001983435831066241,
-      "loss": 0.9676,
-      "step": 1296
-    },
-    {
-      "epoch": 0.6379734382685687,
-      "grad_norm": 0.47036017948332953,
-      "learning_rate": 0.00019833240345803427,
-      "loss": 0.9913,
-      "step": 1297
-    },
-    {
-      "epoch": 0.6384653221839646,
-      "grad_norm": 0.48065973778314525,
-      "learning_rate": 0.00019832118652604727,
-      "loss": 0.8905,
-      "step": 1298
-    },
-    {
-      "epoch": 0.6389572060993606,
-      "grad_norm": 0.46647813373553965,
-      "learning_rate": 0.00019830993231491605,
-      "loss": 0.8878,
-      "step": 1299
-    },
-    {
-      "epoch": 0.6394490900147565,
-      "grad_norm": 0.5114555581077286,
-      "learning_rate": 0.00019829864082890772,
-      "loss": 0.9685,
-      "step": 1300
-    },
-    {
-      "epoch": 0.6399409739301525,
-      "grad_norm": 0.4965084945995025,
-      "learning_rate": 0.00019828731207230355,
-      "loss": 1.0056,
-      "step": 1301
-    },
-    {
-      "epoch": 0.6404328578455485,
-      "grad_norm": 0.5214245592678145,
-      "learning_rate": 0.0001982759460493989,
-      "loss": 1.0231,
-      "step": 1302
-    },
-    {
-      "epoch": 0.6409247417609444,
-      "grad_norm": 0.46714813612874007,
-      "learning_rate": 0.00019826454276450326,
-      "loss": 0.9114,
-      "step": 1303
-    },
-    {
-      "epoch": 0.6414166256763404,
-      "grad_norm": 0.49141817093497603,
-      "learning_rate": 0.00019825310222194026,
-      "loss": 0.9891,
-      "step": 1304
-    },
-    {
-      "epoch": 0.6419085095917364,
-      "grad_norm": 0.460648180573316,
-      "learning_rate": 0.0001982416244260477,
-      "loss": 0.9866,
-      "step": 1305
-    },
-    {
-      "epoch": 0.6424003935071323,
-      "grad_norm": 0.4887121271829397,
-      "learning_rate": 0.00019823010938117743,
-      "loss": 0.9695,
-      "step": 1306
-    },
-    {
-      "epoch": 0.6428922774225283,
-      "grad_norm": 0.5027267845537975,
-      "learning_rate": 0.00019821855709169545,
-      "loss": 0.921,
-      "step": 1307
-    },
-    {
-      "epoch": 0.6433841613379242,
-      "grad_norm": 0.48475309963053514,
-      "learning_rate": 0.00019820696756198193,
-      "loss": 0.9635,
-      "step": 1308
-    },
-    {
-      "epoch": 0.6438760452533202,
-      "grad_norm": 0.48365476545550495,
-      "learning_rate": 0.00019819534079643108,
-      "loss": 0.9205,
-      "step": 1309
-    },
-    {
-      "epoch": 0.6443679291687162,
-      "grad_norm": 0.5501815549047151,
-      "learning_rate": 0.00019818367679945128,
-      "loss": 1.0474,
-      "step": 1310
-    },
-    {
-      "epoch": 0.6448598130841121,
-      "grad_norm": 0.502069360256173,
-      "learning_rate": 0.000198171975575465,
-      "loss": 0.9187,
-      "step": 1311
-    },
-    {
-      "epoch": 0.6453516969995081,
-      "grad_norm": 0.48678101208708974,
-      "learning_rate": 0.0001981602371289089,
-      "loss": 0.9325,
-      "step": 1312
-    },
-    {
-      "epoch": 0.6458435809149041,
-      "grad_norm": 0.506409525085648,
-      "learning_rate": 0.0001981484614642336,
-      "loss": 0.9562,
-      "step": 1313
-    },
-    {
-      "epoch": 0.6463354648303,
-      "grad_norm": 0.4661370633764652,
-      "learning_rate": 0.00019813664858590397,
-      "loss": 0.9183,
-      "step": 1314
-    },
-    {
-      "epoch": 0.646827348745696,
-      "grad_norm": 0.4806886534468273,
-      "learning_rate": 0.00019812479849839893,
-      "loss": 0.9439,
-      "step": 1315
-    },
-    {
-      "epoch": 0.647319232661092,
-      "grad_norm": 0.47126460868823805,
-      "learning_rate": 0.00019811291120621155,
-      "loss": 0.951,
-      "step": 1316
-    },
-    {
-      "epoch": 0.6478111165764879,
-      "grad_norm": 0.4542406666227788,
-      "learning_rate": 0.00019810098671384894,
-      "loss": 0.9383,
-      "step": 1317
-    },
-    {
-      "epoch": 0.6483030004918839,
-      "grad_norm": 0.4994704028630951,
-      "learning_rate": 0.0001980890250258324,
-      "loss": 0.9752,
-      "step": 1318
-    },
-    {
-      "epoch": 0.6487948844072798,
-      "grad_norm": 0.4930725970805956,
-      "learning_rate": 0.00019807702614669723,
-      "loss": 0.9901,
-      "step": 1319
-    },
-    {
-      "epoch": 0.6492867683226758,
-      "grad_norm": 0.45835005141114954,
-      "learning_rate": 0.0001980649900809929,
-      "loss": 0.9519,
-      "step": 1320
-    },
-    {
-      "epoch": 0.6497786522380719,
-      "grad_norm": 0.48788496599573516,
-      "learning_rate": 0.000198052916833283,
-      "loss": 0.8987,
-      "step": 1321
-    },
-    {
-      "epoch": 0.6502705361534677,
-      "grad_norm": 0.5118753458003786,
-      "learning_rate": 0.00019804080640814514,
-      "loss": 0.9839,
-      "step": 1322
-    },
-    {
-      "epoch": 0.6507624200688638,
-      "grad_norm": 0.4831313513667437,
-      "learning_rate": 0.00019802865881017115,
-      "loss": 1.021,
-      "step": 1323
-    },
-    {
-      "epoch": 0.6512543039842598,
-      "grad_norm": 0.5060562265707746,
-      "learning_rate": 0.00019801647404396676,
-      "loss": 0.9929,
-      "step": 1324
-    },
-    {
-      "epoch": 0.6517461878996557,
-      "grad_norm": 0.4782416160343314,
-      "learning_rate": 0.00019800425211415195,
-      "loss": 1.0068,
-      "step": 1325
-    },
-    {
-      "epoch": 0.6522380718150517,
-      "grad_norm": 0.48910673323559956,
-      "learning_rate": 0.0001979919930253608,
-      "loss": 0.9691,
-      "step": 1326
-    },
-    {
-      "epoch": 0.6527299557304476,
-      "grad_norm": 0.548161295704732,
-      "learning_rate": 0.00019797969678224134,
-      "loss": 0.9615,
-      "step": 1327
-    },
-    {
-      "epoch": 0.6532218396458436,
-      "grad_norm": 0.4363558582637419,
-      "learning_rate": 0.00019796736338945588,
-      "loss": 0.8841,
-      "step": 1328
-    },
-    {
-      "epoch": 0.6537137235612396,
-      "grad_norm": 0.4364977172112859,
-      "learning_rate": 0.00019795499285168059,
-      "loss": 1.0105,
-      "step": 1329
-    },
-    {
-      "epoch": 0.6542056074766355,
-      "grad_norm": 0.49734668944721677,
-      "learning_rate": 0.00019794258517360594,
-      "loss": 1.0651,
-      "step": 1330
-    },
-    {
-      "epoch": 0.6546974913920315,
-      "grad_norm": 0.45890224119065426,
-      "learning_rate": 0.00019793014035993634,
-      "loss": 0.9658,
-      "step": 1331
-    },
-    {
-      "epoch": 0.6551893753074275,
-      "grad_norm": 0.47898528377522087,
-      "learning_rate": 0.0001979176584153903,
-      "loss": 1.0042,
-      "step": 1332
-    },
-    {
-      "epoch": 0.6556812592228234,
-      "grad_norm": 0.47836600015020336,
-      "learning_rate": 0.0001979051393447005,
-      "loss": 0.8972,
-      "step": 1333
-    },
-    {
-      "epoch": 0.6561731431382194,
-      "grad_norm": 0.6329808475765831,
-      "learning_rate": 0.0001978925831526136,
-      "loss": 0.9891,
-      "step": 1334
-    },
-    {
-      "epoch": 0.6566650270536154,
-      "grad_norm": 0.5212393325985413,
-      "learning_rate": 0.00019787998984389035,
-      "loss": 1.0377,
-      "step": 1335
-    },
-    {
-      "epoch": 0.6571569109690113,
-      "grad_norm": 0.5012305937786911,
-      "learning_rate": 0.00019786735942330558,
-      "loss": 0.9978,
-      "step": 1336
-    },
-    {
-      "epoch": 0.6576487948844073,
-      "grad_norm": 0.5522986750214255,
-      "learning_rate": 0.00019785469189564827,
-      "loss": 1.0115,
-      "step": 1337
-    },
-    {
-      "epoch": 0.6581406787998032,
-      "grad_norm": 0.6066439726715466,
-      "learning_rate": 0.0001978419872657213,
-      "loss": 1.03,
-      "step": 1338
-    },
-    {
-      "epoch": 0.6586325627151992,
-      "grad_norm": 0.46082593127000604,
-      "learning_rate": 0.00019782924553834178,
-      "loss": 0.959,
-      "step": 1339
-    },
-    {
-      "epoch": 0.6591244466305952,
-      "grad_norm": 0.5021208184636669,
-      "learning_rate": 0.00019781646671834083,
-      "loss": 0.9057,
-      "step": 1340
-    },
-    {
-      "epoch": 0.6596163305459911,
-      "grad_norm": 0.4844685339555825,
-      "learning_rate": 0.00019780365081056357,
-      "loss": 0.9454,
-      "step": 1341
-    },
-    {
-      "epoch": 0.6601082144613871,
-      "grad_norm": 1.9164930910785711,
-      "learning_rate": 0.00019779079781986932,
-      "loss": 0.9429,
-      "step": 1342
-    },
-    {
-      "epoch": 0.6606000983767831,
-      "grad_norm": 0.6216687573184569,
-      "learning_rate": 0.0001977779077511313,
-      "loss": 0.9998,
-      "step": 1343
-    },
-    {
-      "epoch": 0.661091982292179,
-      "grad_norm": 0.4977795406068352,
-      "learning_rate": 0.0001977649806092369,
-      "loss": 0.9701,
-      "step": 1344
-    },
-    {
-      "epoch": 0.661583866207575,
-      "grad_norm": 0.5243996014098755,
-      "learning_rate": 0.00019775201639908753,
-      "loss": 0.8677,
-      "step": 1345
-    },
-    {
-      "epoch": 0.6620757501229709,
-      "grad_norm": 0.4812749828993618,
-      "learning_rate": 0.00019773901512559866,
-      "loss": 0.9539,
-      "step": 1346
-    },
-    {
-      "epoch": 0.6625676340383669,
-      "grad_norm": 0.546259167840415,
-      "learning_rate": 0.00019772597679369982,
-      "loss": 0.9659,
-      "step": 1347
-    },
-    {
-      "epoch": 0.663059517953763,
-      "grad_norm": 0.45627340502765906,
-      "learning_rate": 0.00019771290140833457,
-      "loss": 0.9403,
-      "step": 1348
-    },
-    {
-      "epoch": 0.6635514018691588,
-      "grad_norm": 0.5198080634215432,
-      "learning_rate": 0.00019769978897446052,
-      "loss": 0.9559,
-      "step": 1349
-    },
-    {
-      "epoch": 0.6640432857845548,
-      "grad_norm": 0.5238002127306112,
-      "learning_rate": 0.00019768663949704934,
-      "loss": 1.0228,
-      "step": 1350
-    },
-    {
-      "epoch": 0.6645351696999509,
-      "grad_norm": 0.5208634930014021,
-      "learning_rate": 0.0001976734529810868,
-      "loss": 1.0147,
-      "step": 1351
-    },
-    {
-      "epoch": 0.6650270536153468,
-      "grad_norm": 0.5248961222985504,
-      "learning_rate": 0.0001976602294315726,
-      "loss": 0.9727,
-      "step": 1352
-    },
-    {
-      "epoch": 0.6655189375307428,
-      "grad_norm": 0.6254830439164408,
-      "learning_rate": 0.00019764696885352052,
-      "loss": 0.9649,
-      "step": 1353
-    },
-    {
-      "epoch": 0.6660108214461387,
-      "grad_norm": 0.47110297597909734,
-      "learning_rate": 0.00019763367125195846,
-      "loss": 0.9369,
-      "step": 1354
-    },
-    {
-      "epoch": 0.6665027053615347,
-      "grad_norm": 0.47187579909328614,
-      "learning_rate": 0.00019762033663192825,
-      "loss": 0.94,
-      "step": 1355
-    },
-    {
-      "epoch": 0.6669945892769307,
-      "grad_norm": 0.5772950964597373,
-      "learning_rate": 0.00019760696499848581,
-      "loss": 1.0795,
-      "step": 1356
-    },
-    {
-      "epoch": 0.6674864731923266,
-      "grad_norm": 0.4882232660721926,
-      "learning_rate": 0.00019759355635670108,
-      "loss": 0.95,
-      "step": 1357
-    },
-    {
-      "epoch": 0.6679783571077226,
-      "grad_norm": 0.5487075048985032,
-      "learning_rate": 0.00019758011071165806,
-      "loss": 0.9049,
-      "step": 1358
-    },
-    {
-      "epoch": 0.6684702410231186,
-      "grad_norm": 0.5135018536157103,
-      "learning_rate": 0.00019756662806845476,
-      "loss": 0.9378,
-      "step": 1359
-    },
-    {
-      "epoch": 0.6689621249385145,
-      "grad_norm": 0.5064050586331412,
-      "learning_rate": 0.0001975531084322032,
-      "loss": 1.0298,
-      "step": 1360
-    },
-    {
-      "epoch": 0.6694540088539105,
-      "grad_norm": 0.48251250028109743,
-      "learning_rate": 0.00019753955180802947,
-      "loss": 1.0018,
-      "step": 1361
-    },
-    {
-      "epoch": 0.6699458927693065,
-      "grad_norm": 0.465794161860618,
-      "learning_rate": 0.00019752595820107357,
-      "loss": 0.9575,
-      "step": 1362
-    },
-    {
-      "epoch": 0.6704377766847024,
-      "grad_norm": 0.46197345974106235,
-      "learning_rate": 0.00019751232761648968,
-      "loss": 0.9247,
-      "step": 1363
-    },
-    {
-      "epoch": 0.6709296606000984,
-      "grad_norm": 0.4872691668123498,
-      "learning_rate": 0.00019749866005944596,
-      "loss": 0.9579,
-      "step": 1364
-    },
-    {
-      "epoch": 0.6714215445154943,
-      "grad_norm": 0.5246678631985456,
-      "learning_rate": 0.00019748495553512446,
-      "loss": 0.9364,
-      "step": 1365
-    },
-    {
-      "epoch": 0.6719134284308903,
-      "grad_norm": 0.44793193058858854,
-      "learning_rate": 0.0001974712140487214,
-      "loss": 0.961,
-      "step": 1366
-    },
-    {
-      "epoch": 0.6724053123462863,
-      "grad_norm": 0.448453558256479,
-      "learning_rate": 0.000197457435605447,
-      "loss": 0.8742,
-      "step": 1367
-    },
-    {
-      "epoch": 0.6728971962616822,
-      "grad_norm": 0.43038781161979683,
-      "learning_rate": 0.00019744362021052538,
-      "loss": 0.8682,
-      "step": 1368
-    },
-    {
-      "epoch": 0.6733890801770782,
-      "grad_norm": 0.5163316258068651,
-      "learning_rate": 0.00019742976786919477,
-      "loss": 1.0356,
-      "step": 1369
-    },
-    {
-      "epoch": 0.6738809640924742,
-      "grad_norm": 0.44590154547646654,
-      "learning_rate": 0.00019741587858670734,
-      "loss": 0.8705,
-      "step": 1370
-    },
-    {
-      "epoch": 0.6743728480078701,
-      "grad_norm": 0.4936612873848726,
-      "learning_rate": 0.0001974019523683294,
-      "loss": 1.068,
-      "step": 1371
-    },
-    {
-      "epoch": 0.6748647319232661,
-      "grad_norm": 0.4693563025941603,
-      "learning_rate": 0.00019738798921934106,
-      "loss": 0.9408,
-      "step": 1372
-    },
-    {
-      "epoch": 0.675356615838662,
-      "grad_norm": 0.4809559613105093,
-      "learning_rate": 0.00019737398914503659,
-      "loss": 0.9443,
-      "step": 1373
-    },
-    {
-      "epoch": 0.675848499754058,
-      "grad_norm": 0.5138724513993868,
-      "learning_rate": 0.00019735995215072424,
-      "loss": 1.0152,
-      "step": 1374
-    },
-    {
-      "epoch": 0.676340383669454,
-      "grad_norm": 0.5055424715655393,
-      "learning_rate": 0.0001973458782417262,
-      "loss": 1.0696,
-      "step": 1375
-    },
-    {
-      "epoch": 0.6768322675848499,
-      "grad_norm": 0.4667703894682567,
-      "learning_rate": 0.0001973317674233787,
-      "loss": 0.9734,
-      "step": 1376
-    },
-    {
-      "epoch": 0.6773241515002459,
-      "grad_norm": 0.5384300104335755,
-      "learning_rate": 0.00019731761970103194,
-      "loss": 1.0105,
-      "step": 1377
-    },
-    {
-      "epoch": 0.677816035415642,
-      "grad_norm": 0.4655179511026166,
-      "learning_rate": 0.0001973034350800501,
-      "loss": 0.9076,
-      "step": 1378
-    },
-    {
-      "epoch": 0.6783079193310378,
-      "grad_norm": 0.5347175128994112,
-      "learning_rate": 0.00019728921356581145,
-      "loss": 0.9431,
-      "step": 1379
-    },
-    {
-      "epoch": 0.6787998032464339,
-      "grad_norm": 0.4656927796859899,
-      "learning_rate": 0.0001972749551637081,
-      "loss": 0.9451,
-      "step": 1380
-    },
-    {
-      "epoch": 0.6792916871618299,
-      "grad_norm": 0.48177774646127997,
-      "learning_rate": 0.00019726065987914627,
-      "loss": 0.9401,
-      "step": 1381
-    },
-    {
-      "epoch": 0.6797835710772258,
-      "grad_norm": 0.5123952935408527,
-      "learning_rate": 0.0001972463277175461,
-      "loss": 1.0117,
-      "step": 1382
-    },
-    {
-      "epoch": 0.6802754549926218,
-      "grad_norm": 0.47515439460474146,
-      "learning_rate": 0.00019723195868434173,
-      "loss": 0.8981,
-      "step": 1383
-    },
-    {
-      "epoch": 0.6807673389080177,
-      "grad_norm": 0.4758402685560721,
-      "learning_rate": 0.00019721755278498125,
-      "loss": 1.0036,
-      "step": 1384
-    },
-    {
-      "epoch": 0.6812592228234137,
-      "grad_norm": 0.4368396159224531,
-      "learning_rate": 0.00019720311002492674,
-      "loss": 0.8823,
-      "step": 1385
-    },
-    {
-      "epoch": 0.6817511067388097,
-      "grad_norm": 0.49245664440439657,
-      "learning_rate": 0.00019718863040965433,
-      "loss": 0.9077,
-      "step": 1386
-    },
-    {
-      "epoch": 0.6822429906542056,
-      "grad_norm": 0.45367647317868026,
-      "learning_rate": 0.00019717411394465403,
-      "loss": 0.8926,
-      "step": 1387
-    },
-    {
-      "epoch": 0.6827348745696016,
-      "grad_norm": 0.44195369077293045,
-      "learning_rate": 0.00019715956063542987,
-      "loss": 0.8891,
-      "step": 1388
-    },
-    {
-      "epoch": 0.6832267584849976,
-      "grad_norm": 0.46903923611598164,
-      "learning_rate": 0.00019714497048749983,
-      "loss": 0.9259,
-      "step": 1389
-    },
-    {
-      "epoch": 0.6837186424003935,
-      "grad_norm": 0.465613582505028,
-      "learning_rate": 0.00019713034350639586,
-      "loss": 0.9348,
-      "step": 1390
-    },
-    {
-      "epoch": 0.6842105263157895,
-      "grad_norm": 0.4582440126316995,
-      "learning_rate": 0.00019711567969766386,
-      "loss": 0.985,
-      "step": 1391
-    },
-    {
-      "epoch": 0.6847024102311854,
-      "grad_norm": 0.4392419955464664,
-      "learning_rate": 0.0001971009790668638,
-      "loss": 0.9216,
-      "step": 1392
-    },
-    {
-      "epoch": 0.6851942941465814,
-      "grad_norm": 0.4491326846139963,
-      "learning_rate": 0.00019708624161956937,
-      "loss": 0.961,
-      "step": 1393
-    },
-    {
-      "epoch": 0.6856861780619774,
-      "grad_norm": 0.4696919061376751,
-      "learning_rate": 0.0001970714673613685,
-      "loss": 0.9931,
-      "step": 1394
-    },
-    {
-      "epoch": 0.6861780619773733,
-      "grad_norm": 0.4884458175997585,
-      "learning_rate": 0.0001970566562978629,
-      "loss": 0.9676,
-      "step": 1395
-    },
-    {
-      "epoch": 0.6866699458927693,
-      "grad_norm": 0.525966371118796,
-      "learning_rate": 0.00019704180843466832,
-      "loss": 0.9762,
-      "step": 1396
-    },
-    {
-      "epoch": 0.6871618298081653,
-      "grad_norm": 0.44826665427966983,
-      "learning_rate": 0.00019702692377741437,
-      "loss": 0.9858,
-      "step": 1397
-    },
-    {
-      "epoch": 0.6876537137235612,
-      "grad_norm": 0.46657972521671054,
-      "learning_rate": 0.0001970120023317447,
-      "loss": 0.9516,
-      "step": 1398
-    },
-    {
-      "epoch": 0.6881455976389572,
-      "grad_norm": 0.4902442987303898,
-      "learning_rate": 0.0001969970441033169,
-      "loss": 0.9438,
-      "step": 1399
-    },
-    {
-      "epoch": 0.6886374815543532,
-      "grad_norm": 0.4967055119940703,
-      "learning_rate": 0.00019698204909780244,
-      "loss": 0.9215,
-      "step": 1400
-    },
-    {
-      "epoch": 0.6891293654697491,
-      "grad_norm": 0.4407476049359459,
-      "learning_rate": 0.00019696701732088682,
-      "loss": 0.9008,
-      "step": 1401
-    },
-    {
-      "epoch": 0.6896212493851451,
-      "grad_norm": 0.48602316707693566,
-      "learning_rate": 0.00019695194877826942,
-      "loss": 0.9536,
-      "step": 1402
-    },
-    {
-      "epoch": 0.690113133300541,
-      "grad_norm": 0.4338969018010427,
-      "learning_rate": 0.00019693684347566357,
-      "loss": 0.9118,
-      "step": 1403
-    },
-    {
-      "epoch": 0.690605017215937,
-      "grad_norm": 0.4617062326813443,
-      "learning_rate": 0.00019692170141879655,
-      "loss": 0.962,
-      "step": 1404
-    },
-    {
-      "epoch": 0.691096901131333,
-      "grad_norm": 0.4853395184959136,
-      "learning_rate": 0.0001969065226134096,
-      "loss": 0.9644,
-      "step": 1405
-    },
-    {
-      "epoch": 0.6915887850467289,
-      "grad_norm": 0.45556059903509,
-      "learning_rate": 0.00019689130706525783,
-      "loss": 0.9241,
-      "step": 1406
-    },
-    {
-      "epoch": 0.692080668962125,
-      "grad_norm": 0.4845371013575807,
-      "learning_rate": 0.00019687605478011037,
-      "loss": 0.9443,
-      "step": 1407
-    },
-    {
-      "epoch": 0.692572552877521,
-      "grad_norm": 0.5027071999875199,
-      "learning_rate": 0.00019686076576375016,
-      "loss": 1.0104,
-      "step": 1408
-    },
-    {
-      "epoch": 0.6930644367929168,
-      "grad_norm": 0.49588782762936817,
-      "learning_rate": 0.00019684544002197417,
-      "loss": 0.9522,
-      "step": 1409
-    },
-    {
-      "epoch": 0.6935563207083129,
-      "grad_norm": 0.49351571536552424,
-      "learning_rate": 0.00019683007756059325,
-      "loss": 0.8928,
-      "step": 1410
-    },
-    {
-      "epoch": 0.6940482046237088,
-      "grad_norm": 0.46301918368683015,
-      "learning_rate": 0.00019681467838543224,
-      "loss": 1.0038,
-      "step": 1411
-    },
-    {
-      "epoch": 0.6945400885391048,
-      "grad_norm": 0.45749522584031316,
-      "learning_rate": 0.00019679924250232974,
-      "loss": 0.9363,
-      "step": 1412
-    },
-    {
-      "epoch": 0.6950319724545008,
-      "grad_norm": 0.44525199283556677,
-      "learning_rate": 0.00019678376991713844,
-      "loss": 0.9343,
-      "step": 1413
-    },
-    {
-      "epoch": 0.6955238563698967,
-      "grad_norm": 0.4744116189440004,
-      "learning_rate": 0.00019676826063572483,
-      "loss": 0.978,
-      "step": 1414
-    },
-    {
-      "epoch": 0.6960157402852927,
-      "grad_norm": 0.46033717354531234,
-      "learning_rate": 0.00019675271466396943,
-      "loss": 0.9656,
-      "step": 1415
-    },
-    {
-      "epoch": 0.6965076242006887,
-      "grad_norm": 0.4836554184941867,
-      "learning_rate": 0.00019673713200776653,
-      "loss": 0.9155,
-      "step": 1416
-    },
-    {
-      "epoch": 0.6969995081160846,
-      "grad_norm": 0.46460747714447387,
-      "learning_rate": 0.0001967215126730244,
-      "loss": 0.8926,
-      "step": 1417
-    },
-    {
-      "epoch": 0.6974913920314806,
-      "grad_norm": 0.4583940507740178,
-      "learning_rate": 0.0001967058566656653,
-      "loss": 0.9965,
-      "step": 1418
-    },
-    {
-      "epoch": 0.6979832759468766,
-      "grad_norm": 1.748934833714311,
-      "learning_rate": 0.0001966901639916252,
-      "loss": 1.0506,
-      "step": 1419
-    },
-    {
-      "epoch": 0.6984751598622725,
-      "grad_norm": 0.5196255454458959,
-      "learning_rate": 0.0001966744346568542,
-      "loss": 0.9866,
-      "step": 1420
-    },
-    {
-      "epoch": 0.6989670437776685,
-      "grad_norm": 0.44519596488889035,
-      "learning_rate": 0.00019665866866731604,
-      "loss": 0.9717,
-      "step": 1421
-    },
-    {
-      "epoch": 0.6994589276930644,
-      "grad_norm": 0.4829070905960186,
-      "learning_rate": 0.0001966428660289886,
-      "loss": 0.9013,
-      "step": 1422
-    },
-    {
-      "epoch": 0.6999508116084604,
-      "grad_norm": 0.4924165229937081,
-      "learning_rate": 0.00019662702674786358,
-      "loss": 0.9345,
-      "step": 1423
-    },
-    {
-      "epoch": 0.7004426955238564,
-      "grad_norm": 0.4501450802072641,
-      "learning_rate": 0.00019661115082994648,
-      "loss": 0.9368,
-      "step": 1424
-    },
-    {
-      "epoch": 0.7009345794392523,
-      "grad_norm": 0.44427865956883583,
-      "learning_rate": 0.0001965952382812568,
-      "loss": 0.9414,
-      "step": 1425
-    },
-    {
-      "epoch": 0.7014264633546483,
-      "grad_norm": 0.44543902575339583,
-      "learning_rate": 0.00019657928910782788,
-      "loss": 0.9622,
-      "step": 1426
-    },
-    {
-      "epoch": 0.7019183472700443,
-      "grad_norm": 0.4700780537106658,
-      "learning_rate": 0.00019656330331570696,
-      "loss": 1.0538,
-      "step": 1427
-    },
-    {
-      "epoch": 0.7024102311854402,
-      "grad_norm": 0.43588081453106137,
-      "learning_rate": 0.00019654728091095516,
-      "loss": 0.8978,
-      "step": 1428
-    },
-    {
-      "epoch": 0.7029021151008362,
-      "grad_norm": 0.4843955149609755,
-      "learning_rate": 0.00019653122189964748,
-      "loss": 0.9875,
-      "step": 1429
-    },
-    {
-      "epoch": 0.7033939990162321,
-      "grad_norm": 0.4495943285880614,
-      "learning_rate": 0.00019651512628787284,
-      "loss": 0.9256,
-      "step": 1430
-    },
-    {
-      "epoch": 0.7038858829316281,
-      "grad_norm": 0.5071515618320818,
-      "learning_rate": 0.00019649899408173392,
-      "loss": 0.9805,
-      "step": 1431
-    },
-    {
-      "epoch": 0.7043777668470241,
-      "grad_norm": 0.4946709894083941,
-      "learning_rate": 0.00019648282528734743,
-      "loss": 1.0207,
-      "step": 1432
-    },
-    {
-      "epoch": 0.70486965076242,
-      "grad_norm": 0.5317316907653367,
-      "learning_rate": 0.00019646661991084386,
-      "loss": 0.9049,
-      "step": 1433
-    },
-    {
-      "epoch": 0.705361534677816,
-      "grad_norm": 0.47497952868667054,
-      "learning_rate": 0.00019645037795836757,
-      "loss": 1.0027,
-      "step": 1434
-    },
-    {
-      "epoch": 0.705853418593212,
-      "grad_norm": 0.42811971576461416,
-      "learning_rate": 0.0001964340994360768,
-      "loss": 0.8511,
-      "step": 1435
-    },
-    {
-      "epoch": 0.7063453025086079,
-      "grad_norm": 0.47933110770990467,
-      "learning_rate": 0.00019641778435014367,
-      "loss": 0.9524,
-      "step": 1436
-    },
-    {
-      "epoch": 0.706837186424004,
-      "grad_norm": 0.44032343135979374,
-      "learning_rate": 0.0001964014327067542,
-      "loss": 0.9693,
-      "step": 1437
-    },
-    {
-      "epoch": 0.7073290703394,
-      "grad_norm": 0.45546567885492706,
-      "learning_rate": 0.00019638504451210818,
-      "loss": 0.9352,
-      "step": 1438
-    },
-    {
-      "epoch": 0.7078209542547959,
-      "grad_norm": 0.4367818108711072,
-      "learning_rate": 0.0001963686197724193,
-      "loss": 0.8951,
-      "step": 1439
-    },
-    {
-      "epoch": 0.7083128381701919,
-      "grad_norm": 0.4519902395124329,
-      "learning_rate": 0.00019635215849391513,
-      "loss": 0.8983,
-      "step": 1440
-    },
-    {
-      "epoch": 0.7088047220855878,
-      "grad_norm": 0.49348356075891786,
-      "learning_rate": 0.0001963356606828371,
-      "loss": 0.9997,
-      "step": 1441
-    },
-    {
-      "epoch": 0.7092966060009838,
-      "grad_norm": 0.4313215319354549,
-      "learning_rate": 0.00019631912634544038,
-      "loss": 0.8413,
-      "step": 1442
-    },
-    {
-      "epoch": 0.7097884899163798,
-      "grad_norm": 0.48002828957831445,
-      "learning_rate": 0.00019630255548799418,
-      "loss": 0.9967,
-      "step": 1443
-    },
-    {
-      "epoch": 0.7102803738317757,
-      "grad_norm": 0.4623049897510101,
-      "learning_rate": 0.0001962859481167814,
-      "loss": 0.862,
-      "step": 1444
-    },
-    {
-      "epoch": 0.7107722577471717,
-      "grad_norm": 0.48333998986430504,
-      "learning_rate": 0.00019626930423809883,
-      "loss": 1.0378,
-      "step": 1445
-    },
-    {
-      "epoch": 0.7112641416625677,
-      "grad_norm": 0.4565159520400276,
-      "learning_rate": 0.00019625262385825713,
-      "loss": 0.9619,
-      "step": 1446
-    },
-    {
-      "epoch": 0.7117560255779636,
-      "grad_norm": 0.4620869378230684,
-      "learning_rate": 0.00019623590698358078,
-      "loss": 0.8501,
-      "step": 1447
-    },
-    {
-      "epoch": 0.7122479094933596,
-      "grad_norm": 0.44254477361234446,
-      "learning_rate": 0.0001962191536204081,
-      "loss": 0.9782,
-      "step": 1448
-    },
-    {
-      "epoch": 0.7127397934087555,
-      "grad_norm": 0.4353683138462511,
-      "learning_rate": 0.0001962023637750912,
-      "loss": 0.8796,
-      "step": 1449
-    },
-    {
-      "epoch": 0.7132316773241515,
-      "grad_norm": 0.46120579313640847,
-      "learning_rate": 0.00019618553745399614,
-      "loss": 0.946,
-      "step": 1450
-    },
-    {
-      "epoch": 0.7137235612395475,
-      "grad_norm": 0.47686488482110984,
-      "learning_rate": 0.00019616867466350265,
-      "loss": 0.9315,
-      "step": 1451
-    },
-    {
-      "epoch": 0.7142154451549434,
-      "grad_norm": 0.4633726294481254,
-      "learning_rate": 0.0001961517754100044,
-      "loss": 0.9626,
-      "step": 1452
-    },
-    {
-      "epoch": 0.7147073290703394,
-      "grad_norm": 0.45707485377320534,
-      "learning_rate": 0.00019613483969990888,
-      "loss": 0.9778,
-      "step": 1453
-    },
-    {
-      "epoch": 0.7151992129857354,
-      "grad_norm": 0.45880063101939245,
-      "learning_rate": 0.0001961178675396374,
-      "loss": 0.9926,
-      "step": 1454
-    },
-    {
-      "epoch": 0.7156910969011313,
-      "grad_norm": 0.46807713742194085,
-      "learning_rate": 0.000196100858935625,
-      "loss": 1.0291,
-      "step": 1455
-    },
-    {
-      "epoch": 0.7161829808165273,
-      "grad_norm": 0.4788185630398034,
-      "learning_rate": 0.00019608381389432062,
-      "loss": 0.8905,
-      "step": 1456
-    },
-    {
-      "epoch": 0.7166748647319232,
-      "grad_norm": 0.5116107496857625,
-      "learning_rate": 0.00019606673242218706,
-      "loss": 0.8742,
-      "step": 1457
-    },
-    {
-      "epoch": 0.7171667486473192,
-      "grad_norm": 0.44046305805249486,
-      "learning_rate": 0.00019604961452570084,
-      "loss": 0.8847,
-      "step": 1458
-    },
-    {
-      "epoch": 0.7176586325627152,
-      "grad_norm": 0.4835553328061689,
-      "learning_rate": 0.0001960324602113523,
-      "loss": 0.9298,
-      "step": 1459
-    },
-    {
-      "epoch": 0.7181505164781111,
-      "grad_norm": 0.5134368076985019,
-      "learning_rate": 0.00019601526948564567,
-      "loss": 1.0206,
-      "step": 1460
-    },
-    {
-      "epoch": 0.7186424003935071,
-      "grad_norm": 0.5960705026051694,
-      "learning_rate": 0.00019599804235509888,
-      "loss": 0.9985,
-      "step": 1461
-    },
-    {
-      "epoch": 0.7191342843089031,
-      "grad_norm": 0.621317736308171,
-      "learning_rate": 0.00019598077882624376,
-      "loss": 0.9574,
-      "step": 1462
-    },
-    {
-      "epoch": 0.719626168224299,
-      "grad_norm": 0.489190450109789,
-      "learning_rate": 0.00019596347890562586,
-      "loss": 0.9254,
-      "step": 1463
-    },
-    {
-      "epoch": 0.720118052139695,
-      "grad_norm": 0.6065307822795144,
-      "learning_rate": 0.00019594614259980456,
-      "loss": 0.9965,
-      "step": 1464
-    },
-    {
-      "epoch": 0.720609936055091,
-      "grad_norm": 0.45873329370551547,
-      "learning_rate": 0.00019592876991535304,
-      "loss": 0.9031,
-      "step": 1465
-    },
-    {
-      "epoch": 0.721101819970487,
-      "grad_norm": 0.467556547135862,
-      "learning_rate": 0.0001959113608588583,
-      "loss": 0.9664,
-      "step": 1466
-    },
-    {
-      "epoch": 0.721593703885883,
-      "grad_norm": 0.4758304877114131,
-      "learning_rate": 0.0001958939154369211,
-      "loss": 0.9792,
-      "step": 1467
-    },
-    {
-      "epoch": 0.7220855878012788,
-      "grad_norm": 0.478407499672416,
-      "learning_rate": 0.00019587643365615595,
-      "loss": 0.9107,
-      "step": 1468
-    },
-    {
-      "epoch": 0.7225774717166749,
-      "grad_norm": 0.5020595058918841,
-      "learning_rate": 0.00019585891552319123,
-      "loss": 0.9603,
-      "step": 1469
-    },
-    {
-      "epoch": 0.7230693556320709,
-      "grad_norm": 0.47531218980151596,
-      "learning_rate": 0.000195841361044669,
-      "loss": 0.9525,
-      "step": 1470
-    },
-    {
-      "epoch": 0.7235612395474668,
-      "grad_norm": 0.4309090855161772,
-      "learning_rate": 0.00019582377022724524,
-      "loss": 0.9628,
-      "step": 1471
-    },
-    {
-      "epoch": 0.7240531234628628,
-      "grad_norm": 0.4613319397147935,
-      "learning_rate": 0.00019580614307758952,
-      "loss": 0.972,
-      "step": 1472
-    },
-    {
-      "epoch": 0.7245450073782588,
-      "grad_norm": 0.4714298734153978,
-      "learning_rate": 0.0001957884796023854,
-      "loss": 0.9153,
-      "step": 1473
-    },
-    {
-      "epoch": 0.7250368912936547,
-      "grad_norm": 0.45793842944304836,
-      "learning_rate": 0.00019577077980833006,
-      "loss": 0.9442,
-      "step": 1474
-    },
-    {
-      "epoch": 0.7255287752090507,
-      "grad_norm": 0.4768470879202041,
-      "learning_rate": 0.0001957530437021345,
-      "loss": 0.8783,
-      "step": 1475
-    },
-    {
-      "epoch": 0.7260206591244466,
-      "grad_norm": 0.43647545764853873,
-      "learning_rate": 0.00019573527129052346,
-      "loss": 0.8677,
-      "step": 1476
-    },
-    {
-      "epoch": 0.7265125430398426,
-      "grad_norm": 0.4365078187098896,
-      "learning_rate": 0.00019571746258023555,
-      "loss": 0.967,
-      "step": 1477
-    },
-    {
-      "epoch": 0.7270044269552386,
-      "grad_norm": 0.4545650221688097,
-      "learning_rate": 0.00019569961757802298,
-      "loss": 0.9082,
-      "step": 1478
-    },
-    {
-      "epoch": 0.7274963108706345,
-      "grad_norm": 0.4401653990286907,
-      "learning_rate": 0.00019568173629065183,
-      "loss": 0.909,
-      "step": 1479
-    },
-    {
-      "epoch": 0.7279881947860305,
-      "grad_norm": 0.45995965989755017,
-      "learning_rate": 0.0001956638187249019,
-      "loss": 0.9893,
-      "step": 1480
-    },
-    {
-      "epoch": 0.7284800787014265,
-      "grad_norm": 0.44975805510063777,
-      "learning_rate": 0.0001956458648875668,
-      "loss": 0.9266,
-      "step": 1481
-    },
-    {
-      "epoch": 0.7289719626168224,
-      "grad_norm": 0.4930061220127927,
-      "learning_rate": 0.00019562787478545377,
-      "loss": 1.0045,
-      "step": 1482
-    },
-    {
-      "epoch": 0.7294638465322184,
-      "grad_norm": 0.5347367764789409,
-      "learning_rate": 0.00019560984842538397,
-      "loss": 1.022,
-      "step": 1483
-    },
-    {
-      "epoch": 0.7299557304476144,
-      "grad_norm": 0.4969844454587438,
-      "learning_rate": 0.00019559178581419215,
-      "loss": 0.9755,
-      "step": 1484
-    },
-    {
-      "epoch": 0.7304476143630103,
-      "grad_norm": 0.4771433816539928,
-      "learning_rate": 0.00019557368695872693,
-      "loss": 0.9305,
-      "step": 1485
-    },
-    {
-      "epoch": 0.7309394982784063,
-      "grad_norm": 0.49038254254723584,
-      "learning_rate": 0.00019555555186585055,
-      "loss": 0.8811,
-      "step": 1486
-    },
-    {
-      "epoch": 0.7314313821938022,
-      "grad_norm": 0.4400367538995633,
-      "learning_rate": 0.0001955373805424391,
-      "loss": 0.9098,
-      "step": 1487
-    },
-    {
-      "epoch": 0.7319232661091982,
-      "grad_norm": 0.5074890597051152,
-      "learning_rate": 0.00019551917299538232,
-      "loss": 0.9253,
-      "step": 1488
-    },
-    {
-      "epoch": 0.7324151500245942,
-      "grad_norm": 0.5091464986563263,
-      "learning_rate": 0.00019550092923158378,
-      "loss": 1.0048,
-      "step": 1489
-    },
-    {
-      "epoch": 0.7329070339399901,
-      "grad_norm": 0.472882246063679,
-      "learning_rate": 0.00019548264925796068,
-      "loss": 0.9963,
-      "step": 1490
-    },
-    {
-      "epoch": 0.7333989178553861,
-      "grad_norm": 0.4516101349724856,
-      "learning_rate": 0.00019546433308144403,
-      "loss": 0.9094,
-      "step": 1491
-    },
-    {
-      "epoch": 0.7338908017707821,
-      "grad_norm": 0.5118558111688102,
-      "learning_rate": 0.00019544598070897847,
-      "loss": 0.935,
-      "step": 1492
-    },
-    {
-      "epoch": 0.734382685686178,
-      "grad_norm": 0.45535268218100866,
-      "learning_rate": 0.0001954275921475225,
-      "loss": 0.9147,
-      "step": 1493
-    },
-    {
-      "epoch": 0.734874569601574,
-      "grad_norm": 0.42115777229307744,
-      "learning_rate": 0.0001954091674040482,
-      "loss": 0.91,
-      "step": 1494
-    },
-    {
-      "epoch": 0.7353664535169699,
-      "grad_norm": 0.4832454548127842,
-      "learning_rate": 0.00019539070648554155,
-      "loss": 0.9255,
-      "step": 1495
-    },
-    {
-      "epoch": 0.735858337432366,
-      "grad_norm": 0.4602081923293951,
-      "learning_rate": 0.00019537220939900202,
-      "loss": 0.9809,
-      "step": 1496
-    },
-    {
-      "epoch": 0.736350221347762,
-      "grad_norm": 0.47156152008319946,
-      "learning_rate": 0.0001953536761514429,
-      "loss": 0.9763,
-      "step": 1497
-    },
-    {
-      "epoch": 0.7368421052631579,
-      "grad_norm": 0.5141716716407748,
-      "learning_rate": 0.00019533510674989127,
-      "loss": 0.9673,
-      "step": 1498
-    },
-    {
-      "epoch": 0.7373339891785539,
-      "grad_norm": 0.47149772748268765,
-      "learning_rate": 0.00019531650120138783,
-      "loss": 0.9298,
-      "step": 1499
-    },
-    {
-      "epoch": 0.7378258730939499,
-      "grad_norm": 0.48429056140486454,
-      "learning_rate": 0.00019529785951298697,
-      "loss": 0.9077,
-      "step": 1500
-    },
-    {
-      "epoch": 0.7383177570093458,
-      "grad_norm": 0.5527017998587758,
-      "learning_rate": 0.00019527918169175683,
-      "loss": 1.0596,
-      "step": 1501
-    },
-    {
-      "epoch": 0.7388096409247418,
-      "grad_norm": 0.4619922352446741,
-      "learning_rate": 0.0001952604677447792,
-      "loss": 0.9087,
-      "step": 1502
-    },
-    {
-      "epoch": 0.7393015248401378,
-      "grad_norm": 0.49899778377780707,
-      "learning_rate": 0.00019524171767914967,
-      "loss": 0.9907,
-      "step": 1503
-    },
-    {
-      "epoch": 0.7397934087555337,
-      "grad_norm": 0.4404848899417216,
-      "learning_rate": 0.00019522293150197738,
-      "loss": 0.9308,
-      "step": 1504
-    },
-    {
-      "epoch": 0.7402852926709297,
-      "grad_norm": 0.4562171327754888,
-      "learning_rate": 0.0001952041092203853,
-      "loss": 0.9408,
-      "step": 1505
-    },
-    {
-      "epoch": 0.7407771765863256,
-      "grad_norm": 0.43446031665160695,
-      "learning_rate": 0.00019518525084150995,
-      "loss": 0.926,
-      "step": 1506
-    },
-    {
-      "epoch": 0.7412690605017216,
-      "grad_norm": 0.4900985254829056,
-      "learning_rate": 0.0001951663563725017,
-      "loss": 0.917,
-      "step": 1507
-    },
-    {
-      "epoch": 0.7417609444171176,
-      "grad_norm": 0.5238018635317627,
-      "learning_rate": 0.00019514742582052446,
-      "loss": 1.0288,
-      "step": 1508
-    },
-    {
-      "epoch": 0.7422528283325135,
-      "grad_norm": 0.45483497986719174,
-      "learning_rate": 0.00019512845919275587,
-      "loss": 0.9224,
-      "step": 1509
-    },
-    {
-      "epoch": 0.7427447122479095,
-      "grad_norm": 0.45470957670600526,
-      "learning_rate": 0.0001951094564963873,
-      "loss": 0.9517,
-      "step": 1510
-    },
-    {
-      "epoch": 0.7432365961633055,
-      "grad_norm": 0.4822942918152697,
-      "learning_rate": 0.0001950904177386237,
-      "loss": 1.0147,
-      "step": 1511
-    },
-    {
-      "epoch": 0.7437284800787014,
-      "grad_norm": 0.47794452427478296,
-      "learning_rate": 0.00019507134292668377,
-      "loss": 0.9073,
-      "step": 1512
-    },
-    {
-      "epoch": 0.7442203639940974,
-      "grad_norm": 0.44594834309930764,
-      "learning_rate": 0.00019505223206779987,
-      "loss": 0.8856,
-      "step": 1513
-    },
-    {
-      "epoch": 0.7447122479094933,
-      "grad_norm": 0.46883285255090074,
-      "learning_rate": 0.000195033085169218,
-      "loss": 0.9091,
-      "step": 1514
-    },
-    {
-      "epoch": 0.7452041318248893,
-      "grad_norm": 0.5042459269749029,
-      "learning_rate": 0.00019501390223819782,
-      "loss": 0.9948,
-      "step": 1515
-    },
-    {
-      "epoch": 0.7456960157402853,
-      "grad_norm": 0.5029030858543784,
-      "learning_rate": 0.00019499468328201269,
-      "loss": 0.9345,
-      "step": 1516
-    },
-    {
-      "epoch": 0.7461878996556812,
-      "grad_norm": 0.4252714821508642,
-      "learning_rate": 0.00019497542830794958,
-      "loss": 0.8709,
-      "step": 1517
-    },
-    {
-      "epoch": 0.7466797835710772,
-      "grad_norm": 0.5050335002582659,
-      "learning_rate": 0.00019495613732330917,
-      "loss": 0.9007,
-      "step": 1518
-    },
-    {
-      "epoch": 0.7471716674864732,
-      "grad_norm": 0.4890954632088174,
-      "learning_rate": 0.00019493681033540576,
-      "loss": 0.9823,
-      "step": 1519
-    },
-    {
-      "epoch": 0.7476635514018691,
-      "grad_norm": 0.4553543503953204,
-      "learning_rate": 0.00019491744735156732,
-      "loss": 0.8745,
-      "step": 1520
-    },
-    {
-      "epoch": 0.7481554353172651,
-      "grad_norm": 0.44030988669129767,
-      "learning_rate": 0.00019489804837913543,
-      "loss": 0.9011,
-      "step": 1521
-    },
-    {
-      "epoch": 0.7486473192326611,
-      "grad_norm": 0.4671129225093876,
-      "learning_rate": 0.00019487861342546537,
-      "loss": 0.8703,
-      "step": 1522
-    },
-    {
-      "epoch": 0.749139203148057,
-      "grad_norm": 0.48648288352303504,
-      "learning_rate": 0.00019485914249792603,
-      "loss": 0.9533,
-      "step": 1523
-    },
-    {
-      "epoch": 0.749631087063453,
-      "grad_norm": 0.43543064871374093,
-      "learning_rate": 0.0001948396356038999,
-      "loss": 0.9175,
-      "step": 1524
-    },
-    {
-      "epoch": 0.750122970978849,
-      "grad_norm": 0.48098356204014586,
-      "learning_rate": 0.0001948200927507832,
-      "loss": 0.9555,
-      "step": 1525
-    },
-    {
-      "epoch": 0.750614854894245,
-      "grad_norm": 0.46401849995174804,
-      "learning_rate": 0.00019480051394598572,
-      "loss": 0.9021,
-      "step": 1526
-    },
-    {
-      "epoch": 0.751106738809641,
-      "grad_norm": 0.5480349658843938,
-      "learning_rate": 0.00019478089919693092,
-      "loss": 0.9719,
-      "step": 1527
-    },
-    {
-      "epoch": 0.7515986227250369,
-      "grad_norm": 0.4450166515844905,
-      "learning_rate": 0.00019476124851105578,
-      "loss": 1.0222,
-      "step": 1528
-    },
-    {
-      "epoch": 0.7520905066404329,
-      "grad_norm": 0.46402663986264037,
-      "learning_rate": 0.00019474156189581111,
-      "loss": 0.9842,
-      "step": 1529
-    },
-    {
-      "epoch": 0.7525823905558289,
-      "grad_norm": 0.4983316468313297,
-      "learning_rate": 0.00019472183935866118,
-      "loss": 0.9784,
-      "step": 1530
-    },
-    {
-      "epoch": 0.7530742744712248,
-      "grad_norm": 0.442659315148361,
-      "learning_rate": 0.00019470208090708387,
-      "loss": 0.9043,
-      "step": 1531
-    },
-    {
-      "epoch": 0.7535661583866208,
-      "grad_norm": 0.44801710083887064,
-      "learning_rate": 0.0001946822865485708,
-      "loss": 0.9062,
-      "step": 1532
-    },
-    {
-      "epoch": 0.7540580423020167,
-      "grad_norm": 0.49620844983399426,
-      "learning_rate": 0.0001946624562906271,
-      "loss": 0.9754,
-      "step": 1533
-    },
-    {
-      "epoch": 0.7545499262174127,
-      "grad_norm": 0.4570053805043761,
-      "learning_rate": 0.00019464259014077153,
-      "loss": 1.0158,
-      "step": 1534
-    },
-    {
-      "epoch": 0.7550418101328087,
-      "grad_norm": 0.4604203532938178,
-      "learning_rate": 0.0001946226881065365,
-      "loss": 0.9914,
-      "step": 1535
-    },
-    {
-      "epoch": 0.7555336940482046,
-      "grad_norm": 0.4843548091166105,
-      "learning_rate": 0.00019460275019546802,
-      "loss": 0.997,
-      "step": 1536
-    },
-    {
-      "epoch": 0.7560255779636006,
-      "grad_norm": 0.4539157484540685,
-      "learning_rate": 0.00019458277641512566,
-      "loss": 0.9446,
-      "step": 1537
-    },
-    {
-      "epoch": 0.7565174618789966,
-      "grad_norm": 0.4448221663748266,
-      "learning_rate": 0.00019456276677308262,
-      "loss": 0.8979,
-      "step": 1538
-    },
-    {
-      "epoch": 0.7570093457943925,
-      "grad_norm": 0.45644392016358865,
-      "learning_rate": 0.00019454272127692569,
-      "loss": 0.9854,
-      "step": 1539
-    },
-    {
-      "epoch": 0.7575012297097885,
-      "grad_norm": 0.4623780984854484,
-      "learning_rate": 0.00019452263993425522,
-      "loss": 0.9623,
-      "step": 1540
-    },
-    {
-      "epoch": 0.7579931136251845,
-      "grad_norm": 0.4377138354111479,
-      "learning_rate": 0.00019450252275268528,
-      "loss": 0.9085,
-      "step": 1541
-    },
-    {
-      "epoch": 0.7584849975405804,
-      "grad_norm": 0.4309747845767413,
-      "learning_rate": 0.00019448236973984334,
-      "loss": 0.8849,
-      "step": 1542
-    },
-    {
-      "epoch": 0.7589768814559764,
-      "grad_norm": 0.42781127795203905,
-      "learning_rate": 0.0001944621809033706,
-      "loss": 0.8148,
-      "step": 1543
-    },
-    {
-      "epoch": 0.7594687653713723,
-      "grad_norm": 0.46547586850746064,
-      "learning_rate": 0.0001944419562509218,
-      "loss": 0.9215,
-      "step": 1544
-    },
-    {
-      "epoch": 0.7599606492867683,
-      "grad_norm": 0.4438294386361765,
-      "learning_rate": 0.00019442169579016522,
-      "loss": 0.9195,
-      "step": 1545
-    },
-    {
-      "epoch": 0.7604525332021643,
-      "grad_norm": 0.4403595828515475,
-      "learning_rate": 0.00019440139952878275,
-      "loss": 0.8287,
-      "step": 1546
-    },
-    {
-      "epoch": 0.7609444171175602,
-      "grad_norm": 0.45832092210999603,
-      "learning_rate": 0.0001943810674744699,
-      "loss": 0.8571,
-      "step": 1547
-    },
-    {
-      "epoch": 0.7614363010329562,
-      "grad_norm": 0.49296253953229596,
-      "learning_rate": 0.00019436069963493568,
-      "loss": 0.9385,
-      "step": 1548
-    },
-    {
-      "epoch": 0.7619281849483522,
-      "grad_norm": 0.473185289657026,
-      "learning_rate": 0.0001943402960179027,
-      "loss": 0.933,
-      "step": 1549
-    },
-    {
-      "epoch": 0.7624200688637481,
-      "grad_norm": 0.4296344153592067,
-      "learning_rate": 0.0001943198566311071,
-      "loss": 0.8782,
-      "step": 1550
-    },
-    {
-      "epoch": 0.7629119527791441,
-      "grad_norm": 0.4662749203028031,
-      "learning_rate": 0.00019429938148229865,
-      "loss": 0.8848,
-      "step": 1551
-    },
-    {
-      "epoch": 0.76340383669454,
-      "grad_norm": 0.4513214582091005,
-      "learning_rate": 0.0001942788705792406,
-      "loss": 0.9188,
-      "step": 1552
-    },
-    {
-      "epoch": 0.763895720609936,
-      "grad_norm": 0.4591419564477227,
-      "learning_rate": 0.00019425832392970987,
-      "loss": 0.9115,
-      "step": 1553
-    },
-    {
-      "epoch": 0.764387604525332,
-      "grad_norm": 0.4434159713443017,
-      "learning_rate": 0.0001942377415414968,
-      "loss": 0.9414,
-      "step": 1554
-    },
-    {
-      "epoch": 0.764879488440728,
-      "grad_norm": 0.4792413561377097,
-      "learning_rate": 0.00019421712342240538,
-      "loss": 0.9341,
-      "step": 1555
-    },
-    {
-      "epoch": 0.765371372356124,
-      "grad_norm": 0.4571233966332298,
-      "learning_rate": 0.00019419646958025304,
-      "loss": 0.9736,
-      "step": 1556
-    },
-    {
-      "epoch": 0.76586325627152,
-      "grad_norm": 0.5151264468857435,
-      "learning_rate": 0.0001941757800228709,
-      "loss": 1.0183,
-      "step": 1557
-    },
-    {
-      "epoch": 0.7663551401869159,
-      "grad_norm": 0.4554295322487166,
-      "learning_rate": 0.00019415505475810352,
-      "loss": 0.9483,
-      "step": 1558
-    },
-    {
-      "epoch": 0.7668470241023119,
-      "grad_norm": 0.46158790861470056,
-      "learning_rate": 0.00019413429379380902,
-      "loss": 0.9447,
-      "step": 1559
-    },
-    {
-      "epoch": 0.7673389080177078,
-      "grad_norm": 0.4813083752405416,
-      "learning_rate": 0.00019411349713785905,
-      "loss": 0.9625,
-      "step": 1560
-    },
-    {
-      "epoch": 0.7678307919331038,
-      "grad_norm": 0.46427574383257714,
-      "learning_rate": 0.00019409266479813883,
-      "loss": 0.9808,
-      "step": 1561
-    },
-    {
-      "epoch": 0.7683226758484998,
-      "grad_norm": 0.5031775742968747,
-      "learning_rate": 0.00019407179678254707,
-      "loss": 0.9054,
-      "step": 1562
-    },
-    {
-      "epoch": 0.7688145597638957,
-      "grad_norm": 0.4539929202407643,
-      "learning_rate": 0.00019405089309899602,
-      "loss": 0.9528,
-      "step": 1563
-    },
-    {
-      "epoch": 0.7693064436792917,
-      "grad_norm": 0.46450887484643755,
-      "learning_rate": 0.00019402995375541145,
-      "loss": 0.9818,
-      "step": 1564
-    },
-    {
-      "epoch": 0.7697983275946877,
-      "grad_norm": 0.4737268161140511,
-      "learning_rate": 0.00019400897875973265,
-      "loss": 1.0212,
-      "step": 1565
-    },
-    {
-      "epoch": 0.7702902115100836,
-      "grad_norm": 0.4746804064421646,
-      "learning_rate": 0.00019398796811991243,
-      "loss": 0.9546,
-      "step": 1566
-    },
-    {
-      "epoch": 0.7707820954254796,
-      "grad_norm": 0.4927982579852802,
-      "learning_rate": 0.00019396692184391715,
-      "loss": 0.9783,
-      "step": 1567
-    },
-    {
-      "epoch": 0.7712739793408756,
-      "grad_norm": 0.46978666141723746,
-      "learning_rate": 0.0001939458399397266,
-      "loss": 0.9595,
-      "step": 1568
-    },
-    {
-      "epoch": 0.7717658632562715,
-      "grad_norm": 0.4368343909072726,
-      "learning_rate": 0.00019392472241533417,
-      "loss": 0.9283,
-      "step": 1569
-    },
-    {
-      "epoch": 0.7722577471716675,
-      "grad_norm": 0.4729589519572927,
-      "learning_rate": 0.00019390356927874666,
-      "loss": 0.9241,
-      "step": 1570
-    },
-    {
-      "epoch": 0.7727496310870634,
-      "grad_norm": 0.4680601858523261,
-      "learning_rate": 0.0001938823805379845,
-      "loss": 0.9536,
-      "step": 1571
-    },
-    {
-      "epoch": 0.7732415150024594,
-      "grad_norm": 0.44741493265017857,
-      "learning_rate": 0.00019386115620108148,
-      "loss": 0.9878,
-      "step": 1572
-    },
-    {
-      "epoch": 0.7737333989178554,
-      "grad_norm": 0.45874614310505646,
-      "learning_rate": 0.00019383989627608496,
-      "loss": 0.9403,
-      "step": 1573
-    },
-    {
-      "epoch": 0.7742252828332513,
-      "grad_norm": 0.4783541327168479,
-      "learning_rate": 0.0001938186007710558,
-      "loss": 0.9792,
-      "step": 1574
-    },
-    {
-      "epoch": 0.7747171667486473,
-      "grad_norm": 0.47751385243285904,
-      "learning_rate": 0.00019379726969406832,
-      "loss": 0.9044,
-      "step": 1575
-    },
-    {
-      "epoch": 0.7752090506640433,
-      "grad_norm": 0.4815871742945832,
-      "learning_rate": 0.0001937759030532104,
-      "loss": 0.9999,
-      "step": 1576
-    },
-    {
-      "epoch": 0.7757009345794392,
-      "grad_norm": 0.4578006443117354,
-      "learning_rate": 0.00019375450085658326,
-      "loss": 0.9107,
-      "step": 1577
-    },
-    {
-      "epoch": 0.7761928184948352,
-      "grad_norm": 0.49119482793649694,
-      "learning_rate": 0.00019373306311230176,
-      "loss": 0.9841,
-      "step": 1578
-    },
-    {
-      "epoch": 0.7766847024102311,
-      "grad_norm": 6.823370082793621,
-      "learning_rate": 0.0001937115898284941,
-      "loss": 1.3784,
-      "step": 1579
-    },
-    {
-      "epoch": 0.7771765863256271,
-      "grad_norm": 0.5531959848646097,
-      "learning_rate": 0.0001936900810133021,
-      "loss": 0.9586,
-      "step": 1580
-    },
-    {
-      "epoch": 0.7776684702410231,
-      "grad_norm": 0.48994368995911186,
-      "learning_rate": 0.0001936685366748809,
-      "loss": 0.9523,
-      "step": 1581
-    },
-    {
-      "epoch": 0.778160354156419,
-      "grad_norm": 0.626551897369954,
-      "learning_rate": 0.00019364695682139922,
-      "loss": 0.972,
-      "step": 1582
-    },
-    {
-      "epoch": 0.778652238071815,
-      "grad_norm": 0.45270344719609446,
-      "learning_rate": 0.00019362534146103922,
-      "loss": 0.9204,
-      "step": 1583
-    },
-    {
-      "epoch": 0.7791441219872111,
-      "grad_norm": 0.4691826031541264,
-      "learning_rate": 0.00019360369060199647,
-      "loss": 0.9541,
-      "step": 1584
-    },
-    {
-      "epoch": 0.779636005902607,
-      "grad_norm": 0.5077218371602953,
-      "learning_rate": 0.00019358200425248002,
-      "loss": 0.8997,
-      "step": 1585
-    },
-    {
-      "epoch": 0.780127889818003,
-      "grad_norm": 0.4561702721741504,
-      "learning_rate": 0.00019356028242071248,
-      "loss": 0.891,
-      "step": 1586
-    },
-    {
-      "epoch": 0.780619773733399,
-      "grad_norm": 0.4676087402728596,
-      "learning_rate": 0.0001935385251149298,
-      "loss": 0.9643,
-      "step": 1587
-    },
-    {
-      "epoch": 0.7811116576487949,
-      "grad_norm": 0.47241355479761177,
-      "learning_rate": 0.00019351673234338138,
-      "loss": 0.9294,
-      "step": 1588
-    },
-    {
-      "epoch": 0.7816035415641909,
-      "grad_norm": 0.5681763043851413,
-      "learning_rate": 0.0001934949041143301,
-      "loss": 1.0596,
-      "step": 1589
-    },
-    {
-      "epoch": 0.7820954254795868,
-      "grad_norm": 0.4391500738716643,
-      "learning_rate": 0.0001934730404360523,
-      "loss": 0.8989,
-      "step": 1590
-    },
-    {
-      "epoch": 0.7825873093949828,
-      "grad_norm": 0.46519103493459424,
-      "learning_rate": 0.00019345114131683774,
-      "loss": 0.8926,
-      "step": 1591
-    },
-    {
-      "epoch": 0.7830791933103788,
-      "grad_norm": 0.502320252683488,
-      "learning_rate": 0.0001934292067649896,
-      "loss": 1.0,
-      "step": 1592
-    },
-    {
-      "epoch": 0.7835710772257747,
-      "grad_norm": 0.48300705699795393,
-      "learning_rate": 0.00019340723678882454,
-      "loss": 1.0473,
-      "step": 1593
-    },
-    {
-      "epoch": 0.7840629611411707,
-      "grad_norm": 0.47061023321541257,
-      "learning_rate": 0.00019338523139667262,
-      "loss": 1.0138,
-      "step": 1594
-    },
-    {
-      "epoch": 0.7845548450565667,
-      "grad_norm": 0.4224254111419423,
-      "learning_rate": 0.0001933631905968773,
-      "loss": 0.9177,
-      "step": 1595
-    },
-    {
-      "epoch": 0.7850467289719626,
-      "grad_norm": 0.4452485746247102,
-      "learning_rate": 0.00019334111439779558,
-      "loss": 0.8828,
-      "step": 1596
-    },
-    {
-      "epoch": 0.7855386128873586,
-      "grad_norm": 0.5015343696233722,
-      "learning_rate": 0.0001933190028077977,
-      "loss": 1.0458,
-      "step": 1597
-    },
-    {
-      "epoch": 0.7860304968027545,
-      "grad_norm": 0.4549251969089811,
-      "learning_rate": 0.00019329685583526748,
-      "loss": 0.8869,
-      "step": 1598
-    },
-    {
-      "epoch": 0.7865223807181505,
-      "grad_norm": 0.45312364995771337,
-      "learning_rate": 0.00019327467348860208,
-      "loss": 0.9665,
-      "step": 1599
-    },
-    {
-      "epoch": 0.7870142646335465,
-      "grad_norm": 0.5053126722237422,
-      "learning_rate": 0.00019325245577621209,
-      "loss": 0.9931,
-      "step": 1600
-    },
-    {
-      "epoch": 0.7875061485489424,
-      "grad_norm": 0.48392174160193185,
-      "learning_rate": 0.0001932302027065215,
-      "loss": 0.9856,
-      "step": 1601
-    },
-    {
-      "epoch": 0.7879980324643384,
-      "grad_norm": 0.5986324995620144,
-      "learning_rate": 0.00019320791428796774,
-      "loss": 0.9421,
-      "step": 1602
-    },
-    {
-      "epoch": 0.7884899163797344,
-      "grad_norm": 0.45550074064574225,
-      "learning_rate": 0.00019318559052900155,
-      "loss": 0.9373,
-      "step": 1603
-    },
-    {
-      "epoch": 0.7889818002951303,
-      "grad_norm": 0.48491000695876824,
-      "learning_rate": 0.00019316323143808722,
-      "loss": 1.0153,
-      "step": 1604
-    },
-    {
-      "epoch": 0.7894736842105263,
-      "grad_norm": 0.497246864253746,
-      "learning_rate": 0.00019314083702370226,
-      "loss": 0.9551,
-      "step": 1605
-    },
-    {
-      "epoch": 0.7899655681259223,
-      "grad_norm": 0.517932175705986,
-      "learning_rate": 0.00019311840729433773,
-      "loss": 1.0129,
-      "step": 1606
-    },
-    {
-      "epoch": 0.7904574520413182,
-      "grad_norm": 0.45489433714258243,
-      "learning_rate": 0.000193095942258498,
-      "loss": 0.8971,
-      "step": 1607
-    },
-    {
-      "epoch": 0.7909493359567142,
-      "grad_norm": 0.47133976803694805,
-      "learning_rate": 0.00019307344192470084,
-      "loss": 0.9763,
-      "step": 1608
-    },
-    {
-      "epoch": 0.7914412198721101,
-      "grad_norm": 0.46947040535121726,
-      "learning_rate": 0.00019305090630147742,
-      "loss": 0.9255,
-      "step": 1609
-    },
-    {
-      "epoch": 0.7919331037875061,
-      "grad_norm": 0.4420899586233333,
-      "learning_rate": 0.0001930283353973722,
-      "loss": 0.8474,
-      "step": 1610
-    },
-    {
-      "epoch": 0.7924249877029022,
-      "grad_norm": 0.4480831417460121,
-      "learning_rate": 0.00019300572922094318,
-      "loss": 0.9437,
-      "step": 1611
-    },
-    {
-      "epoch": 0.792916871618298,
-      "grad_norm": 0.4684174784436505,
-      "learning_rate": 0.0001929830877807616,
-      "loss": 0.999,
-      "step": 1612
-    },
-    {
-      "epoch": 0.793408755533694,
-      "grad_norm": 0.48752167381100503,
-      "learning_rate": 0.00019296041108541215,
-      "loss": 0.9927,
-      "step": 1613
-    },
-    {
-      "epoch": 0.7939006394490901,
-      "grad_norm": 0.5276968315456267,
-      "learning_rate": 0.0001929376991434928,
-      "loss": 0.9702,
-      "step": 1614
-    },
-    {
-      "epoch": 0.794392523364486,
-      "grad_norm": 0.4772786470315235,
-      "learning_rate": 0.00019291495196361496,
-      "loss": 1.0118,
-      "step": 1615
-    },
-    {
-      "epoch": 0.794884407279882,
-      "grad_norm": 0.4718448673875022,
-      "learning_rate": 0.00019289216955440338,
-      "loss": 1.0037,
-      "step": 1616
-    },
-    {
-      "epoch": 0.7953762911952779,
-      "grad_norm": 0.4537633482352533,
-      "learning_rate": 0.0001928693519244962,
-      "loss": 0.9617,
-      "step": 1617
-    },
-    {
-      "epoch": 0.7958681751106739,
-      "grad_norm": 0.5164387700241561,
-      "learning_rate": 0.00019284649908254479,
-      "loss": 0.9491,
-      "step": 1618
-    },
-    {
-      "epoch": 0.7963600590260699,
-      "grad_norm": 0.438994288744205,
-      "learning_rate": 0.00019282361103721405,
-      "loss": 0.8788,
-      "step": 1619
-    },
-    {
-      "epoch": 0.7968519429414658,
-      "grad_norm": 0.4631003609801896,
-      "learning_rate": 0.0001928006877971821,
-      "loss": 1.006,
-      "step": 1620
-    },
-    {
-      "epoch": 0.7973438268568618,
-      "grad_norm": 0.432299703741968,
-      "learning_rate": 0.00019277772937114046,
-      "loss": 0.9982,
-      "step": 1621
-    },
-    {
-      "epoch": 0.7978357107722578,
-      "grad_norm": 0.44108770983585777,
-      "learning_rate": 0.00019275473576779395,
-      "loss": 0.951,
-      "step": 1622
-    },
-    {
-      "epoch": 0.7983275946876537,
-      "grad_norm": 0.4342514741553981,
-      "learning_rate": 0.00019273170699586075,
-      "loss": 0.9107,
-      "step": 1623
-    },
-    {
-      "epoch": 0.7988194786030497,
-      "grad_norm": 0.46203229766886594,
-      "learning_rate": 0.00019270864306407238,
-      "loss": 0.9805,
-      "step": 1624
-    },
-    {
-      "epoch": 0.7993113625184457,
-      "grad_norm": 0.43764253671175973,
-      "learning_rate": 0.0001926855439811737,
-      "loss": 0.956,
-      "step": 1625
-    },
-    {
-      "epoch": 0.7998032464338416,
-      "grad_norm": 0.4553303193705595,
-      "learning_rate": 0.00019266240975592287,
-      "loss": 0.9146,
-      "step": 1626
-    },
-    {
-      "epoch": 0.8002951303492376,
-      "grad_norm": 0.43882644054933073,
-      "learning_rate": 0.00019263924039709138,
-      "loss": 0.9142,
-      "step": 1627
-    },
-    {
-      "epoch": 0.8007870142646335,
-      "grad_norm": 0.43967325794082146,
-      "learning_rate": 0.00019261603591346407,
-      "loss": 0.9005,
-      "step": 1628
-    },
-    {
-      "epoch": 0.8012788981800295,
-      "grad_norm": 0.45993695271604523,
-      "learning_rate": 0.00019259279631383906,
-      "loss": 0.9224,
-      "step": 1629
-    },
-    {
-      "epoch": 0.8017707820954255,
-      "grad_norm": 0.41841573061177034,
-      "learning_rate": 0.00019256952160702783,
-      "loss": 0.8089,
-      "step": 1630
-    },
-    {
-      "epoch": 0.8022626660108214,
-      "grad_norm": 0.4517922112489481,
-      "learning_rate": 0.0001925462118018551,
-      "loss": 0.9832,
-      "step": 1631
-    },
-    {
-      "epoch": 0.8027545499262174,
-      "grad_norm": 0.4294363143126728,
-      "learning_rate": 0.000192522866907159,
-      "loss": 0.9678,
-      "step": 1632
-    },
-    {
-      "epoch": 0.8032464338416134,
-      "grad_norm": 0.4655940714688177,
-      "learning_rate": 0.00019249948693179084,
-      "loss": 1.0027,
-      "step": 1633
-    },
-    {
-      "epoch": 0.8037383177570093,
-      "grad_norm": 0.4585318027515355,
-      "learning_rate": 0.00019247607188461533,
-      "loss": 1.0446,
-      "step": 1634
-    },
-    {
-      "epoch": 0.8042302016724053,
-      "grad_norm": 0.4389156114118129,
-      "learning_rate": 0.00019245262177451043,
-      "loss": 0.9155,
-      "step": 1635
-    },
-    {
-      "epoch": 0.8047220855878012,
-      "grad_norm": 0.4371706023446502,
-      "learning_rate": 0.00019242913661036742,
-      "loss": 0.9244,
-      "step": 1636
-    },
-    {
-      "epoch": 0.8052139695031972,
-      "grad_norm": 0.4429513462143878,
-      "learning_rate": 0.00019240561640109088,
-      "loss": 0.9678,
-      "step": 1637
-    },
-    {
-      "epoch": 0.8057058534185932,
-      "grad_norm": 0.45578339516845023,
-      "learning_rate": 0.0001923820611555986,
-      "loss": 0.9286,
-      "step": 1638
-    },
-    {
-      "epoch": 0.8061977373339891,
-      "grad_norm": 0.4438082735917897,
-      "learning_rate": 0.00019235847088282175,
-      "loss": 0.8986,
-      "step": 1639
-    },
-    {
-      "epoch": 0.8066896212493851,
-      "grad_norm": 0.4363231081040358,
-      "learning_rate": 0.00019233484559170474,
-      "loss": 0.9207,
-      "step": 1640
-    },
-    {
-      "epoch": 0.8071815051647812,
-      "grad_norm": 0.47427861827466716,
-      "learning_rate": 0.00019231118529120525,
-      "loss": 1.0304,
-      "step": 1641
-    },
-    {
-      "epoch": 0.807673389080177,
-      "grad_norm": 0.42566058975372983,
-      "learning_rate": 0.00019228748999029424,
-      "loss": 0.8729,
-      "step": 1642
-    },
-    {
-      "epoch": 0.8081652729955731,
-      "grad_norm": 0.47573325704432545,
-      "learning_rate": 0.00019226375969795596,
-      "loss": 0.9169,
-      "step": 1643
-    },
-    {
-      "epoch": 0.8086571569109691,
-      "grad_norm": 4.151933419356274,
-      "learning_rate": 0.00019223999442318792,
-      "loss": 0.9422,
-      "step": 1644
-    },
-    {
-      "epoch": 0.809149040826365,
-      "grad_norm": 0.47818491115512063,
-      "learning_rate": 0.00019221619417500086,
-      "loss": 0.9432,
-      "step": 1645
-    },
-    {
-      "epoch": 0.809640924741761,
-      "grad_norm": 2.5932967494548795,
-      "learning_rate": 0.00019219235896241878,
-      "loss": 0.8862,
-      "step": 1646
-    },
-    {
-      "epoch": 0.8101328086571569,
-      "grad_norm": 0.5538370036659264,
-      "learning_rate": 0.00019216848879447903,
-      "loss": 0.9647,
-      "step": 1647
-    },
-    {
-      "epoch": 0.8106246925725529,
-      "grad_norm": 0.5292785138239161,
-      "learning_rate": 0.00019214458368023213,
-      "loss": 0.9156,
-      "step": 1648
-    },
-    {
-      "epoch": 0.8111165764879489,
-      "grad_norm": 0.5013209288940909,
-      "learning_rate": 0.0001921206436287418,
-      "loss": 0.978,
-      "step": 1649
-    },
-    {
-      "epoch": 0.8116084604033448,
-      "grad_norm": 0.4825509785596944,
-      "learning_rate": 0.00019209666864908513,
-      "loss": 0.9485,
-      "step": 1650
-    },
-    {
-      "epoch": 0.8121003443187408,
-      "grad_norm": 0.4871980247632107,
-      "learning_rate": 0.00019207265875035242,
-      "loss": 0.9772,
-      "step": 1651
-    },
-    {
-      "epoch": 0.8125922282341368,
-      "grad_norm": 0.5304798536348684,
-      "learning_rate": 0.00019204861394164712,
-      "loss": 0.9387,
-      "step": 1652
-    },
-    {
-      "epoch": 0.8130841121495327,
-      "grad_norm": 0.4824297692753066,
-      "learning_rate": 0.000192024534232086,
-      "loss": 0.9482,
-      "step": 1653
-    },
-    {
-      "epoch": 0.8135759960649287,
-      "grad_norm": 0.44976498051806146,
-      "learning_rate": 0.0001920004196307991,
-      "loss": 0.8662,
-      "step": 1654
-    },
-    {
-      "epoch": 0.8140678799803246,
-      "grad_norm": 0.474231975096475,
-      "learning_rate": 0.00019197627014692957,
-      "loss": 0.9841,
-      "step": 1655
-    },
-    {
-      "epoch": 0.8145597638957206,
-      "grad_norm": 0.43054221145474725,
-      "learning_rate": 0.0001919520857896339,
-      "loss": 0.8923,
-      "step": 1656
-    },
-    {
-      "epoch": 0.8150516478111166,
-      "grad_norm": 0.43791924280616396,
-      "learning_rate": 0.00019192786656808172,
-      "loss": 0.9218,
-      "step": 1657
-    },
-    {
-      "epoch": 0.8155435317265125,
-      "grad_norm": 0.4502867626244973,
-      "learning_rate": 0.0001919036124914559,
-      "loss": 0.9307,
-      "step": 1658
-    },
-    {
-      "epoch": 0.8160354156419085,
-      "grad_norm": 0.4601747667702632,
-      "learning_rate": 0.00019187932356895255,
-      "loss": 0.888,
-      "step": 1659
-    },
-    {
-      "epoch": 0.8165272995573045,
-      "grad_norm": 0.41441192313089104,
-      "learning_rate": 0.000191854999809781,
-      "loss": 0.8535,
-      "step": 1660
-    },
-    {
-      "epoch": 0.8170191834727004,
-      "grad_norm": 0.46412142092536074,
-      "learning_rate": 0.00019183064122316377,
-      "loss": 0.9116,
-      "step": 1661
-    },
-    {
-      "epoch": 0.8175110673880964,
-      "grad_norm": 0.4781238738230767,
-      "learning_rate": 0.00019180624781833653,
-      "loss": 0.9667,
-      "step": 1662
-    },
-    {
-      "epoch": 0.8180029513034923,
-      "grad_norm": 0.47809572634639297,
-      "learning_rate": 0.0001917818196045483,
-      "loss": 0.9217,
-      "step": 1663
-    },
-    {
-      "epoch": 0.8184948352188883,
-      "grad_norm": 0.4289354306018643,
-      "learning_rate": 0.0001917573565910611,
-      "loss": 0.9228,
-      "step": 1664
-    },
-    {
-      "epoch": 0.8189867191342843,
-      "grad_norm": 0.4376385534401151,
-      "learning_rate": 0.0001917328587871503,
-      "loss": 0.9193,
-      "step": 1665
-    },
-    {
-      "epoch": 0.8194786030496802,
-      "grad_norm": 0.4602183329184602,
-      "learning_rate": 0.0001917083262021044,
-      "loss": 0.9596,
-      "step": 1666
-    },
-    {
-      "epoch": 0.8199704869650762,
-      "grad_norm": 0.4705010674999434,
-      "learning_rate": 0.0001916837588452251,
-      "loss": 0.9906,
-      "step": 1667
-    },
-    {
-      "epoch": 0.8204623708804722,
-      "grad_norm": 0.4540706040744854,
-      "learning_rate": 0.00019165915672582728,
-      "loss": 0.9111,
-      "step": 1668
-    },
-    {
-      "epoch": 0.8209542547958681,
-      "grad_norm": 0.4399156745771892,
-      "learning_rate": 0.000191634519853239,
-      "loss": 0.8972,
-      "step": 1669
-    },
-    {
-      "epoch": 0.8214461387112642,
-      "grad_norm": 0.4423058953446686,
-      "learning_rate": 0.0001916098482368015,
-      "loss": 0.9503,
-      "step": 1670
-    },
-    {
-      "epoch": 0.8219380226266602,
-      "grad_norm": 0.4505275951880537,
-      "learning_rate": 0.00019158514188586919,
-      "loss": 0.9449,
-      "step": 1671
-    },
-    {
-      "epoch": 0.822429906542056,
-      "grad_norm": 0.4408739893258349,
-      "learning_rate": 0.00019156040080980962,
-      "loss": 0.9479,
-      "step": 1672
-    },
-    {
-      "epoch": 0.8229217904574521,
-      "grad_norm": 0.4380737534669539,
-      "learning_rate": 0.00019153562501800355,
-      "loss": 1.0027,
-      "step": 1673
-    },
-    {
-      "epoch": 0.823413674372848,
-      "grad_norm": 0.4787880601377095,
-      "learning_rate": 0.00019151081451984495,
-      "loss": 0.9935,
-      "step": 1674
-    },
-    {
-      "epoch": 0.823905558288244,
-      "grad_norm": 1.119589826970695,
-      "learning_rate": 0.00019148596932474078,
-      "loss": 1.0879,
-      "step": 1675
-    },
-    {
-      "epoch": 0.82439744220364,
-      "grad_norm": 0.46572840439247004,
-      "learning_rate": 0.00019146108944211138,
-      "loss": 0.8878,
-      "step": 1676
-    },
-    {
-      "epoch": 0.8248893261190359,
-      "grad_norm": 0.4659208296839495,
-      "learning_rate": 0.00019143617488139004,
-      "loss": 0.9279,
-      "step": 1677
-    },
-    {
-      "epoch": 0.8253812100344319,
-      "grad_norm": 0.44553558928537207,
-      "learning_rate": 0.00019141122565202335,
-      "loss": 0.9658,
-      "step": 1678
-    },
-    {
-      "epoch": 0.8258730939498279,
-      "grad_norm": 0.4184334521190072,
-      "learning_rate": 0.0001913862417634709,
-      "loss": 0.9047,
-      "step": 1679
-    },
-    {
-      "epoch": 0.8263649778652238,
-      "grad_norm": 0.467572541474465,
-      "learning_rate": 0.00019136122322520555,
-      "loss": 0.9362,
-      "step": 1680
-    },
-    {
-      "epoch": 0.8268568617806198,
-      "grad_norm": 0.4733217009790865,
-      "learning_rate": 0.00019133617004671324,
-      "loss": 0.9328,
-      "step": 1681
-    },
-    {
-      "epoch": 0.8273487456960157,
-      "grad_norm": 0.4349385845127341,
-      "learning_rate": 0.0001913110822374931,
-      "loss": 0.9613,
-      "step": 1682
-    },
-    {
-      "epoch": 0.8278406296114117,
-      "grad_norm": 0.6962138041986241,
-      "learning_rate": 0.00019128595980705726,
-      "loss": 0.9973,
-      "step": 1683
-    },
-    {
-      "epoch": 0.8283325135268077,
-      "grad_norm": 0.4687083158184151,
-      "learning_rate": 0.0001912608027649311,
-      "loss": 0.919,
-      "step": 1684
-    },
-    {
-      "epoch": 0.8288243974422036,
-      "grad_norm": 0.7325556364348405,
-      "learning_rate": 0.0001912356111206531,
-      "loss": 0.9348,
-      "step": 1685
-    },
-    {
-      "epoch": 0.8293162813575996,
-      "grad_norm": 0.45290954712544573,
-      "learning_rate": 0.00019121038488377481,
-      "loss": 0.941,
-      "step": 1686
-    },
-    {
-      "epoch": 0.8298081652729956,
-      "grad_norm": 0.4638279977461688,
-      "learning_rate": 0.00019118512406386093,
-      "loss": 0.9467,
-      "step": 1687
-    },
-    {
-      "epoch": 0.8303000491883915,
-      "grad_norm": 0.4937722375033083,
-      "learning_rate": 0.0001911598286704893,
-      "loss": 1.0182,
-      "step": 1688
-    },
-    {
-      "epoch": 0.8307919331037875,
-      "grad_norm": 0.4978750826206753,
-      "learning_rate": 0.00019113449871325082,
-      "loss": 0.9416,
-      "step": 1689
-    },
-    {
-      "epoch": 0.8312838170191835,
-      "grad_norm": 0.47674968324113887,
-      "learning_rate": 0.00019110913420174953,
-      "loss": 0.9751,
-      "step": 1690
-    },
-    {
-      "epoch": 0.8317757009345794,
-      "grad_norm": 0.5853175017337767,
-      "learning_rate": 0.00019108373514560253,
-      "loss": 0.9332,
-      "step": 1691
-    },
-    {
-      "epoch": 0.8322675848499754,
-      "grad_norm": 0.8346336216787719,
-      "learning_rate": 0.00019105830155444006,
-      "loss": 0.942,
-      "step": 1692
-    },
-    {
-      "epoch": 0.8327594687653713,
-      "grad_norm": 0.4704554197974211,
-      "learning_rate": 0.00019103283343790544,
-      "loss": 0.9197,
-      "step": 1693
-    },
-    {
-      "epoch": 0.8332513526807673,
-      "grad_norm": 0.44791715506213525,
-      "learning_rate": 0.00019100733080565507,
-      "loss": 0.8747,
-      "step": 1694
-    },
-    {
-      "epoch": 0.8337432365961633,
-      "grad_norm": 0.4785105991498684,
-      "learning_rate": 0.00019098179366735846,
-      "loss": 0.9503,
-      "step": 1695
-    },
-    {
-      "epoch": 0.8342351205115592,
-      "grad_norm": 0.44983205244026786,
-      "learning_rate": 0.00019095622203269818,
-      "loss": 0.9216,
-      "step": 1696
-    },
-    {
-      "epoch": 0.8347270044269552,
-      "grad_norm": 0.4145305835128189,
-      "learning_rate": 0.00019093061591136988,
-      "loss": 0.9095,
-      "step": 1697
-    },
-    {
-      "epoch": 0.8352188883423513,
-      "grad_norm": 0.5481351760124087,
-      "learning_rate": 0.00019090497531308232,
-      "loss": 0.9839,
-      "step": 1698
-    },
-    {
-      "epoch": 0.8357107722577471,
-      "grad_norm": 0.4298929399943066,
-      "learning_rate": 0.00019087930024755729,
-      "loss": 0.8435,
-      "step": 1699
-    },
-    {
-      "epoch": 0.8362026561731432,
-      "grad_norm": 0.5331492927411966,
-      "learning_rate": 0.00019085359072452965,
-      "loss": 0.9511,
-      "step": 1700
-    },
-    {
-      "epoch": 0.836694540088539,
-      "grad_norm": 0.4547815109154883,
-      "learning_rate": 0.00019082784675374738,
-      "loss": 0.9574,
-      "step": 1701
-    },
-    {
-      "epoch": 0.8371864240039351,
-      "grad_norm": 0.4669681706844525,
-      "learning_rate": 0.00019080206834497143,
-      "loss": 0.8932,
-      "step": 1702
-    },
-    {
-      "epoch": 0.8376783079193311,
-      "grad_norm": 0.5294472646601358,
-      "learning_rate": 0.00019077625550797592,
-      "loss": 0.8594,
-      "step": 1703
-    },
-    {
-      "epoch": 0.838170191834727,
-      "grad_norm": 0.49214535071449894,
-      "learning_rate": 0.0001907504082525479,
-      "loss": 0.9738,
-      "step": 1704
-    },
-    {
-      "epoch": 0.838662075750123,
-      "grad_norm": 0.43244479297440763,
-      "learning_rate": 0.00019072452658848755,
-      "loss": 0.942,
-      "step": 1705
-    },
-    {
-      "epoch": 0.839153959665519,
-      "grad_norm": 0.4904340790964634,
-      "learning_rate": 0.00019069861052560812,
-      "loss": 0.9537,
-      "step": 1706
-    },
-    {
-      "epoch": 0.8396458435809149,
-      "grad_norm": 0.43774183775168746,
-      "learning_rate": 0.00019067266007373582,
-      "loss": 0.8301,
-      "step": 1707
-    },
-    {
-      "epoch": 0.8401377274963109,
-      "grad_norm": 0.46050558873106656,
-      "learning_rate": 0.00019064667524270994,
-      "loss": 0.953,
-      "step": 1708
-    },
-    {
-      "epoch": 0.8406296114117069,
-      "grad_norm": 0.4484627683624173,
-      "learning_rate": 0.0001906206560423828,
-      "loss": 0.918,
-      "step": 1709
-    },
-    {
-      "epoch": 0.8411214953271028,
-      "grad_norm": 0.4296532777857049,
-      "learning_rate": 0.00019059460248261977,
-      "loss": 0.824,
-      "step": 1710
-    },
-    {
-      "epoch": 0.8416133792424988,
-      "grad_norm": 0.45532748283712865,
-      "learning_rate": 0.00019056851457329926,
-      "loss": 0.991,
-      "step": 1711
-    },
-    {
-      "epoch": 0.8421052631578947,
-      "grad_norm": 0.4160688152743597,
-      "learning_rate": 0.00019054239232431264,
-      "loss": 0.8287,
-      "step": 1712
-    },
-    {
-      "epoch": 0.8425971470732907,
-      "grad_norm": 0.4187920785854741,
-      "learning_rate": 0.00019051623574556435,
-      "loss": 0.9079,
-      "step": 1713
-    },
-    {
-      "epoch": 0.8430890309886867,
-      "grad_norm": 0.4483790378833997,
-      "learning_rate": 0.00019049004484697184,
-      "loss": 0.9457,
-      "step": 1714
-    },
-    {
-      "epoch": 0.8435809149040826,
-      "grad_norm": 0.4868550243867067,
-      "learning_rate": 0.00019046381963846553,
-      "loss": 0.961,
-      "step": 1715
-    },
-    {
-      "epoch": 0.8440727988194786,
-      "grad_norm": 0.43569359093080523,
-      "learning_rate": 0.00019043756012998895,
-      "loss": 0.9038,
-      "step": 1716
-    },
-    {
-      "epoch": 0.8445646827348746,
-      "grad_norm": 0.44509012324050984,
-      "learning_rate": 0.0001904112663314985,
-      "loss": 0.983,
-      "step": 1717
-    },
-    {
-      "epoch": 0.8450565666502705,
-      "grad_norm": 0.476317419826852,
-      "learning_rate": 0.0001903849382529637,
-      "loss": 0.9143,
-      "step": 1718
-    },
-    {
-      "epoch": 0.8455484505656665,
-      "grad_norm": 0.41731740875023754,
-      "learning_rate": 0.00019035857590436704,
-      "loss": 0.8802,
-      "step": 1719
-    },
-    {
-      "epoch": 0.8460403344810624,
-      "grad_norm": 0.47613499203651155,
-      "learning_rate": 0.00019033217929570391,
-      "loss": 0.9191,
-      "step": 1720
-    },
-    {
-      "epoch": 0.8465322183964584,
-      "grad_norm": 0.4464178651406063,
-      "learning_rate": 0.00019030574843698281,
-      "loss": 0.9648,
-      "step": 1721
-    },
-    {
-      "epoch": 0.8470241023118544,
-      "grad_norm": 0.44463784745952545,
-      "learning_rate": 0.00019027928333822515,
-      "loss": 0.8904,
-      "step": 1722
-    },
-    {
-      "epoch": 0.8475159862272503,
-      "grad_norm": 0.42619905969166394,
-      "learning_rate": 0.00019025278400946542,
-      "loss": 0.8552,
-      "step": 1723
-    },
-    {
-      "epoch": 0.8480078701426463,
-      "grad_norm": 0.4261901301127405,
-      "learning_rate": 0.00019022625046075095,
-      "loss": 0.9629,
-      "step": 1724
-    },
-    {
-      "epoch": 0.8484997540580423,
-      "grad_norm": 0.5796037243857726,
-      "learning_rate": 0.00019019968270214213,
-      "loss": 0.9205,
-      "step": 1725
-    },
-    {
-      "epoch": 0.8489916379734382,
-      "grad_norm": 0.44780252737015724,
-      "learning_rate": 0.00019017308074371233,
-      "loss": 0.9937,
-      "step": 1726
-    },
-    {
-      "epoch": 0.8494835218888342,
-      "grad_norm": 0.4877912151431391,
-      "learning_rate": 0.00019014644459554786,
-      "loss": 0.9661,
-      "step": 1727
-    },
-    {
-      "epoch": 0.8499754058042303,
-      "grad_norm": 0.48469045725544496,
-      "learning_rate": 0.00019011977426774797,
-      "loss": 0.9564,
-      "step": 1728
-    },
-    {
-      "epoch": 0.8504672897196262,
-      "grad_norm": 0.44586207792614124,
-      "learning_rate": 0.00019009306977042488,
-      "loss": 0.9836,
-      "step": 1729
-    },
-    {
-      "epoch": 0.8509591736350222,
-      "grad_norm": 0.41808746204810365,
-      "learning_rate": 0.00019006633111370386,
-      "loss": 0.8683,
-      "step": 1730
-    },
-    {
-      "epoch": 0.8514510575504181,
-      "grad_norm": 0.4524247488232523,
-      "learning_rate": 0.00019003955830772297,
-      "loss": 0.8832,
-      "step": 1731
-    },
-    {
-      "epoch": 0.8519429414658141,
-      "grad_norm": 0.4613344924505022,
-      "learning_rate": 0.00019001275136263333,
-      "loss": 0.9312,
-      "step": 1732
-    },
-    {
-      "epoch": 0.8524348253812101,
-      "grad_norm": 0.45390650900046664,
-      "learning_rate": 0.00018998591028859902,
-      "loss": 0.8647,
-      "step": 1733
-    },
-    {
-      "epoch": 0.852926709296606,
-      "grad_norm": 0.4582601512809806,
-      "learning_rate": 0.00018995903509579694,
-      "loss": 0.9831,
-      "step": 1734
-    },
-    {
-      "epoch": 0.853418593212002,
-      "grad_norm": 0.4711376053055614,
-      "learning_rate": 0.00018993212579441703,
-      "loss": 0.9372,
-      "step": 1735
-    },
-    {
-      "epoch": 0.853910477127398,
-      "grad_norm": 0.44223921031020474,
-      "learning_rate": 0.00018990518239466215,
-      "loss": 0.9191,
-      "step": 1736
-    },
-    {
-      "epoch": 0.8544023610427939,
-      "grad_norm": 0.43119686684735786,
-      "learning_rate": 0.00018987820490674805,
-      "loss": 0.9995,
-      "step": 1737
-    },
-    {
-      "epoch": 0.8548942449581899,
-      "grad_norm": 0.4472573559680049,
-      "learning_rate": 0.00018985119334090348,
-      "loss": 0.9674,
-      "step": 1738
-    },
-    {
-      "epoch": 0.8553861288735858,
-      "grad_norm": 0.44812802548277875,
-      "learning_rate": 0.00018982414770737,
-      "loss": 0.94,
-      "step": 1739
-    },
-    {
-      "epoch": 0.8558780127889818,
-      "grad_norm": 0.4269593800378158,
-      "learning_rate": 0.00018979706801640212,
-      "loss": 0.8899,
-      "step": 1740
-    },
-    {
-      "epoch": 0.8563698967043778,
-      "grad_norm": 0.42748425771216847,
-      "learning_rate": 0.00018976995427826736,
-      "loss": 0.9473,
-      "step": 1741
-    },
-    {
-      "epoch": 0.8568617806197737,
-      "grad_norm": 0.434640751370082,
-      "learning_rate": 0.00018974280650324606,
-      "loss": 0.9424,
-      "step": 1742
-    },
-    {
-      "epoch": 0.8573536645351697,
-      "grad_norm": 0.44851806532954686,
-      "learning_rate": 0.00018971562470163146,
-      "loss": 0.9983,
-      "step": 1743
-    },
-    {
-      "epoch": 0.8578455484505657,
-      "grad_norm": 0.4552608523036954,
-      "learning_rate": 0.00018968840888372972,
-      "loss": 0.9899,
-      "step": 1744
-    },
-    {
-      "epoch": 0.8583374323659616,
-      "grad_norm": 0.4286141547884595,
-      "learning_rate": 0.00018966115905985994,
-      "loss": 0.9454,
-      "step": 1745
-    },
-    {
-      "epoch": 0.8588293162813576,
-      "grad_norm": 0.41057848902217786,
-      "learning_rate": 0.00018963387524035405,
-      "loss": 0.947,
-      "step": 1746
-    },
-    {
-      "epoch": 0.8593212001967536,
-      "grad_norm": 0.4325695053662549,
-      "learning_rate": 0.0001896065574355569,
-      "loss": 0.9442,
-      "step": 1747
-    },
-    {
-      "epoch": 0.8598130841121495,
-      "grad_norm": 0.42886033764967324,
-      "learning_rate": 0.0001895792056558262,
-      "loss": 0.9677,
-      "step": 1748
-    },
-    {
-      "epoch": 0.8603049680275455,
-      "grad_norm": 0.4614553992069052,
-      "learning_rate": 0.00018955181991153262,
-      "loss": 0.9718,
-      "step": 1749
-    },
-    {
-      "epoch": 0.8607968519429414,
-      "grad_norm": 0.5977683144717094,
-      "learning_rate": 0.00018952440021305958,
-      "loss": 0.8984,
-      "step": 1750
-    },
-    {
-      "epoch": 0.8612887358583374,
-      "grad_norm": 0.4444559794221156,
-      "learning_rate": 0.00018949694657080347,
-      "loss": 0.9548,
-      "step": 1751
-    },
-    {
-      "epoch": 0.8617806197737334,
-      "grad_norm": 0.44414338599542635,
-      "learning_rate": 0.00018946945899517355,
-      "loss": 0.923,
-      "step": 1752
-    },
-    {
-      "epoch": 0.8622725036891293,
-      "grad_norm": 0.4367775065694892,
-      "learning_rate": 0.0001894419374965919,
-      "loss": 0.9239,
-      "step": 1753
-    },
-    {
-      "epoch": 0.8627643876045253,
-      "grad_norm": 0.4332821664486575,
-      "learning_rate": 0.00018941438208549348,
-      "loss": 0.9647,
-      "step": 1754
-    },
-    {
-      "epoch": 0.8632562715199213,
-      "grad_norm": 0.442896623249607,
-      "learning_rate": 0.0001893867927723261,
-      "loss": 0.9067,
-      "step": 1755
-    },
-    {
-      "epoch": 0.8637481554353172,
-      "grad_norm": 0.4765982923498218,
-      "learning_rate": 0.00018935916956755049,
-      "loss": 0.8763,
-      "step": 1756
-    },
-    {
-      "epoch": 0.8642400393507133,
-      "grad_norm": 0.5085873054440848,
-      "learning_rate": 0.00018933151248164013,
-      "loss": 0.9125,
-      "step": 1757
-    },
-    {
-      "epoch": 0.8647319232661091,
-      "grad_norm": 0.4450768587114687,
-      "learning_rate": 0.00018930382152508135,
-      "loss": 0.9892,
-      "step": 1758
-    },
-    {
-      "epoch": 0.8652238071815052,
-      "grad_norm": 0.44517932512121955,
-      "learning_rate": 0.00018927609670837345,
-      "loss": 0.9625,
-      "step": 1759
-    },
-    {
-      "epoch": 0.8657156910969012,
-      "grad_norm": 0.48411990596383014,
-      "learning_rate": 0.0001892483380420284,
-      "loss": 1.005,
-      "step": 1760
-    },
-    {
-      "epoch": 0.8662075750122971,
-      "grad_norm": 0.45385066165339627,
-      "learning_rate": 0.0001892205455365712,
-      "loss": 0.9319,
-      "step": 1761
-    },
-    {
-      "epoch": 0.8666994589276931,
-      "grad_norm": 0.45143595633541134,
-      "learning_rate": 0.00018919271920253946,
-      "loss": 0.9441,
-      "step": 1762
-    },
-    {
-      "epoch": 0.8671913428430891,
-      "grad_norm": 0.45260700492803047,
-      "learning_rate": 0.00018916485905048376,
-      "loss": 0.9624,
-      "step": 1763
-    },
-    {
-      "epoch": 0.867683226758485,
-      "grad_norm": 0.5206644890104113,
-      "learning_rate": 0.00018913696509096744,
-      "loss": 0.9982,
-      "step": 1764
-    },
-    {
-      "epoch": 0.868175110673881,
-      "grad_norm": 0.43102045635046177,
-      "learning_rate": 0.00018910903733456675,
-      "loss": 0.8632,
-      "step": 1765
-    },
-    {
-      "epoch": 0.8686669945892769,
-      "grad_norm": 0.45239651443635964,
-      "learning_rate": 0.00018908107579187062,
-      "loss": 0.9662,
-      "step": 1766
-    },
-    {
-      "epoch": 0.8691588785046729,
-      "grad_norm": 0.44902898324976237,
-      "learning_rate": 0.00018905308047348091,
-      "loss": 0.8942,
-      "step": 1767
-    },
-    {
-      "epoch": 0.8696507624200689,
-      "grad_norm": 0.44351856444298865,
-      "learning_rate": 0.00018902505139001217,
-      "loss": 0.8789,
-      "step": 1768
-    },
-    {
-      "epoch": 0.8701426463354648,
-      "grad_norm": 0.4375309473126057,
-      "learning_rate": 0.00018899698855209191,
-      "loss": 0.9314,
-      "step": 1769
-    },
-    {
-      "epoch": 0.8706345302508608,
-      "grad_norm": 0.6783015497558752,
-      "learning_rate": 0.00018896889197036028,
-      "loss": 0.9623,
-      "step": 1770
-    },
-    {
-      "epoch": 0.8711264141662568,
-      "grad_norm": 0.460625183828372,
-      "learning_rate": 0.00018894076165547026,
-      "loss": 0.9257,
-      "step": 1771
-    },
-    {
-      "epoch": 0.8716182980816527,
-      "grad_norm": 0.41840175884135405,
-      "learning_rate": 0.0001889125976180877,
-      "loss": 0.941,
-      "step": 1772
-    },
-    {
-      "epoch": 0.8721101819970487,
-      "grad_norm": 0.4473604481778361,
-      "learning_rate": 0.0001888843998688912,
-      "loss": 0.9419,
-      "step": 1773
-    },
-    {
-      "epoch": 0.8726020659124447,
-      "grad_norm": 0.4575990059010378,
-      "learning_rate": 0.00018885616841857213,
-      "loss": 0.9284,
-      "step": 1774
-    },
-    {
-      "epoch": 0.8730939498278406,
-      "grad_norm": 0.4737021978731937,
-      "learning_rate": 0.00018882790327783456,
-      "loss": 0.9625,
-      "step": 1775
-    },
-    {
-      "epoch": 0.8735858337432366,
-      "grad_norm": 0.4677052281961736,
-      "learning_rate": 0.00018879960445739545,
-      "loss": 0.9286,
-      "step": 1776
-    },
-    {
-      "epoch": 0.8740777176586325,
-      "grad_norm": 0.4466015227284853,
-      "learning_rate": 0.00018877127196798458,
-      "loss": 0.9675,
-      "step": 1777
-    },
-    {
-      "epoch": 0.8745696015740285,
-      "grad_norm": 0.4380609013163242,
-      "learning_rate": 0.00018874290582034426,
-      "loss": 0.8788,
-      "step": 1778
-    },
-    {
-      "epoch": 0.8750614854894245,
-      "grad_norm": 0.42955753585069034,
-      "learning_rate": 0.00018871450602522978,
-      "loss": 0.9219,
-      "step": 1779
-    },
-    {
-      "epoch": 0.8755533694048204,
-      "grad_norm": 0.4673997530348016,
-      "learning_rate": 0.00018868607259340912,
-      "loss": 1.0069,
-      "step": 1780
-    },
-    {
-      "epoch": 0.8760452533202164,
-      "grad_norm": 0.45299558876329443,
-      "learning_rate": 0.00018865760553566296,
-      "loss": 0.8895,
-      "step": 1781
-    },
-    {
-      "epoch": 0.8765371372356124,
-      "grad_norm": 0.43026142358832664,
-      "learning_rate": 0.00018862910486278485,
-      "loss": 0.9287,
-      "step": 1782
-    },
-    {
-      "epoch": 0.8770290211510083,
-      "grad_norm": 0.45346870074831724,
-      "learning_rate": 0.00018860057058558096,
-      "loss": 0.967,
-      "step": 1783
-    },
-    {
-      "epoch": 0.8775209050664043,
-      "grad_norm": 0.41438574406103895,
-      "learning_rate": 0.0001885720027148703,
-      "loss": 0.8696,
-      "step": 1784
-    },
-    {
-      "epoch": 0.8780127889818002,
-      "grad_norm": 0.4209309017198743,
-      "learning_rate": 0.0001885434012614845,
-      "loss": 0.893,
-      "step": 1785
-    },
-    {
-      "epoch": 0.8785046728971962,
-      "grad_norm": 0.45927453652065253,
-      "learning_rate": 0.00018851476623626804,
-      "loss": 0.9263,
-      "step": 1786
-    },
-    {
-      "epoch": 0.8789965568125923,
-      "grad_norm": 0.4428204057844921,
-      "learning_rate": 0.0001884860976500781,
-      "loss": 0.9492,
-      "step": 1787
-    },
-    {
-      "epoch": 0.8794884407279882,
-      "grad_norm": 0.4360707709403154,
-      "learning_rate": 0.00018845739551378454,
-      "loss": 0.9493,
-      "step": 1788
-    },
-    {
-      "epoch": 0.8799803246433842,
-      "grad_norm": 0.42945002857421083,
-      "learning_rate": 0.00018842865983827,
-      "loss": 0.9494,
-      "step": 1789
-    },
-    {
-      "epoch": 0.8804722085587802,
-      "grad_norm": 0.42044826504153504,
-      "learning_rate": 0.00018839989063442974,
-      "loss": 0.9344,
-      "step": 1790
-    },
-    {
-      "epoch": 0.8809640924741761,
-      "grad_norm": 0.4175668456403529,
-      "learning_rate": 0.00018837108791317192,
-      "loss": 0.8691,
-      "step": 1791
-    },
-    {
-      "epoch": 0.8814559763895721,
-      "grad_norm": 0.41589198575210096,
-      "learning_rate": 0.00018834225168541714,
-      "loss": 0.8277,
-      "step": 1792
-    },
-    {
-      "epoch": 0.8819478603049681,
-      "grad_norm": 0.45472823744226093,
-      "learning_rate": 0.00018831338196209898,
-      "loss": 0.9579,
-      "step": 1793
-    },
-    {
-      "epoch": 0.882439744220364,
-      "grad_norm": 0.43642395638749093,
-      "learning_rate": 0.0001882844787541635,
-      "loss": 0.8524,
-      "step": 1794
-    },
-    {
-      "epoch": 0.88293162813576,
-      "grad_norm": 2.101447151660097,
-      "learning_rate": 0.00018825554207256958,
-      "loss": 1.0781,
-      "step": 1795
-    },
-    {
-      "epoch": 0.8834235120511559,
-      "grad_norm": 0.4709293264588742,
-      "learning_rate": 0.00018822657192828877,
-      "loss": 0.944,
-      "step": 1796
-    },
-    {
-      "epoch": 0.8839153959665519,
-      "grad_norm": 0.4847808483734464,
-      "learning_rate": 0.00018819756833230526,
-      "loss": 0.912,
-      "step": 1797
-    },
-    {
-      "epoch": 0.8844072798819479,
-      "grad_norm": 0.4565319301550743,
-      "learning_rate": 0.00018816853129561601,
-      "loss": 0.9549,
-      "step": 1798
-    },
-    {
-      "epoch": 0.8848991637973438,
-      "grad_norm": 0.447116787345367,
-      "learning_rate": 0.00018813946082923058,
-      "loss": 0.9802,
-      "step": 1799
-    },
-    {
-      "epoch": 0.8853910477127398,
-      "grad_norm": 0.45879294419820016,
-      "learning_rate": 0.00018811035694417126,
-      "loss": 0.9126,
-      "step": 1800
-    },
-    {
-      "epoch": 0.8858829316281358,
-      "grad_norm": 0.43693103995260546,
-      "learning_rate": 0.00018808121965147294,
-      "loss": 0.8992,
-      "step": 1801
-    },
-    {
-      "epoch": 0.8863748155435317,
-      "grad_norm": 0.48290914109320904,
-      "learning_rate": 0.00018805204896218326,
-      "loss": 0.9972,
-      "step": 1802
-    },
-    {
-      "epoch": 0.8868666994589277,
-      "grad_norm": 0.4240014058321878,
-      "learning_rate": 0.00018802284488736245,
-      "loss": 0.8836,
-      "step": 1803
-    },
-    {
-      "epoch": 0.8873585833743236,
-      "grad_norm": 0.45171863426094383,
-      "learning_rate": 0.00018799360743808346,
-      "loss": 0.9427,
-      "step": 1804
-    },
-    {
-      "epoch": 0.8878504672897196,
-      "grad_norm": 0.4464676215669015,
-      "learning_rate": 0.0001879643366254319,
-      "loss": 0.9595,
-      "step": 1805
-    },
-    {
-      "epoch": 0.8883423512051156,
-      "grad_norm": 0.4561868913576593,
-      "learning_rate": 0.00018793503246050593,
-      "loss": 0.9077,
-      "step": 1806
-    },
-    {
-      "epoch": 0.8888342351205115,
-      "grad_norm": 0.41949886953612353,
-      "learning_rate": 0.00018790569495441645,
-      "loss": 0.8574,
-      "step": 1807
-    },
-    {
-      "epoch": 0.8893261190359075,
-      "grad_norm": 0.446706485049033,
-      "learning_rate": 0.00018787632411828697,
-      "loss": 0.9026,
-      "step": 1808
-    },
-    {
-      "epoch": 0.8898180029513035,
-      "grad_norm": 0.4697568797443752,
-      "learning_rate": 0.00018784691996325366,
-      "loss": 0.968,
-      "step": 1809
-    },
-    {
-      "epoch": 0.8903098868666994,
-      "grad_norm": 0.44579647844335313,
-      "learning_rate": 0.0001878174825004653,
-      "loss": 1.0486,
-      "step": 1810
-    },
-    {
-      "epoch": 0.8908017707820954,
-      "grad_norm": 0.44467914706379713,
-      "learning_rate": 0.00018778801174108327,
-      "loss": 0.9979,
-      "step": 1811
-    },
-    {
-      "epoch": 0.8912936546974914,
-      "grad_norm": 0.44786684227577395,
-      "learning_rate": 0.00018775850769628164,
-      "loss": 1.013,
-      "step": 1812
-    },
-    {
-      "epoch": 0.8917855386128873,
-      "grad_norm": 0.42923655580706227,
-      "learning_rate": 0.00018772897037724708,
-      "loss": 0.9077,
-      "step": 1813
-    },
-    {
-      "epoch": 0.8922774225282833,
-      "grad_norm": 0.42947719794182665,
-      "learning_rate": 0.00018769939979517883,
-      "loss": 0.9336,
-      "step": 1814
-    },
-    {
-      "epoch": 0.8927693064436792,
-      "grad_norm": 0.4515765543587999,
-      "learning_rate": 0.0001876697959612888,
-      "loss": 0.9822,
-      "step": 1815
-    },
-    {
-      "epoch": 0.8932611903590753,
-      "grad_norm": 0.4433570383113576,
-      "learning_rate": 0.0001876401588868015,
-      "loss": 0.8458,
-      "step": 1816
-    },
-    {
-      "epoch": 0.8937530742744713,
-      "grad_norm": 0.44141775764517255,
-      "learning_rate": 0.00018761048858295396,
-      "loss": 0.9131,
-      "step": 1817
-    },
-    {
-      "epoch": 0.8942449581898672,
-      "grad_norm": 0.4907548622677925,
-      "learning_rate": 0.00018758078506099594,
-      "loss": 0.9734,
-      "step": 1818
-    },
-    {
-      "epoch": 0.8947368421052632,
-      "grad_norm": 0.42816057571922717,
-      "learning_rate": 0.0001875510483321897,
-      "loss": 0.8329,
-      "step": 1819
-    },
-    {
-      "epoch": 0.8952287260206592,
-      "grad_norm": 0.4278009393349627,
-      "learning_rate": 0.00018752127840781016,
-      "loss": 0.9078,
-      "step": 1820
-    },
-    {
-      "epoch": 0.8957206099360551,
-      "grad_norm": 0.4558932669114811,
-      "learning_rate": 0.00018749147529914477,
-      "loss": 0.9961,
-      "step": 1821
-    },
-    {
-      "epoch": 0.8962124938514511,
-      "grad_norm": 0.46760053325802814,
-      "learning_rate": 0.00018746163901749356,
-      "loss": 0.8574,
-      "step": 1822
-    },
-    {
-      "epoch": 0.896704377766847,
-      "grad_norm": 0.44382122126076795,
-      "learning_rate": 0.0001874317695741692,
-      "loss": 0.9047,
-      "step": 1823
-    },
-    {
-      "epoch": 0.897196261682243,
-      "grad_norm": 0.46080332998879403,
-      "learning_rate": 0.00018740186698049682,
-      "loss": 0.8905,
-      "step": 1824
-    },
-    {
-      "epoch": 0.897688145597639,
-      "grad_norm": 0.4685192139052703,
-      "learning_rate": 0.00018737193124781425,
-      "loss": 0.9739,
-      "step": 1825
-    },
-    {
-      "epoch": 0.8981800295130349,
-      "grad_norm": 0.4594300245239511,
-      "learning_rate": 0.00018734196238747185,
-      "loss": 0.9846,
-      "step": 1826
-    },
-    {
-      "epoch": 0.8986719134284309,
-      "grad_norm": 0.5865265442015458,
-      "learning_rate": 0.00018731196041083244,
-      "loss": 0.8612,
-      "step": 1827
-    },
-    {
-      "epoch": 0.8991637973438269,
-      "grad_norm": 0.42629041077369384,
-      "learning_rate": 0.0001872819253292715,
-      "loss": 0.9517,
-      "step": 1828
-    },
-    {
-      "epoch": 0.8996556812592228,
-      "grad_norm": 0.45426128027116996,
-      "learning_rate": 0.00018725185715417708,
-      "loss": 0.953,
-      "step": 1829
-    },
-    {
-      "epoch": 0.9001475651746188,
-      "grad_norm": 0.46208254440984714,
-      "learning_rate": 0.0001872217558969497,
-      "loss": 0.8879,
-      "step": 1830
-    },
-    {
-      "epoch": 0.9006394490900148,
-      "grad_norm": 0.4518760176288391,
-      "learning_rate": 0.00018719162156900242,
-      "loss": 0.9349,
-      "step": 1831
-    },
-    {
-      "epoch": 0.9011313330054107,
-      "grad_norm": 0.4507850113266245,
-      "learning_rate": 0.00018716145418176092,
-      "loss": 0.9039,
-      "step": 1832
-    },
-    {
-      "epoch": 0.9016232169208067,
-      "grad_norm": 0.4257408028936146,
-      "learning_rate": 0.00018713125374666338,
-      "loss": 0.8838,
-      "step": 1833
-    },
-    {
-      "epoch": 0.9021151008362026,
-      "grad_norm": 0.48557527688552893,
-      "learning_rate": 0.00018710102027516047,
-      "loss": 1.0541,
-      "step": 1834
-    },
-    {
-      "epoch": 0.9026069847515986,
-      "grad_norm": 0.4236455716031953,
-      "learning_rate": 0.00018707075377871543,
-      "loss": 0.8758,
-      "step": 1835
-    },
-    {
-      "epoch": 0.9030988686669946,
-      "grad_norm": 0.464174581369836,
-      "learning_rate": 0.00018704045426880397,
-      "loss": 0.8693,
-      "step": 1836
-    },
-    {
-      "epoch": 0.9035907525823905,
-      "grad_norm": 0.42982053434237333,
-      "learning_rate": 0.0001870101217569144,
-      "loss": 0.9359,
-      "step": 1837
-    },
-    {
-      "epoch": 0.9040826364977865,
-      "grad_norm": 0.511779869667633,
-      "learning_rate": 0.0001869797562545475,
-      "loss": 1.0436,
-      "step": 1838
-    },
-    {
-      "epoch": 0.9045745204131825,
-      "grad_norm": 0.4888814299069477,
-      "learning_rate": 0.00018694935777321652,
-      "loss": 0.8595,
-      "step": 1839
-    },
-    {
-      "epoch": 0.9050664043285784,
-      "grad_norm": 0.4508780961225434,
-      "learning_rate": 0.0001869189263244473,
-      "loss": 0.895,
-      "step": 1840
-    },
-    {
-      "epoch": 0.9055582882439744,
-      "grad_norm": 0.4475570054583435,
-      "learning_rate": 0.00018688846191977808,
-      "loss": 0.9599,
-      "step": 1841
-    },
-    {
-      "epoch": 0.9060501721593703,
-      "grad_norm": 0.4866002867345899,
-      "learning_rate": 0.00018685796457075963,
-      "loss": 0.9138,
-      "step": 1842
-    },
-    {
-      "epoch": 0.9065420560747663,
-      "grad_norm": 0.46993297013305496,
-      "learning_rate": 0.0001868274342889553,
-      "loss": 1.0056,
-      "step": 1843
-    },
-    {
-      "epoch": 0.9070339399901624,
-      "grad_norm": 0.4298751366446946,
-      "learning_rate": 0.00018679687108594082,
-      "loss": 0.869,
-      "step": 1844
-    },
-    {
-      "epoch": 0.9075258239055582,
-      "grad_norm": 0.4444019633615243,
-      "learning_rate": 0.0001867662749733044,
-      "loss": 0.9685,
-      "step": 1845
-    },
-    {
-      "epoch": 0.9080177078209543,
-      "grad_norm": 0.4722580187848079,
-      "learning_rate": 0.0001867356459626468,
-      "loss": 0.9801,
-      "step": 1846
-    },
-    {
-      "epoch": 0.9085095917363503,
-      "grad_norm": 0.4628673412343863,
-      "learning_rate": 0.00018670498406558124,
-      "loss": 0.9607,
-      "step": 1847
-    },
-    {
-      "epoch": 0.9090014756517462,
-      "grad_norm": 0.44309659797891776,
-      "learning_rate": 0.00018667428929373335,
-      "loss": 0.9297,
-      "step": 1848
-    },
-    {
-      "epoch": 0.9094933595671422,
-      "grad_norm": 0.46922581862168145,
-      "learning_rate": 0.00018664356165874123,
-      "loss": 0.9994,
-      "step": 1849
-    },
-    {
-      "epoch": 0.9099852434825382,
-      "grad_norm": 0.4398640279326358,
-      "learning_rate": 0.00018661280117225555,
-      "loss": 0.9195,
-      "step": 1850
-    },
-    {
-      "epoch": 0.9104771273979341,
-      "grad_norm": 0.43088492080767665,
-      "learning_rate": 0.00018658200784593928,
-      "loss": 0.8881,
-      "step": 1851
-    },
-    {
-      "epoch": 0.9109690113133301,
-      "grad_norm": 0.46171660251764074,
-      "learning_rate": 0.00018655118169146797,
-      "loss": 0.9336,
-      "step": 1852
-    },
-    {
-      "epoch": 0.911460895228726,
-      "grad_norm": 0.49775393116646227,
-      "learning_rate": 0.00018652032272052958,
-      "loss": 0.9685,
-      "step": 1853
-    },
-    {
-      "epoch": 0.911952779144122,
-      "grad_norm": 0.4944238767925219,
-      "learning_rate": 0.00018648943094482442,
-      "loss": 0.8922,
-      "step": 1854
-    },
-    {
-      "epoch": 0.912444663059518,
-      "grad_norm": 0.4581618552443388,
-      "learning_rate": 0.00018645850637606537,
-      "loss": 0.9662,
-      "step": 1855
-    },
-    {
-      "epoch": 0.9129365469749139,
-      "grad_norm": 0.4489032703552213,
-      "learning_rate": 0.0001864275490259777,
-      "loss": 0.9587,
-      "step": 1856
-    },
-    {
-      "epoch": 0.9134284308903099,
-      "grad_norm": 0.42582539486312315,
-      "learning_rate": 0.00018639655890629909,
-      "loss": 0.919,
-      "step": 1857
-    },
-    {
-      "epoch": 0.9139203148057059,
-      "grad_norm": 0.4377317417163939,
-      "learning_rate": 0.00018636553602877963,
-      "loss": 0.8757,
-      "step": 1858
-    },
-    {
-      "epoch": 0.9144121987211018,
-      "grad_norm": 0.4032185122618743,
-      "learning_rate": 0.00018633448040518186,
-      "loss": 0.845,
-      "step": 1859
-    },
-    {
-      "epoch": 0.9149040826364978,
-      "grad_norm": 0.7158335031811692,
-      "learning_rate": 0.00018630339204728076,
-      "loss": 1.0356,
-      "step": 1860
-    },
-    {
-      "epoch": 0.9153959665518937,
-      "grad_norm": 0.44170501967634584,
-      "learning_rate": 0.00018627227096686366,
-      "loss": 0.9958,
-      "step": 1861
-    },
-    {
-      "epoch": 0.9158878504672897,
-      "grad_norm": 0.416661324416237,
-      "learning_rate": 0.00018624111717573035,
-      "loss": 0.8612,
-      "step": 1862
-    },
-    {
-      "epoch": 0.9163797343826857,
-      "grad_norm": 0.4382768634773933,
-      "learning_rate": 0.000186209930685693,
-      "loss": 0.9949,
-      "step": 1863
-    },
-    {
-      "epoch": 0.9168716182980816,
-      "grad_norm": 0.5154781599152937,
-      "learning_rate": 0.00018617871150857616,
-      "loss": 0.9891,
-      "step": 1864
-    },
-    {
-      "epoch": 0.9173635022134776,
-      "grad_norm": 0.45208056070997843,
-      "learning_rate": 0.00018614745965621677,
-      "loss": 0.8528,
-      "step": 1865
-    },
-    {
-      "epoch": 0.9178553861288736,
-      "grad_norm": 0.5629004632365258,
-      "learning_rate": 0.00018611617514046426,
-      "loss": 0.8792,
-      "step": 1866
-    },
-    {
-      "epoch": 0.9183472700442695,
-      "grad_norm": 0.43089672623691216,
-      "learning_rate": 0.00018608485797318028,
-      "loss": 0.9007,
-      "step": 1867
-    },
-    {
-      "epoch": 0.9188391539596655,
-      "grad_norm": 0.44375858233910803,
-      "learning_rate": 0.00018605350816623903,
-      "loss": 0.9179,
-      "step": 1868
-    },
-    {
-      "epoch": 0.9193310378750615,
-      "grad_norm": 0.4396800077461436,
-      "learning_rate": 0.00018602212573152693,
-      "loss": 0.8481,
-      "step": 1869
-    },
-    {
-      "epoch": 0.9198229217904574,
-      "grad_norm": 0.5170865212268547,
-      "learning_rate": 0.0001859907106809429,
-      "loss": 0.9622,
-      "step": 1870
-    },
-    {
-      "epoch": 0.9203148057058534,
-      "grad_norm": 0.48029214092701894,
-      "learning_rate": 0.00018595926302639813,
-      "loss": 0.9137,
-      "step": 1871
-    },
-    {
-      "epoch": 0.9208066896212493,
-      "grad_norm": 0.42481304631971484,
-      "learning_rate": 0.0001859277827798162,
-      "loss": 0.9244,
-      "step": 1872
-    },
-    {
-      "epoch": 0.9212985735366453,
-      "grad_norm": 0.7570373907074284,
-      "learning_rate": 0.00018589626995313313,
-      "loss": 0.9975,
-      "step": 1873
-    },
-    {
-      "epoch": 0.9217904574520414,
-      "grad_norm": 0.4345765525864509,
-      "learning_rate": 0.00018586472455829714,
-      "loss": 0.9243,
-      "step": 1874
-    },
-    {
-      "epoch": 0.9222823413674373,
-      "grad_norm": 0.6711781365151313,
-      "learning_rate": 0.00018583314660726888,
-      "loss": 0.9892,
-      "step": 1875
-    },
-    {
-      "epoch": 0.9227742252828333,
-      "grad_norm": 0.4348899393531873,
-      "learning_rate": 0.00018580153611202143,
-      "loss": 0.9168,
-      "step": 1876
-    },
-    {
-      "epoch": 0.9232661091982293,
-      "grad_norm": 0.41152523031057653,
-      "learning_rate": 0.00018576989308454004,
-      "loss": 0.8642,
-      "step": 1877
-    },
-    {
-      "epoch": 0.9237579931136252,
-      "grad_norm": 0.43576406709295346,
-      "learning_rate": 0.00018573821753682242,
-      "loss": 0.9849,
-      "step": 1878
-    },
-    {
-      "epoch": 0.9242498770290212,
-      "grad_norm": 0.5755530057660726,
-      "learning_rate": 0.00018570650948087857,
-      "loss": 0.9933,
-      "step": 1879
-    },
-    {
-      "epoch": 0.9247417609444171,
-      "grad_norm": 0.45628544644922775,
-      "learning_rate": 0.0001856747689287308,
-      "loss": 0.9279,
-      "step": 1880
-    },
-    {
-      "epoch": 0.9252336448598131,
-      "grad_norm": 0.49067584406698944,
-      "learning_rate": 0.00018564299589241375,
-      "loss": 1.007,
-      "step": 1881
-    },
-    {
-      "epoch": 0.9257255287752091,
-      "grad_norm": 0.552174844805938,
-      "learning_rate": 0.0001856111903839744,
-      "loss": 0.9696,
-      "step": 1882
-    },
-    {
-      "epoch": 0.926217412690605,
-      "grad_norm": 0.4464444978628804,
-      "learning_rate": 0.000185579352415472,
-      "loss": 0.9474,
-      "step": 1883
-    },
-    {
-      "epoch": 0.926709296606001,
-      "grad_norm": 0.4491417360834481,
-      "learning_rate": 0.00018554748199897813,
-      "loss": 0.9586,
-      "step": 1884
-    },
-    {
-      "epoch": 0.927201180521397,
-      "grad_norm": 0.43677339425529066,
-      "learning_rate": 0.0001855155791465767,
-      "loss": 0.896,
-      "step": 1885
-    },
-    {
-      "epoch": 0.9276930644367929,
-      "grad_norm": 0.43873421867582957,
-      "learning_rate": 0.0001854836438703639,
-      "loss": 0.8738,
-      "step": 1886
-    },
-    {
-      "epoch": 0.9281849483521889,
-      "grad_norm": 0.4324939602446171,
-      "learning_rate": 0.0001854516761824482,
-      "loss": 0.863,
-      "step": 1887
-    },
-    {
-      "epoch": 0.9286768322675848,
-      "grad_norm": 0.4551328220273714,
-      "learning_rate": 0.00018541967609495032,
-      "loss": 0.9598,
-      "step": 1888
-    },
-    {
-      "epoch": 0.9291687161829808,
-      "grad_norm": 0.4478932130948516,
-      "learning_rate": 0.00018538764362000337,
-      "loss": 0.9441,
-      "step": 1889
-    },
-    {
-      "epoch": 0.9296606000983768,
-      "grad_norm": 0.44028936711062305,
-      "learning_rate": 0.00018535557876975264,
-      "loss": 0.9548,
-      "step": 1890
-    },
-    {
-      "epoch": 0.9301524840137727,
-      "grad_norm": 0.43604633499273593,
-      "learning_rate": 0.00018532348155635576,
-      "loss": 0.9036,
-      "step": 1891
-    },
-    {
-      "epoch": 0.9306443679291687,
-      "grad_norm": 0.44310703755069564,
-      "learning_rate": 0.00018529135199198257,
-      "loss": 0.864,
-      "step": 1892
-    },
-    {
-      "epoch": 0.9311362518445647,
-      "grad_norm": 0.4195606656335038,
-      "learning_rate": 0.00018525919008881525,
-      "loss": 0.8824,
-      "step": 1893
-    },
-    {
-      "epoch": 0.9316281357599606,
-      "grad_norm": 0.4752966391769209,
-      "learning_rate": 0.00018522699585904822,
-      "loss": 0.9175,
-      "step": 1894
-    },
-    {
-      "epoch": 0.9321200196753566,
-      "grad_norm": 0.43006861648123,
-      "learning_rate": 0.00018519476931488807,
-      "loss": 0.9666,
-      "step": 1895
-    },
-    {
-      "epoch": 0.9326119035907526,
-      "grad_norm": 0.4441307733494571,
-      "learning_rate": 0.0001851625104685538,
-      "loss": 0.9484,
-      "step": 1896
-    },
-    {
-      "epoch": 0.9331037875061485,
-      "grad_norm": 0.47865967363559037,
-      "learning_rate": 0.00018513021933227647,
-      "loss": 0.9295,
-      "step": 1897
-    },
-    {
-      "epoch": 0.9335956714215445,
-      "grad_norm": 0.5011728461681424,
-      "learning_rate": 0.00018509789591829957,
-      "loss": 0.9939,
-      "step": 1898
-    },
-    {
-      "epoch": 0.9340875553369404,
-      "grad_norm": 0.44450221135964785,
-      "learning_rate": 0.0001850655402388787,
-      "loss": 0.937,
-      "step": 1899
-    },
-    {
-      "epoch": 0.9345794392523364,
-      "grad_norm": 0.47221343746374045,
-      "learning_rate": 0.00018503315230628176,
-      "loss": 0.9073,
-      "step": 1900
-    },
-    {
-      "epoch": 0.9350713231677324,
-      "grad_norm": 0.4526235162335821,
-      "learning_rate": 0.0001850007321327888,
-      "loss": 0.9197,
-      "step": 1901
-    },
-    {
-      "epoch": 0.9355632070831283,
-      "grad_norm": 0.4397491491786641,
-      "learning_rate": 0.00018496827973069223,
-      "loss": 0.9732,
-      "step": 1902
-    },
-    {
-      "epoch": 0.9360550909985244,
-      "grad_norm": 0.42750906504865316,
-      "learning_rate": 0.00018493579511229656,
-      "loss": 0.9582,
-      "step": 1903
-    },
-    {
-      "epoch": 0.9365469749139204,
-      "grad_norm": 0.44320595941036217,
-      "learning_rate": 0.00018490327828991852,
-      "loss": 0.953,
-      "step": 1904
-    },
-    {
-      "epoch": 0.9370388588293163,
-      "grad_norm": 0.42991703773405404,
-      "learning_rate": 0.0001848707292758871,
-      "loss": 0.9255,
-      "step": 1905
-    },
-    {
-      "epoch": 0.9375307427447123,
-      "grad_norm": 0.5780618216337025,
-      "learning_rate": 0.0001848381480825435,
-      "loss": 0.9496,
-      "step": 1906
-    },
-    {
-      "epoch": 0.9380226266601082,
-      "grad_norm": 0.4831354064630574,
-      "learning_rate": 0.00018480553472224114,
-      "loss": 1.0226,
-      "step": 1907
-    },
-    {
-      "epoch": 0.9385145105755042,
-      "grad_norm": 0.7668692313759613,
-      "learning_rate": 0.0001847728892073455,
-      "loss": 0.9747,
-      "step": 1908
-    },
-    {
-      "epoch": 0.9390063944909002,
-      "grad_norm": 0.4471821230833163,
-      "learning_rate": 0.0001847402115502344,
-      "loss": 0.885,
-      "step": 1909
-    },
-    {
-      "epoch": 0.9394982784062961,
-      "grad_norm": 0.49974458096838337,
-      "learning_rate": 0.00018470750176329781,
-      "loss": 0.9677,
-      "step": 1910
-    },
-    {
-      "epoch": 0.9399901623216921,
-      "grad_norm": 0.5012870583762135,
-      "learning_rate": 0.00018467475985893783,
-      "loss": 0.9629,
-      "step": 1911
-    },
-    {
-      "epoch": 0.9404820462370881,
-      "grad_norm": 0.4367049482339108,
-      "learning_rate": 0.0001846419858495688,
-      "loss": 0.8721,
-      "step": 1912
-    },
-    {
-      "epoch": 0.940973930152484,
-      "grad_norm": 0.4699932772670166,
-      "learning_rate": 0.00018460917974761717,
-      "loss": 0.8508,
-      "step": 1913
-    },
-    {
-      "epoch": 0.94146581406788,
-      "grad_norm": 0.4718433943987246,
-      "learning_rate": 0.00018457634156552168,
-      "loss": 0.8985,
-      "step": 1914
-    },
-    {
-      "epoch": 0.941957697983276,
-      "grad_norm": 0.4974602660062723,
-      "learning_rate": 0.00018454347131573306,
-      "loss": 1.0154,
-      "step": 1915
-    },
-    {
-      "epoch": 0.9424495818986719,
-      "grad_norm": 0.4258535297177866,
-      "learning_rate": 0.00018451056901071434,
-      "loss": 0.9385,
-      "step": 1916
-    },
-    {
-      "epoch": 0.9429414658140679,
-      "grad_norm": 0.5211541078048513,
-      "learning_rate": 0.00018447763466294067,
-      "loss": 0.9845,
-      "step": 1917
-    },
-    {
-      "epoch": 0.9434333497294638,
-      "grad_norm": 0.45479187706728935,
-      "learning_rate": 0.00018444466828489923,
-      "loss": 0.9489,
-      "step": 1918
-    },
-    {
-      "epoch": 0.9439252336448598,
-      "grad_norm": 0.46391141014345066,
-      "learning_rate": 0.00018441166988908956,
-      "loss": 0.9351,
-      "step": 1919
-    },
-    {
-      "epoch": 0.9444171175602558,
-      "grad_norm": 0.4745079284455217,
-      "learning_rate": 0.00018437863948802315,
-      "loss": 0.9196,
-      "step": 1920
-    },
-    {
-      "epoch": 0.9449090014756517,
-      "grad_norm": 0.5151632004372109,
-      "learning_rate": 0.00018434557709422376,
-      "loss": 0.9256,
-      "step": 1921
-    },
-    {
-      "epoch": 0.9454008853910477,
-      "grad_norm": 0.47566366606315597,
-      "learning_rate": 0.00018431248272022716,
-      "loss": 0.9531,
-      "step": 1922
-    },
-    {
-      "epoch": 0.9458927693064437,
-      "grad_norm": 0.44407590076749176,
-      "learning_rate": 0.00018427935637858135,
-      "loss": 0.9311,
-      "step": 1923
-    },
-    {
-      "epoch": 0.9463846532218396,
-      "grad_norm": 0.46497905603862427,
-      "learning_rate": 0.0001842461980818464,
-      "loss": 0.916,
-      "step": 1924
-    },
-    {
-      "epoch": 0.9468765371372356,
-      "grad_norm": 0.4074746058828995,
-      "learning_rate": 0.0001842130078425945,
-      "loss": 0.8571,
-      "step": 1925
-    },
-    {
-      "epoch": 0.9473684210526315,
-      "grad_norm": 0.46808976566775873,
-      "learning_rate": 0.00018417978567340996,
-      "loss": 0.9876,
-      "step": 1926
-    },
-    {
-      "epoch": 0.9478603049680275,
-      "grad_norm": 0.4544606752020085,
-      "learning_rate": 0.00018414653158688914,
-      "loss": 0.8897,
-      "step": 1927
-    },
-    {
-      "epoch": 0.9483521888834235,
-      "grad_norm": 0.5237793407899939,
-      "learning_rate": 0.0001841132455956406,
-      "loss": 0.959,
-      "step": 1928
-    },
-    {
-      "epoch": 0.9488440727988194,
-      "grad_norm": 0.4504943544679579,
-      "learning_rate": 0.00018407992771228497,
-      "loss": 0.9106,
-      "step": 1929
-    },
-    {
-      "epoch": 0.9493359567142154,
-      "grad_norm": 0.4247941082073109,
-      "learning_rate": 0.0001840465779494549,
-      "loss": 0.9221,
-      "step": 1930
-    },
-    {
-      "epoch": 0.9498278406296115,
-      "grad_norm": 0.5203584029782751,
-      "learning_rate": 0.0001840131963197952,
-      "loss": 1.0279,
-      "step": 1931
-    },
-    {
-      "epoch": 0.9503197245450073,
-      "grad_norm": 0.462036073986436,
-      "learning_rate": 0.00018397978283596274,
-      "loss": 0.8616,
-      "step": 1932
-    },
-    {
-      "epoch": 0.9508116084604034,
-      "grad_norm": 0.408317546888834,
-      "learning_rate": 0.0001839463375106265,
-      "loss": 0.9045,
-      "step": 1933
-    },
-    {
-      "epoch": 0.9513034923757994,
-      "grad_norm": 0.4588025515575967,
-      "learning_rate": 0.0001839128603564675,
-      "loss": 0.9361,
-      "step": 1934
-    },
-    {
-      "epoch": 0.9517953762911953,
-      "grad_norm": 0.4317585534792301,
-      "learning_rate": 0.00018387935138617875,
-      "loss": 0.8756,
-      "step": 1935
-    },
-    {
-      "epoch": 0.9522872602065913,
-      "grad_norm": 0.4519435724897018,
-      "learning_rate": 0.00018384581061246547,
-      "loss": 0.8793,
-      "step": 1936
-    },
-    {
-      "epoch": 0.9527791441219872,
-      "grad_norm": 0.4520350266397909,
-      "learning_rate": 0.00018381223804804484,
-      "loss": 0.9543,
-      "step": 1937
-    },
-    {
-      "epoch": 0.9532710280373832,
-      "grad_norm": 0.41525151813417516,
-      "learning_rate": 0.00018377863370564617,
-      "loss": 0.8823,
-      "step": 1938
-    },
-    {
-      "epoch": 0.9537629119527792,
-      "grad_norm": 0.4268831399183173,
-      "learning_rate": 0.00018374499759801074,
-      "loss": 0.9312,
-      "step": 1939
-    },
-    {
-      "epoch": 0.9542547958681751,
-      "grad_norm": 0.4253702887268113,
-      "learning_rate": 0.00018371132973789192,
-      "loss": 0.8114,
-      "step": 1940
-    },
-    {
-      "epoch": 0.9547466797835711,
-      "grad_norm": 0.4819668518465043,
-      "learning_rate": 0.00018367763013805508,
-      "loss": 0.9675,
-      "step": 1941
-    },
-    {
-      "epoch": 0.9552385636989671,
-      "grad_norm": 0.4547317992060099,
-      "learning_rate": 0.0001836438988112777,
-      "loss": 0.957,
-      "step": 1942
-    },
-    {
-      "epoch": 0.955730447614363,
-      "grad_norm": 0.44309298006477255,
-      "learning_rate": 0.0001836101357703492,
-      "loss": 0.9799,
-      "step": 1943
-    },
-    {
-      "epoch": 0.956222331529759,
-      "grad_norm": 0.4664756768139931,
-      "learning_rate": 0.00018357634102807112,
-      "loss": 1.0056,
-      "step": 1944
-    },
-    {
-      "epoch": 0.9567142154451549,
-      "grad_norm": 0.4539679694753678,
-      "learning_rate": 0.0001835425145972569,
-      "loss": 0.8856,
-      "step": 1945
-    },
-    {
-      "epoch": 0.9572060993605509,
-      "grad_norm": 0.44164075657025315,
-      "learning_rate": 0.0001835086564907321,
-      "loss": 0.8615,
-      "step": 1946
-    },
-    {
-      "epoch": 0.9576979832759469,
-      "grad_norm": 0.45247831990312676,
-      "learning_rate": 0.00018347476672133422,
-      "loss": 0.9535,
-      "step": 1947
-    },
-    {
-      "epoch": 0.9581898671913428,
-      "grad_norm": 0.44221000588098075,
-      "learning_rate": 0.00018344084530191282,
-      "loss": 0.9398,
-      "step": 1948
-    },
-    {
-      "epoch": 0.9586817511067388,
-      "grad_norm": 0.4213546371708338,
-      "learning_rate": 0.0001834068922453294,
-      "loss": 0.8714,
-      "step": 1949
-    },
-    {
-      "epoch": 0.9591736350221348,
-      "grad_norm": 0.46276619653227,
-      "learning_rate": 0.00018337290756445757,
-      "loss": 0.9773,
-      "step": 1950
-    },
-    {
-      "epoch": 0.9596655189375307,
-      "grad_norm": 0.41336397476424397,
-      "learning_rate": 0.00018333889127218278,
-      "loss": 0.8715,
-      "step": 1951
-    },
-    {
-      "epoch": 0.9601574028529267,
-      "grad_norm": 0.4323107804018267,
-      "learning_rate": 0.0001833048433814026,
-      "loss": 0.9519,
-      "step": 1952
-    },
-    {
-      "epoch": 0.9606492867683227,
-      "grad_norm": 0.4161203428466473,
-      "learning_rate": 0.00018327076390502645,
-      "loss": 0.9167,
-      "step": 1953
-    },
-    {
-      "epoch": 0.9611411706837186,
-      "grad_norm": 0.41943509986690275,
-      "learning_rate": 0.00018323665285597583,
-      "loss": 0.9048,
-      "step": 1954
-    },
-    {
-      "epoch": 0.9616330545991146,
-      "grad_norm": 0.4043101057931535,
-      "learning_rate": 0.00018320251024718414,
-      "loss": 0.8473,
-      "step": 1955
-    },
-    {
-      "epoch": 0.9621249385145105,
-      "grad_norm": 0.42181886666341173,
-      "learning_rate": 0.00018316833609159682,
-      "loss": 0.9289,
-      "step": 1956
-    },
-    {
-      "epoch": 0.9626168224299065,
-      "grad_norm": 0.40080973404412934,
-      "learning_rate": 0.00018313413040217124,
-      "loss": 0.8799,
-      "step": 1957
-    },
-    {
-      "epoch": 0.9631087063453025,
-      "grad_norm": 0.4487507479231779,
-      "learning_rate": 0.00018309989319187673,
-      "loss": 0.9578,
-      "step": 1958
-    },
-    {
-      "epoch": 0.9636005902606984,
-      "grad_norm": 0.44091223877862584,
-      "learning_rate": 0.00018306562447369449,
-      "loss": 1.0044,
-      "step": 1959
-    },
-    {
-      "epoch": 0.9640924741760944,
-      "grad_norm": 0.47143391770110404,
-      "learning_rate": 0.00018303132426061782,
-      "loss": 0.8695,
-      "step": 1960
-    },
-    {
-      "epoch": 0.9645843580914905,
-      "grad_norm": 0.40991211217100143,
-      "learning_rate": 0.0001829969925656518,
-      "loss": 0.8824,
-      "step": 1961
-    },
-    {
-      "epoch": 0.9650762420068864,
-      "grad_norm": 0.44327349800592386,
-      "learning_rate": 0.0001829626294018136,
-      "loss": 0.9499,
-      "step": 1962
-    },
-    {
-      "epoch": 0.9655681259222824,
-      "grad_norm": 0.5725296845392949,
-      "learning_rate": 0.00018292823478213217,
-      "loss": 0.9597,
-      "step": 1963
-    },
-    {
-      "epoch": 0.9660600098376783,
-      "grad_norm": 0.4440215432884469,
-      "learning_rate": 0.0001828938087196485,
-      "loss": 0.9184,
-      "step": 1964
-    },
-    {
-      "epoch": 0.9665518937530743,
-      "grad_norm": 0.42147265987377897,
-      "learning_rate": 0.00018285935122741554,
-      "loss": 0.9281,
-      "step": 1965
-    },
-    {
-      "epoch": 0.9670437776684703,
-      "grad_norm": 0.4188232523231526,
-      "learning_rate": 0.00018282486231849794,
-      "loss": 0.8534,
-      "step": 1966
-    },
-    {
-      "epoch": 0.9675356615838662,
-      "grad_norm": 1.2377315932762718,
-      "learning_rate": 0.00018279034200597248,
-      "loss": 0.9696,
-      "step": 1967
-    },
-    {
-      "epoch": 0.9680275454992622,
-      "grad_norm": 0.4472282474869554,
-      "learning_rate": 0.00018275579030292778,
-      "loss": 0.9089,
-      "step": 1968
-    },
-    {
-      "epoch": 0.9685194294146582,
-      "grad_norm": 0.4191650854370175,
-      "learning_rate": 0.00018272120722246434,
-      "loss": 0.8669,
-      "step": 1969
-    },
-    {
-      "epoch": 0.9690113133300541,
-      "grad_norm": 0.43753161804389906,
-      "learning_rate": 0.00018268659277769457,
-      "loss": 0.8819,
-      "step": 1970
-    },
-    {
-      "epoch": 0.9695031972454501,
-      "grad_norm": 0.4176731671835176,
-      "learning_rate": 0.00018265194698174276,
-      "loss": 0.9417,
-      "step": 1971
-    },
-    {
-      "epoch": 0.9699950811608461,
-      "grad_norm": 0.5380636017973098,
-      "learning_rate": 0.00018261726984774508,
-      "loss": 0.8814,
-      "step": 1972
-    },
-    {
-      "epoch": 0.970486965076242,
-      "grad_norm": 0.4693424487883787,
-      "learning_rate": 0.00018258256138884967,
-      "loss": 1.08,
-      "step": 1973
-    },
-    {
-      "epoch": 0.970978848991638,
-      "grad_norm": 0.44287467230128147,
-      "learning_rate": 0.0001825478216182164,
-      "loss": 0.8724,
-      "step": 1974
-    },
-    {
-      "epoch": 0.9714707329070339,
-      "grad_norm": 0.47401241076558015,
-      "learning_rate": 0.00018251305054901713,
-      "loss": 0.9435,
-      "step": 1975
-    },
-    {
-      "epoch": 0.9719626168224299,
-      "grad_norm": 0.4716995663392478,
-      "learning_rate": 0.00018247824819443554,
-      "loss": 0.8926,
-      "step": 1976
-    },
-    {
-      "epoch": 0.9724545007378259,
-      "grad_norm": 0.4048631732017963,
-      "learning_rate": 0.00018244341456766718,
-      "loss": 0.8727,
-      "step": 1977
-    },
-    {
-      "epoch": 0.9729463846532218,
-      "grad_norm": 0.5501419065196186,
-      "learning_rate": 0.00018240854968191945,
-      "loss": 0.9842,
-      "step": 1978
-    },
-    {
-      "epoch": 0.9734382685686178,
-      "grad_norm": 0.4983078287581272,
-      "learning_rate": 0.0001823736535504116,
-      "loss": 0.9187,
-      "step": 1979
-    },
-    {
-      "epoch": 0.9739301524840138,
-      "grad_norm": 0.456374927682862,
-      "learning_rate": 0.00018233872618637475,
-      "loss": 0.8765,
-      "step": 1980
-    },
-    {
-      "epoch": 0.9744220363994097,
-      "grad_norm": 0.4361771046637676,
-      "learning_rate": 0.00018230376760305185,
-      "loss": 0.9327,
-      "step": 1981
-    },
-    {
-      "epoch": 0.9749139203148057,
-      "grad_norm": 0.5218870199750633,
-      "learning_rate": 0.0001822687778136977,
-      "loss": 0.9182,
-      "step": 1982
-    },
-    {
-      "epoch": 0.9754058042302016,
-      "grad_norm": 0.4377434158553101,
-      "learning_rate": 0.00018223375683157884,
-      "loss": 0.9139,
-      "step": 1983
-    },
-    {
-      "epoch": 0.9758976881455976,
-      "grad_norm": 0.441508679114558,
-      "learning_rate": 0.0001821987046699738,
-      "loss": 0.8847,
-      "step": 1984
-    },
-    {
-      "epoch": 0.9763895720609936,
-      "grad_norm": 0.4231708327844232,
-      "learning_rate": 0.0001821636213421728,
-      "loss": 0.8684,
-      "step": 1985
-    },
-    {
-      "epoch": 0.9768814559763895,
-      "grad_norm": 0.46641621666423816,
-      "learning_rate": 0.00018212850686147793,
-      "loss": 0.9781,
-      "step": 1986
-    },
-    {
-      "epoch": 0.9773733398917855,
-      "grad_norm": 0.47553941657395565,
-      "learning_rate": 0.00018209336124120308,
-      "loss": 0.9948,
-      "step": 1987
-    },
-    {
-      "epoch": 0.9778652238071815,
-      "grad_norm": 0.46381453154697494,
-      "learning_rate": 0.00018205818449467398,
-      "loss": 0.9003,
-      "step": 1988
-    },
-    {
-      "epoch": 0.9783571077225774,
-      "grad_norm": 0.396028159540396,
-      "learning_rate": 0.00018202297663522807,
-      "loss": 0.8598,
-      "step": 1989
-    },
-    {
-      "epoch": 0.9788489916379735,
-      "grad_norm": 0.44853225265759955,
-      "learning_rate": 0.0001819877376762147,
-      "loss": 0.9052,
-      "step": 1990
-    },
-    {
-      "epoch": 0.9793408755533693,
-      "grad_norm": 0.4410343172515614,
-      "learning_rate": 0.00018195246763099494,
-      "loss": 0.8545,
-      "step": 1991
-    },
-    {
-      "epoch": 0.9798327594687654,
-      "grad_norm": 0.445285394442272,
-      "learning_rate": 0.00018191716651294167,
-      "loss": 0.9216,
-      "step": 1992
-    },
-    {
-      "epoch": 0.9803246433841614,
-      "grad_norm": 0.46444107702664084,
-      "learning_rate": 0.0001818818343354396,
-      "loss": 0.9005,
-      "step": 1993
-    },
-    {
-      "epoch": 0.9808165272995573,
-      "grad_norm": 0.4673226564151492,
-      "learning_rate": 0.00018184647111188506,
-      "loss": 0.9788,
-      "step": 1994
-    },
-    {
-      "epoch": 0.9813084112149533,
-      "grad_norm": 0.4338580685874424,
-      "learning_rate": 0.0001818110768556863,
-      "loss": 0.9595,
-      "step": 1995
-    },
-    {
-      "epoch": 0.9818002951303493,
-      "grad_norm": 0.4605529891465894,
-      "learning_rate": 0.00018177565158026334,
-      "loss": 0.8888,
-      "step": 1996
-    },
-    {
-      "epoch": 0.9822921790457452,
-      "grad_norm": 0.461882261804312,
-      "learning_rate": 0.00018174019529904785,
-      "loss": 0.9512,
-      "step": 1997
-    },
-    {
-      "epoch": 0.9827840629611412,
-      "grad_norm": 0.4185592648902373,
-      "learning_rate": 0.00018170470802548336,
-      "loss": 0.8775,
-      "step": 1998
-    },
-    {
-      "epoch": 0.9832759468765372,
-      "grad_norm": 0.46017098695300246,
-      "learning_rate": 0.00018166918977302506,
-      "loss": 0.9257,
-      "step": 1999
-    },
-    {
-      "epoch": 0.9837678307919331,
-      "grad_norm": 0.43770465454669844,
-      "learning_rate": 0.00018163364055514002,
-      "loss": 0.8512,
-      "step": 2000
-    },
-    {
-      "epoch": 0.9842597147073291,
-      "grad_norm": 0.425725999392596,
-      "learning_rate": 0.00018159806038530688,
-      "loss": 0.9506,
-      "step": 2001
-    },
-    {
-      "epoch": 0.984751598622725,
-      "grad_norm": 0.4363217963946529,
-      "learning_rate": 0.00018156244927701616,
-      "loss": 0.8999,
-      "step": 2002
-    },
-    {
-      "epoch": 0.985243482538121,
-      "grad_norm": 0.693829954051338,
-      "learning_rate": 0.00018152680724377004,
-      "loss": 0.8858,
-      "step": 2003
-    },
-    {
-      "epoch": 0.985735366453517,
-      "grad_norm": 0.45623689519322974,
-      "learning_rate": 0.00018149113429908242,
-      "loss": 0.8933,
-      "step": 2004
-    },
-    {
-      "epoch": 0.9862272503689129,
-      "grad_norm": 0.3961063923489086,
-      "learning_rate": 0.00018145543045647896,
-      "loss": 0.8796,
-      "step": 2005
-    },
-    {
-      "epoch": 0.9867191342843089,
-      "grad_norm": 0.44731886453675274,
-      "learning_rate": 0.000181419695729497,
-      "loss": 0.9288,
-      "step": 2006
-    },
-    {
-      "epoch": 0.9872110181997049,
-      "grad_norm": 0.39864669645559797,
-      "learning_rate": 0.0001813839301316856,
-      "loss": 0.8674,
-      "step": 2007
-    },
-    {
-      "epoch": 0.9877029021151008,
-      "grad_norm": 0.4178994922394768,
-      "learning_rate": 0.00018134813367660555,
-      "loss": 0.9006,
-      "step": 2008
-    },
-    {
-      "epoch": 0.9881947860304968,
-      "grad_norm": 0.4998770359493896,
-      "learning_rate": 0.0001813123063778293,
-      "loss": 0.9131,
-      "step": 2009
-    },
-    {
-      "epoch": 0.9886866699458927,
-      "grad_norm": 0.4541264244365027,
-      "learning_rate": 0.00018127644824894102,
-      "loss": 0.9381,
-      "step": 2010
-    },
-    {
-      "epoch": 0.9891785538612887,
-      "grad_norm": 0.45508736170043473,
-      "learning_rate": 0.00018124055930353653,
-      "loss": 0.9072,
-      "step": 2011
-    },
-    {
-      "epoch": 0.9896704377766847,
-      "grad_norm": 0.4873708340449761,
-      "learning_rate": 0.00018120463955522345,
-      "loss": 0.8998,
-      "step": 2012
-    },
-    {
-      "epoch": 0.9901623216920806,
-      "grad_norm": 0.45549724651475265,
-      "learning_rate": 0.00018116868901762092,
-      "loss": 0.934,
-      "step": 2013
-    },
-    {
-      "epoch": 0.9906542056074766,
-      "grad_norm": 0.42652167135627134,
-      "learning_rate": 0.00018113270770435985,
-      "loss": 0.9115,
-      "step": 2014
-    },
-    {
-      "epoch": 0.9911460895228726,
-      "grad_norm": 0.4453477772420515,
-      "learning_rate": 0.0001810966956290828,
-      "loss": 0.9395,
-      "step": 2015
-    },
-    {
-      "epoch": 0.9916379734382685,
-      "grad_norm": 0.4046617832179208,
-      "learning_rate": 0.000181060652805444,
-      "loss": 0.8994,
-      "step": 2016
-    },
-    {
-      "epoch": 0.9921298573536645,
-      "grad_norm": 0.4283315925798483,
-      "learning_rate": 0.00018102457924710935,
-      "loss": 0.8584,
-      "step": 2017
-    },
-    {
-      "epoch": 0.9926217412690606,
-      "grad_norm": 0.3899504326189742,
-      "learning_rate": 0.00018098847496775635,
-      "loss": 0.8959,
-      "step": 2018
-    },
-    {
-      "epoch": 0.9931136251844564,
-      "grad_norm": 1.2193680639418478,
-      "learning_rate": 0.0001809523399810742,
-      "loss": 0.9118,
-      "step": 2019
-    },
-    {
-      "epoch": 0.9936055090998525,
-      "grad_norm": 0.4479688984977468,
-      "learning_rate": 0.0001809161743007637,
-      "loss": 0.9338,
-      "step": 2020
-    },
-    {
-      "epoch": 0.9940973930152484,
-      "grad_norm": 0.44531325059233584,
-      "learning_rate": 0.00018087997794053733,
-      "loss": 1.0175,
-      "step": 2021
-    },
-    {
-      "epoch": 0.9945892769306444,
-      "grad_norm": 0.7929071437715141,
-      "learning_rate": 0.0001808437509141192,
-      "loss": 0.9242,
-      "step": 2022
-    },
-    {
-      "epoch": 0.9950811608460404,
-      "grad_norm": 0.41910134077034905,
-      "learning_rate": 0.000180807493235245,
-      "loss": 0.9019,
-      "step": 2023
-    },
-    {
-      "epoch": 0.9955730447614363,
-      "grad_norm": 0.5899825884926907,
-      "learning_rate": 0.00018077120491766208,
-      "loss": 0.9682,
-      "step": 2024
-    },
-    {
-      "epoch": 0.9960649286768323,
-      "grad_norm": 0.44754835592415726,
-      "learning_rate": 0.00018073488597512944,
-      "loss": 0.944,
-      "step": 2025
-    },
-    {
-      "epoch": 0.9965568125922283,
-      "grad_norm": 0.4195479696566459,
-      "learning_rate": 0.00018069853642141762,
-      "loss": 0.965,
-      "step": 2026
-    },
-    {
-      "epoch": 0.9970486965076242,
-      "grad_norm": 0.4288060668519654,
-      "learning_rate": 0.0001806621562703088,
-      "loss": 0.9673,
-      "step": 2027
-    },
-    {
-      "epoch": 0.9975405804230202,
-      "grad_norm": 0.41813471233399524,
-      "learning_rate": 0.0001806257455355968,
-      "loss": 0.8521,
-      "step": 2028
-    },
-    {
-      "epoch": 0.9980324643384161,
-      "grad_norm": 0.40576008584353307,
-      "learning_rate": 0.00018058930423108693,
-      "loss": 0.9327,
-      "step": 2029
-    },
-    {
-      "epoch": 0.9985243482538121,
-      "grad_norm": 5.737513840425279,
-      "learning_rate": 0.00018055283237059622,
-      "loss": 0.9231,
-      "step": 2030
-    },
-    {
-      "epoch": 0.9990162321692081,
-      "grad_norm": 0.4424306323239711,
-      "learning_rate": 0.00018051632996795317,
-      "loss": 0.9896,
-      "step": 2031
-    },
-    {
-      "epoch": 0.999508116084604,
-      "grad_norm": 0.4472237618121447,
-      "learning_rate": 0.00018047979703699797,
-      "loss": 0.8683,
-      "step": 2032
-    },
-    {
-      "epoch": 1.0,
-      "grad_norm": 0.484997707838406,
-      "learning_rate": 0.0001804432335915823,
-      "loss": 1.0024,
-      "step": 2033
-    },
-    {
-      "epoch": 1.000491883915396,
-      "grad_norm": 0.46317454321941864,
-      "learning_rate": 0.0001804066396455694,
-      "loss": 0.9482,
-      "step": 2034
-    },
-    {
-      "epoch": 1.000491883915396,
-      "eval_loss": 0.8394731283187866,
-      "eval_runtime": 6667.7597,
-      "eval_samples_per_second": 4.274,
-      "eval_steps_per_second": 2.137,
-      "step": 2034
-    },
-    {
-      "epoch": 1.000983767830792,
-      "grad_norm": 0.44238663483109564,
-      "learning_rate": 0.00018037001521283418,
-      "loss": 0.8669,
-      "step": 2035
-    },
-    {
-      "epoch": 1.0014756517461878,
-      "grad_norm": 0.4357808574170002,
-      "learning_rate": 0.000180333360307263,
-      "loss": 0.9221,
-      "step": 2036
-    },
-    {
-      "epoch": 1.0019675356615838,
-      "grad_norm": 0.4211852718407382,
-      "learning_rate": 0.0001802966749427538,
-      "loss": 0.8529,
-      "step": 2037
-    },
-    {
-      "epoch": 1.000184535892231,
-      "grad_norm": 0.4965038022203159,
-      "learning_rate": 0.00018025995913321615,
-      "loss": 0.8953,
-      "step": 2038
-    },
-    {
-      "epoch": 1.0006766316048472,
-      "grad_norm": 0.554681476096924,
-      "learning_rate": 0.00018022321289257103,
-      "loss": 0.716,
-      "step": 2039
-    },
-    {
-      "epoch": 1.0011687273174632,
-      "grad_norm": 0.556727123533325,
-      "learning_rate": 0.00018018643623475105,
-      "loss": 0.7354,
-      "step": 2040
-    },
-    {
-      "epoch": 1.0016608230300794,
-      "grad_norm": 0.485911946754523,
-      "learning_rate": 0.00018014962917370036,
-      "loss": 0.7437,
-      "step": 2041
-    },
-    {
-      "epoch": 1.0021529187426954,
-      "grad_norm": 0.701277557366079,
-      "learning_rate": 0.00018011279172337456,
-      "loss": 0.7734,
-      "step": 2042
-    },
-    {
-      "epoch": 1.0026450144553116,
-      "grad_norm": 0.5987959982751151,
-      "learning_rate": 0.00018007592389774086,
-      "loss": 0.7576,
-      "step": 2043
-    },
-    {
-      "epoch": 1.0031371101679276,
-      "grad_norm": 0.5011515957192303,
-      "learning_rate": 0.00018003902571077794,
-      "loss": 0.7445,
-      "step": 2044
-    },
-    {
-      "epoch": 1.0036292058805438,
-      "grad_norm": 0.5326564253251685,
-      "learning_rate": 0.00018000209717647595,
-      "loss": 0.7925,
-      "step": 2045
-    },
-    {
-      "epoch": 1.0041213015931598,
-      "grad_norm": 0.5289613239491519,
-      "learning_rate": 0.00017996513830883664,
-      "loss": 0.7474,
-      "step": 2046
-    },
-    {
-      "epoch": 1.004613397305776,
-      "grad_norm": 0.48940192303125907,
-      "learning_rate": 0.00017992814912187318,
-      "loss": 0.7219,
-      "step": 2047
-    },
-    {
-      "epoch": 1.005105493018392,
-      "grad_norm": 0.5272442447192495,
-      "learning_rate": 0.00017989112962961033,
-      "loss": 0.7195,
-      "step": 2048
-    },
-    {
-      "epoch": 1.0055975887310082,
-      "grad_norm": 0.4363098304248932,
-      "learning_rate": 0.0001798540798460842,
-      "loss": 0.6887,
-      "step": 2049
-    },
-    {
-      "epoch": 1.0060896844436242,
-      "grad_norm": 0.46195532336117584,
-      "learning_rate": 0.00017981699978534256,
-      "loss": 0.7519,
-      "step": 2050
-    },
-    {
-      "epoch": 1.0065817801562404,
-      "grad_norm": 0.46424481755052327,
-      "learning_rate": 0.0001797798894614445,
-      "loss": 0.6281,
-      "step": 2051
-    },
-    {
-      "epoch": 1.0070738758688564,
-      "grad_norm": 1.671733380094309,
-      "learning_rate": 0.00017974274888846065,
-      "loss": 0.6882,
-      "step": 2052
-    },
-    {
-      "epoch": 1.0075659715814727,
-      "grad_norm": 0.4677554918316789,
-      "learning_rate": 0.00017970557808047314,
-      "loss": 0.7027,
-      "step": 2053
-    },
-    {
-      "epoch": 1.0080580672940886,
-      "grad_norm": 0.5008453388603415,
-      "learning_rate": 0.0001796683770515755,
-      "loss": 0.7026,
-      "step": 2054
-    },
-    {
-      "epoch": 1.0085501630067049,
-      "grad_norm": 0.44121797690454334,
-      "learning_rate": 0.0001796311458158728,
-      "loss": 0.7218,
-      "step": 2055
-    },
-    {
-      "epoch": 1.0090422587193209,
-      "grad_norm": 0.47635340403842863,
-      "learning_rate": 0.00017959388438748151,
-      "loss": 0.7384,
-      "step": 2056
-    },
-    {
-      "epoch": 1.009534354431937,
-      "grad_norm": 0.4765597636860264,
-      "learning_rate": 0.0001795565927805295,
-      "loss": 0.778,
-      "step": 2057
-    },
-    {
-      "epoch": 1.010026450144553,
-      "grad_norm": 0.5194102087512645,
-      "learning_rate": 0.00017951927100915617,
-      "loss": 0.7989,
-      "step": 2058
-    },
-    {
-      "epoch": 1.0105185458571693,
-      "grad_norm": 0.4455080918099978,
-      "learning_rate": 0.00017948191908751234,
-      "loss": 0.683,
-      "step": 2059
-    },
-    {
-      "epoch": 1.0110106415697853,
-      "grad_norm": 0.46978662112903474,
-      "learning_rate": 0.00017944453702976022,
-      "loss": 0.753,
-      "step": 2060
-    },
-    {
-      "epoch": 1.0115027372824015,
-      "grad_norm": 0.4894148552681891,
-      "learning_rate": 0.00017940712485007347,
-      "loss": 0.7841,
-      "step": 2061
-    },
-    {
-      "epoch": 1.0119948329950175,
-      "grad_norm": 0.49302032266441115,
-      "learning_rate": 0.00017936968256263719,
-      "loss": 0.8665,
-      "step": 2062
-    },
-    {
-      "epoch": 1.0124869287076337,
-      "grad_norm": 0.4643383424097789,
-      "learning_rate": 0.00017933221018164784,
-      "loss": 0.7664,
-      "step": 2063
-    },
-    {
-      "epoch": 1.0129790244202497,
-      "grad_norm": 0.43714680751204893,
-      "learning_rate": 0.00017929470772131336,
-      "loss": 0.686,
-      "step": 2064
-    },
-    {
-      "epoch": 1.0134711201328659,
-      "grad_norm": 0.4343407075088194,
-      "learning_rate": 0.00017925717519585305,
-      "loss": 0.6869,
-      "step": 2065
-    },
-    {
-      "epoch": 1.0139632158454819,
-      "grad_norm": 0.4373287916799852,
-      "learning_rate": 0.00017921961261949763,
-      "loss": 0.7374,
-      "step": 2066
-    },
-    {
-      "epoch": 1.014455311558098,
-      "grad_norm": 0.46816481061573334,
-      "learning_rate": 0.00017918202000648917,
-      "loss": 0.7419,
-      "step": 2067
-    },
-    {
-      "epoch": 1.014947407270714,
-      "grad_norm": 0.49053088889668656,
-      "learning_rate": 0.00017914439737108128,
-      "loss": 0.6899,
-      "step": 2068
-    },
-    {
-      "epoch": 1.0154395029833303,
-      "grad_norm": 0.4802762766167413,
-      "learning_rate": 0.00017910674472753865,
-      "loss": 0.7384,
-      "step": 2069
-    },
-    {
-      "epoch": 1.0159315986959463,
-      "grad_norm": 0.44872570424600855,
-      "learning_rate": 0.0001790690620901377,
-      "loss": 0.7102,
-      "step": 2070
-    },
-    {
-      "epoch": 1.0164236944085625,
-      "grad_norm": 0.4705487092947291,
-      "learning_rate": 0.00017903134947316594,
-      "loss": 0.741,
-      "step": 2071
-    },
-    {
-      "epoch": 1.0169157901211785,
-      "grad_norm": 0.4321593935650233,
-      "learning_rate": 0.00017899360689092248,
-      "loss": 0.6556,
-      "step": 2072
-    },
-    {
-      "epoch": 1.0174078858337947,
-      "grad_norm": 0.5212546002206624,
-      "learning_rate": 0.00017895583435771758,
-      "loss": 0.8173,
-      "step": 2073
-    },
-    {
-      "epoch": 1.0178999815464107,
-      "grad_norm": 0.4635438789335315,
-      "learning_rate": 0.000178918031887873,
-      "loss": 0.7334,
-      "step": 2074
-    },
-    {
-      "epoch": 1.018392077259027,
-      "grad_norm": 0.4436241029043569,
-      "learning_rate": 0.00017888019949572178,
-      "loss": 0.7896,
-      "step": 2075
-    },
-    {
-      "epoch": 1.018884172971643,
-      "grad_norm": 0.45448345835544507,
-      "learning_rate": 0.00017884233719560832,
-      "loss": 0.8205,
-      "step": 2076
-    },
-    {
-      "epoch": 1.0193762686842591,
-      "grad_norm": 0.46201987319120674,
-      "learning_rate": 0.0001788044450018884,
-      "loss": 0.7084,
-      "step": 2077
-    },
-    {
-      "epoch": 1.0198683643968751,
-      "grad_norm": 0.47161403666785723,
-      "learning_rate": 0.0001787665229289291,
-      "loss": 0.714,
-      "step": 2078
-    },
-    {
-      "epoch": 1.0203604601094913,
-      "grad_norm": 0.4755426078626228,
-      "learning_rate": 0.0001787285709911088,
-      "loss": 0.7326,
-      "step": 2079
-    },
-    {
-      "epoch": 1.0208525558221073,
-      "grad_norm": 0.4505436214042156,
-      "learning_rate": 0.00017869058920281727,
-      "loss": 0.699,
-      "step": 2080
-    },
-    {
-      "epoch": 1.0213446515347235,
-      "grad_norm": 0.4579915340549243,
-      "learning_rate": 0.0001786525775784555,
-      "loss": 0.6847,
-      "step": 2081
-    },
-    {
-      "epoch": 1.0218367472473395,
-      "grad_norm": 0.4577831593591981,
-      "learning_rate": 0.00017861453613243593,
-      "loss": 0.7078,
-      "step": 2082
-    },
-    {
-      "epoch": 1.0223288429599557,
-      "grad_norm": 0.5072246200909621,
-      "learning_rate": 0.0001785764648791822,
-      "loss": 0.8272,
-      "step": 2083
-    },
-    {
-      "epoch": 1.0228209386725717,
-      "grad_norm": 0.47134968568286834,
-      "learning_rate": 0.0001785383638331293,
-      "loss": 0.7323,
-      "step": 2084
-    },
-    {
-      "epoch": 1.023313034385188,
-      "grad_norm": 0.45264240486960794,
-      "learning_rate": 0.00017850023300872346,
-      "loss": 0.7033,
-      "step": 2085
-    },
-    {
-      "epoch": 1.023805130097804,
-      "grad_norm": 0.47031111770245215,
-      "learning_rate": 0.00017846207242042228,
-      "loss": 0.7706,
-      "step": 2086
-    },
-    {
-      "epoch": 1.0242972258104202,
-      "grad_norm": 0.4658262557406207,
-      "learning_rate": 0.00017842388208269457,
-      "loss": 0.7325,
-      "step": 2087
-    },
-    {
-      "epoch": 1.0247893215230361,
-      "grad_norm": 0.46871993189108374,
-      "learning_rate": 0.00017838566201002046,
-      "loss": 0.7374,
-      "step": 2088
-    },
-    {
-      "epoch": 1.0252814172356524,
-      "grad_norm": 0.4754992124043651,
-      "learning_rate": 0.00017834741221689143,
-      "loss": 0.7363,
-      "step": 2089
-    },
-    {
-      "epoch": 1.0257735129482684,
-      "grad_norm": 0.4538167742547551,
-      "learning_rate": 0.00017830913271781005,
-      "loss": 0.7165,
-      "step": 2090
-    },
-    {
-      "epoch": 1.0262656086608846,
-      "grad_norm": 0.46410781780592864,
-      "learning_rate": 0.00017827082352729026,
-      "loss": 0.7135,
-      "step": 2091
-    },
-    {
-      "epoch": 1.0267577043735006,
-      "grad_norm": 0.4420478759408918,
-      "learning_rate": 0.00017823248465985732,
-      "loss": 0.7796,
-      "step": 2092
-    },
-    {
-      "epoch": 1.0272498000861168,
-      "grad_norm": 0.4443775997398954,
-      "learning_rate": 0.0001781941161300476,
-      "loss": 0.7298,
-      "step": 2093
-    },
-    {
-      "epoch": 1.0277418957987328,
-      "grad_norm": 0.46999968781222573,
-      "learning_rate": 0.0001781557179524088,
-      "loss": 0.7263,
-      "step": 2094
-    },
-    {
-      "epoch": 1.028233991511349,
-      "grad_norm": 0.4482118121027141,
-      "learning_rate": 0.0001781172901414999,
-      "loss": 0.722,
-      "step": 2095
-    },
-    {
-      "epoch": 1.028726087223965,
-      "grad_norm": 0.4601441480915822,
-      "learning_rate": 0.00017807883271189098,
-      "loss": 0.7515,
-      "step": 2096
-    },
-    {
-      "epoch": 1.0292181829365812,
-      "grad_norm": 0.46318942282534875,
-      "learning_rate": 0.0001780403456781635,
-      "loss": 0.7406,
-      "step": 2097
-    },
-    {
-      "epoch": 1.0297102786491972,
-      "grad_norm": 0.4350085824417785,
-      "learning_rate": 0.0001780018290549101,
-      "loss": 0.6987,
-      "step": 2098
-    },
-    {
-      "epoch": 1.0302023743618134,
-      "grad_norm": 0.45798874605425643,
-      "learning_rate": 0.00017796328285673454,
-      "loss": 0.6562,
-      "step": 2099
-    },
-    {
-      "epoch": 1.0306944700744294,
-      "grad_norm": 0.4656180646396438,
-      "learning_rate": 0.00017792470709825193,
-      "loss": 0.7056,
-      "step": 2100
-    },
-    {
-      "epoch": 1.0311865657870456,
-      "grad_norm": 0.5069432109096865,
-      "learning_rate": 0.00017788610179408852,
-      "loss": 0.7901,
-      "step": 2101
-    },
-    {
-      "epoch": 1.0316786614996616,
-      "grad_norm": 0.4717718604355321,
-      "learning_rate": 0.00017784746695888174,
-      "loss": 0.8137,
-      "step": 2102
-    },
-    {
-      "epoch": 1.0321707572122778,
-      "grad_norm": 0.43441187391989783,
-      "learning_rate": 0.0001778088026072803,
-      "loss": 0.6866,
-      "step": 2103
-    },
-    {
-      "epoch": 1.0326628529248938,
-      "grad_norm": 0.45221372848824265,
-      "learning_rate": 0.00017777010875394403,
-      "loss": 0.672,
-      "step": 2104
-    },
-    {
-      "epoch": 1.03315494863751,
-      "grad_norm": 0.4328549218543093,
-      "learning_rate": 0.00017773138541354397,
-      "loss": 0.6551,
-      "step": 2105
-    },
-    {
-      "epoch": 1.033647044350126,
-      "grad_norm": 0.45683272510185197,
-      "learning_rate": 0.00017769263260076232,
-      "loss": 0.7069,
-      "step": 2106
-    },
-    {
-      "epoch": 1.0341391400627422,
-      "grad_norm": 0.5068602570947087,
-      "learning_rate": 0.00017765385033029248,
-      "loss": 0.6768,
-      "step": 2107
-    },
-    {
-      "epoch": 1.0346312357753582,
-      "grad_norm": 0.45976586253535545,
-      "learning_rate": 0.000177615038616839,
-      "loss": 0.7623,
-      "step": 2108
-    },
-    {
-      "epoch": 1.0351233314879744,
-      "grad_norm": 0.45961453783662415,
-      "learning_rate": 0.00017757619747511765,
-      "loss": 0.7822,
-      "step": 2109
-    },
-    {
-      "epoch": 1.0356154272005904,
-      "grad_norm": 0.4663721653624339,
-      "learning_rate": 0.00017753732691985525,
-      "loss": 0.7308,
-      "step": 2110
-    },
-    {
-      "epoch": 1.0361075229132066,
-      "grad_norm": 0.4374999127844936,
-      "learning_rate": 0.00017749842696578987,
-      "loss": 0.7453,
-      "step": 2111
-    },
-    {
-      "epoch": 1.0365996186258226,
-      "grad_norm": 0.4549081742466566,
-      "learning_rate": 0.00017745949762767072,
-      "loss": 0.83,
-      "step": 2112
-    },
-    {
-      "epoch": 1.0370917143384388,
-      "grad_norm": 0.44504919310170415,
-      "learning_rate": 0.00017742053892025802,
-      "loss": 0.727,
-      "step": 2113
-    },
-    {
-      "epoch": 1.0375838100510548,
-      "grad_norm": 0.4609188149655354,
-      "learning_rate": 0.00017738155085832337,
-      "loss": 0.6878,
-      "step": 2114
-    },
-    {
-      "epoch": 1.038075905763671,
-      "grad_norm": 0.47085891920119277,
-      "learning_rate": 0.00017734253345664925,
-      "loss": 0.7279,
-      "step": 2115
-    },
-    {
-      "epoch": 1.0385680014762873,
-      "grad_norm": 0.45176435553856,
-      "learning_rate": 0.0001773034867300294,
-      "loss": 0.7784,
-      "step": 2116
-    },
-    {
-      "epoch": 1.0390600971889032,
-      "grad_norm": 0.42587095371218864,
-      "learning_rate": 0.00017726441069326865,
-      "loss": 0.6972,
-      "step": 2117
-    },
-    {
-      "epoch": 1.0395521929015192,
-      "grad_norm": 0.44384491365038603,
-      "learning_rate": 0.000177225305361183,
-      "loss": 0.7523,
-      "step": 2118
-    },
-    {
-      "epoch": 1.0400442886141354,
-      "grad_norm": 0.43432218603855066,
-      "learning_rate": 0.0001771861707485994,
-      "loss": 0.7433,
-      "step": 2119
-    },
-    {
-      "epoch": 1.0405363843267517,
-      "grad_norm": 0.4729564357018369,
-      "learning_rate": 0.00017714700687035607,
-      "loss": 0.7847,
-      "step": 2120
-    },
-    {
-      "epoch": 1.0410284800393677,
-      "grad_norm": 0.4516088185913565,
-      "learning_rate": 0.00017710781374130226,
-      "loss": 0.7404,
-      "step": 2121
-    },
-    {
-      "epoch": 1.0415205757519839,
-      "grad_norm": 0.5076092053877148,
-      "learning_rate": 0.00017706859137629825,
-      "loss": 0.7635,
-      "step": 2122
-    },
-    {
-      "epoch": 1.0420126714645999,
-      "grad_norm": 0.43121740216834487,
-      "learning_rate": 0.00017702933979021554,
-      "loss": 0.7523,
-      "step": 2123
-    },
-    {
-      "epoch": 1.042504767177216,
-      "grad_norm": 0.4377234039381213,
-      "learning_rate": 0.0001769900589979366,
-      "loss": 0.7238,
-      "step": 2124
-    },
-    {
-      "epoch": 1.042996862889832,
-      "grad_norm": 0.44984533226658263,
-      "learning_rate": 0.00017695074901435506,
-      "loss": 0.731,
-      "step": 2125
-    },
-    {
-      "epoch": 1.0434889586024483,
-      "grad_norm": 0.45407476366180155,
-      "learning_rate": 0.0001769114098543755,
-      "loss": 0.7172,
-      "step": 2126
-    },
-    {
-      "epoch": 1.0439810543150643,
-      "grad_norm": 0.4727108415085289,
-      "learning_rate": 0.00017687204153291365,
-      "loss": 0.7908,
-      "step": 2127
-    },
-    {
-      "epoch": 1.0444731500276805,
-      "grad_norm": 0.405208877396721,
-      "learning_rate": 0.00017683264406489625,
-      "loss": 0.6784,
-      "step": 2128
-    },
-    {
-      "epoch": 1.0449652457402965,
-      "grad_norm": 0.434904148515732,
-      "learning_rate": 0.00017679321746526117,
-      "loss": 0.6561,
-      "step": 2129
-    },
-    {
-      "epoch": 1.0454573414529127,
-      "grad_norm": 0.4592627747583997,
-      "learning_rate": 0.00017675376174895724,
-      "loss": 0.7214,
-      "step": 2130
-    },
-    {
-      "epoch": 1.0459494371655287,
-      "grad_norm": 0.44081523246731175,
-      "learning_rate": 0.00017671427693094437,
-      "loss": 0.7354,
-      "step": 2131
-    },
-    {
-      "epoch": 1.046441532878145,
-      "grad_norm": 0.4423562297857163,
-      "learning_rate": 0.00017667476302619354,
-      "loss": 0.6852,
-      "step": 2132
-    },
-    {
-      "epoch": 1.046933628590761,
-      "grad_norm": 0.485161027908596,
-      "learning_rate": 0.00017663522004968658,
-      "loss": 0.7069,
-      "step": 2133
-    },
-    {
-      "epoch": 1.047425724303377,
-      "grad_norm": 0.49260707389446934,
-      "learning_rate": 0.00017659564801641664,
-      "loss": 0.752,
-      "step": 2134
-    },
-    {
-      "epoch": 1.047917820015993,
-      "grad_norm": 0.45045532516594605,
-      "learning_rate": 0.00017655604694138762,
-      "loss": 0.7502,
-      "step": 2135
-    },
-    {
-      "epoch": 1.0484099157286093,
-      "grad_norm": 0.46244390432668536,
-      "learning_rate": 0.00017651641683961457,
-      "loss": 0.7466,
-      "step": 2136
-    },
-    {
-      "epoch": 1.0489020114412253,
-      "grad_norm": 0.47524874462793565,
-      "learning_rate": 0.00017647675772612353,
-      "loss": 0.7272,
-      "step": 2137
-    },
-    {
-      "epoch": 1.0493941071538415,
-      "grad_norm": 0.4493309339046456,
-      "learning_rate": 0.00017643706961595148,
-      "loss": 0.7626,
-      "step": 2138
-    },
-    {
-      "epoch": 1.0498862028664575,
-      "grad_norm": 0.48658266275674944,
-      "learning_rate": 0.00017639735252414647,
-      "loss": 0.7389,
-      "step": 2139
-    },
-    {
-      "epoch": 1.0503782985790737,
-      "grad_norm": 0.4842574407728781,
-      "learning_rate": 0.00017635760646576748,
-      "loss": 0.8179,
-      "step": 2140
-    },
-    {
-      "epoch": 1.0508703942916897,
-      "grad_norm": 0.48790494389203287,
-      "learning_rate": 0.00017631783145588453,
-      "loss": 0.7801,
-      "step": 2141
-    },
-    {
-      "epoch": 1.051362490004306,
-      "grad_norm": 0.43366852253965477,
-      "learning_rate": 0.00017627802750957853,
-      "loss": 0.6612,
-      "step": 2142
-    },
-    {
-      "epoch": 1.051854585716922,
-      "grad_norm": 0.43545977277735415,
-      "learning_rate": 0.00017623819464194148,
-      "loss": 0.7348,
-      "step": 2143
-    },
-    {
-      "epoch": 1.0523466814295381,
-      "grad_norm": 0.4494947095246579,
-      "learning_rate": 0.00017619833286807625,
-      "loss": 0.6911,
-      "step": 2144
-    },
-    {
-      "epoch": 1.0528387771421541,
-      "grad_norm": 0.4832898099383395,
-      "learning_rate": 0.00017615844220309667,
-      "loss": 0.7399,
-      "step": 2145
-    },
-    {
-      "epoch": 1.0533308728547703,
-      "grad_norm": 0.4845604328962755,
-      "learning_rate": 0.00017611852266212762,
-      "loss": 0.7519,
-      "step": 2146
-    },
-    {
-      "epoch": 1.0538229685673863,
-      "grad_norm": 0.5155619319445711,
-      "learning_rate": 0.00017607857426030484,
-      "loss": 0.7451,
-      "step": 2147
-    },
-    {
-      "epoch": 1.0543150642800025,
-      "grad_norm": 0.4203321493512875,
-      "learning_rate": 0.00017603859701277502,
-      "loss": 0.7028,
-      "step": 2148
-    },
-    {
-      "epoch": 1.0548071599926185,
-      "grad_norm": 0.4389995266081142,
-      "learning_rate": 0.0001759985909346958,
-      "loss": 0.711,
-      "step": 2149
-    },
-    {
-      "epoch": 1.0552992557052348,
-      "grad_norm": 0.5154157780916582,
-      "learning_rate": 0.0001759585560412358,
-      "loss": 0.7303,
-      "step": 2150
-    },
-    {
-      "epoch": 1.0557913514178507,
-      "grad_norm": 0.42620801001840763,
-      "learning_rate": 0.00017591849234757447,
-      "loss": 0.7073,
-      "step": 2151
-    },
-    {
-      "epoch": 1.056283447130467,
-      "grad_norm": 0.4419775781650784,
-      "learning_rate": 0.00017587839986890228,
-      "loss": 0.6977,
-      "step": 2152
-    },
-    {
-      "epoch": 1.056775542843083,
-      "grad_norm": 0.45640857304759624,
-      "learning_rate": 0.00017583827862042054,
-      "loss": 0.7826,
-      "step": 2153
-    },
-    {
-      "epoch": 1.0572676385556992,
-      "grad_norm": 0.4681387690418811,
-      "learning_rate": 0.00017579812861734143,
-      "loss": 0.7432,
-      "step": 2154
-    },
-    {
-      "epoch": 1.0577597342683152,
-      "grad_norm": 0.42792426567868647,
-      "learning_rate": 0.00017575794987488824,
-      "loss": 0.6719,
-      "step": 2155
-    },
-    {
-      "epoch": 1.0582518299809314,
-      "grad_norm": 0.43220590496957695,
-      "learning_rate": 0.00017571774240829487,
-      "loss": 0.7404,
-      "step": 2156
-    },
-    {
-      "epoch": 1.0587439256935474,
-      "grad_norm": 0.45791478079235115,
-      "learning_rate": 0.00017567750623280633,
-      "loss": 0.6881,
-      "step": 2157
-    },
-    {
-      "epoch": 1.0592360214061636,
-      "grad_norm": 0.43224341363914714,
-      "learning_rate": 0.00017563724136367842,
-      "loss": 0.7255,
-      "step": 2158
-    },
-    {
-      "epoch": 1.0597281171187796,
-      "grad_norm": 0.4473423163991253,
-      "learning_rate": 0.00017559694781617787,
-      "loss": 0.7639,
-      "step": 2159
-    },
-    {
-      "epoch": 1.0602202128313958,
-      "grad_norm": 0.4471786879502586,
-      "learning_rate": 0.0001755566256055822,
-      "loss": 0.6978,
-      "step": 2160
-    },
-    {
-      "epoch": 1.0607123085440118,
-      "grad_norm": 0.5020232214769531,
-      "learning_rate": 0.00017551627474717986,
-      "loss": 0.7834,
-      "step": 2161
-    },
-    {
-      "epoch": 1.061204404256628,
-      "grad_norm": 0.5026469782707866,
-      "learning_rate": 0.00017547589525627018,
-      "loss": 0.8847,
-      "step": 2162
-    },
-    {
-      "epoch": 1.061696499969244,
-      "grad_norm": 0.47665535206322823,
-      "learning_rate": 0.0001754354871481633,
-      "loss": 0.6764,
-      "step": 2163
-    },
-    {
-      "epoch": 1.0621885956818602,
-      "grad_norm": 0.4420774179727459,
-      "learning_rate": 0.0001753950504381802,
-      "loss": 0.7618,
-      "step": 2164
-    },
-    {
-      "epoch": 1.0626806913944762,
-      "grad_norm": 0.4582259735972832,
-      "learning_rate": 0.00017535458514165278,
-      "loss": 0.7019,
-      "step": 2165
-    },
-    {
-      "epoch": 1.0631727871070924,
-      "grad_norm": 0.4776584769453669,
-      "learning_rate": 0.00017531409127392373,
-      "loss": 0.6775,
-      "step": 2166
-    },
-    {
-      "epoch": 1.0636648828197084,
-      "grad_norm": 0.4363715883984462,
-      "learning_rate": 0.00017527356885034653,
-      "loss": 0.7295,
-      "step": 2167
-    },
-    {
-      "epoch": 1.0641569785323246,
-      "grad_norm": 0.45831574764188365,
-      "learning_rate": 0.00017523301788628556,
-      "loss": 0.7827,
-      "step": 2168
-    },
-    {
-      "epoch": 1.0646490742449406,
-      "grad_norm": 0.4590135901022549,
-      "learning_rate": 0.000175192438397116,
-      "loss": 0.7199,
-      "step": 2169
-    },
-    {
-      "epoch": 1.0651411699575568,
-      "grad_norm": 0.46030473454145515,
-      "learning_rate": 0.00017515183039822383,
-      "loss": 0.787,
-      "step": 2170
-    },
-    {
-      "epoch": 1.0656332656701728,
-      "grad_norm": 0.46878407586423176,
-      "learning_rate": 0.00017511119390500586,
-      "loss": 0.7143,
-      "step": 2171
-    },
-    {
-      "epoch": 1.066125361382789,
-      "grad_norm": 0.42573502630114696,
-      "learning_rate": 0.00017507052893286966,
-      "loss": 0.6712,
-      "step": 2172
-    },
-    {
-      "epoch": 1.066617457095405,
-      "grad_norm": 0.4370551028018308,
-      "learning_rate": 0.00017502983549723365,
-      "loss": 0.7221,
-      "step": 2173
-    },
-    {
-      "epoch": 1.0671095528080212,
-      "grad_norm": 0.4231519476593116,
-      "learning_rate": 0.00017498911361352702,
-      "loss": 0.6729,
-      "step": 2174
-    },
-    {
-      "epoch": 1.0676016485206372,
-      "grad_norm": 0.4891257396776434,
-      "learning_rate": 0.00017494836329718974,
-      "loss": 0.7522,
-      "step": 2175
-    },
-    {
-      "epoch": 1.0680937442332534,
-      "grad_norm": 0.4236571353722022,
-      "learning_rate": 0.0001749075845636726,
-      "loss": 0.6895,
-      "step": 2176
-    },
-    {
-      "epoch": 1.0685858399458694,
-      "grad_norm": 0.4899655849378574,
-      "learning_rate": 0.0001748667774284371,
-      "loss": 0.7899,
-      "step": 2177
-    },
-    {
-      "epoch": 1.0690779356584856,
-      "grad_norm": 0.5124751631506923,
-      "learning_rate": 0.00017482594190695557,
-      "loss": 0.7775,
-      "step": 2178
-    },
-    {
-      "epoch": 1.0695700313711016,
-      "grad_norm": 0.4604651269621623,
-      "learning_rate": 0.000174785078014711,
-      "loss": 0.6805,
-      "step": 2179
-    },
-    {
-      "epoch": 1.0700621270837178,
-      "grad_norm": 0.5027190232775441,
-      "learning_rate": 0.00017474418576719734,
-      "loss": 0.7402,
-      "step": 2180
-    },
-    {
-      "epoch": 1.0705542227963338,
-      "grad_norm": 0.4384068034829691,
-      "learning_rate": 0.00017470326517991905,
-      "loss": 0.7724,
-      "step": 2181
-    },
-    {
-      "epoch": 1.07104631850895,
-      "grad_norm": 0.45884553767656433,
-      "learning_rate": 0.0001746623162683915,
-      "loss": 0.7247,
-      "step": 2182
-    },
-    {
-      "epoch": 1.071538414221566,
-      "grad_norm": 0.496978133299866,
-      "learning_rate": 0.00017462133904814074,
-      "loss": 0.7243,
-      "step": 2183
-    },
-    {
-      "epoch": 1.0720305099341823,
-      "grad_norm": 0.4465078830394886,
-      "learning_rate": 0.00017458033353470354,
-      "loss": 0.6879,
-      "step": 2184
-    },
-    {
-      "epoch": 1.0725226056467982,
-      "grad_norm": 0.4536245142154501,
-      "learning_rate": 0.0001745392997436275,
-      "loss": 0.781,
-      "step": 2185
-    },
-    {
-      "epoch": 1.0730147013594145,
-      "grad_norm": 2.9669028314572117,
-      "learning_rate": 0.00017449823769047072,
-      "loss": 0.8559,
-      "step": 2186
-    },
-    {
-      "epoch": 1.0735067970720304,
-      "grad_norm": 0.5012689594444691,
-      "learning_rate": 0.00017445714739080227,
-      "loss": 0.8166,
-      "step": 2187
-    },
-    {
-      "epoch": 1.0739988927846467,
-      "grad_norm": 0.4543299719609067,
-      "learning_rate": 0.0001744160288602018,
-      "loss": 0.7688,
-      "step": 2188
-    },
-    {
-      "epoch": 1.0744909884972627,
-      "grad_norm": 0.49085748053135525,
-      "learning_rate": 0.00017437488211425957,
-      "loss": 0.7885,
-      "step": 2189
-    },
-    {
-      "epoch": 1.0749830842098789,
-      "grad_norm": 0.4799641811424035,
-      "learning_rate": 0.0001743337071685768,
-      "loss": 0.6744,
-      "step": 2190
-    },
-    {
-      "epoch": 1.0754751799224949,
-      "grad_norm": 0.47759641698512745,
-      "learning_rate": 0.0001742925040387652,
-      "loss": 0.7843,
-      "step": 2191
-    },
-    {
-      "epoch": 1.075967275635111,
-      "grad_norm": 0.4896965131557718,
-      "learning_rate": 0.00017425127274044714,
-      "loss": 0.7888,
-      "step": 2192
-    },
-    {
-      "epoch": 1.076459371347727,
-      "grad_norm": 0.4795673015714201,
-      "learning_rate": 0.00017421001328925585,
-      "loss": 0.772,
-      "step": 2193
-    },
-    {
-      "epoch": 1.0769514670603433,
-      "grad_norm": 0.4311158106248663,
-      "learning_rate": 0.00017416872570083508,
-      "loss": 0.7267,
-      "step": 2194
-    },
-    {
-      "epoch": 1.0774435627729593,
-      "grad_norm": 0.43624552001017447,
-      "learning_rate": 0.00017412740999083931,
-      "loss": 0.6747,
-      "step": 2195
-    },
-    {
-      "epoch": 1.0779356584855755,
-      "grad_norm": 0.4627150853495084,
-      "learning_rate": 0.00017408606617493367,
-      "loss": 0.726,
-      "step": 2196
-    },
-    {
-      "epoch": 1.0784277541981915,
-      "grad_norm": 0.47247549042808046,
-      "learning_rate": 0.00017404469426879392,
-      "loss": 0.7389,
-      "step": 2197
-    },
-    {
-      "epoch": 1.0789198499108077,
-      "grad_norm": 0.4475268862699268,
-      "learning_rate": 0.00017400329428810655,
-      "loss": 0.7291,
-      "step": 2198
-    },
-    {
-      "epoch": 1.0794119456234237,
-      "grad_norm": 0.46337366071213587,
-      "learning_rate": 0.00017396186624856863,
-      "loss": 0.8107,
-      "step": 2199
-    },
-    {
-      "epoch": 1.07990404133604,
-      "grad_norm": 0.4282607673568657,
-      "learning_rate": 0.00017392041016588781,
-      "loss": 0.7312,
-      "step": 2200
-    },
-    {
-      "epoch": 1.080396137048656,
-      "grad_norm": 0.44883945305298184,
-      "learning_rate": 0.00017387892605578257,
-      "loss": 0.7327,
-      "step": 2201
-    },
-    {
-      "epoch": 1.080888232761272,
-      "grad_norm": 0.4622041375900797,
-      "learning_rate": 0.00017383741393398177,
-      "loss": 0.7729,
-      "step": 2202
-    },
-    {
-      "epoch": 1.081380328473888,
-      "grad_norm": 0.447719502426236,
-      "learning_rate": 0.00017379587381622513,
-      "loss": 0.7054,
-      "step": 2203
-    },
-    {
-      "epoch": 1.0818724241865043,
-      "grad_norm": 0.43936405353026553,
-      "learning_rate": 0.00017375430571826277,
-      "loss": 0.7353,
-      "step": 2204
-    },
-    {
-      "epoch": 1.0823645198991203,
-      "grad_norm": 0.45444989271124786,
-      "learning_rate": 0.00017371270965585556,
-      "loss": 0.7765,
-      "step": 2205
-    },
-    {
-      "epoch": 1.0828566156117365,
-      "grad_norm": 0.4333094373678045,
-      "learning_rate": 0.0001736710856447749,
-      "loss": 0.7195,
-      "step": 2206
-    },
-    {
-      "epoch": 1.0833487113243525,
-      "grad_norm": 0.4426912048826289,
-      "learning_rate": 0.00017362943370080282,
-      "loss": 0.7756,
-      "step": 2207
-    },
-    {
-      "epoch": 1.0838408070369687,
-      "grad_norm": 0.46608728261291205,
-      "learning_rate": 0.00017358775383973201,
-      "loss": 0.7369,
-      "step": 2208
-    },
-    {
-      "epoch": 1.0843329027495847,
-      "grad_norm": 0.5499731215025803,
-      "learning_rate": 0.00017354604607736556,
-      "loss": 0.7295,
-      "step": 2209
-    },
-    {
-      "epoch": 1.084824998462201,
-      "grad_norm": 0.47059985575331065,
-      "learning_rate": 0.00017350431042951735,
-      "loss": 0.8143,
-      "step": 2210
-    },
-    {
-      "epoch": 1.085317094174817,
-      "grad_norm": 0.4987450515535781,
-      "learning_rate": 0.00017346254691201165,
-      "loss": 0.7546,
-      "step": 2211
-    },
-    {
-      "epoch": 1.0858091898874331,
-      "grad_norm": 0.4157649272895208,
-      "learning_rate": 0.00017342075554068343,
-      "loss": 0.7087,
-      "step": 2212
-    },
-    {
-      "epoch": 1.0863012856000491,
-      "grad_norm": 0.4394968554901881,
-      "learning_rate": 0.00017337893633137817,
-      "loss": 0.7286,
-      "step": 2213
-    },
-    {
-      "epoch": 1.0867933813126653,
-      "grad_norm": 0.4454638368328683,
-      "learning_rate": 0.00017333708929995192,
-      "loss": 0.7267,
-      "step": 2214
-    },
-    {
-      "epoch": 1.0872854770252813,
-      "grad_norm": 0.463833356574586,
-      "learning_rate": 0.00017329521446227122,
-      "loss": 0.673,
-      "step": 2215
-    },
-    {
-      "epoch": 1.0877775727378975,
-      "grad_norm": 0.5488935729149627,
-      "learning_rate": 0.00017325331183421324,
-      "loss": 0.8265,
-      "step": 2216
-    },
-    {
-      "epoch": 1.0882696684505135,
-      "grad_norm": 0.4388729476804382,
-      "learning_rate": 0.0001732113814316656,
-      "loss": 0.7175,
-      "step": 2217
-    },
-    {
-      "epoch": 1.0887617641631298,
-      "grad_norm": 0.4445625165505587,
-      "learning_rate": 0.00017316942327052652,
-      "loss": 0.7486,
-      "step": 2218
-    },
-    {
-      "epoch": 1.0892538598757457,
-      "grad_norm": 0.44796055484436026,
-      "learning_rate": 0.00017312743736670473,
-      "loss": 0.7858,
-      "step": 2219
-    },
-    {
-      "epoch": 1.089745955588362,
-      "grad_norm": 0.44343202198450404,
-      "learning_rate": 0.00017308542373611948,
-      "loss": 0.7668,
-      "step": 2220
-    },
-    {
-      "epoch": 1.090238051300978,
-      "grad_norm": 0.40910960991257106,
-      "learning_rate": 0.00017304338239470052,
-      "loss": 0.6735,
-      "step": 2221
-    },
-    {
-      "epoch": 1.0907301470135942,
-      "grad_norm": 0.46148660147352416,
-      "learning_rate": 0.00017300131335838806,
-      "loss": 0.7881,
-      "step": 2222
-    },
-    {
-      "epoch": 1.0912222427262102,
-      "grad_norm": 0.43129583307473374,
-      "learning_rate": 0.0001729592166431329,
-      "loss": 0.7698,
-      "step": 2223
-    },
-    {
-      "epoch": 1.0917143384388264,
-      "grad_norm": 0.4424498838899368,
-      "learning_rate": 0.00017291709226489635,
-      "loss": 0.6895,
-      "step": 2224
-    },
-    {
-      "epoch": 1.0922064341514424,
-      "grad_norm": 0.49562791398681033,
-      "learning_rate": 0.00017287494023965003,
-      "loss": 0.8248,
-      "step": 2225
-    },
-    {
-      "epoch": 1.0926985298640586,
-      "grad_norm": 0.46740114333129823,
-      "learning_rate": 0.00017283276058337624,
-      "loss": 0.7096,
-      "step": 2226
-    },
-    {
-      "epoch": 1.0931906255766746,
-      "grad_norm": 0.4376232396500971,
-      "learning_rate": 0.0001727905533120677,
-      "loss": 0.6524,
-      "step": 2227
-    },
-    {
-      "epoch": 1.0936827212892908,
-      "grad_norm": 0.4575713395407467,
-      "learning_rate": 0.00017274831844172757,
-      "loss": 0.7172,
-      "step": 2228
-    },
-    {
-      "epoch": 1.0941748170019068,
-      "grad_norm": 0.43953153703339076,
-      "learning_rate": 0.00017270605598836944,
-      "loss": 0.6764,
-      "step": 2229
-    },
-    {
-      "epoch": 1.094666912714523,
-      "grad_norm": 0.445750420266202,
-      "learning_rate": 0.00017266376596801743,
-      "loss": 0.7335,
-      "step": 2230
-    },
-    {
-      "epoch": 1.095159008427139,
-      "grad_norm": 0.4611985883315243,
-      "learning_rate": 0.00017262144839670613,
-      "loss": 0.7505,
-      "step": 2231
-    },
-    {
-      "epoch": 1.0956511041397552,
-      "grad_norm": 0.45842080565617677,
-      "learning_rate": 0.0001725791032904805,
-      "loss": 0.7346,
-      "step": 2232
-    },
-    {
-      "epoch": 1.0961431998523712,
-      "grad_norm": 0.5740138281698376,
-      "learning_rate": 0.00017253673066539596,
-      "loss": 0.8259,
-      "step": 2233
-    },
-    {
-      "epoch": 1.0966352955649874,
-      "grad_norm": 0.42582203968776655,
-      "learning_rate": 0.0001724943305375184,
-      "loss": 0.6761,
-      "step": 2234
-    },
-    {
-      "epoch": 1.0971273912776034,
-      "grad_norm": 0.4652739229260749,
-      "learning_rate": 0.00017245190292292412,
-      "loss": 0.6901,
-      "step": 2235
-    },
-    {
-      "epoch": 1.0976194869902196,
-      "grad_norm": 0.432367305195393,
-      "learning_rate": 0.0001724094478376998,
-      "loss": 0.7304,
-      "step": 2236
-    },
-    {
-      "epoch": 1.0981115827028356,
-      "grad_norm": 0.5989411933659401,
-      "learning_rate": 0.00017236696529794262,
-      "loss": 0.7272,
-      "step": 2237
-    },
-    {
-      "epoch": 1.0986036784154518,
-      "grad_norm": 0.4698301123236689,
-      "learning_rate": 0.00017232445531976012,
-      "loss": 0.7513,
-      "step": 2238
-    },
-    {
-      "epoch": 1.099095774128068,
-      "grad_norm": 0.46462850411082196,
-      "learning_rate": 0.0001722819179192702,
-      "loss": 0.7998,
-      "step": 2239
-    },
-    {
-      "epoch": 1.099587869840684,
-      "grad_norm": 0.4548416372221925,
-      "learning_rate": 0.00017223935311260125,
-      "loss": 0.6745,
-      "step": 2240
-    },
-    {
-      "epoch": 1.1000799655533,
-      "grad_norm": 0.45439739200697815,
-      "learning_rate": 0.00017219676091589198,
-      "loss": 0.6473,
-      "step": 2241
-    },
-    {
-      "epoch": 1.1005720612659162,
-      "grad_norm": 0.5158247670472813,
-      "learning_rate": 0.00017215414134529153,
-      "loss": 0.7285,
-      "step": 2242
-    },
-    {
-      "epoch": 1.1010641569785324,
-      "grad_norm": 0.44207593480252416,
-      "learning_rate": 0.00017211149441695938,
-      "loss": 0.7564,
-      "step": 2243
-    },
-    {
-      "epoch": 1.1015562526911484,
-      "grad_norm": 0.44620600073109734,
-      "learning_rate": 0.0001720688201470654,
-      "loss": 0.7439,
-      "step": 2244
-    },
-    {
-      "epoch": 1.1020483484037644,
-      "grad_norm": 0.48831775008135214,
-      "learning_rate": 0.00017202611855178987,
-      "loss": 0.7847,
-      "step": 2245
-    },
-    {
-      "epoch": 1.1025404441163806,
-      "grad_norm": 0.45163132354543045,
-      "learning_rate": 0.00017198338964732334,
-      "loss": 0.6993,
-      "step": 2246
-    },
-    {
-      "epoch": 1.1030325398289968,
-      "grad_norm": 0.47662254693220946,
-      "learning_rate": 0.00017194063344986676,
-      "loss": 0.7711,
-      "step": 2247
-    },
-    {
-      "epoch": 1.1035246355416128,
-      "grad_norm": 0.46584053994009733,
-      "learning_rate": 0.00017189784997563147,
-      "loss": 0.7479,
-      "step": 2248
-    },
-    {
-      "epoch": 1.1040167312542288,
-      "grad_norm": 0.4581977072685601,
-      "learning_rate": 0.0001718550392408391,
-      "loss": 0.7518,
-      "step": 2249
-    },
-    {
-      "epoch": 1.104508826966845,
-      "grad_norm": 1.106498959634675,
-      "learning_rate": 0.00017181220126172164,
-      "loss": 0.938,
-      "step": 2250
-    },
-    {
-      "epoch": 1.1050009226794613,
-      "grad_norm": 0.4374225120852344,
-      "learning_rate": 0.00017176933605452137,
-      "loss": 0.7394,
-      "step": 2251
-    },
-    {
-      "epoch": 1.1054930183920773,
-      "grad_norm": 0.46203521006258097,
-      "learning_rate": 0.00017172644363549092,
-      "loss": 0.7336,
-      "step": 2252
-    },
-    {
-      "epoch": 1.1059851141046932,
-      "grad_norm": 0.4646136658884963,
-      "learning_rate": 0.0001716835240208933,
-      "loss": 0.7741,
-      "step": 2253
-    },
-    {
-      "epoch": 1.1064772098173095,
-      "grad_norm": 0.4505833272163009,
-      "learning_rate": 0.00017164057722700174,
-      "loss": 0.6399,
-      "step": 2254
-    },
-    {
-      "epoch": 1.1069693055299257,
-      "grad_norm": 0.46665113137328695,
-      "learning_rate": 0.00017159760327009976,
-      "loss": 0.7365,
-      "step": 2255
-    },
-    {
-      "epoch": 1.1074614012425417,
-      "grad_norm": 0.4598719270091006,
-      "learning_rate": 0.00017155460216648131,
-      "loss": 0.7157,
-      "step": 2256
-    },
-    {
-      "epoch": 1.1079534969551579,
-      "grad_norm": 1.7020270183409554,
-      "learning_rate": 0.0001715115739324505,
-      "loss": 0.7406,
-      "step": 2257
-    },
-    {
-      "epoch": 1.1084455926677739,
-      "grad_norm": 0.45547883137071377,
-      "learning_rate": 0.0001714685185843218,
-      "loss": 0.7748,
-      "step": 2258
-    },
-    {
-      "epoch": 1.10893768838039,
-      "grad_norm": 0.48366139004901476,
-      "learning_rate": 0.0001714254361384199,
-      "loss": 0.778,
-      "step": 2259
-    },
-    {
-      "epoch": 1.109429784093006,
-      "grad_norm": 0.5294775630886707,
-      "learning_rate": 0.00017138232661107982,
-      "loss": 0.789,
-      "step": 2260
-    },
-    {
-      "epoch": 1.1099218798056223,
-      "grad_norm": 0.46851869269796753,
-      "learning_rate": 0.00017133919001864687,
-      "loss": 0.8326,
-      "step": 2261
-    },
-    {
-      "epoch": 1.1104139755182383,
-      "grad_norm": 0.46248523358403254,
-      "learning_rate": 0.0001712960263774765,
-      "loss": 0.6937,
-      "step": 2262
-    },
-    {
-      "epoch": 1.1109060712308545,
-      "grad_norm": 0.49243432547808386,
-      "learning_rate": 0.00017125283570393457,
-      "loss": 0.6992,
-      "step": 2263
-    },
-    {
-      "epoch": 1.1113981669434705,
-      "grad_norm": 0.4284797536044024,
-      "learning_rate": 0.0001712096180143971,
-      "loss": 0.8002,
-      "step": 2264
-    },
-    {
-      "epoch": 1.1118902626560867,
-      "grad_norm": 0.46352338618065314,
-      "learning_rate": 0.00017116637332525035,
-      "loss": 0.7287,
-      "step": 2265
-    },
-    {
-      "epoch": 1.1123823583687027,
-      "grad_norm": 0.4245556645074505,
-      "learning_rate": 0.00017112310165289082,
-      "loss": 0.6463,
-      "step": 2266
-    },
-    {
-      "epoch": 1.112874454081319,
-      "grad_norm": 0.48546026679680565,
-      "learning_rate": 0.00017107980301372532,
-      "loss": 0.7225,
-      "step": 2267
-    },
-    {
-      "epoch": 1.113366549793935,
-      "grad_norm": 0.46099006053184366,
-      "learning_rate": 0.0001710364774241708,
-      "loss": 0.7582,
-      "step": 2268
-    },
-    {
-      "epoch": 1.1138586455065511,
-      "grad_norm": 0.45945252544408977,
-      "learning_rate": 0.0001709931249006544,
-      "loss": 0.7532,
-      "step": 2269
-    },
-    {
-      "epoch": 1.114350741219167,
-      "grad_norm": 0.478142870545973,
-      "learning_rate": 0.00017094974545961357,
-      "loss": 0.7544,
-      "step": 2270
-    },
-    {
-      "epoch": 1.1148428369317833,
-      "grad_norm": 0.46784284146501,
-      "learning_rate": 0.0001709063391174959,
-      "loss": 0.7667,
-      "step": 2271
-    },
-    {
-      "epoch": 1.1153349326443993,
-      "grad_norm": 0.4652646918442406,
-      "learning_rate": 0.00017086290589075925,
-      "loss": 0.7029,
-      "step": 2272
-    },
-    {
-      "epoch": 1.1158270283570155,
-      "grad_norm": 0.42024375595041674,
-      "learning_rate": 0.00017081944579587152,
-      "loss": 0.6997,
-      "step": 2273
-    },
-    {
-      "epoch": 1.1163191240696315,
-      "grad_norm": 0.4588599547020526,
-      "learning_rate": 0.000170775958849311,
-      "loss": 0.7094,
-      "step": 2274
-    },
-    {
-      "epoch": 1.1168112197822477,
-      "grad_norm": 0.5138911216762359,
-      "learning_rate": 0.00017073244506756602,
-      "loss": 0.7752,
-      "step": 2275
-    },
-    {
-      "epoch": 1.1173033154948637,
-      "grad_norm": 0.4537775613774461,
-      "learning_rate": 0.00017068890446713512,
-      "loss": 0.7158,
-      "step": 2276
-    },
-    {
-      "epoch": 1.11779541120748,
-      "grad_norm": 0.5499498245682529,
-      "learning_rate": 0.00017064533706452704,
-      "loss": 0.7772,
-      "step": 2277
-    },
-    {
-      "epoch": 1.118287506920096,
-      "grad_norm": 0.476587121108361,
-      "learning_rate": 0.0001706017428762606,
-      "loss": 0.7724,
-      "step": 2278
-    },
-    {
-      "epoch": 1.1187796026327121,
-      "grad_norm": 0.6875046666894814,
-      "learning_rate": 0.0001705581219188649,
-      "loss": 0.8251,
-      "step": 2279
-    },
-    {
-      "epoch": 1.1192716983453281,
-      "grad_norm": 0.434734476091617,
-      "learning_rate": 0.00017051447420887906,
-      "loss": 0.7132,
-      "step": 2280
-    },
-    {
-      "epoch": 1.1197637940579444,
-      "grad_norm": 0.46175851533503126,
-      "learning_rate": 0.00017047079976285247,
-      "loss": 0.7632,
-      "step": 2281
-    },
-    {
-      "epoch": 1.1202558897705603,
-      "grad_norm": 0.4750203299707701,
-      "learning_rate": 0.00017042709859734455,
-      "loss": 0.7548,
-      "step": 2282
-    },
-    {
-      "epoch": 1.1207479854831766,
-      "grad_norm": 0.6847769258262533,
-      "learning_rate": 0.00017038337072892485,
-      "loss": 0.8399,
-      "step": 2283
-    },
-    {
-      "epoch": 1.1212400811957925,
-      "grad_norm": 0.42056091819107644,
-      "learning_rate": 0.0001703396161741732,
-      "loss": 0.7052,
-      "step": 2284
-    },
-    {
-      "epoch": 1.1217321769084088,
-      "grad_norm": 0.5053689475536761,
-      "learning_rate": 0.00017029583494967935,
-      "loss": 0.8291,
-      "step": 2285
-    },
-    {
-      "epoch": 1.1222242726210248,
-      "grad_norm": 0.4662819700684409,
-      "learning_rate": 0.00017025202707204325,
-      "loss": 0.715,
-      "step": 2286
-    },
-    {
-      "epoch": 1.122716368333641,
-      "grad_norm": 0.44534392370815595,
-      "learning_rate": 0.000170208192557875,
-      "loss": 0.7347,
-      "step": 2287
-    },
-    {
-      "epoch": 1.123208464046257,
-      "grad_norm": 0.6510306626472884,
-      "learning_rate": 0.00017016433142379473,
-      "loss": 0.8045,
-      "step": 2288
-    },
-    {
-      "epoch": 1.1237005597588732,
-      "grad_norm": 0.4419822695979343,
-      "learning_rate": 0.00017012044368643268,
-      "loss": 0.714,
-      "step": 2289
-    },
-    {
-      "epoch": 1.1241926554714892,
-      "grad_norm": 0.5317833744497776,
-      "learning_rate": 0.0001700765293624292,
-      "loss": 0.8055,
-      "step": 2290
-    },
-    {
-      "epoch": 1.1246847511841054,
-      "grad_norm": 0.4491957822634459,
-      "learning_rate": 0.00017003258846843465,
-      "loss": 0.7083,
-      "step": 2291
-    },
-    {
-      "epoch": 1.1251768468967214,
-      "grad_norm": 0.4861262392437043,
-      "learning_rate": 0.0001699886210211096,
-      "loss": 0.8051,
-      "step": 2292
-    },
-    {
-      "epoch": 1.1256689426093376,
-      "grad_norm": 0.4586144821407561,
-      "learning_rate": 0.00016994462703712456,
-      "loss": 0.7624,
-      "step": 2293
-    },
-    {
-      "epoch": 1.1261610383219536,
-      "grad_norm": 0.44717349121927574,
-      "learning_rate": 0.00016990060653316013,
-      "loss": 0.7486,
-      "step": 2294
-    },
-    {
-      "epoch": 1.1266531340345698,
-      "grad_norm": 0.47667487702788763,
-      "learning_rate": 0.00016985655952590702,
-      "loss": 0.7568,
-      "step": 2295
-    },
-    {
-      "epoch": 1.1271452297471858,
-      "grad_norm": 0.5202189184782849,
-      "learning_rate": 0.00016981248603206592,
-      "loss": 0.877,
-      "step": 2296
-    },
-    {
-      "epoch": 1.127637325459802,
-      "grad_norm": 0.45493093827344977,
-      "learning_rate": 0.00016976838606834764,
-      "loss": 0.7834,
-      "step": 2297
-    },
-    {
-      "epoch": 1.128129421172418,
-      "grad_norm": 0.48036913116061,
-      "learning_rate": 0.00016972425965147293,
-      "loss": 0.7749,
-      "step": 2298
-    },
-    {
-      "epoch": 1.1286215168850342,
-      "grad_norm": 0.4359946394576743,
-      "learning_rate": 0.00016968010679817264,
-      "loss": 0.7249,
-      "step": 2299
-    },
-    {
-      "epoch": 1.1291136125976502,
-      "grad_norm": 0.4377335882015854,
-      "learning_rate": 0.00016963592752518763,
-      "loss": 0.7379,
-      "step": 2300
-    },
-    {
-      "epoch": 1.1296057083102664,
-      "grad_norm": 0.43918112206406407,
-      "learning_rate": 0.00016959172184926875,
-      "loss": 0.7775,
-      "step": 2301
-    },
-    {
-      "epoch": 1.1300978040228824,
-      "grad_norm": 0.46085230534250393,
-      "learning_rate": 0.0001695474897871769,
-      "loss": 0.7272,
-      "step": 2302
-    },
-    {
-      "epoch": 1.1305898997354986,
-      "grad_norm": 0.47015545707857204,
-      "learning_rate": 0.00016950323135568298,
-      "loss": 0.7234,
-      "step": 2303
-    },
-    {
-      "epoch": 1.1310819954481146,
-      "grad_norm": 0.4944887088290653,
-      "learning_rate": 0.00016945894657156784,
-      "loss": 0.759,
-      "step": 2304
-    },
-    {
-      "epoch": 1.1315740911607308,
-      "grad_norm": 0.456194455456828,
-      "learning_rate": 0.00016941463545162234,
-      "loss": 0.7579,
-      "step": 2305
-    },
-    {
-      "epoch": 1.1320661868733468,
-      "grad_norm": 0.5019289572944563,
-      "learning_rate": 0.00016937029801264742,
-      "loss": 0.7301,
-      "step": 2306
-    },
-    {
-      "epoch": 1.132558282585963,
-      "grad_norm": 0.44457249428378054,
-      "learning_rate": 0.00016932593427145386,
-      "loss": 0.7618,
-      "step": 2307
-    },
-    {
-      "epoch": 1.133050378298579,
-      "grad_norm": 0.4319289977770867,
-      "learning_rate": 0.0001692815442448625,
-      "loss": 0.7327,
-      "step": 2308
-    },
-    {
-      "epoch": 1.1335424740111952,
-      "grad_norm": 0.4883241687960278,
-      "learning_rate": 0.0001692371279497041,
-      "loss": 0.7523,
-      "step": 2309
-    },
-    {
-      "epoch": 1.1340345697238112,
-      "grad_norm": 0.45780220449658393,
-      "learning_rate": 0.00016919268540281936,
-      "loss": 0.6821,
-      "step": 2310
-    },
-    {
-      "epoch": 1.1345266654364274,
-      "grad_norm": 0.4753729627358104,
-      "learning_rate": 0.00016914821662105908,
-      "loss": 0.8173,
-      "step": 2311
-    },
-    {
-      "epoch": 1.1350187611490434,
-      "grad_norm": 0.44447953352027786,
-      "learning_rate": 0.00016910372162128382,
-      "loss": 0.7236,
-      "step": 2312
-    },
-    {
-      "epoch": 1.1355108568616596,
-      "grad_norm": 0.4243710961328533,
-      "learning_rate": 0.00016905920042036417,
-      "loss": 0.7329,
-      "step": 2313
-    },
-    {
-      "epoch": 1.1360029525742756,
-      "grad_norm": 0.4468629999914307,
-      "learning_rate": 0.00016901465303518064,
-      "loss": 0.7685,
-      "step": 2314
-    },
-    {
-      "epoch": 1.1364950482868919,
-      "grad_norm": 0.45134253816361103,
-      "learning_rate": 0.00016897007948262372,
-      "loss": 0.7661,
-      "step": 2315
-    },
-    {
-      "epoch": 1.1369871439995078,
-      "grad_norm": 0.4478372330374271,
-      "learning_rate": 0.0001689254797795937,
-      "loss": 0.8289,
-      "step": 2316
-    },
-    {
-      "epoch": 1.137479239712124,
-      "grad_norm": 0.45223770169916977,
-      "learning_rate": 0.00016888085394300094,
-      "loss": 0.7879,
-      "step": 2317
-    },
-    {
-      "epoch": 1.13797133542474,
-      "grad_norm": 0.4183224861655834,
-      "learning_rate": 0.00016883620198976558,
-      "loss": 0.6802,
-      "step": 2318
-    },
-    {
-      "epoch": 1.1384634311373563,
-      "grad_norm": 0.44475791786138535,
-      "learning_rate": 0.0001687915239368177,
-      "loss": 0.7196,
-      "step": 2319
-    },
-    {
-      "epoch": 1.1389555268499723,
-      "grad_norm": 0.4389263002912342,
-      "learning_rate": 0.00016874681980109734,
-      "loss": 0.7083,
-      "step": 2320
-    },
-    {
-      "epoch": 1.1394476225625885,
-      "grad_norm": 0.4313633988441347,
-      "learning_rate": 0.00016870208959955435,
-      "loss": 0.8626,
-      "step": 2321
-    },
-    {
-      "epoch": 1.1399397182752045,
-      "grad_norm": 0.44536307722028046,
-      "learning_rate": 0.00016865733334914845,
-      "loss": 0.7838,
-      "step": 2322
-    },
-    {
-      "epoch": 1.1404318139878207,
-      "grad_norm": 0.4308906082632619,
-      "learning_rate": 0.00016861255106684933,
-      "loss": 0.7292,
-      "step": 2323
-    },
-    {
-      "epoch": 1.1409239097004367,
-      "grad_norm": 0.4715337438609384,
-      "learning_rate": 0.00016856774276963646,
-      "loss": 0.7318,
-      "step": 2324
-    },
-    {
-      "epoch": 1.1414160054130529,
-      "grad_norm": 0.5015199873249059,
-      "learning_rate": 0.0001685229084744992,
-      "loss": 0.7696,
-      "step": 2325
-    },
-    {
-      "epoch": 1.1419081011256689,
-      "grad_norm": 0.44433966071888925,
-      "learning_rate": 0.00016847804819843684,
-      "loss": 0.66,
-      "step": 2326
-    },
-    {
-      "epoch": 1.142400196838285,
-      "grad_norm": 0.46345689407796764,
-      "learning_rate": 0.00016843316195845842,
-      "loss": 0.7575,
-      "step": 2327
-    },
-    {
-      "epoch": 1.142892292550901,
-      "grad_norm": 0.43189372122649544,
-      "learning_rate": 0.00016838824977158284,
-      "loss": 0.7627,
-      "step": 2328
-    },
-    {
-      "epoch": 1.1433843882635173,
-      "grad_norm": 0.47553138873250456,
-      "learning_rate": 0.00016834331165483887,
-      "loss": 0.7917,
-      "step": 2329
-    },
-    {
-      "epoch": 1.1438764839761333,
-      "grad_norm": 0.43652638702069657,
-      "learning_rate": 0.00016829834762526513,
-      "loss": 0.7457,
-      "step": 2330
-    },
-    {
-      "epoch": 1.1443685796887495,
-      "grad_norm": 0.43384112294714045,
-      "learning_rate": 0.00016825335769991002,
-      "loss": 0.7283,
-      "step": 2331
-    },
-    {
-      "epoch": 1.1448606754013655,
-      "grad_norm": 0.46679399150765144,
-      "learning_rate": 0.00016820834189583175,
-      "loss": 0.7201,
-      "step": 2332
-    },
-    {
-      "epoch": 1.1453527711139817,
-      "grad_norm": 0.4471805264967156,
-      "learning_rate": 0.00016816330023009841,
-      "loss": 0.7751,
-      "step": 2333
-    },
-    {
-      "epoch": 1.1458448668265977,
-      "grad_norm": 0.5097809916933203,
-      "learning_rate": 0.00016811823271978784,
-      "loss": 0.7356,
-      "step": 2334
-    },
-    {
-      "epoch": 1.146336962539214,
-      "grad_norm": 0.45399847525617715,
-      "learning_rate": 0.00016807313938198767,
-      "loss": 0.7642,
-      "step": 2335
-    },
-    {
-      "epoch": 1.14682905825183,
-      "grad_norm": 0.44216215398989483,
-      "learning_rate": 0.00016802802023379538,
-      "loss": 0.7776,
-      "step": 2336
-    },
-    {
-      "epoch": 1.1473211539644461,
-      "grad_norm": 0.47391952554629074,
-      "learning_rate": 0.00016798287529231815,
-      "loss": 0.7241,
-      "step": 2337
-    },
-    {
-      "epoch": 1.147813249677062,
-      "grad_norm": 0.45151083393427893,
-      "learning_rate": 0.00016793770457467302,
-      "loss": 0.7254,
-      "step": 2338
-    },
-    {
-      "epoch": 1.1483053453896783,
-      "grad_norm": 0.44787964084182247,
-      "learning_rate": 0.0001678925080979868,
-      "loss": 0.7138,
-      "step": 2339
-    },
-    {
-      "epoch": 1.1487974411022943,
-      "grad_norm": 0.5878711015458571,
-      "learning_rate": 0.00016784728587939602,
-      "loss": 0.761,
-      "step": 2340
-    },
-    {
-      "epoch": 1.1492895368149105,
-      "grad_norm": 0.5067852039768288,
-      "learning_rate": 0.00016780203793604695,
-      "loss": 0.8009,
-      "step": 2341
-    },
-    {
-      "epoch": 1.1497816325275265,
-      "grad_norm": 0.4948494649935818,
-      "learning_rate": 0.0001677567642850957,
-      "loss": 0.7566,
-      "step": 2342
-    },
-    {
-      "epoch": 1.1502737282401427,
-      "grad_norm": 0.4445892575995358,
-      "learning_rate": 0.00016771146494370812,
-      "loss": 0.7102,
-      "step": 2343
-    },
-    {
-      "epoch": 1.1507658239527587,
-      "grad_norm": 0.4415227933373556,
-      "learning_rate": 0.00016766613992905965,
-      "loss": 0.787,
-      "step": 2344
-    },
-    {
-      "epoch": 1.151257919665375,
-      "grad_norm": 0.4369592312691189,
-      "learning_rate": 0.0001676207892583357,
-      "loss": 0.7441,
-      "step": 2345
-    },
-    {
-      "epoch": 1.151750015377991,
-      "grad_norm": 0.45450469510797614,
-      "learning_rate": 0.00016757541294873117,
-      "loss": 0.7359,
-      "step": 2346
-    },
-    {
-      "epoch": 1.1522421110906071,
-      "grad_norm": 0.4612742978346103,
-      "learning_rate": 0.00016753001101745088,
-      "loss": 0.743,
-      "step": 2347
-    },
-    {
-      "epoch": 1.1527342068032231,
-      "grad_norm": 0.4612558288498186,
-      "learning_rate": 0.00016748458348170924,
-      "loss": 0.7068,
-      "step": 2348
-    },
-    {
-      "epoch": 1.1532263025158394,
-      "grad_norm": 0.44262806429187956,
-      "learning_rate": 0.00016743913035873042,
-      "loss": 0.7073,
-      "step": 2349
-    },
-    {
-      "epoch": 1.1537183982284553,
-      "grad_norm": 0.49501560308668063,
-      "learning_rate": 0.00016739365166574827,
-      "loss": 0.7744,
-      "step": 2350
-    },
-    {
-      "epoch": 1.1542104939410716,
-      "grad_norm": 0.48938371898703736,
-      "learning_rate": 0.00016734814742000635,
-      "loss": 0.8085,
-      "step": 2351
-    },
-    {
-      "epoch": 1.1547025896536875,
-      "grad_norm": 0.44161795904827694,
-      "learning_rate": 0.0001673026176387579,
-      "loss": 0.7402,
-      "step": 2352
-    },
-    {
-      "epoch": 1.1551946853663038,
-      "grad_norm": 0.47192240761441423,
-      "learning_rate": 0.00016725706233926589,
-      "loss": 0.779,
-      "step": 2353
-    },
-    {
-      "epoch": 1.1556867810789198,
-      "grad_norm": 0.4462355788643901,
-      "learning_rate": 0.00016721148153880285,
-      "loss": 0.7466,
-      "step": 2354
-    },
-    {
-      "epoch": 1.156178876791536,
-      "grad_norm": 0.49915936011147427,
-      "learning_rate": 0.00016716587525465108,
-      "loss": 0.8115,
-      "step": 2355
-    },
-    {
-      "epoch": 1.156670972504152,
-      "grad_norm": 0.43363277540110245,
-      "learning_rate": 0.00016712024350410253,
-      "loss": 0.6893,
-      "step": 2356
-    },
-    {
-      "epoch": 1.1571630682167682,
-      "grad_norm": 0.48002361737554866,
-      "learning_rate": 0.00016707458630445875,
-      "loss": 0.7675,
-      "step": 2357
-    },
-    {
-      "epoch": 1.1576551639293844,
-      "grad_norm": 0.44353609318569737,
-      "learning_rate": 0.00016702890367303102,
-      "loss": 0.6764,
-      "step": 2358
-    },
-    {
-      "epoch": 1.1581472596420004,
-      "grad_norm": 0.41941593553602335,
-      "learning_rate": 0.0001669831956271402,
-      "loss": 0.692,
-      "step": 2359
-    },
-    {
-      "epoch": 1.1586393553546164,
-      "grad_norm": 0.42467355773936255,
-      "learning_rate": 0.00016693746218411677,
-      "loss": 0.7411,
-      "step": 2360
-    },
-    {
-      "epoch": 1.1591314510672326,
-      "grad_norm": 0.4282108620441184,
-      "learning_rate": 0.00016689170336130088,
-      "loss": 0.7182,
-      "step": 2361
-    },
-    {
-      "epoch": 1.1596235467798488,
-      "grad_norm": 0.3833084268608547,
-      "learning_rate": 0.0001668459191760424,
-      "loss": 0.643,
-      "step": 2362
-    },
-    {
-      "epoch": 1.1601156424924648,
-      "grad_norm": 0.4286783736712506,
-      "learning_rate": 0.00016680010964570058,
-      "loss": 0.8028,
-      "step": 2363
-    },
-    {
-      "epoch": 1.1606077382050808,
-      "grad_norm": 0.41664458769042473,
-      "learning_rate": 0.00016675427478764448,
-      "loss": 0.7111,
-      "step": 2364
-    },
-    {
-      "epoch": 1.161099833917697,
-      "grad_norm": 0.4283241918209358,
-      "learning_rate": 0.0001667084146192527,
-      "loss": 0.7747,
-      "step": 2365
-    },
-    {
-      "epoch": 1.1615919296303132,
-      "grad_norm": 0.45885220296525653,
-      "learning_rate": 0.00016666252915791346,
-      "loss": 0.6824,
-      "step": 2366
-    },
-    {
-      "epoch": 1.1620840253429292,
-      "grad_norm": 0.46532975154226963,
-      "learning_rate": 0.00016661661842102445,
-      "loss": 0.7824,
-      "step": 2367
-    },
-    {
-      "epoch": 1.1625761210555452,
-      "grad_norm": 0.575659023474276,
-      "learning_rate": 0.00016657068242599313,
-      "loss": 0.7705,
-      "step": 2368
-    },
-    {
-      "epoch": 1.1630682167681614,
-      "grad_norm": 0.44966093797615586,
-      "learning_rate": 0.00016652472119023636,
-      "loss": 0.7729,
-      "step": 2369
-    },
-    {
-      "epoch": 1.1635603124807776,
-      "grad_norm": 0.4393724975779391,
-      "learning_rate": 0.00016647873473118075,
-      "loss": 0.7316,
-      "step": 2370
-    },
-    {
-      "epoch": 1.1640524081933936,
-      "grad_norm": 0.4325600374647357,
-      "learning_rate": 0.0001664327230662623,
-      "loss": 0.7517,
-      "step": 2371
-    },
-    {
-      "epoch": 1.1645445039060096,
-      "grad_norm": 0.4338406490012966,
-      "learning_rate": 0.00016638668621292668,
-      "loss": 0.7504,
-      "step": 2372
-    },
-    {
-      "epoch": 1.1650365996186258,
-      "grad_norm": 0.43878306722290084,
-      "learning_rate": 0.00016634062418862907,
-      "loss": 0.6998,
-      "step": 2373
-    },
-    {
-      "epoch": 1.165528695331242,
-      "grad_norm": 0.46726108298167734,
-      "learning_rate": 0.0001662945370108342,
-      "loss": 0.7676,
-      "step": 2374
-    },
-    {
-      "epoch": 1.166020791043858,
-      "grad_norm": 0.4082613099682847,
-      "learning_rate": 0.00016624842469701632,
-      "loss": 0.7274,
-      "step": 2375
-    },
-    {
-      "epoch": 1.166512886756474,
-      "grad_norm": 0.4343343585815525,
-      "learning_rate": 0.00016620228726465922,
-      "loss": 0.7877,
-      "step": 2376
-    },
-    {
-      "epoch": 1.1670049824690902,
-      "grad_norm": 0.4424821203300492,
-      "learning_rate": 0.0001661561247312563,
-      "loss": 0.8563,
-      "step": 2377
-    },
-    {
-      "epoch": 1.1674970781817064,
-      "grad_norm": 0.42277985017173697,
-      "learning_rate": 0.00016610993711431028,
-      "loss": 0.7049,
-      "step": 2378
-    },
-    {
-      "epoch": 1.1679891738943224,
-      "grad_norm": 0.4095320909773662,
-      "learning_rate": 0.0001660637244313336,
-      "loss": 0.7574,
-      "step": 2379
-    },
-    {
-      "epoch": 1.1684812696069384,
-      "grad_norm": 0.44118747830725574,
-      "learning_rate": 0.00016601748669984806,
-      "loss": 0.7645,
-      "step": 2380
-    },
-    {
-      "epoch": 1.1689733653195546,
-      "grad_norm": 0.4247096205640213,
-      "learning_rate": 0.00016597122393738505,
-      "loss": 0.7665,
-      "step": 2381
-    },
-    {
-      "epoch": 1.1694654610321709,
-      "grad_norm": 0.4414360555795094,
-      "learning_rate": 0.00016592493616148535,
-      "loss": 0.7603,
-      "step": 2382
-    },
-    {
-      "epoch": 1.1699575567447869,
-      "grad_norm": 0.42371789178715363,
-      "learning_rate": 0.00016587862338969934,
-      "loss": 0.7234,
-      "step": 2383
-    },
-    {
-      "epoch": 1.1704496524574028,
-      "grad_norm": 0.4814483832167124,
-      "learning_rate": 0.00016583228563958678,
-      "loss": 0.7139,
-      "step": 2384
-    },
-    {
-      "epoch": 1.170941748170019,
-      "grad_norm": 0.44822704115743667,
-      "learning_rate": 0.00016578592292871698,
-      "loss": 0.7532,
-      "step": 2385
-    },
-    {
-      "epoch": 1.1714338438826353,
-      "grad_norm": 0.4535158060643928,
-      "learning_rate": 0.00016573953527466864,
-      "loss": 0.7691,
-      "step": 2386
-    },
-    {
-      "epoch": 1.1719259395952513,
-      "grad_norm": 0.40666153379654235,
-      "learning_rate": 0.00016569312269503,
-      "loss": 0.7385,
-      "step": 2387
-    },
-    {
-      "epoch": 1.1724180353078673,
-      "grad_norm": 0.39869523953436276,
-      "learning_rate": 0.00016564668520739867,
-      "loss": 0.715,
-      "step": 2388
-    },
-    {
-      "epoch": 1.1729101310204835,
-      "grad_norm": 0.4576211301504304,
-      "learning_rate": 0.00016560022282938172,
-      "loss": 0.799,
-      "step": 2389
-    },
-    {
-      "epoch": 1.1734022267330997,
-      "grad_norm": 0.42050802997000103,
-      "learning_rate": 0.00016555373557859573,
-      "loss": 0.7401,
-      "step": 2390
-    },
-    {
-      "epoch": 1.1738943224457157,
-      "grad_norm": 0.4058734036272197,
-      "learning_rate": 0.00016550722347266663,
-      "loss": 0.7186,
-      "step": 2391
-    },
-    {
-      "epoch": 1.174386418158332,
-      "grad_norm": 0.4460714696577042,
-      "learning_rate": 0.00016546068652922976,
-      "loss": 0.7377,
-      "step": 2392
-    },
-    {
-      "epoch": 1.1748785138709479,
-      "grad_norm": 0.4610839703524595,
-      "learning_rate": 0.00016541412476593,
-      "loss": 0.7252,
-      "step": 2393
-    },
-    {
-      "epoch": 1.175370609583564,
-      "grad_norm": 0.43171886607653037,
-      "learning_rate": 0.0001653675382004215,
-      "loss": 0.7036,
-      "step": 2394
-    },
-    {
-      "epoch": 1.17586270529618,
-      "grad_norm": 0.5673192512943801,
-      "learning_rate": 0.00016532092685036785,
-      "loss": 0.7262,
-      "step": 2395
-    },
-    {
-      "epoch": 1.1763548010087963,
-      "grad_norm": 0.45432051430363635,
-      "learning_rate": 0.0001652742907334421,
-      "loss": 0.7917,
-      "step": 2396
-    },
-    {
-      "epoch": 1.1768468967214123,
-      "grad_norm": 0.43763563327576555,
-      "learning_rate": 0.00016522762986732664,
-      "loss": 0.7891,
-      "step": 2397
-    },
-    {
-      "epoch": 1.1773389924340285,
-      "grad_norm": 0.4368731350206338,
-      "learning_rate": 0.0001651809442697133,
-      "loss": 0.7782,
-      "step": 2398
-    },
-    {
-      "epoch": 1.1778310881466445,
-      "grad_norm": 0.44500281421300364,
-      "learning_rate": 0.00016513423395830316,
-      "loss": 0.7669,
-      "step": 2399
-    },
-    {
-      "epoch": 1.1783231838592607,
-      "grad_norm": 0.41061630531196897,
-      "learning_rate": 0.0001650874989508068,
-      "loss": 0.7342,
-      "step": 2400
-    },
-    {
-      "epoch": 1.1788152795718767,
-      "grad_norm": 0.44385162575035897,
-      "learning_rate": 0.0001650407392649441,
-      "loss": 0.7506,
-      "step": 2401
-    },
-    {
-      "epoch": 1.179307375284493,
-      "grad_norm": 0.419968995168489,
-      "learning_rate": 0.0001649939549184443,
-      "loss": 0.7116,
-      "step": 2402
-    },
-    {
-      "epoch": 1.179799470997109,
-      "grad_norm": 0.45280405961816544,
-      "learning_rate": 0.00016494714592904606,
-      "loss": 0.7031,
-      "step": 2403
-    },
-    {
-      "epoch": 1.1802915667097251,
-      "grad_norm": 0.44911412973344544,
-      "learning_rate": 0.00016490031231449726,
-      "loss": 0.7743,
-      "step": 2404
-    },
-    {
-      "epoch": 1.1807836624223411,
-      "grad_norm": 0.43994851601899965,
-      "learning_rate": 0.0001648534540925552,
-      "loss": 0.7139,
-      "step": 2405
-    },
-    {
-      "epoch": 1.1812757581349573,
-      "grad_norm": 0.42151538382286186,
-      "learning_rate": 0.0001648065712809865,
-      "loss": 0.7426,
-      "step": 2406
-    },
-    {
-      "epoch": 1.1817678538475733,
-      "grad_norm": 0.421772510581625,
-      "learning_rate": 0.0001647596638975671,
-      "loss": 0.7238,
-      "step": 2407
-    },
-    {
-      "epoch": 1.1822599495601895,
-      "grad_norm": 0.43466246290054783,
-      "learning_rate": 0.00016471273196008224,
-      "loss": 0.7355,
-      "step": 2408
-    },
-    {
-      "epoch": 1.1827520452728055,
-      "grad_norm": 0.429120524397614,
-      "learning_rate": 0.00016466577548632648,
-      "loss": 0.7545,
-      "step": 2409
-    },
-    {
-      "epoch": 1.1832441409854217,
-      "grad_norm": 0.436247531580134,
-      "learning_rate": 0.0001646187944941037,
-      "loss": 0.7201,
-      "step": 2410
-    },
-    {
-      "epoch": 1.1837362366980377,
-      "grad_norm": 0.49303383774654047,
-      "learning_rate": 0.00016457178900122706,
-      "loss": 0.7482,
-      "step": 2411
-    },
-    {
-      "epoch": 1.184228332410654,
-      "grad_norm": 0.44407875153581605,
-      "learning_rate": 0.00016452475902551902,
-      "loss": 0.749,
-      "step": 2412
-    },
-    {
-      "epoch": 1.18472042812327,
-      "grad_norm": 0.3942366518830162,
-      "learning_rate": 0.00016447770458481124,
-      "loss": 0.6715,
-      "step": 2413
-    },
-    {
-      "epoch": 1.1852125238358862,
-      "grad_norm": 0.46875812009328377,
-      "learning_rate": 0.00016443062569694483,
-      "loss": 0.7764,
-      "step": 2414
-    },
-    {
-      "epoch": 1.1857046195485021,
-      "grad_norm": 0.4521465433035599,
-      "learning_rate": 0.00016438352237977007,
-      "loss": 0.736,
-      "step": 2415
-    },
-    {
-      "epoch": 1.1861967152611184,
-      "grad_norm": 0.431950611879082,
-      "learning_rate": 0.00016433639465114638,
-      "loss": 0.756,
-      "step": 2416
-    },
-    {
-      "epoch": 1.1866888109737344,
-      "grad_norm": 0.4258989569384757,
-      "learning_rate": 0.0001642892425289427,
-      "loss": 0.6816,
-      "step": 2417
-    },
-    {
-      "epoch": 1.1871809066863506,
-      "grad_norm": 0.4301938931562309,
-      "learning_rate": 0.000164242066031037,
-      "loss": 0.7712,
-      "step": 2418
-    },
-    {
-      "epoch": 1.1876730023989666,
-      "grad_norm": 0.433492593276431,
-      "learning_rate": 0.00016419486517531658,
-      "loss": 0.708,
-      "step": 2419
-    },
-    {
-      "epoch": 1.1881650981115828,
-      "grad_norm": 0.46069856340259263,
-      "learning_rate": 0.00016414763997967793,
-      "loss": 0.759,
-      "step": 2420
-    },
-    {
-      "epoch": 1.1886571938241988,
-      "grad_norm": 0.4239294138071813,
-      "learning_rate": 0.0001641003904620269,
-      "loss": 0.801,
-      "step": 2421
-    },
-    {
-      "epoch": 1.189149289536815,
-      "grad_norm": 0.4921042860737761,
-      "learning_rate": 0.00016405311664027838,
-      "loss": 0.8156,
-      "step": 2422
-    },
-    {
-      "epoch": 1.189641385249431,
-      "grad_norm": 0.4196707896059516,
-      "learning_rate": 0.0001640058185323566,
-      "loss": 0.6957,
-      "step": 2423
-    },
-    {
-      "epoch": 1.1901334809620472,
-      "grad_norm": 0.4578294751734398,
-      "learning_rate": 0.00016395849615619495,
-      "loss": 0.7897,
-      "step": 2424
-    },
-    {
-      "epoch": 1.1906255766746632,
-      "grad_norm": 0.4964683375347343,
-      "learning_rate": 0.00016391114952973602,
-      "loss": 0.7925,
-      "step": 2425
-    },
-    {
-      "epoch": 1.1911176723872794,
-      "grad_norm": 0.4711647247164641,
-      "learning_rate": 0.00016386377867093157,
-      "loss": 0.7544,
-      "step": 2426
-    },
-    {
-      "epoch": 1.1916097680998954,
-      "grad_norm": 0.433824246721515,
-      "learning_rate": 0.0001638163835977427,
-      "loss": 0.7688,
-      "step": 2427
-    },
-    {
-      "epoch": 1.1921018638125116,
-      "grad_norm": 0.4716366493461372,
-      "learning_rate": 0.00016376896432813943,
-      "loss": 0.7886,
-      "step": 2428
-    },
-    {
-      "epoch": 1.1925939595251276,
-      "grad_norm": 0.5575845841328761,
-      "learning_rate": 0.00016372152088010118,
-      "loss": 0.753,
-      "step": 2429
-    },
-    {
-      "epoch": 1.1930860552377438,
-      "grad_norm": 0.4323134443913653,
-      "learning_rate": 0.0001636740532716164,
-      "loss": 0.7244,
-      "step": 2430
-    },
-    {
-      "epoch": 1.1935781509503598,
-      "grad_norm": 0.43494712206575137,
-      "learning_rate": 0.00016362656152068287,
-      "loss": 0.7102,
-      "step": 2431
-    },
-    {
-      "epoch": 1.194070246662976,
-      "grad_norm": 0.4310915793828525,
-      "learning_rate": 0.00016357904564530726,
-      "loss": 0.7172,
-      "step": 2432
-    },
-    {
-      "epoch": 1.194562342375592,
-      "grad_norm": 0.49717458394673475,
-      "learning_rate": 0.00016353150566350563,
-      "loss": 0.7615,
-      "step": 2433
-    },
-    {
-      "epoch": 1.1950544380882082,
-      "grad_norm": 0.4609230308039873,
-      "learning_rate": 0.00016348394159330308,
-      "loss": 0.7557,
-      "step": 2434
-    },
-    {
-      "epoch": 1.1955465338008242,
-      "grad_norm": 0.46317593356743825,
-      "learning_rate": 0.0001634363534527338,
-      "loss": 0.7888,
-      "step": 2435
-    },
-    {
-      "epoch": 1.1960386295134404,
-      "grad_norm": 0.43803936604089877,
-      "learning_rate": 0.0001633887412598412,
-      "loss": 0.7291,
-      "step": 2436
-    },
-    {
-      "epoch": 1.1965307252260564,
-      "grad_norm": 0.4583230627898153,
-      "learning_rate": 0.00016334110503267777,
-      "loss": 0.7471,
-      "step": 2437
-    },
-    {
-      "epoch": 1.1970228209386726,
-      "grad_norm": 0.4465355586005098,
-      "learning_rate": 0.0001632934447893051,
-      "loss": 0.6913,
-      "step": 2438
-    },
-    {
-      "epoch": 1.1975149166512886,
-      "grad_norm": 0.4284055764535819,
-      "learning_rate": 0.00016324576054779386,
-      "loss": 0.7395,
-      "step": 2439
-    },
-    {
-      "epoch": 1.1980070123639048,
-      "grad_norm": 0.4336490156541325,
-      "learning_rate": 0.0001631980523262239,
-      "loss": 0.7994,
-      "step": 2440
-    },
-    {
-      "epoch": 1.1984991080765208,
-      "grad_norm": 0.4362318709859781,
-      "learning_rate": 0.0001631503201426841,
-      "loss": 0.675,
-      "step": 2441
-    },
-    {
-      "epoch": 1.198991203789137,
-      "grad_norm": 0.4972034285587136,
-      "learning_rate": 0.00016310256401527243,
-      "loss": 0.7645,
-      "step": 2442
-    },
-    {
-      "epoch": 1.199483299501753,
-      "grad_norm": 0.43066554321273104,
-      "learning_rate": 0.00016305478396209594,
-      "loss": 0.7103,
-      "step": 2443
-    },
-    {
-      "epoch": 1.1999753952143692,
-      "grad_norm": 0.4434164205535384,
-      "learning_rate": 0.0001630069800012708,
-      "loss": 0.7312,
-      "step": 2444
-    },
-    {
-      "epoch": 1.2004674909269852,
-      "grad_norm": 0.4419446094796589,
-      "learning_rate": 0.00016295915215092216,
-      "loss": 0.7484,
-      "step": 2445
-    },
-    {
-      "epoch": 1.2009595866396015,
-      "grad_norm": 0.436222360431337,
-      "learning_rate": 0.0001629113004291843,
-      "loss": 0.725,
-      "step": 2446
-    },
-    {
-      "epoch": 1.2014516823522174,
-      "grad_norm": 0.4381375698311324,
-      "learning_rate": 0.00016286342485420056,
-      "loss": 0.7175,
-      "step": 2447
-    },
-    {
-      "epoch": 1.2019437780648337,
-      "grad_norm": 0.4219271108947798,
-      "learning_rate": 0.0001628155254441232,
-      "loss": 0.7172,
-      "step": 2448
-    },
-    {
-      "epoch": 1.2024358737774496,
-      "grad_norm": 0.48131877200504575,
-      "learning_rate": 0.00016276760221711368,
-      "loss": 0.7589,
-      "step": 2449
-    },
-    {
-      "epoch": 1.2029279694900659,
-      "grad_norm": 0.4416067404334387,
-      "learning_rate": 0.0001627196551913424,
-      "loss": 0.7681,
-      "step": 2450
-    },
-    {
-      "epoch": 1.2034200652026819,
-      "grad_norm": 0.451485208062328,
-      "learning_rate": 0.00016267168438498879,
-      "loss": 0.7628,
-      "step": 2451
-    },
-    {
-      "epoch": 1.203912160915298,
-      "grad_norm": 0.4399384763222408,
-      "learning_rate": 0.0001626236898162413,
-      "loss": 0.7273,
-      "step": 2452
-    },
-    {
-      "epoch": 1.204404256627914,
-      "grad_norm": 0.4522320174695235,
-      "learning_rate": 0.0001625756715032974,
-      "loss": 0.7437,
-      "step": 2453
-    },
-    {
-      "epoch": 1.2048963523405303,
-      "grad_norm": 0.4509852633870999,
-      "learning_rate": 0.00016252762946436357,
-      "loss": 0.7971,
-      "step": 2454
-    },
-    {
-      "epoch": 1.2053884480531463,
-      "grad_norm": 0.4467131917702599,
-      "learning_rate": 0.00016247956371765528,
-      "loss": 0.7466,
-      "step": 2455
-    },
-    {
-      "epoch": 1.2058805437657625,
-      "grad_norm": 0.45212991474422043,
-      "learning_rate": 0.00016243147428139694,
-      "loss": 0.7864,
-      "step": 2456
-    },
-    {
-      "epoch": 1.2063726394783785,
-      "grad_norm": 0.4753890663625771,
-      "learning_rate": 0.00016238336117382204,
-      "loss": 0.8267,
-      "step": 2457
-    },
-    {
-      "epoch": 1.2068647351909947,
-      "grad_norm": 0.4779672497083118,
-      "learning_rate": 0.00016233522441317296,
-      "loss": 0.7139,
-      "step": 2458
-    },
-    {
-      "epoch": 1.2073568309036107,
-      "grad_norm": 0.4569575903984626,
-      "learning_rate": 0.00016228706401770108,
-      "loss": 0.7367,
-      "step": 2459
-    },
-    {
-      "epoch": 1.207848926616227,
-      "grad_norm": 0.45729088174785243,
-      "learning_rate": 0.00016223888000566677,
-      "loss": 0.7817,
-      "step": 2460
-    },
-    {
-      "epoch": 1.2083410223288429,
-      "grad_norm": 1.243983620542089,
-      "learning_rate": 0.0001621906723953393,
-      "loss": 0.8791,
-      "step": 2461
-    },
-    {
-      "epoch": 1.208833118041459,
-      "grad_norm": 0.4655000992948183,
-      "learning_rate": 0.0001621424412049969,
-      "loss": 0.7644,
-      "step": 2462
-    },
-    {
-      "epoch": 1.209325213754075,
-      "grad_norm": 0.45367678123478067,
-      "learning_rate": 0.00016209418645292674,
-      "loss": 0.7059,
-      "step": 2463
-    },
-    {
-      "epoch": 1.2098173094666913,
-      "grad_norm": 0.4402459317799849,
-      "learning_rate": 0.00016204590815742503,
-      "loss": 0.7171,
-      "step": 2464
-    },
-    {
-      "epoch": 1.2103094051793073,
-      "grad_norm": 0.41688045725088835,
-      "learning_rate": 0.00016199760633679668,
-      "loss": 0.7157,
-      "step": 2465
-    },
-    {
-      "epoch": 1.2108015008919235,
-      "grad_norm": 0.41866447525231854,
-      "learning_rate": 0.00016194928100935575,
-      "loss": 0.6409,
-      "step": 2466
-    },
-    {
-      "epoch": 1.2112935966045395,
-      "grad_norm": 0.4547058457916331,
-      "learning_rate": 0.0001619009321934251,
-      "loss": 0.7391,
-      "step": 2467
-    },
-    {
-      "epoch": 1.2117856923171557,
-      "grad_norm": 0.465123914545746,
-      "learning_rate": 0.00016185255990733649,
-      "loss": 0.7336,
-      "step": 2468
-    },
-    {
-      "epoch": 1.2122777880297717,
-      "grad_norm": 0.44621528747575384,
-      "learning_rate": 0.00016180416416943056,
-      "loss": 0.7793,
-      "step": 2469
-    },
-    {
-      "epoch": 1.212769883742388,
-      "grad_norm": 0.4547779940734016,
-      "learning_rate": 0.00016175574499805698,
-      "loss": 0.7979,
-      "step": 2470
-    },
-    {
-      "epoch": 1.213261979455004,
-      "grad_norm": 0.44229750375713783,
-      "learning_rate": 0.00016170730241157414,
-      "loss": 0.8386,
-      "step": 2471
-    },
-    {
-      "epoch": 1.2137540751676201,
-      "grad_norm": 0.41456072260825083,
-      "learning_rate": 0.00016165883642834937,
-      "loss": 0.7462,
-      "step": 2472
-    },
-    {
-      "epoch": 1.2142461708802361,
-      "grad_norm": 0.4664640247816799,
-      "learning_rate": 0.00016161034706675892,
-      "loss": 0.7187,
-      "step": 2473
-    },
-    {
-      "epoch": 1.2147382665928523,
-      "grad_norm": 0.438875039135234,
-      "learning_rate": 0.0001615618343451878,
-      "loss": 0.7352,
-      "step": 2474
-    },
-    {
-      "epoch": 1.2152303623054683,
-      "grad_norm": 0.43223943483733024,
-      "learning_rate": 0.00016151329828203,
-      "loss": 0.7417,
-      "step": 2475
-    },
-    {
-      "epoch": 1.2157224580180845,
-      "grad_norm": 0.5025039703358414,
-      "learning_rate": 0.00016146473889568828,
-      "loss": 0.7799,
-      "step": 2476
-    },
-    {
-      "epoch": 1.2162145537307005,
-      "grad_norm": 1.0461634593702156,
-      "learning_rate": 0.00016141615620457423,
-      "loss": 0.888,
-      "step": 2477
-    },
-    {
-      "epoch": 1.2167066494433167,
-      "grad_norm": 0.4425790122445545,
-      "learning_rate": 0.00016136755022710836,
-      "loss": 0.7175,
-      "step": 2478
-    },
-    {
-      "epoch": 1.2171987451559327,
-      "grad_norm": 0.40985986664650526,
-      "learning_rate": 0.00016131892098171987,
-      "loss": 0.7551,
-      "step": 2479
-    },
-    {
-      "epoch": 1.217690840868549,
-      "grad_norm": 0.43781332426632397,
-      "learning_rate": 0.00016127026848684696,
-      "loss": 0.7048,
-      "step": 2480
-    },
-    {
-      "epoch": 1.218182936581165,
-      "grad_norm": 0.4595583384205723,
-      "learning_rate": 0.0001612215927609365,
-      "loss": 0.7895,
-      "step": 2481
-    },
-    {
-      "epoch": 1.2186750322937812,
-      "grad_norm": 0.4446513017453665,
-      "learning_rate": 0.00016117289382244424,
-      "loss": 0.6593,
-      "step": 2482
-    },
-    {
-      "epoch": 1.2191671280063971,
-      "grad_norm": 0.4397360591458268,
-      "learning_rate": 0.00016112417168983472,
-      "loss": 0.7679,
-      "step": 2483
-    },
-    {
-      "epoch": 1.2196592237190134,
-      "grad_norm": 0.4188480806178211,
-      "learning_rate": 0.00016107542638158122,
-      "loss": 0.7232,
-      "step": 2484
-    },
-    {
-      "epoch": 1.2201513194316294,
-      "grad_norm": 0.5717598128986419,
-      "learning_rate": 0.0001610266579161659,
-      "loss": 0.7683,
-      "step": 2485
-    },
-    {
-      "epoch": 1.2206434151442456,
-      "grad_norm": 0.5937203025328015,
-      "learning_rate": 0.00016097786631207966,
-      "loss": 0.7768,
-      "step": 2486
-    },
-    {
-      "epoch": 1.2211355108568616,
-      "grad_norm": 0.43135713049256946,
-      "learning_rate": 0.00016092905158782215,
-      "loss": 0.7328,
-      "step": 2487
-    },
-    {
-      "epoch": 1.2216276065694778,
-      "grad_norm": 0.44961496046256094,
-      "learning_rate": 0.00016088021376190175,
-      "loss": 0.7784,
-      "step": 2488
-    },
-    {
-      "epoch": 1.222119702282094,
-      "grad_norm": 0.4459949098102091,
-      "learning_rate": 0.00016083135285283577,
-      "loss": 0.816,
-      "step": 2489
-    },
-    {
-      "epoch": 1.22261179799471,
-      "grad_norm": 0.44488660493938537,
-      "learning_rate": 0.00016078246887915007,
-      "loss": 0.7453,
-      "step": 2490
-    },
-    {
-      "epoch": 1.223103893707326,
-      "grad_norm": 0.45206552513800424,
-      "learning_rate": 0.00016073356185937934,
-      "loss": 0.7711,
-      "step": 2491
-    },
-    {
-      "epoch": 1.2235959894199422,
-      "grad_norm": 0.4449658085773436,
-      "learning_rate": 0.00016068463181206707,
-      "loss": 0.7194,
-      "step": 2492
-    },
-    {
-      "epoch": 1.2240880851325584,
-      "grad_norm": 0.43733253215838264,
-      "learning_rate": 0.00016063567875576536,
-      "loss": 0.7599,
-      "step": 2493
-    },
-    {
-      "epoch": 1.2245801808451744,
-      "grad_norm": 0.4470056028033283,
-      "learning_rate": 0.00016058670270903514,
-      "loss": 0.8031,
-      "step": 2494
-    },
-    {
-      "epoch": 1.2250722765577904,
-      "grad_norm": 0.41171495343876024,
-      "learning_rate": 0.00016053770369044593,
-      "loss": 0.7445,
-      "step": 2495
-    },
-    {
-      "epoch": 1.2255643722704066,
-      "grad_norm": 0.4786555526848234,
-      "learning_rate": 0.00016048868171857612,
-      "loss": 0.7908,
-      "step": 2496
-    },
-    {
-      "epoch": 1.2260564679830228,
-      "grad_norm": 0.4396705191785551,
-      "learning_rate": 0.0001604396368120127,
-      "loss": 0.7545,
-      "step": 2497
-    },
-    {
-      "epoch": 1.2265485636956388,
-      "grad_norm": 0.4525018260389032,
-      "learning_rate": 0.00016039056898935132,
-      "loss": 0.7737,
-      "step": 2498
-    },
-    {
-      "epoch": 1.2270406594082548,
-      "grad_norm": 0.5308962668568638,
-      "learning_rate": 0.00016034147826919648,
-      "loss": 0.855,
-      "step": 2499
-    },
-    {
-      "epoch": 1.227532755120871,
-      "grad_norm": 0.4557709197330377,
-      "learning_rate": 0.0001602923646701612,
-      "loss": 0.7744,
-      "step": 2500
-    },
-    {
-      "epoch": 1.2280248508334872,
-      "grad_norm": 0.4265830603196255,
-      "learning_rate": 0.00016024322821086724,
-      "loss": 0.7284,
-      "step": 2501
-    },
-    {
-      "epoch": 1.2285169465461032,
-      "grad_norm": 0.44895742693998336,
-      "learning_rate": 0.00016019406890994503,
-      "loss": 0.6931,
-      "step": 2502
-    },
-    {
-      "epoch": 1.2290090422587192,
-      "grad_norm": 0.47371664963505233,
-      "learning_rate": 0.00016014488678603366,
-      "loss": 0.8013,
-      "step": 2503
-    },
-    {
-      "epoch": 1.2295011379713354,
-      "grad_norm": 0.4581579926376679,
-      "learning_rate": 0.00016009568185778084,
-      "loss": 0.7289,
-      "step": 2504
-    },
-    {
-      "epoch": 1.2299932336839516,
-      "grad_norm": 0.4739863590086033,
-      "learning_rate": 0.00016004645414384293,
-      "loss": 0.8198,
-      "step": 2505
-    },
-    {
-      "epoch": 1.2304853293965676,
-      "grad_norm": 0.46902170356954354,
-      "learning_rate": 0.00015999720366288503,
-      "loss": 0.7792,
-      "step": 2506
-    },
-    {
-      "epoch": 1.2309774251091836,
-      "grad_norm": 0.4388140815686061,
-      "learning_rate": 0.00015994793043358073,
-      "loss": 0.7442,
-      "step": 2507
-    },
-    {
-      "epoch": 1.2314695208217998,
-      "grad_norm": 0.46388307849058275,
-      "learning_rate": 0.00015989863447461234,
-      "loss": 0.6854,
-      "step": 2508
-    },
-    {
-      "epoch": 1.231961616534416,
-      "grad_norm": 0.445047008840957,
-      "learning_rate": 0.00015984931580467074,
-      "loss": 0.7868,
-      "step": 2509
-    },
-    {
-      "epoch": 1.232453712247032,
-      "grad_norm": 0.4120760036892778,
-      "learning_rate": 0.00015979997444245543,
-      "loss": 0.7572,
-      "step": 2510
-    },
-    {
-      "epoch": 1.232945807959648,
-      "grad_norm": 0.4506870700950525,
-      "learning_rate": 0.00015975061040667454,
-      "loss": 0.8236,
-      "step": 2511
-    },
-    {
-      "epoch": 1.2334379036722642,
-      "grad_norm": 0.43353881339164957,
-      "learning_rate": 0.00015970122371604476,
-      "loss": 0.7513,
-      "step": 2512
-    },
-    {
-      "epoch": 1.2339299993848805,
-      "grad_norm": 0.4197091516443743,
-      "learning_rate": 0.00015965181438929146,
-      "loss": 0.7263,
-      "step": 2513
-    },
-    {
-      "epoch": 1.2344220950974965,
-      "grad_norm": 0.4573628112829376,
-      "learning_rate": 0.00015960238244514842,
-      "loss": 0.7554,
-      "step": 2514
-    },
-    {
-      "epoch": 1.2349141908101124,
-      "grad_norm": 0.4446582717598241,
-      "learning_rate": 0.00015955292790235815,
-      "loss": 0.7417,
-      "step": 2515
-    },
-    {
-      "epoch": 1.2354062865227287,
-      "grad_norm": 0.4382642106035462,
-      "learning_rate": 0.00015950345077967167,
-      "loss": 0.7643,
-      "step": 2516
-    },
-    {
-      "epoch": 1.2358983822353449,
-      "grad_norm": 0.46496882337561524,
-      "learning_rate": 0.00015945395109584856,
-      "loss": 0.6811,
-      "step": 2517
-    },
-    {
-      "epoch": 1.2363904779479609,
-      "grad_norm": 0.41513919237903263,
-      "learning_rate": 0.00015940442886965694,
-      "loss": 0.707,
-      "step": 2518
-    },
-    {
-      "epoch": 1.236882573660577,
-      "grad_norm": 0.456625249700212,
-      "learning_rate": 0.00015935488411987354,
-      "loss": 0.6953,
-      "step": 2519
-    },
-    {
-      "epoch": 1.237374669373193,
-      "grad_norm": 0.4234628275490277,
-      "learning_rate": 0.00015930531686528357,
-      "loss": 0.7408,
-      "step": 2520
-    },
-    {
-      "epoch": 1.2378667650858093,
-      "grad_norm": 0.43349572800507596,
-      "learning_rate": 0.00015925572712468074,
-      "loss": 0.6897,
-      "step": 2521
-    },
-    {
-      "epoch": 1.2383588607984253,
-      "grad_norm": 0.46079156509028607,
-      "learning_rate": 0.00015920611491686745,
-      "loss": 0.7461,
-      "step": 2522
-    },
-    {
-      "epoch": 1.2388509565110415,
-      "grad_norm": 0.44076547659732174,
-      "learning_rate": 0.0001591564802606544,
-      "loss": 0.7291,
-      "step": 2523
-    },
-    {
-      "epoch": 1.2393430522236575,
-      "grad_norm": 0.4482904356132328,
-      "learning_rate": 0.00015910682317486096,
-      "loss": 0.7393,
-      "step": 2524
-    },
-    {
-      "epoch": 1.2398351479362737,
-      "grad_norm": 0.44724083589325964,
-      "learning_rate": 0.00015905714367831495,
-      "loss": 0.7942,
-      "step": 2525
-    },
-    {
-      "epoch": 1.2403272436488897,
-      "grad_norm": 0.4400022233662369,
-      "learning_rate": 0.00015900744178985266,
-      "loss": 0.7723,
-      "step": 2526
-    },
-    {
-      "epoch": 1.240819339361506,
-      "grad_norm": 0.4405444597335698,
-      "learning_rate": 0.00015895771752831892,
-      "loss": 0.7166,
-      "step": 2527
-    },
-    {
-      "epoch": 1.241311435074122,
-      "grad_norm": 0.4766422904958864,
-      "learning_rate": 0.00015890797091256697,
-      "loss": 0.7676,
-      "step": 2528
-    },
-    {
-      "epoch": 1.241803530786738,
-      "grad_norm": 0.4096156751962372,
-      "learning_rate": 0.00015885820196145865,
-      "loss": 0.714,
-      "step": 2529
-    },
-    {
-      "epoch": 1.242295626499354,
-      "grad_norm": 0.4141158344096702,
-      "learning_rate": 0.00015880841069386417,
-      "loss": 0.7124,
-      "step": 2530
-    },
-    {
-      "epoch": 1.2427877222119703,
-      "grad_norm": 0.4383880333252046,
-      "learning_rate": 0.00015875859712866224,
-      "loss": 0.7772,
-      "step": 2531
-    },
-    {
-      "epoch": 1.2432798179245863,
-      "grad_norm": 0.4458031566093325,
-      "learning_rate": 0.00015870876128473994,
-      "loss": 0.8147,
-      "step": 2532
-    },
-    {
-      "epoch": 1.2437719136372025,
-      "grad_norm": 0.4421373461075213,
-      "learning_rate": 0.00015865890318099296,
-      "loss": 0.7479,
-      "step": 2533
-    },
-    {
-      "epoch": 1.2442640093498185,
-      "grad_norm": 0.4451710441613152,
-      "learning_rate": 0.0001586090228363253,
-      "loss": 0.685,
-      "step": 2534
-    },
-    {
-      "epoch": 1.2447561050624347,
-      "grad_norm": 0.43236320468734696,
-      "learning_rate": 0.0001585591202696494,
-      "loss": 0.7551,
-      "step": 2535
-    },
-    {
-      "epoch": 1.2452482007750507,
-      "grad_norm": 0.44693080770780086,
-      "learning_rate": 0.0001585091954998862,
-      "loss": 0.6972,
-      "step": 2536
-    },
-    {
-      "epoch": 1.245740296487667,
-      "grad_norm": 0.4090323440422212,
-      "learning_rate": 0.000158459248545965,
-      "loss": 0.6781,
-      "step": 2537
-    },
-    {
-      "epoch": 1.246232392200283,
-      "grad_norm": 0.417463605139135,
-      "learning_rate": 0.0001584092794268235,
-      "loss": 0.7371,
-      "step": 2538
-    },
-    {
-      "epoch": 1.2467244879128991,
-      "grad_norm": 0.4283222722616072,
-      "learning_rate": 0.0001583592881614079,
-      "loss": 0.7516,
-      "step": 2539
-    },
-    {
-      "epoch": 1.2472165836255151,
-      "grad_norm": 0.47081669612513066,
-      "learning_rate": 0.00015830927476867266,
-      "loss": 0.7918,
-      "step": 2540
-    },
-    {
-      "epoch": 1.2477086793381313,
-      "grad_norm": 0.46696184974319493,
-      "learning_rate": 0.00015825923926758071,
-      "loss": 0.8245,
-      "step": 2541
-    },
-    {
-      "epoch": 1.2482007750507473,
-      "grad_norm": 0.4222758370165774,
-      "learning_rate": 0.00015820918167710338,
-      "loss": 0.7826,
-      "step": 2542
-    },
-    {
-      "epoch": 1.2486928707633635,
-      "grad_norm": 0.41071507656104234,
-      "learning_rate": 0.00015815910201622032,
-      "loss": 0.6986,
-      "step": 2543
-    },
-    {
-      "epoch": 1.2491849664759795,
-      "grad_norm": 0.42879530840772173,
-      "learning_rate": 0.00015810900030391954,
-      "loss": 0.776,
-      "step": 2544
-    },
-    {
-      "epoch": 1.2496770621885958,
-      "grad_norm": 0.41098861871762604,
-      "learning_rate": 0.0001580588765591975,
-      "loss": 0.6994,
-      "step": 2545
-    },
-    {
-      "epoch": 1.2501691579012117,
-      "grad_norm": 0.4315909924016978,
-      "learning_rate": 0.00015800873080105895,
-      "loss": 0.7169,
-      "step": 2546
-    },
-    {
-      "epoch": 1.250661253613828,
-      "grad_norm": 0.4293799610847144,
-      "learning_rate": 0.00015795856304851697,
-      "loss": 0.7805,
-      "step": 2547
-    },
-    {
-      "epoch": 1.251153349326444,
-      "grad_norm": 0.4263069892019798,
-      "learning_rate": 0.000157908373320593,
-      "loss": 0.759,
-      "step": 2548
-    },
-    {
-      "epoch": 1.2516454450390602,
-      "grad_norm": 0.4616225247037991,
-      "learning_rate": 0.00015785816163631686,
-      "loss": 0.7394,
-      "step": 2549
-    },
-    {
-      "epoch": 1.2521375407516762,
-      "grad_norm": 0.4236887316755888,
-      "learning_rate": 0.0001578079280147266,
-      "loss": 0.7259,
-      "step": 2550
-    },
-    {
-      "epoch": 1.2526296364642924,
-      "grad_norm": 0.41721242240395623,
-      "learning_rate": 0.00015775767247486867,
-      "loss": 0.7236,
-      "step": 2551
-    },
-    {
-      "epoch": 1.2531217321769084,
-      "grad_norm": 0.43083139186312563,
-      "learning_rate": 0.0001577073950357978,
-      "loss": 0.7252,
-      "step": 2552
-    },
-    {
-      "epoch": 1.2536138278895246,
-      "grad_norm": 0.4812351771240485,
-      "learning_rate": 0.00015765709571657698,
-      "loss": 0.7246,
-      "step": 2553
-    },
-    {
-      "epoch": 1.2541059236021406,
-      "grad_norm": 0.4998599187138841,
-      "learning_rate": 0.00015760677453627756,
-      "loss": 0.7791,
-      "step": 2554
-    },
-    {
-      "epoch": 1.2545980193147568,
-      "grad_norm": 0.4951475330728904,
-      "learning_rate": 0.00015755643151397922,
-      "loss": 0.6723,
-      "step": 2555
-    },
-    {
-      "epoch": 1.2550901150273728,
-      "grad_norm": 0.46110719532733896,
-      "learning_rate": 0.00015750606666876978,
-      "loss": 0.7506,
-      "step": 2556
-    },
-    {
-      "epoch": 1.255582210739989,
-      "grad_norm": 0.41538497146655773,
-      "learning_rate": 0.00015745568001974538,
-      "loss": 0.7298,
-      "step": 2557
-    },
-    {
-      "epoch": 1.256074306452605,
-      "grad_norm": 0.4354878103310598,
-      "learning_rate": 0.00015740527158601055,
-      "loss": 0.7703,
-      "step": 2558
-    },
-    {
-      "epoch": 1.2565664021652212,
-      "grad_norm": 0.5616558173336792,
-      "learning_rate": 0.00015735484138667798,
-      "loss": 0.7723,
-      "step": 2559
-    },
-    {
-      "epoch": 1.2570584978778372,
-      "grad_norm": 0.4102438659567689,
-      "learning_rate": 0.00015730438944086856,
-      "loss": 0.7432,
-      "step": 2560
-    },
-    {
-      "epoch": 1.2575505935904534,
-      "grad_norm": 0.48142001034247683,
-      "learning_rate": 0.00015725391576771155,
-      "loss": 0.7573,
-      "step": 2561
-    },
-    {
-      "epoch": 1.2580426893030694,
-      "grad_norm": 0.4630310952308724,
-      "learning_rate": 0.0001572034203863443,
-      "loss": 0.7605,
-      "step": 2562
-    },
-    {
-      "epoch": 1.2585347850156856,
-      "grad_norm": 0.4395908066141942,
-      "learning_rate": 0.00015715290331591258,
-      "loss": 0.746,
-      "step": 2563
-    },
-    {
-      "epoch": 1.2590268807283016,
-      "grad_norm": 0.5529886102587573,
-      "learning_rate": 0.00015710236457557023,
-      "loss": 0.7502,
-      "step": 2564
-    },
-    {
-      "epoch": 1.2595189764409178,
-      "grad_norm": 0.42981905900486217,
-      "learning_rate": 0.00015705180418447933,
-      "loss": 0.7287,
-      "step": 2565
-    },
-    {
-      "epoch": 1.2600110721535338,
-      "grad_norm": 0.4216863490800534,
-      "learning_rate": 0.00015700122216181028,
-      "loss": 0.7159,
-      "step": 2566
-    },
-    {
-      "epoch": 1.26050316786615,
-      "grad_norm": 0.46652529393818115,
-      "learning_rate": 0.00015695061852674148,
-      "loss": 0.7915,
-      "step": 2567
-    },
-    {
-      "epoch": 1.260995263578766,
-      "grad_norm": 0.4500385023800325,
-      "learning_rate": 0.00015689999329845974,
-      "loss": 0.7151,
-      "step": 2568
-    },
-    {
-      "epoch": 1.2614873592913822,
-      "grad_norm": 0.4572461898834975,
-      "learning_rate": 0.00015684934649615986,
-      "loss": 0.8097,
-      "step": 2569
-    },
-    {
-      "epoch": 1.2619794550039982,
-      "grad_norm": 0.4059854077300595,
-      "learning_rate": 0.00015679867813904504,
-      "loss": 0.7113,
-      "step": 2570
-    },
-    {
-      "epoch": 1.2624715507166144,
-      "grad_norm": 0.4432979874271413,
-      "learning_rate": 0.00015674798824632646,
-      "loss": 0.7716,
-      "step": 2571
-    },
-    {
-      "epoch": 1.2629636464292304,
-      "grad_norm": 0.40674965149502756,
-      "learning_rate": 0.00015669727683722355,
-      "loss": 0.7182,
-      "step": 2572
-    },
-    {
-      "epoch": 1.2634557421418466,
-      "grad_norm": 0.4107488757313884,
-      "learning_rate": 0.00015664654393096386,
-      "loss": 0.727,
-      "step": 2573
-    },
-    {
-      "epoch": 1.2639478378544626,
-      "grad_norm": 0.4044172220207941,
-      "learning_rate": 0.00015659578954678314,
-      "loss": 0.658,
-      "step": 2574
-    },
-    {
-      "epoch": 1.2644399335670788,
-      "grad_norm": 0.47038385597153354,
-      "learning_rate": 0.00015654501370392524,
-      "loss": 0.7226,
-      "step": 2575
-    },
-    {
-      "epoch": 1.2649320292796948,
-      "grad_norm": 0.4349551722546035,
-      "learning_rate": 0.00015649421642164221,
-      "loss": 0.7548,
-      "step": 2576
-    },
-    {
-      "epoch": 1.265424124992311,
-      "grad_norm": 0.4729920720959477,
-      "learning_rate": 0.00015644339771919415,
-      "loss": 0.7762,
-      "step": 2577
-    },
-    {
-      "epoch": 1.265916220704927,
-      "grad_norm": 0.4288585025886129,
-      "learning_rate": 0.0001563925576158493,
-      "loss": 0.7939,
-      "step": 2578
-    },
-    {
-      "epoch": 1.2664083164175433,
-      "grad_norm": 0.412815562650632,
-      "learning_rate": 0.00015634169613088403,
-      "loss": 0.734,
-      "step": 2579
-    },
-    {
-      "epoch": 1.2669004121301592,
-      "grad_norm": 0.4468556521950253,
-      "learning_rate": 0.00015629081328358285,
-      "loss": 0.7285,
-      "step": 2580
-    },
-    {
-      "epoch": 1.2673925078427755,
-      "grad_norm": 0.44179881487038664,
-      "learning_rate": 0.00015623990909323834,
-      "loss": 0.7738,
-      "step": 2581
-    },
-    {
-      "epoch": 1.2678846035553915,
-      "grad_norm": 0.4285827365625489,
-      "learning_rate": 0.00015618898357915115,
-      "loss": 0.724,
-      "step": 2582
-    },
-    {
-      "epoch": 1.2683766992680077,
-      "grad_norm": 0.45272459710874907,
-      "learning_rate": 0.00015613803676063,
-      "loss": 0.81,
-      "step": 2583
-    },
-    {
-      "epoch": 1.2688687949806237,
-      "grad_norm": 0.41262992071549875,
-      "learning_rate": 0.00015608706865699178,
-      "loss": 0.7521,
-      "step": 2584
-    },
-    {
-      "epoch": 1.2693608906932399,
-      "grad_norm": 0.45924739219356164,
-      "learning_rate": 0.00015603607928756137,
-      "loss": 0.8915,
-      "step": 2585
-    },
-    {
-      "epoch": 1.2698529864058559,
-      "grad_norm": 0.4415343245957275,
-      "learning_rate": 0.0001559850686716717,
-      "loss": 0.7632,
-      "step": 2586
-    },
-    {
-      "epoch": 1.270345082118472,
-      "grad_norm": 0.4224491218804207,
-      "learning_rate": 0.00015593403682866389,
-      "loss": 0.7221,
-      "step": 2587
-    },
-    {
-      "epoch": 1.270837177831088,
-      "grad_norm": 0.432127519328177,
-      "learning_rate": 0.00015588298377788688,
-      "loss": 0.7502,
-      "step": 2588
-    },
-    {
-      "epoch": 1.2713292735437043,
-      "grad_norm": 0.42226438361882757,
-      "learning_rate": 0.00015583190953869785,
-      "loss": 0.7434,
-      "step": 2589
-    },
-    {
-      "epoch": 1.2718213692563203,
-      "grad_norm": 0.4201903008300504,
-      "learning_rate": 0.00015578081413046191,
-      "loss": 0.7284,
-      "step": 2590
-    },
-    {
-      "epoch": 1.2723134649689365,
-      "grad_norm": 0.4242112061271643,
-      "learning_rate": 0.0001557296975725523,
-      "loss": 0.7589,
-      "step": 2591
-    },
-    {
-      "epoch": 1.2728055606815525,
-      "grad_norm": 0.42544311018827574,
-      "learning_rate": 0.0001556785598843502,
-      "loss": 0.7441,
-      "step": 2592
-    },
-    {
-      "epoch": 1.2732976563941687,
-      "grad_norm": 0.41125085727573196,
-      "learning_rate": 0.00015562740108524472,
-      "loss": 0.7414,
-      "step": 2593
-    },
-    {
-      "epoch": 1.2737897521067847,
-      "grad_norm": 0.44852864466061376,
-      "learning_rate": 0.00015557622119463313,
-      "loss": 0.8235,
-      "step": 2594
-    },
-    {
-      "epoch": 1.274281847819401,
-      "grad_norm": 0.4539169694183089,
-      "learning_rate": 0.00015552502023192063,
-      "loss": 0.7368,
-      "step": 2595
-    },
-    {
-      "epoch": 1.274773943532017,
-      "grad_norm": 0.4166987356758154,
-      "learning_rate": 0.00015547379821652038,
-      "loss": 0.7476,
-      "step": 2596
-    },
-    {
-      "epoch": 1.275266039244633,
-      "grad_norm": 0.42703010260322033,
-      "learning_rate": 0.00015542255516785361,
-      "loss": 0.7314,
-      "step": 2597
-    },
-    {
-      "epoch": 1.275758134957249,
-      "grad_norm": 0.4769045049464421,
-      "learning_rate": 0.00015537129110534945,
-      "loss": 0.74,
-      "step": 2598
-    },
-    {
-      "epoch": 1.2762502306698653,
-      "grad_norm": 0.39678421251550167,
-      "learning_rate": 0.000155320006048445,
-      "loss": 0.7511,
-      "step": 2599
-    },
-    {
-      "epoch": 1.2767423263824815,
-      "grad_norm": 0.5044737420814177,
-      "learning_rate": 0.00015526870001658532,
-      "loss": 0.7248,
-      "step": 2600
-    },
-    {
-      "epoch": 1.2772344220950975,
-      "grad_norm": 0.40143263726257994,
-      "learning_rate": 0.00015521737302922348,
-      "loss": 0.7398,
-      "step": 2601
-    },
-    {
-      "epoch": 1.2777265178077135,
-      "grad_norm": 0.43710518643214913,
-      "learning_rate": 0.00015516602510582043,
-      "loss": 0.7526,
-      "step": 2602
-    },
-    {
-      "epoch": 1.2782186135203297,
-      "grad_norm": 0.4662321654183513,
-      "learning_rate": 0.00015511465626584509,
-      "loss": 0.7682,
-      "step": 2603
-    },
-    {
-      "epoch": 1.278710709232946,
-      "grad_norm": 0.4964709774050139,
-      "learning_rate": 0.00015506326652877433,
-      "loss": 0.8085,
-      "step": 2604
-    },
-    {
-      "epoch": 1.279202804945562,
-      "grad_norm": 0.44658978575127173,
-      "learning_rate": 0.00015501185591409287,
-      "loss": 0.7278,
-      "step": 2605
-    },
-    {
-      "epoch": 1.279694900658178,
-      "grad_norm": 0.4175225364618719,
-      "learning_rate": 0.00015496042444129338,
-      "loss": 0.7228,
-      "step": 2606
-    },
-    {
-      "epoch": 1.2801869963707941,
-      "grad_norm": 0.43449038536138557,
-      "learning_rate": 0.00015490897212987656,
-      "loss": 0.7747,
-      "step": 2607
-    },
-    {
-      "epoch": 1.2806790920834104,
-      "grad_norm": 0.41695126357161955,
-      "learning_rate": 0.0001548574989993508,
-      "loss": 0.7668,
-      "step": 2608
-    },
-    {
-      "epoch": 1.2811711877960263,
-      "grad_norm": 0.4381103385716402,
-      "learning_rate": 0.00015480600506923248,
-      "loss": 0.6944,
-      "step": 2609
-    },
-    {
-      "epoch": 1.2816632835086423,
-      "grad_norm": 0.4514121424525538,
-      "learning_rate": 0.00015475449035904596,
-      "loss": 0.7744,
-      "step": 2610
-    },
-    {
-      "epoch": 1.2821553792212586,
-      "grad_norm": 0.40863780724558096,
-      "learning_rate": 0.00015470295488832333,
-      "loss": 0.7081,
-      "step": 2611
-    },
-    {
-      "epoch": 1.2826474749338748,
-      "grad_norm": 0.5372182678341804,
-      "learning_rate": 0.0001546513986766046,
-      "loss": 0.7893,
-      "step": 2612
-    },
-    {
-      "epoch": 1.2831395706464908,
-      "grad_norm": 0.4379451533982603,
-      "learning_rate": 0.00015459982174343767,
-      "loss": 0.7394,
-      "step": 2613
-    },
-    {
-      "epoch": 1.2836316663591067,
-      "grad_norm": 0.4257835176263186,
-      "learning_rate": 0.00015454822410837832,
-      "loss": 0.7494,
-      "step": 2614
-    },
-    {
-      "epoch": 1.284123762071723,
-      "grad_norm": 0.47811432081825217,
-      "learning_rate": 0.00015449660579099013,
-      "loss": 0.7512,
-      "step": 2615
-    },
-    {
-      "epoch": 1.2846158577843392,
-      "grad_norm": 0.42179771281231876,
-      "learning_rate": 0.0001544449668108445,
-      "loss": 0.7946,
-      "step": 2616
-    },
-    {
-      "epoch": 1.2851079534969552,
-      "grad_norm": 0.44306737117407474,
-      "learning_rate": 0.0001543933071875207,
-      "loss": 0.6966,
-      "step": 2617
-    },
-    {
-      "epoch": 1.2856000492095712,
-      "grad_norm": 0.4833762980530231,
-      "learning_rate": 0.0001543416269406059,
-      "loss": 0.7088,
-      "step": 2618
-    },
-    {
-      "epoch": 1.2860921449221874,
-      "grad_norm": 0.3972083842652231,
-      "learning_rate": 0.0001542899260896949,
-      "loss": 0.7038,
-      "step": 2619
-    },
-    {
-      "epoch": 1.2865842406348036,
-      "grad_norm": 0.43506703855597095,
-      "learning_rate": 0.0001542382046543905,
-      "loss": 0.7748,
-      "step": 2620
-    },
-    {
-      "epoch": 1.2870763363474196,
-      "grad_norm": 0.4392513447268944,
-      "learning_rate": 0.00015418646265430327,
-      "loss": 0.7561,
-      "step": 2621
-    },
-    {
-      "epoch": 1.2875684320600356,
-      "grad_norm": 0.46418470143332413,
-      "learning_rate": 0.00015413470010905146,
-      "loss": 0.7978,
-      "step": 2622
-    },
-    {
-      "epoch": 1.2880605277726518,
-      "grad_norm": 0.43135991816412844,
-      "learning_rate": 0.00015408291703826127,
-      "loss": 0.7525,
-      "step": 2623
-    },
-    {
-      "epoch": 1.288552623485268,
-      "grad_norm": 0.4109210674724624,
-      "learning_rate": 0.00015403111346156648,
-      "loss": 0.7428,
-      "step": 2624
-    },
-    {
-      "epoch": 1.289044719197884,
-      "grad_norm": 0.43396204884755263,
-      "learning_rate": 0.00015397928939860886,
-      "loss": 0.7443,
-      "step": 2625
-    },
-    {
-      "epoch": 1.2895368149105,
-      "grad_norm": 0.40773554323909833,
-      "learning_rate": 0.00015392744486903787,
-      "loss": 0.6975,
-      "step": 2626
-    },
-    {
-      "epoch": 1.2900289106231162,
-      "grad_norm": 0.44892980909023017,
-      "learning_rate": 0.00015387557989251068,
-      "loss": 0.7925,
-      "step": 2627
-    },
-    {
-      "epoch": 1.2905210063357324,
-      "grad_norm": 0.4220792652483258,
-      "learning_rate": 0.00015382369448869226,
-      "loss": 0.6736,
-      "step": 2628
-    },
-    {
-      "epoch": 1.2910131020483484,
-      "grad_norm": 0.4209547035227594,
-      "learning_rate": 0.00015377178867725526,
-      "loss": 0.7182,
-      "step": 2629
-    },
-    {
-      "epoch": 1.2915051977609644,
-      "grad_norm": 0.42210710631459336,
-      "learning_rate": 0.00015371986247788018,
-      "loss": 0.786,
-      "step": 2630
-    },
-    {
-      "epoch": 1.2919972934735806,
-      "grad_norm": 0.4233103738421945,
-      "learning_rate": 0.0001536679159102552,
-      "loss": 0.6988,
-      "step": 2631
-    },
-    {
-      "epoch": 1.2924893891861968,
-      "grad_norm": 0.4403054673592731,
-      "learning_rate": 0.00015361594899407615,
-      "loss": 0.7675,
-      "step": 2632
-    },
-    {
-      "epoch": 1.2929814848988128,
-      "grad_norm": 0.42557645989762755,
-      "learning_rate": 0.00015356396174904665,
-      "loss": 0.7679,
-      "step": 2633
-    },
-    {
-      "epoch": 1.2934735806114288,
-      "grad_norm": 0.4236263081120599,
-      "learning_rate": 0.0001535119541948781,
-      "loss": 0.7581,
-      "step": 2634
-    },
-    {
-      "epoch": 1.293965676324045,
-      "grad_norm": 0.43803156837979607,
-      "learning_rate": 0.00015345992635128943,
-      "loss": 0.7325,
-      "step": 2635
-    },
-    {
-      "epoch": 1.2944577720366612,
-      "grad_norm": 0.4420668030812656,
-      "learning_rate": 0.00015340787823800737,
-      "loss": 0.6998,
-      "step": 2636
-    },
-    {
-      "epoch": 1.2949498677492772,
-      "grad_norm": 0.4209700553026789,
-      "learning_rate": 0.00015335580987476633,
-      "loss": 0.7103,
-      "step": 2637
-    },
-    {
-      "epoch": 1.2954419634618932,
-      "grad_norm": 0.4390333251021473,
-      "learning_rate": 0.0001533037212813084,
-      "loss": 0.8193,
-      "step": 2638
-    },
-    {
-      "epoch": 1.2959340591745094,
-      "grad_norm": 0.4397623052427127,
-      "learning_rate": 0.00015325161247738324,
-      "loss": 0.7609,
-      "step": 2639
-    },
-    {
-      "epoch": 1.2964261548871256,
-      "grad_norm": 0.4202037110563578,
-      "learning_rate": 0.00015319948348274835,
-      "loss": 0.7672,
-      "step": 2640
-    },
-    {
-      "epoch": 1.2969182505997416,
-      "grad_norm": 0.46863029685362373,
-      "learning_rate": 0.00015314733431716877,
-      "loss": 0.7904,
-      "step": 2641
-    },
-    {
-      "epoch": 1.2974103463123576,
-      "grad_norm": 0.45916684501587546,
-      "learning_rate": 0.00015309516500041718,
-      "loss": 0.7986,
-      "step": 2642
-    },
-    {
-      "epoch": 1.2979024420249738,
-      "grad_norm": 0.4146209464048312,
-      "learning_rate": 0.00015304297555227393,
-      "loss": 0.7309,
-      "step": 2643
-    },
-    {
-      "epoch": 1.29839453773759,
-      "grad_norm": 0.4345550879236155,
-      "learning_rate": 0.00015299076599252701,
-      "loss": 0.7794,
-      "step": 2644
-    },
-    {
-      "epoch": 1.298886633450206,
-      "grad_norm": 0.44004376957854224,
-      "learning_rate": 0.00015293853634097207,
-      "loss": 0.731,
-      "step": 2645
-    },
-    {
-      "epoch": 1.299378729162822,
-      "grad_norm": 0.42114874972899674,
-      "learning_rate": 0.00015288628661741229,
-      "loss": 0.6931,
-      "step": 2646
-    },
-    {
-      "epoch": 1.2998708248754383,
-      "grad_norm": 0.43650945760271986,
-      "learning_rate": 0.00015283401684165851,
-      "loss": 0.7435,
-      "step": 2647
-    },
-    {
-      "epoch": 1.3003629205880545,
-      "grad_norm": 0.49219684391170015,
-      "learning_rate": 0.00015278172703352916,
-      "loss": 0.7146,
-      "step": 2648
-    },
-    {
-      "epoch": 1.3008550163006705,
-      "grad_norm": 0.4322023793096788,
-      "learning_rate": 0.00015272941721285033,
-      "loss": 0.8173,
-      "step": 2649
-    },
-    {
-      "epoch": 1.3013471120132865,
-      "grad_norm": 0.44082095384624576,
-      "learning_rate": 0.00015267708739945558,
-      "loss": 0.8049,
-      "step": 2650
-    },
-    {
-      "epoch": 1.3018392077259027,
-      "grad_norm": 0.45126265501229934,
-      "learning_rate": 0.00015262473761318618,
-      "loss": 0.759,
-      "step": 2651
-    },
-    {
-      "epoch": 1.3023313034385189,
-      "grad_norm": 0.4483440426939856,
-      "learning_rate": 0.00015257236787389084,
-      "loss": 0.6709,
-      "step": 2652
-    },
-    {
-      "epoch": 1.3028233991511349,
-      "grad_norm": 0.42136193587078413,
-      "learning_rate": 0.00015251997820142594,
-      "loss": 0.722,
-      "step": 2653
-    },
-    {
-      "epoch": 1.3033154948637509,
-      "grad_norm": 0.4288100440936685,
-      "learning_rate": 0.00015246756861565536,
-      "loss": 0.7478,
-      "step": 2654
-    },
-    {
-      "epoch": 1.303807590576367,
-      "grad_norm": 0.43267631712947824,
-      "learning_rate": 0.00015241513913645055,
-      "loss": 0.751,
-      "step": 2655
-    },
-    {
-      "epoch": 1.3042996862889833,
-      "grad_norm": 0.42818314355417175,
-      "learning_rate": 0.00015236268978369052,
-      "loss": 0.7629,
-      "step": 2656
-    },
-    {
-      "epoch": 1.3047917820015993,
-      "grad_norm": 0.4099356351781274,
-      "learning_rate": 0.0001523102205772618,
-      "loss": 0.7288,
-      "step": 2657
-    },
-    {
-      "epoch": 1.3052838777142153,
-      "grad_norm": 0.4236925245280607,
-      "learning_rate": 0.0001522577315370584,
-      "loss": 0.7601,
-      "step": 2658
-    },
-    {
-      "epoch": 1.3057759734268315,
-      "grad_norm": 0.40563395419484544,
-      "learning_rate": 0.00015220522268298193,
-      "loss": 0.6793,
-      "step": 2659
-    },
-    {
-      "epoch": 1.3062680691394477,
-      "grad_norm": 0.40432994975635483,
-      "learning_rate": 0.0001521526940349415,
-      "loss": 0.6969,
-      "step": 2660
-    },
-    {
-      "epoch": 1.3067601648520637,
-      "grad_norm": 0.42299007827818075,
-      "learning_rate": 0.00015210014561285365,
-      "loss": 0.7626,
-      "step": 2661
-    },
-    {
-      "epoch": 1.30725226056468,
-      "grad_norm": 0.4537197460382443,
-      "learning_rate": 0.00015204757743664252,
-      "loss": 0.8275,
-      "step": 2662
-    },
-    {
-      "epoch": 1.307744356277296,
-      "grad_norm": 0.42102164306599055,
-      "learning_rate": 0.00015199498952623967,
-      "loss": 0.7387,
-      "step": 2663
-    },
-    {
-      "epoch": 1.3082364519899121,
-      "grad_norm": 0.40736146587821015,
-      "learning_rate": 0.00015194238190158416,
-      "loss": 0.76,
-      "step": 2664
-    },
-    {
-      "epoch": 1.3087285477025281,
-      "grad_norm": 0.4839056190832484,
-      "learning_rate": 0.00015188975458262256,
-      "loss": 0.7263,
-      "step": 2665
-    },
-    {
-      "epoch": 1.3092206434151443,
-      "grad_norm": 0.4228987201039662,
-      "learning_rate": 0.0001518371075893088,
-      "loss": 0.7212,
-      "step": 2666
-    },
-    {
-      "epoch": 1.3097127391277603,
-      "grad_norm": 0.4281709483909016,
-      "learning_rate": 0.0001517844409416044,
-      "loss": 0.7377,
-      "step": 2667
-    },
-    {
-      "epoch": 1.3102048348403765,
-      "grad_norm": 0.4202265635833236,
-      "learning_rate": 0.00015173175465947827,
-      "loss": 0.7236,
-      "step": 2668
-    },
-    {
-      "epoch": 1.3106969305529925,
-      "grad_norm": 0.41013205706430633,
-      "learning_rate": 0.00015167904876290677,
-      "loss": 0.698,
-      "step": 2669
-    },
-    {
-      "epoch": 1.3111890262656087,
-      "grad_norm": 0.4559053248434141,
-      "learning_rate": 0.00015162632327187368,
-      "loss": 0.6948,
-      "step": 2670
-    },
-    {
-      "epoch": 1.3116811219782247,
-      "grad_norm": 0.47860523913959824,
-      "learning_rate": 0.00015157357820637021,
-      "loss": 0.793,
-      "step": 2671
-    },
-    {
-      "epoch": 1.312173217690841,
-      "grad_norm": 0.43126462460049897,
-      "learning_rate": 0.0001515208135863951,
-      "loss": 0.781,
-      "step": 2672
-    },
-    {
-      "epoch": 1.312665313403457,
-      "grad_norm": 0.45629304452154124,
-      "learning_rate": 0.00015146802943195433,
-      "loss": 0.8345,
-      "step": 2673
-    },
-    {
-      "epoch": 1.3131574091160731,
-      "grad_norm": 0.41588679447347654,
-      "learning_rate": 0.00015141522576306136,
-      "loss": 0.7517,
-      "step": 2674
-    },
-    {
-      "epoch": 1.3136495048286891,
-      "grad_norm": 0.4430729681470319,
-      "learning_rate": 0.00015136240259973715,
-      "loss": 0.7445,
-      "step": 2675
-    },
-    {
-      "epoch": 1.3141416005413054,
-      "grad_norm": 0.45830387363452274,
-      "learning_rate": 0.0001513095599620099,
-      "loss": 0.78,
-      "step": 2676
-    },
-    {
-      "epoch": 1.3146336962539213,
-      "grad_norm": 0.43554428453591837,
-      "learning_rate": 0.0001512566978699152,
-      "loss": 0.748,
-      "step": 2677
-    },
-    {
-      "epoch": 1.3151257919665376,
-      "grad_norm": 0.4369067135485014,
-      "learning_rate": 0.00015120381634349617,
-      "loss": 0.7524,
-      "step": 2678
-    },
-    {
-      "epoch": 1.3156178876791536,
-      "grad_norm": 0.4438122596677527,
-      "learning_rate": 0.00015115091540280316,
-      "loss": 0.7754,
-      "step": 2679
-    },
-    {
-      "epoch": 1.3161099833917698,
-      "grad_norm": 0.4695610067237712,
-      "learning_rate": 0.0001510979950678939,
-      "loss": 0.7698,
-      "step": 2680
-    },
-    {
-      "epoch": 1.3166020791043858,
-      "grad_norm": 0.4290852887766823,
-      "learning_rate": 0.0001510450553588335,
-      "loss": 0.7156,
-      "step": 2681
-    },
-    {
-      "epoch": 1.317094174817002,
-      "grad_norm": 0.42483568878517364,
-      "learning_rate": 0.00015099209629569442,
-      "loss": 0.7312,
-      "step": 2682
-    },
-    {
-      "epoch": 1.317586270529618,
-      "grad_norm": 0.46614704246217564,
-      "learning_rate": 0.00015093911789855645,
-      "loss": 0.722,
-      "step": 2683
-    },
-    {
-      "epoch": 1.3180783662422342,
-      "grad_norm": 0.40866774536870926,
-      "learning_rate": 0.0001508861201875067,
-      "loss": 0.6605,
-      "step": 2684
-    },
-    {
-      "epoch": 1.3185704619548502,
-      "grad_norm": 0.47439495732029524,
-      "learning_rate": 0.00015083310318263964,
-      "loss": 0.7408,
-      "step": 2685
-    },
-    {
-      "epoch": 1.3190625576674664,
-      "grad_norm": 0.395423912110938,
-      "learning_rate": 0.000150780066904057,
-      "loss": 0.698,
-      "step": 2686
-    },
-    {
-      "epoch": 1.3195546533800824,
-      "grad_norm": 0.45237785162409755,
-      "learning_rate": 0.00015072701137186785,
-      "loss": 0.7701,
-      "step": 2687
-    },
-    {
-      "epoch": 1.3200467490926986,
-      "grad_norm": 0.4436148748238658,
-      "learning_rate": 0.00015067393660618853,
-      "loss": 0.7413,
-      "step": 2688
-    },
-    {
-      "epoch": 1.3205388448053146,
-      "grad_norm": 0.4026463831862003,
-      "learning_rate": 0.00015062084262714275,
-      "loss": 0.7118,
-      "step": 2689
-    },
-    {
-      "epoch": 1.3210309405179308,
-      "grad_norm": 0.41205806933136185,
-      "learning_rate": 0.0001505677294548614,
-      "loss": 0.6963,
-      "step": 2690
-    },
-    {
-      "epoch": 1.3215230362305468,
-      "grad_norm": 0.4355776162302929,
-      "learning_rate": 0.0001505145971094827,
-      "loss": 0.7035,
-      "step": 2691
-    },
-    {
-      "epoch": 1.322015131943163,
-      "grad_norm": 0.4265899734182896,
-      "learning_rate": 0.0001504614456111522,
-      "loss": 0.7754,
-      "step": 2692
-    },
-    {
-      "epoch": 1.322507227655779,
-      "grad_norm": 0.41098003634808444,
-      "learning_rate": 0.0001504082749800226,
-      "loss": 0.7134,
-      "step": 2693
-    },
-    {
-      "epoch": 1.3229993233683952,
-      "grad_norm": 0.4326632723797512,
-      "learning_rate": 0.0001503550852362539,
-      "loss": 0.7819,
-      "step": 2694
-    },
-    {
-      "epoch": 1.3234914190810112,
-      "grad_norm": 0.4223330074094377,
-      "learning_rate": 0.0001503018764000134,
-      "loss": 0.7096,
-      "step": 2695
-    },
-    {
-      "epoch": 1.3239835147936274,
-      "grad_norm": 0.42211742162701826,
-      "learning_rate": 0.00015024864849147554,
-      "loss": 0.7337,
-      "step": 2696
-    },
-    {
-      "epoch": 1.3244756105062434,
-      "grad_norm": 0.4143320500928967,
-      "learning_rate": 0.00015019540153082201,
-      "loss": 0.6731,
-      "step": 2697
-    },
-    {
-      "epoch": 1.3249677062188596,
-      "grad_norm": 0.418157045554795,
-      "learning_rate": 0.00015014213553824187,
-      "loss": 0.8011,
-      "step": 2698
-    },
-    {
-      "epoch": 1.3254598019314756,
-      "grad_norm": 0.43450430706254045,
-      "learning_rate": 0.0001500888505339312,
-      "loss": 0.7006,
-      "step": 2699
-    },
-    {
-      "epoch": 1.3259518976440918,
-      "grad_norm": 0.411237308827399,
-      "learning_rate": 0.00015003554653809342,
-      "loss": 0.7208,
-      "step": 2700
-    },
-    {
-      "epoch": 1.3264439933567078,
-      "grad_norm": 0.3991299352012469,
-      "learning_rate": 0.00014998222357093903,
-      "loss": 0.7275,
-      "step": 2701
-    },
-    {
-      "epoch": 1.326936089069324,
-      "grad_norm": 0.41991732320264646,
-      "learning_rate": 0.00014992888165268583,
-      "loss": 0.6953,
-      "step": 2702
-    },
-    {
-      "epoch": 1.32742818478194,
-      "grad_norm": 0.44469660282099244,
-      "learning_rate": 0.0001498755208035588,
-      "loss": 0.7119,
-      "step": 2703
-    },
-    {
-      "epoch": 1.3279202804945562,
-      "grad_norm": 0.4501658533700222,
-      "learning_rate": 0.00014982214104379,
-      "loss": 0.7995,
-      "step": 2704
-    },
-    {
-      "epoch": 1.3284123762071722,
-      "grad_norm": 0.42220874843132583,
-      "learning_rate": 0.00014976874239361882,
-      "loss": 0.7337,
-      "step": 2705
-    },
-    {
-      "epoch": 1.3289044719197884,
-      "grad_norm": 0.4381833508887948,
-      "learning_rate": 0.00014971532487329165,
-      "loss": 0.7355,
-      "step": 2706
-    },
-    {
-      "epoch": 1.3293965676324044,
-      "grad_norm": 0.41296393354127864,
-      "learning_rate": 0.0001496618885030621,
-      "loss": 0.6702,
-      "step": 2707
-    },
-    {
-      "epoch": 1.3298886633450206,
-      "grad_norm": 0.4449986243348617,
-      "learning_rate": 0.000149608433303191,
-      "loss": 0.7757,
-      "step": 2708
-    },
-    {
-      "epoch": 1.3303807590576366,
-      "grad_norm": 0.4125510892416481,
-      "learning_rate": 0.0001495549592939462,
-      "loss": 0.7235,
-      "step": 2709
-    },
-    {
-      "epoch": 1.3308728547702529,
-      "grad_norm": 0.4159871936786702,
-      "learning_rate": 0.00014950146649560274,
-      "loss": 0.7513,
-      "step": 2710
-    },
-    {
-      "epoch": 1.3313649504828688,
-      "grad_norm": 0.4562920096613522,
-      "learning_rate": 0.00014944795492844278,
-      "loss": 0.7282,
-      "step": 2711
-    },
-    {
-      "epoch": 1.331857046195485,
-      "grad_norm": 0.4236634548305959,
-      "learning_rate": 0.00014939442461275556,
-      "loss": 0.7675,
-      "step": 2712
-    },
-    {
-      "epoch": 1.332349141908101,
-      "grad_norm": 0.4483689418369436,
-      "learning_rate": 0.00014934087556883754,
-      "loss": 0.8156,
-      "step": 2713
-    },
-    {
-      "epoch": 1.3328412376207173,
-      "grad_norm": 0.41471430318775926,
-      "learning_rate": 0.00014928730781699212,
-      "loss": 0.7828,
-      "step": 2714
-    },
-    {
-      "epoch": 1.3333333333333333,
-      "grad_norm": 0.45280773074731756,
-      "learning_rate": 0.00014923372137752992,
-      "loss": 0.7223,
-      "step": 2715
-    },
-    {
-      "epoch": 1.3338254290459495,
-      "grad_norm": 0.4433372619402007,
-      "learning_rate": 0.0001491801162707686,
-      "loss": 0.7417,
-      "step": 2716
-    },
-    {
-      "epoch": 1.3343175247585655,
-      "grad_norm": 0.4218838171767951,
-      "learning_rate": 0.00014912649251703288,
-      "loss": 0.724,
-      "step": 2717
-    },
-    {
-      "epoch": 1.3348096204711817,
-      "grad_norm": 0.44886560272394316,
-      "learning_rate": 0.0001490728501366546,
-      "loss": 0.7547,
-      "step": 2718
-    },
-    {
-      "epoch": 1.3353017161837977,
-      "grad_norm": 0.5339889527003661,
-      "learning_rate": 0.0001490191891499726,
-      "loss": 0.7936,
-      "step": 2719
-    },
-    {
-      "epoch": 1.3357938118964139,
-      "grad_norm": 0.4508003935947812,
-      "learning_rate": 0.00014896550957733284,
-      "loss": 0.775,
-      "step": 2720
-    },
-    {
-      "epoch": 1.3362859076090299,
-      "grad_norm": 0.4299488485447041,
-      "learning_rate": 0.00014891181143908826,
-      "loss": 0.7989,
-      "step": 2721
-    },
-    {
-      "epoch": 1.336778003321646,
-      "grad_norm": 0.46604109703082197,
-      "learning_rate": 0.0001488580947555989,
-      "loss": 0.7592,
-      "step": 2722
-    },
-    {
-      "epoch": 1.337270099034262,
-      "grad_norm": 0.4696760045210466,
-      "learning_rate": 0.0001488043595472318,
-      "loss": 0.7481,
-      "step": 2723
-    },
-    {
-      "epoch": 1.3377621947468783,
-      "grad_norm": 0.43535020769769234,
-      "learning_rate": 0.00014875060583436101,
-      "loss": 0.7488,
-      "step": 2724
-    },
-    {
-      "epoch": 1.3382542904594943,
-      "grad_norm": 0.42759257210065515,
-      "learning_rate": 0.00014869683363736763,
-      "loss": 0.7461,
-      "step": 2725
-    },
-    {
-      "epoch": 1.3387463861721105,
-      "grad_norm": 0.4163764624439068,
-      "learning_rate": 0.00014864304297663975,
-      "loss": 0.7547,
-      "step": 2726
-    },
-    {
-      "epoch": 1.3392384818847267,
-      "grad_norm": 0.40831939882603613,
-      "learning_rate": 0.00014858923387257245,
-      "loss": 0.6954,
-      "step": 2727
-    },
-    {
-      "epoch": 1.3397305775973427,
-      "grad_norm": 0.4691562198083282,
-      "learning_rate": 0.00014853540634556783,
-      "loss": 0.7478,
-      "step": 2728
-    },
-    {
-      "epoch": 1.3402226733099587,
-      "grad_norm": 0.43429507939103096,
-      "learning_rate": 0.000148481560416035,
-      "loss": 0.7302,
-      "step": 2729
-    },
-    {
-      "epoch": 1.340714769022575,
-      "grad_norm": 0.4302405186388293,
-      "learning_rate": 0.00014842769610438992,
-      "loss": 0.7908,
-      "step": 2730
-    },
-    {
-      "epoch": 1.3412068647351911,
-      "grad_norm": 0.41467307101897777,
-      "learning_rate": 0.00014837381343105568,
-      "loss": 0.7113,
-      "step": 2731
-    },
-    {
-      "epoch": 1.3416989604478071,
-      "grad_norm": 0.4396330083425095,
-      "learning_rate": 0.00014831991241646226,
-      "loss": 0.7036,
-      "step": 2732
-    },
-    {
-      "epoch": 1.3421910561604231,
-      "grad_norm": 0.4273104827213855,
-      "learning_rate": 0.00014826599308104653,
-      "loss": 0.7805,
-      "step": 2733
-    },
-    {
-      "epoch": 1.3426831518730393,
-      "grad_norm": 0.4660312695309136,
-      "learning_rate": 0.00014821205544525244,
-      "loss": 0.7577,
-      "step": 2734
-    },
-    {
-      "epoch": 1.3431752475856555,
-      "grad_norm": 0.44721064984702363,
-      "learning_rate": 0.00014815809952953082,
-      "loss": 0.7386,
-      "step": 2735
-    },
-    {
-      "epoch": 1.3436673432982715,
-      "grad_norm": 0.44093357124612287,
-      "learning_rate": 0.00014810412535433935,
-      "loss": 0.6815,
-      "step": 2736
-    },
-    {
-      "epoch": 1.3441594390108875,
-      "grad_norm": 0.4320067264334088,
-      "learning_rate": 0.00014805013294014275,
-      "loss": 0.703,
-      "step": 2737
-    },
-    {
-      "epoch": 1.3446515347235037,
-      "grad_norm": 0.48853666968241427,
-      "learning_rate": 0.00014799612230741258,
-      "loss": 0.8311,
-      "step": 2738
-    },
-    {
-      "epoch": 1.34514363043612,
-      "grad_norm": 0.42564282859711905,
-      "learning_rate": 0.0001479420934766274,
-      "loss": 0.75,
-      "step": 2739
-    },
-    {
-      "epoch": 1.345635726148736,
-      "grad_norm": 0.4551383735721082,
-      "learning_rate": 0.0001478880464682725,
-      "loss": 0.7773,
-      "step": 2740
-    },
-    {
-      "epoch": 1.346127821861352,
-      "grad_norm": 0.4823949036191055,
-      "learning_rate": 0.0001478339813028403,
-      "loss": 0.7904,
-      "step": 2741
-    },
-    {
-      "epoch": 1.3466199175739682,
-      "grad_norm": 0.45454536255046707,
-      "learning_rate": 0.0001477798980008299,
-      "loss": 0.7397,
-      "step": 2742
-    },
-    {
-      "epoch": 1.3471120132865844,
-      "grad_norm": 0.4896659331141192,
-      "learning_rate": 0.00014772579658274733,
-      "loss": 0.7963,
-      "step": 2743
-    },
-    {
-      "epoch": 1.3476041089992004,
-      "grad_norm": 0.43717123793847806,
-      "learning_rate": 0.00014767167706910555,
-      "loss": 0.7434,
-      "step": 2744
-    },
-    {
-      "epoch": 1.3480962047118163,
-      "grad_norm": 0.44713356814795013,
-      "learning_rate": 0.00014761753948042434,
-      "loss": 0.7242,
-      "step": 2745
-    },
-    {
-      "epoch": 1.3485883004244326,
-      "grad_norm": 0.4453704215510461,
-      "learning_rate": 0.00014756338383723033,
-      "loss": 0.7451,
-      "step": 2746
-    },
-    {
-      "epoch": 1.3490803961370488,
-      "grad_norm": 0.43980454771431327,
-      "learning_rate": 0.000147509210160057,
-      "loss": 0.7768,
-      "step": 2747
-    },
-    {
-      "epoch": 1.3495724918496648,
-      "grad_norm": 0.4196286931419831,
-      "learning_rate": 0.00014745501846944462,
-      "loss": 0.6887,
-      "step": 2748
-    },
-    {
-      "epoch": 1.3500645875622808,
-      "grad_norm": 0.47158978701859167,
-      "learning_rate": 0.00014740080878594043,
-      "loss": 0.8065,
-      "step": 2749
-    },
-    {
-      "epoch": 1.350556683274897,
-      "grad_norm": 0.42575906936646557,
-      "learning_rate": 0.0001473465811300983,
-      "loss": 0.7393,
-      "step": 2750
-    },
-    {
-      "epoch": 1.3510487789875132,
-      "grad_norm": 0.46702662801495093,
-      "learning_rate": 0.00014729233552247906,
-      "loss": 0.7982,
-      "step": 2751
-    },
-    {
-      "epoch": 1.3515408747001292,
-      "grad_norm": 0.4687994252429321,
-      "learning_rate": 0.00014723807198365033,
-      "loss": 0.7859,
-      "step": 2752
-    },
-    {
-      "epoch": 1.3520329704127452,
-      "grad_norm": 0.40948468726502346,
-      "learning_rate": 0.00014718379053418643,
-      "loss": 0.739,
-      "step": 2753
-    },
-    {
-      "epoch": 1.3525250661253614,
-      "grad_norm": 0.45155120830076073,
-      "learning_rate": 0.0001471294911946686,
-      "loss": 0.773,
-      "step": 2754
-    },
-    {
-      "epoch": 1.3530171618379776,
-      "grad_norm": 0.43327535587765165,
-      "learning_rate": 0.00014707517398568478,
-      "loss": 0.7334,
-      "step": 2755
-    },
-    {
-      "epoch": 1.3535092575505936,
-      "grad_norm": 0.423007154634038,
-      "learning_rate": 0.0001470208389278297,
-      "loss": 0.724,
-      "step": 2756
-    },
-    {
-      "epoch": 1.3540013532632096,
-      "grad_norm": 0.4392988409483355,
-      "learning_rate": 0.00014696648604170488,
-      "loss": 0.7741,
-      "step": 2757
-    },
-    {
-      "epoch": 1.3544934489758258,
-      "grad_norm": 0.429697872741243,
-      "learning_rate": 0.00014691211534791857,
-      "loss": 0.7331,
-      "step": 2758
-    },
-    {
-      "epoch": 1.354985544688442,
-      "grad_norm": 0.463454284334484,
-      "learning_rate": 0.00014685772686708577,
-      "loss": 0.705,
-      "step": 2759
-    },
-    {
-      "epoch": 1.355477640401058,
-      "grad_norm": 0.40727469892993934,
-      "learning_rate": 0.0001468033206198283,
-      "loss": 0.7576,
-      "step": 2760
-    },
-    {
-      "epoch": 1.355969736113674,
-      "grad_norm": 0.4446907558160837,
-      "learning_rate": 0.00014674889662677463,
-      "loss": 0.7904,
-      "step": 2761
-    },
-    {
-      "epoch": 1.3564618318262902,
-      "grad_norm": 0.44088645970811163,
-      "learning_rate": 0.00014669445490855996,
-      "loss": 0.7535,
-      "step": 2762
-    },
-    {
-      "epoch": 1.3569539275389064,
-      "grad_norm": 0.41702807251086554,
-      "learning_rate": 0.00014663999548582623,
-      "loss": 0.7527,
-      "step": 2763
-    },
-    {
-      "epoch": 1.3574460232515224,
-      "grad_norm": 0.4401593184803354,
-      "learning_rate": 0.00014658551837922214,
-      "loss": 0.7713,
-      "step": 2764
-    },
-    {
-      "epoch": 1.3579381189641384,
-      "grad_norm": 0.41800494517405506,
-      "learning_rate": 0.00014653102360940304,
-      "loss": 0.767,
-      "step": 2765
-    },
-    {
-      "epoch": 1.3584302146767546,
-      "grad_norm": 0.4213390147614005,
-      "learning_rate": 0.00014647651119703098,
-      "loss": 0.7639,
-      "step": 2766
-    },
-    {
-      "epoch": 1.3589223103893708,
-      "grad_norm": 0.4264935640399931,
-      "learning_rate": 0.00014642198116277477,
-      "loss": 0.753,
-      "step": 2767
-    },
-    {
-      "epoch": 1.3594144061019868,
-      "grad_norm": 0.39343467342051713,
-      "learning_rate": 0.00014636743352730976,
-      "loss": 0.7162,
-      "step": 2768
-    },
-    {
-      "epoch": 1.3599065018146028,
-      "grad_norm": 0.4308535571127244,
-      "learning_rate": 0.00014631286831131806,
-      "loss": 0.7264,
-      "step": 2769
-    },
-    {
-      "epoch": 1.360398597527219,
-      "grad_norm": 0.4149011577472118,
-      "learning_rate": 0.00014625828553548853,
-      "loss": 0.6912,
-      "step": 2770
-    },
-    {
-      "epoch": 1.3608906932398352,
-      "grad_norm": 0.4099517014433799,
-      "learning_rate": 0.0001462036852205165,
-      "loss": 0.7752,
-      "step": 2771
-    },
-    {
-      "epoch": 1.3613827889524512,
-      "grad_norm": 0.4092203740757273,
-      "learning_rate": 0.00014614906738710408,
-      "loss": 0.7293,
-      "step": 2772
-    },
-    {
-      "epoch": 1.3618748846650672,
-      "grad_norm": 0.4549383754358427,
-      "learning_rate": 0.00014609443205596,
-      "loss": 0.7842,
-      "step": 2773
-    },
-    {
-      "epoch": 1.3623669803776834,
-      "grad_norm": 0.4306240858525313,
-      "learning_rate": 0.00014603977924779963,
-      "loss": 0.7812,
-      "step": 2774
-    },
-    {
-      "epoch": 1.3628590760902997,
-      "grad_norm": 0.45801146218745514,
-      "learning_rate": 0.00014598510898334498,
-      "loss": 0.7374,
-      "step": 2775
-    },
-    {
-      "epoch": 1.3633511718029157,
-      "grad_norm": 0.43859349099215705,
-      "learning_rate": 0.00014593042128332453,
-      "loss": 0.6934,
-      "step": 2776
-    },
-    {
-      "epoch": 1.3638432675155316,
-      "grad_norm": 0.4542107179550836,
-      "learning_rate": 0.00014587571616847363,
-      "loss": 0.7173,
-      "step": 2777
-    },
-    {
-      "epoch": 1.3643353632281479,
-      "grad_norm": 0.44054827069162666,
-      "learning_rate": 0.00014582099365953398,
-      "loss": 0.7642,
-      "step": 2778
-    },
-    {
-      "epoch": 1.364827458940764,
-      "grad_norm": 0.415389898067285,
-      "learning_rate": 0.0001457662537772541,
-      "loss": 0.7136,
-      "step": 2779
-    },
-    {
-      "epoch": 1.36531955465338,
-      "grad_norm": 0.4508678812762311,
-      "learning_rate": 0.0001457114965423889,
-      "loss": 0.679,
-      "step": 2780
-    },
-    {
-      "epoch": 1.365811650365996,
-      "grad_norm": 0.4502740150907887,
-      "learning_rate": 0.00014565672197570004,
-      "loss": 0.7419,
-      "step": 2781
-    },
-    {
-      "epoch": 1.3663037460786123,
-      "grad_norm": 0.42787059612739786,
-      "learning_rate": 0.00014560193009795555,
-      "loss": 0.6888,
-      "step": 2782
-    },
-    {
-      "epoch": 1.3667958417912285,
-      "grad_norm": 0.4194007248160413,
-      "learning_rate": 0.00014554712092993026,
-      "loss": 0.7562,
-      "step": 2783
-    },
-    {
-      "epoch": 1.3672879375038445,
-      "grad_norm": 0.46160780550536346,
-      "learning_rate": 0.0001454922944924054,
-      "loss": 0.7636,
-      "step": 2784
-    },
-    {
-      "epoch": 1.3677800332164605,
-      "grad_norm": 0.5217135724226263,
-      "learning_rate": 0.00014543745080616876,
-      "loss": 0.7046,
-      "step": 2785
-    },
-    {
-      "epoch": 1.3682721289290767,
-      "grad_norm": 2.621954437128583,
-      "learning_rate": 0.00014538258989201466,
-      "loss": 0.7991,
-      "step": 2786
-    },
-    {
-      "epoch": 1.368764224641693,
-      "grad_norm": 0.4167833300901098,
-      "learning_rate": 0.0001453277117707441,
-      "loss": 0.7142,
-      "step": 2787
-    },
-    {
-      "epoch": 1.3692563203543089,
-      "grad_norm": 0.423599670071847,
-      "learning_rate": 0.00014527281646316438,
-      "loss": 0.7201,
-      "step": 2788
-    },
-    {
-      "epoch": 1.3697484160669249,
-      "grad_norm": 0.4360138295144438,
-      "learning_rate": 0.0001452179039900895,
-      "loss": 0.7964,
-      "step": 2789
-    },
-    {
-      "epoch": 1.370240511779541,
-      "grad_norm": 0.39593837473454535,
-      "learning_rate": 0.00014516297437233987,
-      "loss": 0.6846,
-      "step": 2790
-    },
-    {
-      "epoch": 1.3707326074921573,
-      "grad_norm": 0.39349976449439356,
-      "learning_rate": 0.00014510802763074241,
-      "loss": 0.7268,
-      "step": 2791
-    },
-    {
-      "epoch": 1.3712247032047733,
-      "grad_norm": 0.46673403957282444,
-      "learning_rate": 0.00014505306378613062,
-      "loss": 0.7293,
-      "step": 2792
-    },
-    {
-      "epoch": 1.3717167989173895,
-      "grad_norm": 0.4306549813980889,
-      "learning_rate": 0.00014499808285934433,
-      "loss": 0.7627,
-      "step": 2793
-    },
-    {
-      "epoch": 1.3722088946300055,
-      "grad_norm": 0.426697419320064,
-      "learning_rate": 0.00014494308487123,
-      "loss": 0.7223,
-      "step": 2794
-    },
-    {
-      "epoch": 1.3727009903426217,
-      "grad_norm": 0.42243755494952273,
-      "learning_rate": 0.00014488806984264038,
-      "loss": 0.7061,
-      "step": 2795
-    },
-    {
-      "epoch": 1.3731930860552377,
-      "grad_norm": 0.42098719038053645,
-      "learning_rate": 0.0001448330377944349,
-      "loss": 0.74,
-      "step": 2796
-    },
-    {
-      "epoch": 1.373685181767854,
-      "grad_norm": 0.4381137328460556,
-      "learning_rate": 0.00014477798874747933,
-      "loss": 0.6986,
-      "step": 2797
-    },
-    {
-      "epoch": 1.37417727748047,
-      "grad_norm": 0.4377979248161887,
-      "learning_rate": 0.00014472292272264584,
-      "loss": 0.6892,
-      "step": 2798
-    },
-    {
-      "epoch": 1.3746693731930861,
-      "grad_norm": 0.4628397260824491,
-      "learning_rate": 0.0001446678397408131,
-      "loss": 0.8194,
-      "step": 2799
-    },
-    {
-      "epoch": 1.3751614689057021,
-      "grad_norm": 0.4105910026261328,
-      "learning_rate": 0.00014461273982286618,
-      "loss": 0.7701,
-      "step": 2800
-    },
-    {
-      "epoch": 1.3756535646183183,
-      "grad_norm": 0.4052498504584658,
-      "learning_rate": 0.00014455762298969663,
-      "loss": 0.7477,
-      "step": 2801
-    },
-    {
-      "epoch": 1.3761456603309343,
-      "grad_norm": 0.4301997582693896,
-      "learning_rate": 0.00014450248926220236,
-      "loss": 0.7358,
-      "step": 2802
-    },
-    {
-      "epoch": 1.3766377560435505,
-      "grad_norm": 0.40260835089540903,
-      "learning_rate": 0.00014444733866128765,
-      "loss": 0.7107,
-      "step": 2803
-    },
-    {
-      "epoch": 1.3771298517561665,
-      "grad_norm": 0.4319017994896772,
-      "learning_rate": 0.0001443921712078633,
-      "loss": 0.7418,
-      "step": 2804
-    },
-    {
-      "epoch": 1.3776219474687827,
-      "grad_norm": 0.4364926584592031,
-      "learning_rate": 0.00014433698692284635,
-      "loss": 0.7856,
-      "step": 2805
-    },
-    {
-      "epoch": 1.3781140431813987,
-      "grad_norm": 0.4118175977294697,
-      "learning_rate": 0.00014428178582716035,
-      "loss": 0.7492,
-      "step": 2806
-    },
-    {
-      "epoch": 1.378606138894015,
-      "grad_norm": 0.4102828788466361,
-      "learning_rate": 0.00014422656794173513,
-      "loss": 0.7743,
-      "step": 2807
-    },
-    {
-      "epoch": 1.379098234606631,
-      "grad_norm": 0.429336876318765,
-      "learning_rate": 0.00014417133328750693,
-      "loss": 0.8042,
-      "step": 2808
-    },
-    {
-      "epoch": 1.3795903303192472,
-      "grad_norm": 0.40914285687385166,
-      "learning_rate": 0.00014411608188541838,
-      "loss": 0.7258,
-      "step": 2809
-    },
-    {
-      "epoch": 1.3800824260318632,
-      "grad_norm": 0.4592385228100208,
-      "learning_rate": 0.0001440608137564184,
-      "loss": 0.7655,
-      "step": 2810
-    },
-    {
-      "epoch": 1.3805745217444794,
-      "grad_norm": 0.40524311159529497,
-      "learning_rate": 0.0001440055289214623,
-      "loss": 0.7451,
-      "step": 2811
-    },
-    {
-      "epoch": 1.3810666174570954,
-      "grad_norm": 0.42070709054702543,
-      "learning_rate": 0.00014395022740151163,
-      "loss": 0.7721,
-      "step": 2812
-    },
-    {
-      "epoch": 1.3815587131697116,
-      "grad_norm": 0.4398688206824279,
-      "learning_rate": 0.0001438949092175344,
-      "loss": 0.7318,
-      "step": 2813
-    },
-    {
-      "epoch": 1.3820508088823276,
-      "grad_norm": 0.46987366202392017,
-      "learning_rate": 0.00014383957439050485,
-      "loss": 0.7147,
-      "step": 2814
-    },
-    {
-      "epoch": 1.3825429045949438,
-      "grad_norm": 0.44418643849007833,
-      "learning_rate": 0.00014378422294140358,
-      "loss": 0.7462,
-      "step": 2815
-    },
-    {
-      "epoch": 1.3830350003075598,
-      "grad_norm": 0.4402508658792176,
-      "learning_rate": 0.00014372885489121744,
-      "loss": 0.7566,
-      "step": 2816
-    },
-    {
-      "epoch": 1.383527096020176,
-      "grad_norm": 0.46917452097201523,
-      "learning_rate": 0.00014367347026093965,
-      "loss": 0.8237,
-      "step": 2817
-    },
-    {
-      "epoch": 1.384019191732792,
-      "grad_norm": 0.4308436511988418,
-      "learning_rate": 0.00014361806907156957,
-      "loss": 0.7602,
-      "step": 2818
-    },
-    {
-      "epoch": 1.3845112874454082,
-      "grad_norm": 0.43453500343053675,
-      "learning_rate": 0.000143562651344113,
-      "loss": 0.7695,
-      "step": 2819
-    },
-    {
-      "epoch": 1.3850033831580242,
-      "grad_norm": 0.40580720584426183,
-      "learning_rate": 0.00014350721709958196,
-      "loss": 0.7215,
-      "step": 2820
-    },
-    {
-      "epoch": 1.3854954788706404,
-      "grad_norm": 0.4304814434001006,
-      "learning_rate": 0.00014345176635899466,
-      "loss": 0.7581,
-      "step": 2821
-    },
-    {
-      "epoch": 1.3859875745832564,
-      "grad_norm": 0.43924056411036916,
-      "learning_rate": 0.00014339629914337571,
-      "loss": 0.7778,
-      "step": 2822
-    },
-    {
-      "epoch": 1.3864796702958726,
-      "grad_norm": 0.4318623964899569,
-      "learning_rate": 0.00014334081547375584,
-      "loss": 0.773,
-      "step": 2823
-    },
-    {
-      "epoch": 1.3869717660084886,
-      "grad_norm": 0.415662725821242,
-      "learning_rate": 0.00014328531537117204,
-      "loss": 0.7721,
-      "step": 2824
-    },
-    {
-      "epoch": 1.3874638617211048,
-      "grad_norm": 0.4232390291083941,
-      "learning_rate": 0.00014322979885666756,
-      "loss": 0.7137,
-      "step": 2825
-    },
-    {
-      "epoch": 1.3879559574337208,
-      "grad_norm": 0.43954365193076456,
-      "learning_rate": 0.0001431742659512919,
-      "loss": 0.7975,
-      "step": 2826
-    },
-    {
-      "epoch": 1.388448053146337,
-      "grad_norm": 0.42311931954421095,
-      "learning_rate": 0.00014311871667610067,
-      "loss": 0.703,
-      "step": 2827
-    },
-    {
-      "epoch": 1.388940148858953,
-      "grad_norm": 0.40842248284601124,
-      "learning_rate": 0.00014306315105215578,
-      "loss": 0.7509,
-      "step": 2828
-    },
-    {
-      "epoch": 1.3894322445715692,
-      "grad_norm": 0.417261173584133,
-      "learning_rate": 0.00014300756910052534,
-      "loss": 0.6907,
-      "step": 2829
-    },
-    {
-      "epoch": 1.3899243402841852,
-      "grad_norm": 0.4056303091245692,
-      "learning_rate": 0.0001429519708422836,
-      "loss": 0.756,
-      "step": 2830
-    },
-    {
-      "epoch": 1.3904164359968014,
-      "grad_norm": 0.4087623746617413,
-      "learning_rate": 0.000142896356298511,
-      "loss": 0.6662,
-      "step": 2831
-    },
-    {
-      "epoch": 1.3909085317094174,
-      "grad_norm": 0.41966032125022756,
-      "learning_rate": 0.00014284072549029423,
-      "loss": 0.7825,
-      "step": 2832
-    },
-    {
-      "epoch": 1.3914006274220336,
-      "grad_norm": 0.40466274275782793,
-      "learning_rate": 0.00014278507843872604,
-      "loss": 0.7306,
-      "step": 2833
-    },
-    {
-      "epoch": 1.3918927231346496,
-      "grad_norm": 0.46833029296545736,
-      "learning_rate": 0.0001427294151649054,
-      "loss": 0.8599,
-      "step": 2834
-    },
-    {
-      "epoch": 1.3923848188472658,
-      "grad_norm": 0.4247522298363245,
-      "learning_rate": 0.00014267373568993742,
-      "loss": 0.7241,
-      "step": 2835
-    },
-    {
-      "epoch": 1.3928769145598818,
-      "grad_norm": 0.4720901230149452,
-      "learning_rate": 0.00014261804003493333,
-      "loss": 0.7737,
-      "step": 2836
-    },
-    {
-      "epoch": 1.393369010272498,
-      "grad_norm": 0.47239587469693006,
-      "learning_rate": 0.00014256232822101053,
-      "loss": 0.7604,
-      "step": 2837
-    },
-    {
-      "epoch": 1.393861105985114,
-      "grad_norm": 0.48548609967352985,
-      "learning_rate": 0.00014250660026929256,
-      "loss": 0.7775,
-      "step": 2838
-    },
-    {
-      "epoch": 1.3943532016977302,
-      "grad_norm": 0.4443374125802025,
-      "learning_rate": 0.00014245085620090902,
-      "loss": 0.7482,
-      "step": 2839
-    },
-    {
-      "epoch": 1.3948452974103462,
-      "grad_norm": 0.4111894277871565,
-      "learning_rate": 0.00014239509603699562,
-      "loss": 0.7135,
-      "step": 2840
-    },
-    {
-      "epoch": 1.3953373931229625,
-      "grad_norm": 0.44508601998564107,
-      "learning_rate": 0.00014233931979869426,
-      "loss": 0.8062,
-      "step": 2841
-    },
-    {
-      "epoch": 1.3958294888355784,
-      "grad_norm": 0.43154416020043423,
-      "learning_rate": 0.00014228352750715286,
-      "loss": 0.7587,
-      "step": 2842
-    },
-    {
-      "epoch": 1.3963215845481947,
-      "grad_norm": 0.48828566395571754,
-      "learning_rate": 0.00014222771918352543,
-      "loss": 0.6708,
-      "step": 2843
-    },
-    {
-      "epoch": 1.3968136802608107,
-      "grad_norm": 0.4353818605607434,
-      "learning_rate": 0.0001421718948489721,
-      "loss": 0.7412,
-      "step": 2844
-    },
-    {
-      "epoch": 1.3973057759734269,
-      "grad_norm": 0.42988995848635425,
-      "learning_rate": 0.000142116054524659,
-      "loss": 0.6936,
-      "step": 2845
-    },
-    {
-      "epoch": 1.3977978716860429,
-      "grad_norm": 0.4167451646577703,
-      "learning_rate": 0.00014206019823175843,
-      "loss": 0.7439,
-      "step": 2846
-    },
-    {
-      "epoch": 1.398289967398659,
-      "grad_norm": 0.4194091490877988,
-      "learning_rate": 0.00014200432599144867,
-      "loss": 0.7945,
-      "step": 2847
-    },
-    {
-      "epoch": 1.398782063111275,
-      "grad_norm": 0.40440317420121163,
-      "learning_rate": 0.00014194843782491402,
-      "loss": 0.7532,
-      "step": 2848
-    },
-    {
-      "epoch": 1.3992741588238913,
-      "grad_norm": 0.42397793268354184,
-      "learning_rate": 0.00014189253375334487,
-      "loss": 0.6868,
-      "step": 2849
-    },
-    {
-      "epoch": 1.3997662545365073,
-      "grad_norm": 0.4348167217634709,
-      "learning_rate": 0.00014183661379793764,
-      "loss": 0.7858,
-      "step": 2850
-    },
-    {
-      "epoch": 1.4002583502491235,
-      "grad_norm": 0.39554208001514635,
-      "learning_rate": 0.00014178067797989473,
-      "loss": 0.7384,
-      "step": 2851
-    },
-    {
-      "epoch": 1.4007504459617395,
-      "grad_norm": 0.42910325009244044,
-      "learning_rate": 0.00014172472632042465,
-      "loss": 0.7856,
-      "step": 2852
-    },
-    {
-      "epoch": 1.4012425416743557,
-      "grad_norm": 0.40517337832430766,
-      "learning_rate": 0.0001416687588407418,
-      "loss": 0.7217,
-      "step": 2853
-    },
-    {
-      "epoch": 1.401734637386972,
-      "grad_norm": 0.4348992153207263,
-      "learning_rate": 0.0001416127755620666,
-      "loss": 0.7943,
-      "step": 2854
-    },
-    {
-      "epoch": 1.402226733099588,
-      "grad_norm": 0.4120607437044399,
-      "learning_rate": 0.00014155677650562554,
-      "loss": 0.7881,
-      "step": 2855
-    },
-    {
-      "epoch": 1.4027188288122039,
-      "grad_norm": 0.4208840397746761,
-      "learning_rate": 0.00014150076169265106,
-      "loss": 0.7014,
-      "step": 2856
-    },
-    {
-      "epoch": 1.40321092452482,
-      "grad_norm": 0.44500297874848843,
-      "learning_rate": 0.0001414447311443815,
-      "loss": 0.7544,
-      "step": 2857
-    },
-    {
-      "epoch": 1.4037030202374363,
-      "grad_norm": 0.4518710615369706,
-      "learning_rate": 0.00014138868488206127,
-      "loss": 0.6722,
-      "step": 2858
-    },
-    {
-      "epoch": 1.4041951159500523,
-      "grad_norm": 0.4579671823678311,
-      "learning_rate": 0.00014133262292694067,
-      "loss": 0.7084,
-      "step": 2859
-    },
-    {
-      "epoch": 1.4046872116626683,
-      "grad_norm": 0.46593325559466814,
-      "learning_rate": 0.00014127654530027596,
-      "loss": 0.7568,
-      "step": 2860
-    },
-    {
-      "epoch": 1.4051793073752845,
-      "grad_norm": 0.42065468000896555,
-      "learning_rate": 0.0001412204520233294,
-      "loss": 0.7637,
-      "step": 2861
-    },
-    {
-      "epoch": 1.4056714030879007,
-      "grad_norm": 0.4351068289757589,
-      "learning_rate": 0.00014116434311736904,
-      "loss": 0.7539,
-      "step": 2862
-    },
-    {
-      "epoch": 1.4061634988005167,
-      "grad_norm": 0.4371758917451206,
-      "learning_rate": 0.00014110821860366903,
-      "loss": 0.7676,
-      "step": 2863
-    },
-    {
-      "epoch": 1.4066555945131327,
-      "grad_norm": 0.4219230357329842,
-      "learning_rate": 0.00014105207850350932,
-      "loss": 0.838,
-      "step": 2864
-    },
-    {
-      "epoch": 1.407147690225749,
-      "grad_norm": 0.4216272512315438,
-      "learning_rate": 0.00014099592283817585,
-      "loss": 0.7818,
-      "step": 2865
-    },
-    {
-      "epoch": 1.4076397859383651,
-      "grad_norm": 0.4396059343719567,
-      "learning_rate": 0.00014093975162896038,
-      "loss": 0.7467,
-      "step": 2866
-    },
-    {
-      "epoch": 1.4081318816509811,
-      "grad_norm": 0.42006314652049326,
-      "learning_rate": 0.00014088356489716064,
-      "loss": 0.7703,
-      "step": 2867
-    },
-    {
-      "epoch": 1.4086239773635971,
-      "grad_norm": 0.4224713073924144,
-      "learning_rate": 0.0001408273626640802,
-      "loss": 0.7167,
-      "step": 2868
-    },
-    {
-      "epoch": 1.4091160730762133,
-      "grad_norm": 0.4343128075983513,
-      "learning_rate": 0.00014077114495102849,
-      "loss": 0.7962,
-      "step": 2869
-    },
-    {
-      "epoch": 1.4096081687888296,
-      "grad_norm": 0.40606316906369183,
-      "learning_rate": 0.0001407149117793209,
-      "loss": 0.694,
-      "step": 2870
-    },
-    {
-      "epoch": 1.4101002645014455,
-      "grad_norm": 0.4508075061782764,
-      "learning_rate": 0.00014065866317027853,
-      "loss": 0.8228,
-      "step": 2871
-    },
-    {
-      "epoch": 1.4105923602140615,
-      "grad_norm": 0.4070111815672579,
-      "learning_rate": 0.0001406023991452285,
-      "loss": 0.7674,
-      "step": 2872
-    },
-    {
-      "epoch": 1.4110844559266777,
-      "grad_norm": 0.4171067618610642,
-      "learning_rate": 0.00014054611972550365,
-      "loss": 0.7455,
-      "step": 2873
-    },
-    {
-      "epoch": 1.411576551639294,
-      "grad_norm": 0.43382133531872263,
-      "learning_rate": 0.00014048982493244268,
-      "loss": 0.7755,
-      "step": 2874
-    },
-    {
-      "epoch": 1.41206864735191,
-      "grad_norm": 0.4363850746655965,
-      "learning_rate": 0.00014043351478739024,
-      "loss": 0.743,
-      "step": 2875
-    },
-    {
-      "epoch": 1.412560743064526,
-      "grad_norm": 0.4360042913510568,
-      "learning_rate": 0.00014037718931169662,
-      "loss": 0.7641,
-      "step": 2876
-    },
-    {
-      "epoch": 1.4130528387771422,
-      "grad_norm": 0.45730795402467167,
-      "learning_rate": 0.00014032084852671803,
-      "loss": 0.7505,
-      "step": 2877
-    },
-    {
-      "epoch": 1.4135449344897584,
-      "grad_norm": 0.4022597947003752,
-      "learning_rate": 0.00014026449245381647,
-      "loss": 0.6883,
-      "step": 2878
-    },
-    {
-      "epoch": 1.4140370302023744,
-      "grad_norm": 0.44552551333264356,
-      "learning_rate": 0.0001402081211143597,
-      "loss": 0.717,
-      "step": 2879
-    },
-    {
-      "epoch": 1.4145291259149904,
-      "grad_norm": 0.47377949637491557,
-      "learning_rate": 0.0001401517345297213,
-      "loss": 0.7866,
-      "step": 2880
-    },
-    {
-      "epoch": 1.4150212216276066,
-      "grad_norm": 0.4371521549322124,
-      "learning_rate": 0.00014009533272128067,
-      "loss": 0.7803,
-      "step": 2881
-    },
-    {
-      "epoch": 1.4155133173402228,
-      "grad_norm": 0.4200894722489111,
-      "learning_rate": 0.0001400389157104229,
-      "loss": 0.6691,
-      "step": 2882
-    },
-    {
-      "epoch": 1.4160054130528388,
-      "grad_norm": 0.4165235110057634,
-      "learning_rate": 0.0001399824835185389,
-      "loss": 0.746,
-      "step": 2883
-    },
-    {
-      "epoch": 1.4164975087654548,
-      "grad_norm": 0.431662953834368,
-      "learning_rate": 0.00013992603616702525,
-      "loss": 0.6527,
-      "step": 2884
-    },
-    {
-      "epoch": 1.416989604478071,
-      "grad_norm": 0.43850081677439445,
-      "learning_rate": 0.00013986957367728445,
-      "loss": 0.8324,
-      "step": 2885
-    },
-    {
-      "epoch": 1.4174817001906872,
-      "grad_norm": 0.4187732141177877,
-      "learning_rate": 0.0001398130960707246,
-      "loss": 0.7149,
-      "step": 2886
-    },
-    {
-      "epoch": 1.4179737959033032,
-      "grad_norm": 0.40634513468351635,
-      "learning_rate": 0.0001397566033687595,
-      "loss": 0.6807,
-      "step": 2887
-    },
-    {
-      "epoch": 1.4184658916159192,
-      "grad_norm": 0.4583126497511633,
-      "learning_rate": 0.00013970009559280882,
-      "loss": 0.7505,
-      "step": 2888
-    },
-    {
-      "epoch": 1.4189579873285354,
-      "grad_norm": 0.4891573664815128,
-      "learning_rate": 0.00013964357276429788,
-      "loss": 0.7477,
-      "step": 2889
-    },
-    {
-      "epoch": 1.4194500830411516,
-      "grad_norm": 0.4137821213859619,
-      "learning_rate": 0.00013958703490465758,
-      "loss": 0.7773,
-      "step": 2890
-    },
-    {
-      "epoch": 1.4199421787537676,
-      "grad_norm": 0.4460968509297009,
-      "learning_rate": 0.00013953048203532476,
-      "loss": 0.6913,
-      "step": 2891
-    },
-    {
-      "epoch": 1.4204342744663836,
-      "grad_norm": 0.4005465440971991,
-      "learning_rate": 0.00013947391417774176,
-      "loss": 0.7472,
-      "step": 2892
-    },
-    {
-      "epoch": 1.4209263701789998,
-      "grad_norm": 0.44251907002578933,
-      "learning_rate": 0.0001394173313533567,
-      "loss": 0.7393,
-      "step": 2893
-    },
-    {
-      "epoch": 1.421418465891616,
-      "grad_norm": 0.41224985680157783,
-      "learning_rate": 0.00013936073358362328,
-      "loss": 0.7421,
-      "step": 2894
-    },
-    {
-      "epoch": 1.421910561604232,
-      "grad_norm": 0.42618771413527917,
-      "learning_rate": 0.00013930412089000098,
-      "loss": 0.7587,
-      "step": 2895
-    },
-    {
-      "epoch": 1.422402657316848,
-      "grad_norm": 0.45540669872859213,
-      "learning_rate": 0.00013924749329395487,
-      "loss": 0.8174,
-      "step": 2896
-    },
-    {
-      "epoch": 1.4228947530294642,
-      "grad_norm": 0.4359599281544839,
-      "learning_rate": 0.0001391908508169557,
-      "loss": 0.7311,
-      "step": 2897
-    },
-    {
-      "epoch": 1.4233868487420804,
-      "grad_norm": 0.43010465875363985,
-      "learning_rate": 0.00013913419348047983,
-      "loss": 0.7768,
-      "step": 2898
-    },
-    {
-      "epoch": 1.4238789444546964,
-      "grad_norm": 0.435110098412238,
-      "learning_rate": 0.0001390775213060093,
-      "loss": 0.7449,
-      "step": 2899
-    },
-    {
-      "epoch": 1.4243710401673124,
-      "grad_norm": 0.40177833018764403,
-      "learning_rate": 0.0001390208343150317,
-      "loss": 0.7227,
-      "step": 2900
-    },
-    {
-      "epoch": 1.4248631358799286,
-      "grad_norm": 0.8512405579703662,
-      "learning_rate": 0.00013896413252904036,
-      "loss": 0.7107,
-      "step": 2901
-    },
-    {
-      "epoch": 1.4253552315925448,
-      "grad_norm": 0.403879588650031,
-      "learning_rate": 0.00013890741596953406,
-      "loss": 0.712,
-      "step": 2902
-    },
-    {
-      "epoch": 1.4258473273051608,
-      "grad_norm": 0.3964710248800138,
-      "learning_rate": 0.00013885068465801733,
-      "loss": 0.6513,
-      "step": 2903
-    },
-    {
-      "epoch": 1.4263394230177768,
-      "grad_norm": 0.4318649470861009,
-      "learning_rate": 0.00013879393861600023,
-      "loss": 0.7263,
-      "step": 2904
-    },
-    {
-      "epoch": 1.426831518730393,
-      "grad_norm": 0.442240871740813,
-      "learning_rate": 0.0001387371778649984,
-      "loss": 0.7094,
-      "step": 2905
-    },
-    {
-      "epoch": 1.4273236144430093,
-      "grad_norm": 0.4552714815809044,
-      "learning_rate": 0.000138680402426533,
-      "loss": 0.7924,
-      "step": 2906
-    },
-    {
-      "epoch": 1.4278157101556253,
-      "grad_norm": 0.481478122154512,
-      "learning_rate": 0.00013862361232213095,
-      "loss": 0.7331,
-      "step": 2907
-    },
-    {
-      "epoch": 1.4283078058682412,
-      "grad_norm": 0.42710383729499984,
-      "learning_rate": 0.0001385668075733245,
-      "loss": 0.7276,
-      "step": 2908
-    },
-    {
-      "epoch": 1.4287999015808575,
-      "grad_norm": 0.4280297934009003,
-      "learning_rate": 0.00013850998820165157,
-      "loss": 0.7365,
-      "step": 2909
-    },
-    {
-      "epoch": 1.4292919972934737,
-      "grad_norm": 0.46622341113853055,
-      "learning_rate": 0.00013845315422865561,
-      "loss": 0.8161,
-      "step": 2910
-    },
-    {
-      "epoch": 1.4297840930060897,
-      "grad_norm": 0.46592536759745623,
-      "learning_rate": 0.00013839630567588564,
-      "loss": 0.7298,
-      "step": 2911
-    },
-    {
-      "epoch": 1.4302761887187057,
-      "grad_norm": 0.49640539779225473,
-      "learning_rate": 0.00013833944256489615,
-      "loss": 0.7563,
-      "step": 2912
-    },
-    {
-      "epoch": 1.4307682844313219,
-      "grad_norm": 0.45185106365996336,
-      "learning_rate": 0.0001382825649172472,
-      "loss": 0.8016,
-      "step": 2913
-    },
-    {
-      "epoch": 1.431260380143938,
-      "grad_norm": 0.4428266892150101,
-      "learning_rate": 0.00013822567275450427,
-      "loss": 0.794,
-      "step": 2914
-    },
-    {
-      "epoch": 1.431752475856554,
-      "grad_norm": 0.45759508801690096,
-      "learning_rate": 0.00013816876609823849,
-      "loss": 0.765,
-      "step": 2915
-    },
-    {
-      "epoch": 1.43224457156917,
-      "grad_norm": 0.544503508687506,
-      "learning_rate": 0.00013811184497002635,
-      "loss": 0.7612,
-      "step": 2916
-    },
-    {
-      "epoch": 1.4327366672817863,
-      "grad_norm": 0.42121930764428955,
-      "learning_rate": 0.00013805490939144992,
-      "loss": 0.7171,
-      "step": 2917
-    },
-    {
-      "epoch": 1.4332287629944025,
-      "grad_norm": 0.4381253918038311,
-      "learning_rate": 0.0001379979593840967,
-      "loss": 0.7915,
-      "step": 2918
-    },
-    {
-      "epoch": 1.4337208587070185,
-      "grad_norm": 0.4145705065574621,
-      "learning_rate": 0.0001379409949695596,
-      "loss": 0.7264,
-      "step": 2919
-    },
-    {
-      "epoch": 1.4342129544196347,
-      "grad_norm": 0.42024556353558573,
-      "learning_rate": 0.00013788401616943716,
-      "loss": 0.6993,
-      "step": 2920
-    },
-    {
-      "epoch": 1.4347050501322507,
-      "grad_norm": 0.43110288012934284,
-      "learning_rate": 0.00013782702300533324,
-      "loss": 0.7126,
-      "step": 2921
-    },
-    {
-      "epoch": 1.435197145844867,
-      "grad_norm": 0.4349493312847473,
-      "learning_rate": 0.0001377700154988572,
-      "loss": 0.7444,
-      "step": 2922
-    },
-    {
-      "epoch": 1.435689241557483,
-      "grad_norm": 0.41663983120576736,
-      "learning_rate": 0.00013771299367162378,
-      "loss": 0.7253,
-      "step": 2923
-    },
-    {
-      "epoch": 1.4361813372700991,
-      "grad_norm": 0.4303844427833147,
-      "learning_rate": 0.00013765595754525325,
-      "loss": 0.7297,
-      "step": 2924
-    },
-    {
-      "epoch": 1.436673432982715,
-      "grad_norm": 0.48581012699068177,
-      "learning_rate": 0.0001375989071413712,
-      "loss": 0.7889,
-      "step": 2925
-    },
-    {
-      "epoch": 1.4371655286953313,
-      "grad_norm": 0.4284137641486597,
-      "learning_rate": 0.00013754184248160868,
-      "loss": 0.7166,
-      "step": 2926
-    },
-    {
-      "epoch": 1.4376576244079473,
-      "grad_norm": 0.7831283240149731,
-      "learning_rate": 0.0001374847635876022,
-      "loss": 0.8392,
-      "step": 2927
-    },
-    {
-      "epoch": 1.4381497201205635,
-      "grad_norm": 0.4580485811164869,
-      "learning_rate": 0.00013742767048099353,
-      "loss": 0.7546,
-      "step": 2928
-    },
-    {
-      "epoch": 1.4386418158331795,
-      "grad_norm": 0.42604106634937855,
-      "learning_rate": 0.00013737056318342995,
-      "loss": 0.7532,
-      "step": 2929
-    },
-    {
-      "epoch": 1.4391339115457957,
-      "grad_norm": 0.4308574750819755,
-      "learning_rate": 0.0001373134417165641,
-      "loss": 0.6814,
-      "step": 2930
-    },
-    {
-      "epoch": 1.4396260072584117,
-      "grad_norm": 0.5332424690162715,
-      "learning_rate": 0.0001372563061020539,
-      "loss": 0.8342,
-      "step": 2931
-    },
-    {
-      "epoch": 1.440118102971028,
-      "grad_norm": 0.46080850919005695,
-      "learning_rate": 0.00013719915636156276,
-      "loss": 0.7292,
-      "step": 2932
-    },
-    {
-      "epoch": 1.440610198683644,
-      "grad_norm": 0.4261657382522858,
-      "learning_rate": 0.0001371419925167594,
-      "loss": 0.7263,
-      "step": 2933
-    },
-    {
-      "epoch": 1.4411022943962601,
-      "grad_norm": 0.4248726170760586,
-      "learning_rate": 0.00013708481458931784,
-      "loss": 0.7546,
-      "step": 2934
-    },
-    {
-      "epoch": 1.4415943901088761,
-      "grad_norm": 0.4410965585873861,
-      "learning_rate": 0.0001370276226009175,
-      "loss": 0.8106,
-      "step": 2935
-    },
-    {
-      "epoch": 1.4420864858214923,
-      "grad_norm": 0.45333982658390326,
-      "learning_rate": 0.0001369704165732431,
-      "loss": 0.758,
-      "step": 2936
-    },
-    {
-      "epoch": 1.4425785815341083,
-      "grad_norm": 0.39423850287590273,
-      "learning_rate": 0.0001369131965279847,
-      "loss": 0.6921,
-      "step": 2937
-    },
-    {
-      "epoch": 1.4430706772467246,
-      "grad_norm": 0.44768422955250525,
-      "learning_rate": 0.00013685596248683772,
-      "loss": 0.7737,
-      "step": 2938
-    },
-    {
-      "epoch": 1.4435627729593405,
-      "grad_norm": 0.42327666978599154,
-      "learning_rate": 0.00013679871447150275,
-      "loss": 0.7247,
-      "step": 2939
-    },
-    {
-      "epoch": 1.4440548686719568,
-      "grad_norm": 0.43022591934590015,
-      "learning_rate": 0.0001367414525036858,
-      "loss": 0.7282,
-      "step": 2940
-    },
-    {
-      "epoch": 1.4445469643845728,
-      "grad_norm": 0.5328276382722564,
-      "learning_rate": 0.00013668417660509812,
-      "loss": 0.86,
-      "step": 2941
-    },
-    {
-      "epoch": 1.445039060097189,
-      "grad_norm": 0.42019164310891927,
-      "learning_rate": 0.00013662688679745626,
-      "loss": 0.7572,
-      "step": 2942
-    },
-    {
-      "epoch": 1.445531155809805,
-      "grad_norm": 0.45317888939026624,
-      "learning_rate": 0.00013656958310248206,
-      "loss": 0.7606,
-      "step": 2943
-    },
-    {
-      "epoch": 1.4460232515224212,
-      "grad_norm": 0.4159749206257165,
-      "learning_rate": 0.00013651226554190258,
-      "loss": 0.7374,
-      "step": 2944
-    },
-    {
-      "epoch": 1.4465153472350372,
-      "grad_norm": 0.4167892666138694,
-      "learning_rate": 0.00013645493413745016,
-      "loss": 0.7534,
-      "step": 2945
-    },
-    {
-      "epoch": 1.4470074429476534,
-      "grad_norm": 0.4180296030672964,
-      "learning_rate": 0.0001363975889108624,
-      "loss": 0.683,
-      "step": 2946
-    },
-    {
-      "epoch": 1.4474995386602694,
-      "grad_norm": 0.4097577876076059,
-      "learning_rate": 0.00013634022988388215,
-      "loss": 0.713,
-      "step": 2947
-    },
-    {
-      "epoch": 1.4479916343728856,
-      "grad_norm": 0.3988887096595805,
-      "learning_rate": 0.00013628285707825745,
-      "loss": 0.696,
-      "step": 2948
-    },
-    {
-      "epoch": 1.4484837300855016,
-      "grad_norm": 0.43352472144513926,
-      "learning_rate": 0.00013622547051574157,
-      "loss": 0.7573,
-      "step": 2949
-    },
-    {
-      "epoch": 1.4489758257981178,
-      "grad_norm": 0.43484806320664127,
-      "learning_rate": 0.00013616807021809305,
-      "loss": 0.7093,
-      "step": 2950
-    },
-    {
-      "epoch": 1.4494679215107338,
-      "grad_norm": 0.3975537927628502,
-      "learning_rate": 0.00013611065620707563,
-      "loss": 0.6672,
-      "step": 2951
-    },
-    {
-      "epoch": 1.44996001722335,
-      "grad_norm": 0.4149696034740369,
-      "learning_rate": 0.00013605322850445813,
-      "loss": 0.7438,
-      "step": 2952
-    },
-    {
-      "epoch": 1.450452112935966,
-      "grad_norm": 0.4096247946570659,
-      "learning_rate": 0.00013599578713201473,
-      "loss": 0.7311,
-      "step": 2953
-    },
-    {
-      "epoch": 1.4509442086485822,
-      "grad_norm": 0.45218011553335336,
-      "learning_rate": 0.0001359383321115247,
-      "loss": 0.8454,
-      "step": 2954
-    },
-    {
-      "epoch": 1.4514363043611982,
-      "grad_norm": 0.41679761193630593,
-      "learning_rate": 0.00013588086346477245,
-      "loss": 0.7589,
-      "step": 2955
-    },
-    {
-      "epoch": 1.4519284000738144,
-      "grad_norm": 0.42096420998370043,
-      "learning_rate": 0.00013582338121354768,
-      "loss": 0.8439,
-      "step": 2956
-    },
-    {
-      "epoch": 1.4524204957864304,
-      "grad_norm": 0.4213610590768471,
-      "learning_rate": 0.00013576588537964513,
-      "loss": 0.771,
-      "step": 2957
-    },
-    {
-      "epoch": 1.4529125914990466,
-      "grad_norm": 0.4346284009249143,
-      "learning_rate": 0.00013570837598486475,
-      "loss": 0.8156,
-      "step": 2958
-    },
-    {
-      "epoch": 1.4534046872116626,
-      "grad_norm": 0.42241315752609315,
-      "learning_rate": 0.0001356508530510116,
-      "loss": 0.7452,
-      "step": 2959
-    },
-    {
-      "epoch": 1.4538967829242788,
-      "grad_norm": 0.42923311696960353,
-      "learning_rate": 0.0001355933165998959,
-      "loss": 0.7415,
-      "step": 2960
-    },
-    {
-      "epoch": 1.4543888786368948,
-      "grad_norm": 0.4669393188409626,
-      "learning_rate": 0.00013553576665333302,
-      "loss": 0.8085,
-      "step": 2961
-    },
-    {
-      "epoch": 1.454880974349511,
-      "grad_norm": 0.47428268942239843,
-      "learning_rate": 0.00013547820323314336,
-      "loss": 0.7517,
-      "step": 2962
-    },
-    {
-      "epoch": 1.455373070062127,
-      "grad_norm": 0.4167283198912779,
-      "learning_rate": 0.0001354206263611525,
-      "loss": 0.6868,
-      "step": 2963
-    },
-    {
-      "epoch": 1.4558651657747432,
-      "grad_norm": 0.41549655527914137,
-      "learning_rate": 0.0001353630360591911,
-      "loss": 0.7469,
-      "step": 2964
-    },
-    {
-      "epoch": 1.4563572614873592,
-      "grad_norm": 0.3832641047494742,
-      "learning_rate": 0.00013530543234909493,
-      "loss": 0.7121,
-      "step": 2965
-    },
-    {
-      "epoch": 1.4568493571999754,
-      "grad_norm": 0.4294927014503728,
-      "learning_rate": 0.0001352478152527048,
-      "loss": 0.7719,
-      "step": 2966
-    },
-    {
-      "epoch": 1.4573414529125914,
-      "grad_norm": 2.7573070317840536,
-      "learning_rate": 0.00013519018479186666,
-      "loss": 0.7394,
-      "step": 2967
-    },
-    {
-      "epoch": 1.4578335486252076,
-      "grad_norm": 0.4299412409158559,
-      "learning_rate": 0.00013513254098843143,
-      "loss": 0.7342,
-      "step": 2968
-    },
-    {
-      "epoch": 1.4583256443378236,
-      "grad_norm": 0.43663358819270803,
-      "learning_rate": 0.00013507488386425522,
-      "loss": 0.7065,
-      "step": 2969
-    },
-    {
-      "epoch": 1.4588177400504398,
-      "grad_norm": 0.41952467955875783,
-      "learning_rate": 0.00013501721344119907,
-      "loss": 0.7074,
-      "step": 2970
-    },
-    {
-      "epoch": 1.4593098357630558,
-      "grad_norm": 0.4018966441407418,
-      "learning_rate": 0.00013495952974112914,
-      "loss": 0.7468,
-      "step": 2971
-    },
-    {
-      "epoch": 1.459801931475672,
-      "grad_norm": 0.4460412406846254,
-      "learning_rate": 0.00013490183278591652,
-      "loss": 0.7875,
-      "step": 2972
-    },
-    {
-      "epoch": 1.460294027188288,
-      "grad_norm": 0.3995218655936246,
-      "learning_rate": 0.00013484412259743753,
-      "loss": 0.7061,
-      "step": 2973
-    },
-    {
-      "epoch": 1.4607861229009043,
-      "grad_norm": 0.3902152859646245,
-      "learning_rate": 0.0001347863991975733,
-      "loss": 0.7033,
-      "step": 2974
-    },
-    {
-      "epoch": 1.4612782186135203,
-      "grad_norm": 0.45560542563149636,
-      "learning_rate": 0.00013472866260821006,
-      "loss": 0.7665,
-      "step": 2975
-    },
-    {
-      "epoch": 1.4617703143261365,
-      "grad_norm": 1.0765117525058105,
-      "learning_rate": 0.00013467091285123903,
-      "loss": 0.7584,
-      "step": 2976
-    },
-    {
-      "epoch": 1.4622624100387525,
-      "grad_norm": 0.588424029730406,
-      "learning_rate": 0.0001346131499485564,
-      "loss": 0.7788,
-      "step": 2977
-    },
-    {
-      "epoch": 1.4627545057513687,
-      "grad_norm": 0.42299824952303533,
-      "learning_rate": 0.00013455537392206339,
-      "loss": 0.7062,
-      "step": 2978
-    },
-    {
-      "epoch": 1.4632466014639847,
-      "grad_norm": 0.42037663156928856,
-      "learning_rate": 0.00013449758479366618,
-      "loss": 0.7981,
-      "step": 2979
-    },
-    {
-      "epoch": 1.4637386971766009,
-      "grad_norm": 0.46299269963475975,
-      "learning_rate": 0.0001344397825852759,
-      "loss": 0.7638,
-      "step": 2980
-    },
-    {
-      "epoch": 1.4642307928892169,
-      "grad_norm": 0.39361158967557547,
-      "learning_rate": 0.00013438196731880867,
-      "loss": 0.7406,
-      "step": 2981
-    },
-    {
-      "epoch": 1.464722888601833,
-      "grad_norm": 0.42228860720601463,
-      "learning_rate": 0.00013432413901618548,
-      "loss": 0.7183,
-      "step": 2982
-    },
-    {
-      "epoch": 1.465214984314449,
-      "grad_norm": 0.42235631862864903,
-      "learning_rate": 0.00013426629769933238,
-      "loss": 0.7768,
-      "step": 2983
-    },
-    {
-      "epoch": 1.4657070800270653,
-      "grad_norm": 0.43676713475216566,
-      "learning_rate": 0.00013420844339018028,
-      "loss": 0.6861,
-      "step": 2984
-    },
-    {
-      "epoch": 1.4661991757396815,
-      "grad_norm": 0.4250906194418104,
-      "learning_rate": 0.00013415057611066504,
-      "loss": 0.6991,
-      "step": 2985
-    },
-    {
-      "epoch": 1.4666912714522975,
-      "grad_norm": 0.4800896798565744,
-      "learning_rate": 0.0001340926958827274,
-      "loss": 0.7703,
-      "step": 2986
-    },
-    {
-      "epoch": 1.4671833671649135,
-      "grad_norm": 0.4387660488504194,
-      "learning_rate": 0.0001340348027283131,
-      "loss": 0.7642,
-      "step": 2987
-    },
-    {
-      "epoch": 1.4676754628775297,
-      "grad_norm": 0.4185327189884391,
-      "learning_rate": 0.00013397689666937266,
-      "loss": 0.7397,
-      "step": 2988
-    },
-    {
-      "epoch": 1.468167558590146,
-      "grad_norm": 0.40837776294476474,
-      "learning_rate": 0.00013391897772786158,
-      "loss": 0.7363,
-      "step": 2989
-    },
-    {
-      "epoch": 1.468659654302762,
-      "grad_norm": 0.4043044621180528,
-      "learning_rate": 0.00013386104592574022,
-      "loss": 0.7346,
-      "step": 2990
-    },
-    {
-      "epoch": 1.469151750015378,
-      "grad_norm": 0.41509835595502276,
-      "learning_rate": 0.0001338031012849738,
-      "loss": 0.7581,
-      "step": 2991
-    },
-    {
-      "epoch": 1.4696438457279941,
-      "grad_norm": 0.40084265083532755,
-      "learning_rate": 0.00013374514382753246,
-      "loss": 0.7346,
-      "step": 2992
-    },
-    {
-      "epoch": 1.4701359414406103,
-      "grad_norm": 0.4364298452244065,
-      "learning_rate": 0.00013368717357539116,
-      "loss": 0.6604,
-      "step": 2993
-    },
-    {
-      "epoch": 1.4706280371532263,
-      "grad_norm": 0.4305777408560004,
-      "learning_rate": 0.00013362919055052966,
-      "loss": 0.6825,
-      "step": 2994
-    },
-    {
-      "epoch": 1.4711201328658423,
-      "grad_norm": 0.4096676372200456,
-      "learning_rate": 0.00013357119477493265,
-      "loss": 0.7291,
-      "step": 2995
-    },
-    {
-      "epoch": 1.4716122285784585,
-      "grad_norm": 0.4398855482135969,
-      "learning_rate": 0.00013351318627058964,
-      "loss": 0.7369,
-      "step": 2996
-    },
-    {
-      "epoch": 1.4721043242910747,
-      "grad_norm": 0.42490468698913875,
-      "learning_rate": 0.00013345516505949492,
-      "loss": 0.7583,
-      "step": 2997
-    },
-    {
-      "epoch": 1.4725964200036907,
-      "grad_norm": 0.4521087201321453,
-      "learning_rate": 0.00013339713116364768,
-      "loss": 0.731,
-      "step": 2998
-    },
-    {
-      "epoch": 1.4730885157163067,
-      "grad_norm": 0.4558298349166806,
-      "learning_rate": 0.00013333908460505178,
-      "loss": 0.7345,
-      "step": 2999
-    },
-    {
-      "epoch": 1.473580611428923,
-      "grad_norm": 0.43819505834952743,
-      "learning_rate": 0.000133281025405716,
-      "loss": 0.7375,
-      "step": 3000
-    },
-    {
-      "epoch": 1.4740727071415392,
-      "grad_norm": 0.41865292773313334,
-      "learning_rate": 0.0001332229535876539,
-      "loss": 0.6631,
-      "step": 3001
-    },
-    {
-      "epoch": 1.4745648028541551,
-      "grad_norm": 0.4120333157523723,
-      "learning_rate": 0.00013316486917288377,
-      "loss": 0.6795,
-      "step": 3002
-    },
-    {
-      "epoch": 1.4750568985667711,
-      "grad_norm": 0.4021473198346083,
-      "learning_rate": 0.00013310677218342874,
-      "loss": 0.6411,
-      "step": 3003
-    },
-    {
-      "epoch": 1.4755489942793873,
-      "grad_norm": 0.4728250527841678,
-      "learning_rate": 0.00013304866264131669,
-      "loss": 0.7827,
-      "step": 3004
-    },
-    {
-      "epoch": 1.4760410899920036,
-      "grad_norm": 0.44633656756790946,
-      "learning_rate": 0.0001329905405685802,
-      "loss": 0.7199,
-      "step": 3005
-    },
-    {
-      "epoch": 1.4765331857046196,
-      "grad_norm": 0.4948855099340994,
-      "learning_rate": 0.00013293240598725666,
-      "loss": 0.7533,
-      "step": 3006
-    },
-    {
-      "epoch": 1.4770252814172355,
-      "grad_norm": 0.4162448508051315,
-      "learning_rate": 0.00013287425891938824,
-      "loss": 0.7467,
-      "step": 3007
-    },
-    {
-      "epoch": 1.4775173771298518,
-      "grad_norm": 0.4543701745347057,
-      "learning_rate": 0.00013281609938702173,
-      "loss": 0.7436,
-      "step": 3008
-    },
-    {
-      "epoch": 1.478009472842468,
-      "grad_norm": 0.457882069485982,
-      "learning_rate": 0.00013275792741220875,
-      "loss": 0.7581,
-      "step": 3009
-    },
-    {
-      "epoch": 1.478501568555084,
-      "grad_norm": 0.39603707426970053,
-      "learning_rate": 0.00013269974301700557,
-      "loss": 0.7293,
-      "step": 3010
-    },
-    {
-      "epoch": 1.4789936642677,
-      "grad_norm": 0.43248034847450656,
-      "learning_rate": 0.0001326415462234732,
-      "loss": 0.7342,
-      "step": 3011
-    },
-    {
-      "epoch": 1.4794857599803162,
-      "grad_norm": 0.46342484194924166,
-      "learning_rate": 0.0001325833370536774,
-      "loss": 0.769,
-      "step": 3012
-    },
-    {
-      "epoch": 1.4799778556929324,
-      "grad_norm": 0.4066229020918127,
-      "learning_rate": 0.00013252511552968853,
-      "loss": 0.719,
-      "step": 3013
-    },
-    {
-      "epoch": 1.4804699514055484,
-      "grad_norm": 0.41616605393800105,
-      "learning_rate": 0.00013246688167358164,
-      "loss": 0.7519,
-      "step": 3014
-    },
-    {
-      "epoch": 1.4809620471181644,
-      "grad_norm": 0.4219271358945579,
-      "learning_rate": 0.00013240863550743653,
-      "loss": 0.7224,
-      "step": 3015
-    },
-    {
-      "epoch": 1.4814541428307806,
-      "grad_norm": 0.36729337510638593,
-      "learning_rate": 0.00013235037705333765,
-      "loss": 0.716,
-      "step": 3016
-    },
-    {
-      "epoch": 1.4819462385433968,
-      "grad_norm": 0.4215228606656191,
-      "learning_rate": 0.00013229210633337404,
-      "loss": 0.741,
-      "step": 3017
-    },
-    {
-      "epoch": 1.4824383342560128,
-      "grad_norm": 0.39460507590391153,
-      "learning_rate": 0.00013223382336963952,
-      "loss": 0.7464,
-      "step": 3018
-    },
-    {
-      "epoch": 1.4829304299686288,
-      "grad_norm": 0.39184879004971424,
-      "learning_rate": 0.00013217552818423238,
-      "loss": 0.7545,
-      "step": 3019
-    },
-    {
-      "epoch": 1.483422525681245,
-      "grad_norm": 0.41510153203765787,
-      "learning_rate": 0.00013211722079925568,
-      "loss": 0.7773,
-      "step": 3020
-    },
-    {
-      "epoch": 1.4839146213938612,
-      "grad_norm": 0.4255275634596563,
-      "learning_rate": 0.00013205890123681706,
-      "loss": 0.7504,
-      "step": 3021
-    },
-    {
-      "epoch": 1.4844067171064772,
-      "grad_norm": 0.466119135799236,
-      "learning_rate": 0.00013200056951902876,
-      "loss": 0.7264,
-      "step": 3022
-    },
-    {
-      "epoch": 1.4848988128190932,
-      "grad_norm": 0.42155715395882765,
-      "learning_rate": 0.00013194222566800766,
-      "loss": 0.758,
-      "step": 3023
-    },
-    {
-      "epoch": 1.4853909085317094,
-      "grad_norm": 0.4180660786457265,
-      "learning_rate": 0.00013188386970587517,
-      "loss": 0.7597,
-      "step": 3024
-    },
-    {
-      "epoch": 1.4858830042443256,
-      "grad_norm": 0.4257204204565261,
-      "learning_rate": 0.00013182550165475744,
-      "loss": 0.7086,
-      "step": 3025
-    },
-    {
-      "epoch": 1.4863750999569416,
-      "grad_norm": 0.43132732910529703,
-      "learning_rate": 0.00013176712153678509,
-      "loss": 0.7341,
-      "step": 3026
-    },
-    {
-      "epoch": 1.4868671956695576,
-      "grad_norm": 0.42438460081333756,
-      "learning_rate": 0.00013170872937409326,
-      "loss": 0.7845,
-      "step": 3027
-    },
-    {
-      "epoch": 1.4873592913821738,
-      "grad_norm": 0.44307121545989486,
-      "learning_rate": 0.00013165032518882184,
-      "loss": 0.782,
-      "step": 3028
-    },
-    {
-      "epoch": 1.48785138709479,
-      "grad_norm": 0.4044677607381762,
-      "learning_rate": 0.00013159190900311512,
-      "loss": 0.6836,
-      "step": 3029
-    },
-    {
-      "epoch": 1.488343482807406,
-      "grad_norm": 0.3973183220334357,
-      "learning_rate": 0.000131533480839122,
-      "loss": 0.6734,
-      "step": 3030
-    },
-    {
-      "epoch": 1.488835578520022,
-      "grad_norm": 0.40780023312514757,
-      "learning_rate": 0.00013147504071899587,
-      "loss": 0.727,
-      "step": 3031
-    },
-    {
-      "epoch": 1.4893276742326382,
-      "grad_norm": 0.46018741993093354,
-      "learning_rate": 0.00013141658866489477,
-      "loss": 0.8059,
-      "step": 3032
-    },
-    {
-      "epoch": 1.4898197699452544,
-      "grad_norm": 0.42135487675244954,
-      "learning_rate": 0.00013135812469898116,
-      "loss": 0.694,
-      "step": 3033
-    },
-    {
-      "epoch": 1.4903118656578704,
-      "grad_norm": 0.4068258129416411,
-      "learning_rate": 0.00013129964884342206,
-      "loss": 0.6834,
-      "step": 3034
-    },
-    {
-      "epoch": 1.4908039613704864,
-      "grad_norm": 0.4095911790891875,
-      "learning_rate": 0.00013124116112038896,
-      "loss": 0.6842,
-      "step": 3035
-    },
-    {
-      "epoch": 1.4912960570831026,
-      "grad_norm": 0.38707934469090177,
-      "learning_rate": 0.0001311826615520579,
-      "loss": 0.6721,
-      "step": 3036
-    },
-    {
-      "epoch": 1.4917881527957189,
-      "grad_norm": 0.4227456592517028,
-      "learning_rate": 0.00013112415016060938,
-      "loss": 0.6808,
-      "step": 3037
-    },
-    {
-      "epoch": 1.4922802485083349,
-      "grad_norm": 0.40541024584666174,
-      "learning_rate": 0.0001310656269682284,
-      "loss": 0.6862,
-      "step": 3038
-    },
-    {
-      "epoch": 1.4927723442209508,
-      "grad_norm": 0.3940813292989956,
-      "learning_rate": 0.00013100709199710442,
-      "loss": 0.7146,
-      "step": 3039
-    },
-    {
-      "epoch": 1.493264439933567,
-      "grad_norm": 0.42566679643736666,
-      "learning_rate": 0.00013094854526943134,
-      "loss": 0.796,
-      "step": 3040
-    },
-    {
-      "epoch": 1.4937565356461833,
-      "grad_norm": 0.46101186943081,
-      "learning_rate": 0.0001308899868074076,
-      "loss": 0.761,
-      "step": 3041
-    },
-    {
-      "epoch": 1.4942486313587993,
-      "grad_norm": 0.4181233516858649,
-      "learning_rate": 0.00013083141663323603,
-      "loss": 0.6625,
-      "step": 3042
-    },
-    {
-      "epoch": 1.4947407270714153,
-      "grad_norm": 0.41896694413954866,
-      "learning_rate": 0.0001307728347691239,
-      "loss": 0.8141,
-      "step": 3043
-    },
-    {
-      "epoch": 1.4952328227840315,
-      "grad_norm": 0.4091474992052696,
-      "learning_rate": 0.0001307142412372829,
-      "loss": 0.7523,
-      "step": 3044
-    },
-    {
-      "epoch": 1.4957249184966477,
-      "grad_norm": 0.4365804852716206,
-      "learning_rate": 0.00013065563605992918,
-      "loss": 0.7722,
-      "step": 3045
-    },
-    {
-      "epoch": 1.4962170142092637,
-      "grad_norm": 0.45798639122756435,
-      "learning_rate": 0.00013059701925928328,
-      "loss": 0.7476,
-      "step": 3046
-    },
-    {
-      "epoch": 1.4967091099218799,
-      "grad_norm": 0.39903783066877924,
-      "learning_rate": 0.00013053839085757013,
-      "loss": 0.6797,
-      "step": 3047
-    },
-    {
-      "epoch": 1.4972012056344959,
-      "grad_norm": 0.41403278268558885,
-      "learning_rate": 0.00013047975087701917,
-      "loss": 0.7381,
-      "step": 3048
-    },
-    {
-      "epoch": 1.497693301347112,
-      "grad_norm": 0.41042336037569394,
-      "learning_rate": 0.0001304210993398641,
-      "loss": 0.7862,
-      "step": 3049
-    },
-    {
-      "epoch": 1.498185397059728,
-      "grad_norm": 0.39440513575421626,
-      "learning_rate": 0.000130362436268343,
-      "loss": 0.7004,
-      "step": 3050
-    },
-    {
-      "epoch": 1.4986774927723443,
-      "grad_norm": 0.40701229150870666,
-      "learning_rate": 0.00013030376168469848,
-      "loss": 0.6817,
-      "step": 3051
-    },
-    {
-      "epoch": 1.4986774927723443,
-      "eval_loss": 0.8062564730644226,
-      "eval_runtime": 6668.7735,
-      "eval_samples_per_second": 4.273,
-      "eval_steps_per_second": 2.137,
-      "step": 3051
-    },
-    {
-      "epoch": 1.4991695884849603,
-      "grad_norm": 0.4012959059324339,
-      "learning_rate": 0.0001302450756111773,
-      "loss": 0.697,
-      "step": 3052
-    },
-    {
-      "epoch": 1.4996616841975765,
-      "grad_norm": 0.42031391099745824,
-      "learning_rate": 0.00013018637807003078,
-      "loss": 0.75,
-      "step": 3053
-    },
-    {
-      "epoch": 1.5001537799101925,
-      "grad_norm": 0.3785034515960198,
-      "learning_rate": 0.00013012766908351438,
-      "loss": 0.7043,
-      "step": 3054
-    },
-    {
-      "epoch": 1.5006458756228085,
-      "grad_norm": 0.4330691443923283,
-      "learning_rate": 0.0001300689486738881,
-      "loss": 0.8017,
-      "step": 3055
-    },
-    {
-      "epoch": 1.5011379713354247,
-      "grad_norm": 0.39835898188657426,
-      "learning_rate": 0.00013001021686341615,
-      "loss": 0.7541,
-      "step": 3056
-    },
-    {
-      "epoch": 1.501630067048041,
-      "grad_norm": 0.4766175640336095,
-      "learning_rate": 0.00012995147367436704,
-      "loss": 0.7813,
-      "step": 3057
-    },
-    {
-      "epoch": 1.502122162760657,
-      "grad_norm": 0.41451668275342507,
-      "learning_rate": 0.00012989271912901374,
-      "loss": 0.7684,
-      "step": 3058
-    },
-    {
-      "epoch": 1.502614258473273,
-      "grad_norm": 0.4153085818390717,
-      "learning_rate": 0.0001298339532496334,
-      "loss": 0.7648,
-      "step": 3059
-    },
-    {
-      "epoch": 1.5031063541858891,
-      "grad_norm": 0.40939606435324155,
-      "learning_rate": 0.00012977517605850745,
-      "loss": 0.7622,
-      "step": 3060
-    },
-    {
-      "epoch": 1.5035984498985053,
-      "grad_norm": 0.4097088626024703,
-      "learning_rate": 0.00012971638757792176,
-      "loss": 0.7138,
-      "step": 3061
-    },
-    {
-      "epoch": 1.5040905456111213,
-      "grad_norm": 0.40471514832795313,
-      "learning_rate": 0.00012965758783016633,
-      "loss": 0.7048,
-      "step": 3062
-    },
-    {
-      "epoch": 1.5045826413237373,
-      "grad_norm": 0.42223220004450407,
-      "learning_rate": 0.00012959877683753544,
-      "loss": 0.7556,
-      "step": 3063
-    },
-    {
-      "epoch": 1.5050747370363535,
-      "grad_norm": 0.43361994525438585,
-      "learning_rate": 0.00012953995462232771,
-      "loss": 0.7584,
-      "step": 3064
-    },
-    {
-      "epoch": 1.5055668327489697,
-      "grad_norm": 0.4123106850196711,
-      "learning_rate": 0.00012948112120684602,
-      "loss": 0.7816,
-      "step": 3065
-    },
-    {
-      "epoch": 1.5060589284615857,
-      "grad_norm": 0.4254681923207696,
-      "learning_rate": 0.00012942227661339744,
-      "loss": 0.7001,
-      "step": 3066
-    },
-    {
-      "epoch": 1.5065510241742017,
-      "grad_norm": 0.4118680636606478,
-      "learning_rate": 0.00012936342086429325,
-      "loss": 0.752,
-      "step": 3067
-    },
-    {
-      "epoch": 1.507043119886818,
-      "grad_norm": 0.42605811058729165,
-      "learning_rate": 0.00012930455398184904,
-      "loss": 0.7462,
-      "step": 3068
-    },
-    {
-      "epoch": 1.5075352155994342,
-      "grad_norm": 0.44241327862069835,
-      "learning_rate": 0.0001292456759883846,
-      "loss": 0.7251,
-      "step": 3069
-    },
-    {
-      "epoch": 1.5080273113120501,
-      "grad_norm": 0.41674748185876725,
-      "learning_rate": 0.00012918678690622388,
-      "loss": 0.8003,
-      "step": 3070
-    },
-    {
-      "epoch": 1.5085194070246661,
-      "grad_norm": 0.47726405246653525,
-      "learning_rate": 0.00012912788675769512,
-      "loss": 0.7392,
-      "step": 3071
-    },
-    {
-      "epoch": 1.5090115027372824,
-      "grad_norm": 0.4012505019853829,
-      "learning_rate": 0.0001290689755651307,
-      "loss": 0.6901,
-      "step": 3072
-    },
-    {
-      "epoch": 1.5095035984498986,
-      "grad_norm": 0.4152165138620225,
-      "learning_rate": 0.00012901005335086717,
-      "loss": 0.7833,
-      "step": 3073
-    },
-    {
-      "epoch": 1.5099956941625146,
-      "grad_norm": 0.3854227271538217,
-      "learning_rate": 0.00012895112013724532,
-      "loss": 0.6685,
-      "step": 3074
-    },
-    {
-      "epoch": 1.5104877898751305,
-      "grad_norm": 0.3968242727990472,
-      "learning_rate": 0.00012889217594661006,
-      "loss": 0.7435,
-      "step": 3075
-    },
-    {
-      "epoch": 1.5109798855877468,
-      "grad_norm": 0.40764086246986103,
-      "learning_rate": 0.00012883322080131047,
-      "loss": 0.6891,
-      "step": 3076
-    },
-    {
-      "epoch": 1.511471981300363,
-      "grad_norm": 0.460514775082263,
-      "learning_rate": 0.00012877425472369983,
-      "loss": 0.7618,
-      "step": 3077
-    },
-    {
-      "epoch": 1.511964077012979,
-      "grad_norm": 0.40059181460704496,
-      "learning_rate": 0.00012871527773613547,
-      "loss": 0.7208,
-      "step": 3078
-    },
-    {
-      "epoch": 1.5124561727255952,
-      "grad_norm": 0.44988655894234547,
-      "learning_rate": 0.00012865628986097897,
-      "loss": 0.778,
-      "step": 3079
-    },
-    {
-      "epoch": 1.5129482684382114,
-      "grad_norm": 0.4048849756938812,
-      "learning_rate": 0.00012859729112059596,
-      "loss": 0.7492,
-      "step": 3080
-    },
-    {
-      "epoch": 1.5134403641508274,
-      "grad_norm": 0.4191796085659502,
-      "learning_rate": 0.00012853828153735618,
-      "loss": 0.7296,
-      "step": 3081
-    },
-    {
-      "epoch": 1.5139324598634434,
-      "grad_norm": 0.41812779771448,
-      "learning_rate": 0.0001284792611336336,
-      "loss": 0.7048,
-      "step": 3082
-    },
-    {
-      "epoch": 1.5144245555760596,
-      "grad_norm": 0.4797066515574714,
-      "learning_rate": 0.00012842022993180612,
-      "loss": 0.828,
-      "step": 3083
-    },
-    {
-      "epoch": 1.5149166512886758,
-      "grad_norm": 0.41106092005251393,
-      "learning_rate": 0.00012836118795425585,
-      "loss": 0.7813,
-      "step": 3084
-    },
-    {
-      "epoch": 1.5154087470012918,
-      "grad_norm": 0.40817359434708045,
-      "learning_rate": 0.00012830213522336897,
-      "loss": 0.686,
-      "step": 3085
-    },
-    {
-      "epoch": 1.5159008427139078,
-      "grad_norm": 0.4076667748465361,
-      "learning_rate": 0.0001282430717615357,
-      "loss": 0.7094,
-      "step": 3086
-    },
-    {
-      "epoch": 1.516392938426524,
-      "grad_norm": 0.4386133393504196,
-      "learning_rate": 0.0001281839975911504,
-      "loss": 0.7551,
-      "step": 3087
-    },
-    {
-      "epoch": 1.5168850341391402,
-      "grad_norm": 0.429260417303817,
-      "learning_rate": 0.00012812491273461136,
-      "loss": 0.6982,
-      "step": 3088
-    },
-    {
-      "epoch": 1.5173771298517562,
-      "grad_norm": 0.41918902030986416,
-      "learning_rate": 0.00012806581721432108,
-      "loss": 0.7249,
-      "step": 3089
-    },
-    {
-      "epoch": 1.5178692255643722,
-      "grad_norm": 0.4215972943876984,
-      "learning_rate": 0.00012800671105268598,
-      "loss": 0.7339,
-      "step": 3090
-    },
-    {
-      "epoch": 1.5183613212769884,
-      "grad_norm": 0.4038371270111055,
-      "learning_rate": 0.00012794759427211657,
-      "loss": 0.7138,
-      "step": 3091
-    },
-    {
-      "epoch": 1.5188534169896046,
-      "grad_norm": 0.41956476832990003,
-      "learning_rate": 0.0001278884668950274,
-      "loss": 0.7477,
-      "step": 3092
-    },
-    {
-      "epoch": 1.5193455127022206,
-      "grad_norm": 0.44100240467913016,
-      "learning_rate": 0.000127829328943837,
-      "loss": 0.7658,
-      "step": 3093
-    },
-    {
-      "epoch": 1.5198376084148366,
-      "grad_norm": 0.4178678387214205,
-      "learning_rate": 0.00012777018044096792,
-      "loss": 0.6944,
-      "step": 3094
-    },
-    {
-      "epoch": 1.5203297041274528,
-      "grad_norm": 0.4053607057062803,
-      "learning_rate": 0.00012771102140884675,
-      "loss": 0.6927,
-      "step": 3095
-    },
-    {
-      "epoch": 1.520821799840069,
-      "grad_norm": 0.4350538092655992,
-      "learning_rate": 0.00012765185186990396,
-      "loss": 0.6907,
-      "step": 3096
-    },
-    {
-      "epoch": 1.521313895552685,
-      "grad_norm": 0.40854114917475026,
-      "learning_rate": 0.00012759267184657416,
-      "loss": 0.7095,
-      "step": 3097
-    },
-    {
-      "epoch": 1.521805991265301,
-      "grad_norm": 0.4724031030864346,
-      "learning_rate": 0.00012753348136129583,
-      "loss": 0.7489,
-      "step": 3098
-    },
-    {
-      "epoch": 1.5222980869779172,
-      "grad_norm": 0.4337096882195965,
-      "learning_rate": 0.00012747428043651145,
-      "loss": 0.7324,
-      "step": 3099
-    },
-    {
-      "epoch": 1.5227901826905335,
-      "grad_norm": 0.40855131995965477,
-      "learning_rate": 0.00012741506909466743,
-      "loss": 0.7754,
-      "step": 3100
-    },
-    {
-      "epoch": 1.5232822784031494,
-      "grad_norm": 0.3953499045725943,
-      "learning_rate": 0.0001273558473582142,
-      "loss": 0.6775,
-      "step": 3101
-    },
-    {
-      "epoch": 1.5237743741157654,
-      "grad_norm": 0.41147652055069633,
-      "learning_rate": 0.00012729661524960598,
-      "loss": 0.752,
-      "step": 3102
-    },
-    {
-      "epoch": 1.5242664698283817,
-      "grad_norm": 0.44204125317371185,
-      "learning_rate": 0.00012723737279130115,
-      "loss": 0.6875,
-      "step": 3103
-    },
-    {
-      "epoch": 1.5247585655409979,
-      "grad_norm": 0.4012087670532406,
-      "learning_rate": 0.00012717812000576182,
-      "loss": 0.7021,
-      "step": 3104
-    },
-    {
-      "epoch": 1.5252506612536139,
-      "grad_norm": 0.4071975517738948,
-      "learning_rate": 0.0001271188569154541,
-      "loss": 0.7156,
-      "step": 3105
-    },
-    {
-      "epoch": 1.5257427569662299,
-      "grad_norm": 0.48279921604761133,
-      "learning_rate": 0.00012705958354284797,
-      "loss": 0.8309,
-      "step": 3106
-    },
-    {
-      "epoch": 1.526234852678846,
-      "grad_norm": 0.4496843301558887,
-      "learning_rate": 0.00012700029991041738,
-      "loss": 0.7971,
-      "step": 3107
-    },
-    {
-      "epoch": 1.5267269483914623,
-      "grad_norm": 0.41853871282158156,
-      "learning_rate": 0.0001269410060406401,
-      "loss": 0.7736,
-      "step": 3108
-    },
-    {
-      "epoch": 1.5272190441040783,
-      "grad_norm": 0.394324717785068,
-      "learning_rate": 0.00012688170195599774,
-      "loss": 0.7316,
-      "step": 3109
-    },
-    {
-      "epoch": 1.5277111398166943,
-      "grad_norm": 0.4361928198770574,
-      "learning_rate": 0.00012682238767897596,
-      "loss": 0.7639,
-      "step": 3110
-    },
-    {
-      "epoch": 1.5282032355293105,
-      "grad_norm": 0.4032868357652532,
-      "learning_rate": 0.0001267630632320641,
-      "loss": 0.7107,
-      "step": 3111
-    },
-    {
-      "epoch": 1.5286953312419267,
-      "grad_norm": 0.43093832486121686,
-      "learning_rate": 0.00012670372863775545,
-      "loss": 0.7767,
-      "step": 3112
-    },
-    {
-      "epoch": 1.5291874269545427,
-      "grad_norm": 0.3990443330477586,
-      "learning_rate": 0.00012664438391854708,
-      "loss": 0.7084,
-      "step": 3113
-    },
-    {
-      "epoch": 1.5296795226671587,
-      "grad_norm": 0.4078595316209392,
-      "learning_rate": 0.00012658502909694,
-      "loss": 0.7367,
-      "step": 3114
-    },
-    {
-      "epoch": 1.530171618379775,
-      "grad_norm": 0.4842332587104113,
-      "learning_rate": 0.0001265256641954389,
-      "loss": 0.7251,
-      "step": 3115
-    },
-    {
-      "epoch": 1.530663714092391,
-      "grad_norm": 0.41205273022195266,
-      "learning_rate": 0.00012646628923655253,
-      "loss": 0.7531,
-      "step": 3116
-    },
-    {
-      "epoch": 1.531155809805007,
-      "grad_norm": 0.4129857433946101,
-      "learning_rate": 0.0001264069042427932,
-      "loss": 0.7627,
-      "step": 3117
-    },
-    {
-      "epoch": 1.531647905517623,
-      "grad_norm": 0.4608841180298976,
-      "learning_rate": 0.00012634750923667717,
-      "loss": 0.7522,
-      "step": 3118
-    },
-    {
-      "epoch": 1.5321400012302393,
-      "grad_norm": 0.4007114571605506,
-      "learning_rate": 0.0001262881042407244,
-      "loss": 0.7688,
-      "step": 3119
-    },
-    {
-      "epoch": 1.5326320969428555,
-      "grad_norm": 0.40875220062558654,
-      "learning_rate": 0.00012622868927745882,
-      "loss": 0.7977,
-      "step": 3120
-    },
-    {
-      "epoch": 1.5331241926554715,
-      "grad_norm": 0.4363642818563895,
-      "learning_rate": 0.00012616926436940793,
-      "loss": 0.8101,
-      "step": 3121
-    },
-    {
-      "epoch": 1.5336162883680875,
-      "grad_norm": 0.3932820988657897,
-      "learning_rate": 0.00012610982953910308,
-      "loss": 0.6781,
-      "step": 3122
-    },
-    {
-      "epoch": 1.5341083840807037,
-      "grad_norm": 0.4351803023940728,
-      "learning_rate": 0.00012605038480907943,
-      "loss": 0.7682,
-      "step": 3123
-    },
-    {
-      "epoch": 1.53460047979332,
-      "grad_norm": 0.3946445702138645,
-      "learning_rate": 0.00012599093020187582,
-      "loss": 0.6606,
-      "step": 3124
-    },
-    {
-      "epoch": 1.535092575505936,
-      "grad_norm": 0.4513592470917633,
-      "learning_rate": 0.00012593146574003486,
-      "loss": 0.7405,
-      "step": 3125
-    },
-    {
-      "epoch": 1.535584671218552,
-      "grad_norm": 0.4341125634658012,
-      "learning_rate": 0.00012587199144610292,
-      "loss": 0.75,
-      "step": 3126
-    },
-    {
-      "epoch": 1.5360767669311681,
-      "grad_norm": 0.9598121480047209,
-      "learning_rate": 0.00012581250734263011,
-      "loss": 0.7155,
-      "step": 3127
-    },
-    {
-      "epoch": 1.5365688626437843,
-      "grad_norm": 0.4059330316147248,
-      "learning_rate": 0.00012575301345217022,
-      "loss": 0.7371,
-      "step": 3128
-    },
-    {
-      "epoch": 1.5370609583564003,
-      "grad_norm": 0.41905133678894035,
-      "learning_rate": 0.00012569350979728072,
-      "loss": 0.7472,
-      "step": 3129
-    },
-    {
-      "epoch": 1.5375530540690163,
-      "grad_norm": 0.39863870202992024,
-      "learning_rate": 0.00012563399640052288,
-      "loss": 0.6779,
-      "step": 3130
-    },
-    {
-      "epoch": 1.5380451497816325,
-      "grad_norm": 0.4372611193678232,
-      "learning_rate": 0.00012557447328446155,
-      "loss": 0.6678,
-      "step": 3131
-    },
-    {
-      "epoch": 1.5385372454942488,
-      "grad_norm": 0.44284328012292046,
-      "learning_rate": 0.00012551494047166533,
-      "loss": 0.7023,
-      "step": 3132
-    },
-    {
-      "epoch": 1.5390293412068647,
-      "grad_norm": 0.4368631008394386,
-      "learning_rate": 0.00012545539798470654,
-      "loss": 0.775,
-      "step": 3133
-    },
-    {
-      "epoch": 1.5395214369194807,
-      "grad_norm": 0.40955854086198185,
-      "learning_rate": 0.00012539584584616109,
-      "loss": 0.6759,
-      "step": 3134
-    },
-    {
-      "epoch": 1.540013532632097,
-      "grad_norm": 0.399361340268841,
-      "learning_rate": 0.00012533628407860857,
-      "loss": 0.6799,
-      "step": 3135
-    },
-    {
-      "epoch": 1.5405056283447132,
-      "grad_norm": 0.44816478436128276,
-      "learning_rate": 0.00012527671270463225,
-      "loss": 0.7643,
-      "step": 3136
-    },
-    {
-      "epoch": 1.5409977240573292,
-      "grad_norm": 0.4469735527379651,
-      "learning_rate": 0.000125217131746819,
-      "loss": 0.8086,
-      "step": 3137
-    },
-    {
-      "epoch": 1.5414898197699451,
-      "grad_norm": 0.44165998397296513,
-      "learning_rate": 0.00012515754122775931,
-      "loss": 0.7714,
-      "step": 3138
-    },
-    {
-      "epoch": 1.5419819154825614,
-      "grad_norm": 0.4009003684739982,
-      "learning_rate": 0.00012509794117004737,
-      "loss": 0.6879,
-      "step": 3139
-    },
-    {
-      "epoch": 1.5424740111951776,
-      "grad_norm": 0.41411732986544114,
-      "learning_rate": 0.000125038331596281,
-      "loss": 0.7548,
-      "step": 3140
-    },
-    {
-      "epoch": 1.5429661069077936,
-      "grad_norm": 0.40094591741964813,
-      "learning_rate": 0.00012497871252906148,
-      "loss": 0.7728,
-      "step": 3141
-    },
-    {
-      "epoch": 1.5434582026204096,
-      "grad_norm": 0.3836291781594014,
-      "learning_rate": 0.0001249190839909938,
-      "loss": 0.6485,
-      "step": 3142
-    },
-    {
-      "epoch": 1.5439502983330258,
-      "grad_norm": 0.4181964897836226,
-      "learning_rate": 0.00012485944600468658,
-      "loss": 0.7176,
-      "step": 3143
-    },
-    {
-      "epoch": 1.544442394045642,
-      "grad_norm": 0.3900476893534145,
-      "learning_rate": 0.0001247997985927519,
-      "loss": 0.6686,
-      "step": 3144
-    },
-    {
-      "epoch": 1.544934489758258,
-      "grad_norm": 0.4125968693070065,
-      "learning_rate": 0.00012474014177780553,
-      "loss": 0.7701,
-      "step": 3145
-    },
-    {
-      "epoch": 1.545426585470874,
-      "grad_norm": 0.40477972245023536,
-      "learning_rate": 0.00012468047558246675,
-      "loss": 0.7071,
-      "step": 3146
-    },
-    {
-      "epoch": 1.5459186811834902,
-      "grad_norm": 0.4831825795027475,
-      "learning_rate": 0.00012462080002935836,
-      "loss": 0.7199,
-      "step": 3147
-    },
-    {
-      "epoch": 1.5464107768961064,
-      "grad_norm": 0.41862299991321594,
-      "learning_rate": 0.00012456111514110677,
-      "loss": 0.722,
-      "step": 3148
-    },
-    {
-      "epoch": 1.5469028726087224,
-      "grad_norm": 0.4123363747171097,
-      "learning_rate": 0.00012450142094034194,
-      "loss": 0.7545,
-      "step": 3149
-    },
-    {
-      "epoch": 1.5473949683213384,
-      "grad_norm": 0.40525575830553767,
-      "learning_rate": 0.00012444171744969732,
-      "loss": 0.7153,
-      "step": 3150
-    },
-    {
-      "epoch": 1.5478870640339546,
-      "grad_norm": 0.4424817544085473,
-      "learning_rate": 0.00012438200469180985,
-      "loss": 0.7219,
-      "step": 3151
-    },
-    {
-      "epoch": 1.5483791597465708,
-      "grad_norm": 0.4624706482322748,
-      "learning_rate": 0.0001243222826893201,
-      "loss": 0.7537,
-      "step": 3152
-    },
-    {
-      "epoch": 1.5488712554591868,
-      "grad_norm": 0.3996180166604787,
-      "learning_rate": 0.00012426255146487201,
-      "loss": 0.75,
-      "step": 3153
-    },
-    {
-      "epoch": 1.5493633511718028,
-      "grad_norm": 0.4002862540823509,
-      "learning_rate": 0.0001242028110411131,
-      "loss": 0.7402,
-      "step": 3154
-    },
-    {
-      "epoch": 1.549855446884419,
-      "grad_norm": 0.4052397801704882,
-      "learning_rate": 0.00012414306144069436,
-      "loss": 0.7441,
-      "step": 3155
-    },
-    {
-      "epoch": 1.5503475425970352,
-      "grad_norm": 0.39188769814813246,
-      "learning_rate": 0.00012408330268627027,
-      "loss": 0.6907,
-      "step": 3156
-    },
-    {
-      "epoch": 1.5508396383096512,
-      "grad_norm": 0.4703358633800996,
-      "learning_rate": 0.00012402353480049874,
-      "loss": 0.8066,
-      "step": 3157
-    },
-    {
-      "epoch": 1.5513317340222672,
-      "grad_norm": 0.49725706722828167,
-      "learning_rate": 0.00012396375780604116,
-      "loss": 0.7568,
-      "step": 3158
-    },
-    {
-      "epoch": 1.5518238297348834,
-      "grad_norm": 0.3889181561017834,
-      "learning_rate": 0.00012390397172556242,
-      "loss": 0.7377,
-      "step": 3159
-    },
-    {
-      "epoch": 1.5523159254474996,
-      "grad_norm": 0.4146720624581034,
-      "learning_rate": 0.0001238441765817308,
-      "loss": 0.7054,
-      "step": 3160
-    },
-    {
-      "epoch": 1.5528080211601156,
-      "grad_norm": 0.40078899808712837,
-      "learning_rate": 0.000123784372397218,
-      "loss": 0.7031,
-      "step": 3161
-    },
-    {
-      "epoch": 1.5533001168727316,
-      "grad_norm": 0.39964674244438475,
-      "learning_rate": 0.00012372455919469925,
-      "loss": 0.7717,
-      "step": 3162
-    },
-    {
-      "epoch": 1.5537922125853478,
-      "grad_norm": 0.39330947490961,
-      "learning_rate": 0.00012366473699685309,
-      "loss": 0.7092,
-      "step": 3163
-    },
-    {
-      "epoch": 1.554284308297964,
-      "grad_norm": 0.38671419115056166,
-      "learning_rate": 0.0001236049058263615,
-      "loss": 0.6819,
-      "step": 3164
-    },
-    {
-      "epoch": 1.55477640401058,
-      "grad_norm": 0.41233618348051926,
-      "learning_rate": 0.00012354506570590992,
-      "loss": 0.751,
-      "step": 3165
-    },
-    {
-      "epoch": 1.555268499723196,
-      "grad_norm": 0.40554561359199665,
-      "learning_rate": 0.00012348521665818708,
-      "loss": 0.7965,
-      "step": 3166
-    },
-    {
-      "epoch": 1.5557605954358122,
-      "grad_norm": 0.4255750503354243,
-      "learning_rate": 0.0001234253587058852,
-      "loss": 0.7032,
-      "step": 3167
-    },
-    {
-      "epoch": 1.5562526911484285,
-      "grad_norm": 0.4243341887658289,
-      "learning_rate": 0.00012336549187169982,
-      "loss": 0.7741,
-      "step": 3168
-    },
-    {
-      "epoch": 1.5567447868610444,
-      "grad_norm": 0.41913378225345127,
-      "learning_rate": 0.00012330561617832984,
-      "loss": 0.7246,
-      "step": 3169
-    },
-    {
-      "epoch": 1.5572368825736604,
-      "grad_norm": 0.39463951391533864,
-      "learning_rate": 0.0001232457316484775,
-      "loss": 0.6937,
-      "step": 3170
-    },
-    {
-      "epoch": 1.5577289782862767,
-      "grad_norm": 0.4432901611924005,
-      "learning_rate": 0.0001231858383048485,
-      "loss": 0.7259,
-      "step": 3171
-    },
-    {
-      "epoch": 1.5582210739988929,
-      "grad_norm": 0.42505712154884673,
-      "learning_rate": 0.00012312593617015176,
-      "loss": 0.7723,
-      "step": 3172
-    },
-    {
-      "epoch": 1.5587131697115089,
-      "grad_norm": 0.4017494696724237,
-      "learning_rate": 0.0001230660252670996,
-      "loss": 0.7465,
-      "step": 3173
-    },
-    {
-      "epoch": 1.5592052654241249,
-      "grad_norm": 0.4457221849906673,
-      "learning_rate": 0.00012300610561840762,
-      "loss": 0.7829,
-      "step": 3174
-    },
-    {
-      "epoch": 1.559697361136741,
-      "grad_norm": 0.43366350686788174,
-      "learning_rate": 0.0001229461772467948,
-      "loss": 0.7266,
-      "step": 3175
-    },
-    {
-      "epoch": 1.5601894568493573,
-      "grad_norm": 0.4133768951049731,
-      "learning_rate": 0.00012288624017498336,
-      "loss": 0.7523,
-      "step": 3176
-    },
-    {
-      "epoch": 1.5606815525619733,
-      "grad_norm": 0.4226671272286967,
-      "learning_rate": 0.00012282629442569886,
-      "loss": 0.7377,
-      "step": 3177
-    },
-    {
-      "epoch": 1.5611736482745893,
-      "grad_norm": 0.4233081981379191,
-      "learning_rate": 0.0001227663400216701,
-      "loss": 0.8065,
-      "step": 3178
-    },
-    {
-      "epoch": 1.5616657439872055,
-      "grad_norm": 0.4116049168948208,
-      "learning_rate": 0.00012270637698562925,
-      "loss": 0.7369,
-      "step": 3179
-    },
-    {
-      "epoch": 1.5621578396998217,
-      "grad_norm": 0.4085878673645208,
-      "learning_rate": 0.0001226464053403117,
-      "loss": 0.7834,
-      "step": 3180
-    },
-    {
-      "epoch": 1.5626499354124377,
-      "grad_norm": 0.44664794329108926,
-      "learning_rate": 0.00012258642510845608,
-      "loss": 0.8174,
-      "step": 3181
-    },
-    {
-      "epoch": 1.5631420311250537,
-      "grad_norm": 0.414396505533701,
-      "learning_rate": 0.0001225264363128043,
-      "loss": 0.7368,
-      "step": 3182
-    },
-    {
-      "epoch": 1.56363412683767,
-      "grad_norm": 0.4337292987659127,
-      "learning_rate": 0.00012246643897610154,
-      "loss": 0.7606,
-      "step": 3183
-    },
-    {
-      "epoch": 1.564126222550286,
-      "grad_norm": 0.37977316865869715,
-      "learning_rate": 0.00012240643312109615,
-      "loss": 0.6468,
-      "step": 3184
-    },
-    {
-      "epoch": 1.564618318262902,
-      "grad_norm": 0.42540973031554663,
-      "learning_rate": 0.0001223464187705398,
-      "loss": 0.7432,
-      "step": 3185
-    },
-    {
-      "epoch": 1.565110413975518,
-      "grad_norm": 0.40031097511476715,
-      "learning_rate": 0.00012228639594718735,
-      "loss": 0.6835,
-      "step": 3186
-    },
-    {
-      "epoch": 1.5656025096881343,
-      "grad_norm": 0.4423994790045654,
-      "learning_rate": 0.0001222263646737968,
-      "loss": 0.7433,
-      "step": 3187
-    },
-    {
-      "epoch": 1.5660946054007505,
-      "grad_norm": 0.39284839259840126,
-      "learning_rate": 0.00012216632497312948,
-      "loss": 0.7109,
-      "step": 3188
-    },
-    {
-      "epoch": 1.5665867011133665,
-      "grad_norm": 0.38096019470032755,
-      "learning_rate": 0.00012210627686794982,
-      "loss": 0.7217,
-      "step": 3189
-    },
-    {
-      "epoch": 1.5670787968259825,
-      "grad_norm": 0.43208719127571726,
-      "learning_rate": 0.00012204622038102547,
-      "loss": 0.7379,
-      "step": 3190
-    },
-    {
-      "epoch": 1.5675708925385987,
-      "grad_norm": 0.5125978311629931,
-      "learning_rate": 0.00012198615553512724,
-      "loss": 0.7665,
-      "step": 3191
-    },
-    {
-      "epoch": 1.568062988251215,
-      "grad_norm": 0.39537945211758196,
-      "learning_rate": 0.00012192608235302914,
-      "loss": 0.7575,
-      "step": 3192
-    },
-    {
-      "epoch": 1.568555083963831,
-      "grad_norm": 0.40283597554945677,
-      "learning_rate": 0.00012186600085750832,
-      "loss": 0.6771,
-      "step": 3193
-    },
-    {
-      "epoch": 1.569047179676447,
-      "grad_norm": 0.445836777284328,
-      "learning_rate": 0.00012180591107134507,
-      "loss": 0.7641,
-      "step": 3194
-    },
-    {
-      "epoch": 1.5695392753890631,
-      "grad_norm": 0.4066461003156354,
-      "learning_rate": 0.00012174581301732288,
-      "loss": 0.7844,
-      "step": 3195
-    },
-    {
-      "epoch": 1.5700313711016793,
-      "grad_norm": 0.4013926668665078,
-      "learning_rate": 0.0001216857067182283,
-      "loss": 0.7591,
-      "step": 3196
-    },
-    {
-      "epoch": 1.5705234668142953,
-      "grad_norm": 0.43761053795693694,
-      "learning_rate": 0.00012162559219685103,
-      "loss": 0.7691,
-      "step": 3197
-    },
-    {
-      "epoch": 1.5710155625269113,
-      "grad_norm": 0.4364570057829297,
-      "learning_rate": 0.00012156546947598393,
-      "loss": 0.7213,
-      "step": 3198
-    },
-    {
-      "epoch": 1.5715076582395275,
-      "grad_norm": 0.40217805746179286,
-      "learning_rate": 0.00012150533857842294,
-      "loss": 0.6977,
-      "step": 3199
-    },
-    {
-      "epoch": 1.5719997539521438,
-      "grad_norm": 0.4059245290376753,
-      "learning_rate": 0.00012144519952696707,
-      "loss": 0.7809,
-      "step": 3200
-    },
-    {
-      "epoch": 1.5724918496647597,
-      "grad_norm": 0.43773026575758556,
-      "learning_rate": 0.00012138505234441846,
-      "loss": 0.7461,
-      "step": 3201
-    },
-    {
-      "epoch": 1.5729839453773757,
-      "grad_norm": 0.5257924412805737,
-      "learning_rate": 0.00012132489705358234,
-      "loss": 0.6759,
-      "step": 3202
-    },
-    {
-      "epoch": 1.573476041089992,
-      "grad_norm": 0.4015383996173517,
-      "learning_rate": 0.00012126473367726697,
-      "loss": 0.7324,
-      "step": 3203
-    },
-    {
-      "epoch": 1.5739681368026082,
-      "grad_norm": 0.4228274459204497,
-      "learning_rate": 0.00012120456223828371,
-      "loss": 0.6825,
-      "step": 3204
-    },
-    {
-      "epoch": 1.5744602325152242,
-      "grad_norm": 0.41065820678004206,
-      "learning_rate": 0.00012114438275944697,
-      "loss": 0.7421,
-      "step": 3205
-    },
-    {
-      "epoch": 1.5749523282278404,
-      "grad_norm": 0.4096908137103904,
-      "learning_rate": 0.00012108419526357421,
-      "loss": 0.6982,
-      "step": 3206
-    },
-    {
-      "epoch": 1.5754444239404566,
-      "grad_norm": 0.3902471204861861,
-      "learning_rate": 0.0001210239997734859,
-      "loss": 0.7206,
-      "step": 3207
-    },
-    {
-      "epoch": 1.5759365196530726,
-      "grad_norm": 0.394550442380344,
-      "learning_rate": 0.00012096379631200563,
-      "loss": 0.7049,
-      "step": 3208
-    },
-    {
-      "epoch": 1.5764286153656886,
-      "grad_norm": 0.3980228960206703,
-      "learning_rate": 0.0001209035849019599,
-      "loss": 0.7079,
-      "step": 3209
-    },
-    {
-      "epoch": 1.5769207110783048,
-      "grad_norm": 0.3884921384266705,
-      "learning_rate": 0.00012084336556617826,
-      "loss": 0.7638,
-      "step": 3210
-    },
-    {
-      "epoch": 1.577412806790921,
-      "grad_norm": 0.388694547737168,
-      "learning_rate": 0.00012078313832749335,
-      "loss": 0.6976,
-      "step": 3211
-    },
-    {
-      "epoch": 1.577904902503537,
-      "grad_norm": 0.4106725701987002,
-      "learning_rate": 0.00012072290320874067,
-      "loss": 0.7455,
-      "step": 3212
-    },
-    {
-      "epoch": 1.578396998216153,
-      "grad_norm": 0.379571478980294,
-      "learning_rate": 0.00012066266023275881,
-      "loss": 0.7182,
-      "step": 3213
-    },
-    {
-      "epoch": 1.5788890939287692,
-      "grad_norm": 0.3781914515938951,
-      "learning_rate": 0.00012060240942238927,
-      "loss": 0.746,
-      "step": 3214
-    },
-    {
-      "epoch": 1.5793811896413854,
-      "grad_norm": 0.4094939873193998,
-      "learning_rate": 0.0001205421508004766,
-      "loss": 0.7251,
-      "step": 3215
-    },
-    {
-      "epoch": 1.5798732853540014,
-      "grad_norm": 0.406974633221615,
-      "learning_rate": 0.00012048188438986821,
-      "loss": 0.725,
-      "step": 3216
-    },
-    {
-      "epoch": 1.5803653810666174,
-      "grad_norm": 0.41874649043239853,
-      "learning_rate": 0.00012042161021341454,
-      "loss": 0.7302,
-      "step": 3217
-    },
-    {
-      "epoch": 1.5808574767792336,
-      "grad_norm": 0.4372105347499653,
-      "learning_rate": 0.00012036132829396895,
-      "loss": 0.7779,
-      "step": 3218
-    },
-    {
-      "epoch": 1.5813495724918498,
-      "grad_norm": 0.4613183652990205,
-      "learning_rate": 0.00012030103865438778,
-      "loss": 0.7261,
-      "step": 3219
-    },
-    {
-      "epoch": 1.5818416682044658,
-      "grad_norm": 0.42093631785720165,
-      "learning_rate": 0.00012024074131753018,
-      "loss": 0.7896,
-      "step": 3220
-    },
-    {
-      "epoch": 1.5823337639170818,
-      "grad_norm": 0.4302448082721251,
-      "learning_rate": 0.00012018043630625835,
-      "loss": 0.6971,
-      "step": 3221
-    },
-    {
-      "epoch": 1.582825859629698,
-      "grad_norm": 0.46152397247781207,
-      "learning_rate": 0.00012012012364343735,
-      "loss": 0.7125,
-      "step": 3222
-    },
-    {
-      "epoch": 1.5833179553423142,
-      "grad_norm": 0.39258700347927744,
-      "learning_rate": 0.00012005980335193507,
-      "loss": 0.7238,
-      "step": 3223
-    },
-    {
-      "epoch": 1.5838100510549302,
-      "grad_norm": 0.42986712350839656,
-      "learning_rate": 0.00011999947545462242,
-      "loss": 0.745,
-      "step": 3224
-    },
-    {
-      "epoch": 1.5843021467675462,
-      "grad_norm": 0.4333873903179454,
-      "learning_rate": 0.00011993913997437313,
-      "loss": 0.7584,
-      "step": 3225
-    },
-    {
-      "epoch": 1.5847942424801624,
-      "grad_norm": 0.4131821348341676,
-      "learning_rate": 0.00011987879693406379,
-      "loss": 0.7382,
-      "step": 3226
-    },
-    {
-      "epoch": 1.5852863381927786,
-      "grad_norm": 0.4036706507775948,
-      "learning_rate": 0.00011981844635657386,
-      "loss": 0.7521,
-      "step": 3227
-    },
-    {
-      "epoch": 1.5857784339053946,
-      "grad_norm": 0.43449693180274446,
-      "learning_rate": 0.00011975808826478567,
-      "loss": 0.7233,
-      "step": 3228
-    },
-    {
-      "epoch": 1.5862705296180106,
-      "grad_norm": 0.4115064959380051,
-      "learning_rate": 0.00011969772268158443,
-      "loss": 0.6833,
-      "step": 3229
-    },
-    {
-      "epoch": 1.5867626253306268,
-      "grad_norm": 0.47992814892736324,
-      "learning_rate": 0.00011963734962985811,
-      "loss": 0.7624,
-      "step": 3230
-    },
-    {
-      "epoch": 1.587254721043243,
-      "grad_norm": 0.4157417072109434,
-      "learning_rate": 0.00011957696913249761,
-      "loss": 0.7168,
-      "step": 3231
-    },
-    {
-      "epoch": 1.587746816755859,
-      "grad_norm": 0.4044420894643676,
-      "learning_rate": 0.0001195165812123966,
-      "loss": 0.7936,
-      "step": 3232
-    },
-    {
-      "epoch": 1.588238912468475,
-      "grad_norm": 0.39319196370554976,
-      "learning_rate": 0.00011945618589245151,
-      "loss": 0.7117,
-      "step": 3233
-    },
-    {
-      "epoch": 1.5887310081810913,
-      "grad_norm": 0.4432426510736256,
-      "learning_rate": 0.00011939578319556173,
-      "loss": 0.7836,
-      "step": 3234
-    },
-    {
-      "epoch": 1.5892231038937075,
-      "grad_norm": 0.4244378022492606,
-      "learning_rate": 0.00011933537314462929,
-      "loss": 0.6452,
-      "step": 3235
-    },
-    {
-      "epoch": 1.5897151996063235,
-      "grad_norm": 0.4241822686502179,
-      "learning_rate": 0.00011927495576255907,
-      "loss": 0.7163,
-      "step": 3236
-    },
-    {
-      "epoch": 1.5902072953189395,
-      "grad_norm": 0.43134795646055973,
-      "learning_rate": 0.00011921453107225877,
-      "loss": 0.7867,
-      "step": 3237
-    },
-    {
-      "epoch": 1.5906993910315557,
-      "grad_norm": 0.3737104615007287,
-      "learning_rate": 0.00011915409909663878,
-      "loss": 0.6735,
-      "step": 3238
-    },
-    {
-      "epoch": 1.5911914867441719,
-      "grad_norm": 0.4068521661983082,
-      "learning_rate": 0.0001190936598586123,
-      "loss": 0.7293,
-      "step": 3239
-    },
-    {
-      "epoch": 1.5916835824567879,
-      "grad_norm": 0.3995051886846135,
-      "learning_rate": 0.00011903321338109527,
-      "loss": 0.653,
-      "step": 3240
-    },
-    {
-      "epoch": 1.5921756781694039,
-      "grad_norm": 0.4328633281994684,
-      "learning_rate": 0.0001189727596870064,
-      "loss": 0.7238,
-      "step": 3241
-    },
-    {
-      "epoch": 1.59266777388202,
-      "grad_norm": 0.390485099531083,
-      "learning_rate": 0.00011891229879926715,
-      "loss": 0.6991,
-      "step": 3242
-    },
-    {
-      "epoch": 1.5931598695946363,
-      "grad_norm": 0.4334795235601448,
-      "learning_rate": 0.0001188518307408016,
-      "loss": 0.7513,
-      "step": 3243
-    },
-    {
-      "epoch": 1.5936519653072523,
-      "grad_norm": 0.3978157461790713,
-      "learning_rate": 0.00011879135553453666,
-      "loss": 0.7426,
-      "step": 3244
-    },
-    {
-      "epoch": 1.5941440610198683,
-      "grad_norm": 0.450012737682133,
-      "learning_rate": 0.00011873087320340194,
-      "loss": 0.795,
-      "step": 3245
-    },
-    {
-      "epoch": 1.5946361567324845,
-      "grad_norm": 0.4339573027096241,
-      "learning_rate": 0.00011867038377032968,
-      "loss": 0.7899,
-      "step": 3246
-    },
-    {
-      "epoch": 1.5951282524451007,
-      "grad_norm": 0.4315192870884105,
-      "learning_rate": 0.0001186098872582549,
-      "loss": 0.7264,
-      "step": 3247
-    },
-    {
-      "epoch": 1.5956203481577167,
-      "grad_norm": 0.4356228528840322,
-      "learning_rate": 0.00011854938369011524,
-      "loss": 0.7265,
-      "step": 3248
-    },
-    {
-      "epoch": 1.5961124438703327,
-      "grad_norm": 0.4186532600742828,
-      "learning_rate": 0.00011848887308885103,
-      "loss": 0.8307,
-      "step": 3249
-    },
-    {
-      "epoch": 1.596604539582949,
-      "grad_norm": 0.4389404768825083,
-      "learning_rate": 0.00011842835547740532,
-      "loss": 0.6989,
-      "step": 3250
-    },
-    {
-      "epoch": 1.5970966352955651,
-      "grad_norm": 0.4282987131053194,
-      "learning_rate": 0.00011836783087872372,
-      "loss": 0.8298,
-      "step": 3251
-    },
-    {
-      "epoch": 1.597588731008181,
-      "grad_norm": 0.3709300892850735,
-      "learning_rate": 0.00011830729931575455,
-      "loss": 0.7213,
-      "step": 3252
-    },
-    {
-      "epoch": 1.598080826720797,
-      "grad_norm": 0.39262352567043607,
-      "learning_rate": 0.00011824676081144876,
-      "loss": 0.7402,
-      "step": 3253
-    },
-    {
-      "epoch": 1.5985729224334133,
-      "grad_norm": 0.42026054831734694,
-      "learning_rate": 0.00011818621538875998,
-      "loss": 0.7269,
-      "step": 3254
-    },
-    {
-      "epoch": 1.5990650181460295,
-      "grad_norm": 0.4423013018477197,
-      "learning_rate": 0.00011812566307064437,
-      "loss": 0.7616,
-      "step": 3255
-    },
-    {
-      "epoch": 1.5995571138586455,
-      "grad_norm": 0.38771548614502505,
-      "learning_rate": 0.00011806510388006074,
-      "loss": 0.6695,
-      "step": 3256
-    },
-    {
-      "epoch": 1.6000492095712615,
-      "grad_norm": 0.4293862443775147,
-      "learning_rate": 0.00011800453783997056,
-      "loss": 0.7205,
-      "step": 3257
-    },
-    {
-      "epoch": 1.6005413052838777,
-      "grad_norm": 0.40091041932693333,
-      "learning_rate": 0.00011794396497333787,
-      "loss": 0.7765,
-      "step": 3258
-    },
-    {
-      "epoch": 1.601033400996494,
-      "grad_norm": 0.5402956068930469,
-      "learning_rate": 0.00011788338530312921,
-      "loss": 0.7701,
-      "step": 3259
-    },
-    {
-      "epoch": 1.60152549670911,
-      "grad_norm": 0.3947668363909619,
-      "learning_rate": 0.00011782279885231385,
-      "loss": 0.6962,
-      "step": 3260
-    },
-    {
-      "epoch": 1.602017592421726,
-      "grad_norm": 0.4243497046931594,
-      "learning_rate": 0.00011776220564386348,
-      "loss": 0.7828,
-      "step": 3261
-    },
-    {
-      "epoch": 1.6025096881343421,
-      "grad_norm": 0.3909881441241604,
-      "learning_rate": 0.00011770160570075248,
-      "loss": 0.7384,
-      "step": 3262
-    },
-    {
-      "epoch": 1.6030017838469584,
-      "grad_norm": 0.4110875557112792,
-      "learning_rate": 0.00011764099904595772,
-      "loss": 0.7701,
-      "step": 3263
-    },
-    {
-      "epoch": 1.6034938795595743,
-      "grad_norm": 0.3912785966690083,
-      "learning_rate": 0.0001175803857024586,
-      "loss": 0.7465,
-      "step": 3264
-    },
-    {
-      "epoch": 1.6039859752721903,
-      "grad_norm": 0.3911070561680225,
-      "learning_rate": 0.00011751976569323711,
-      "loss": 0.6896,
-      "step": 3265
-    },
-    {
-      "epoch": 1.6044780709848065,
-      "grad_norm": 0.4177901406196027,
-      "learning_rate": 0.00011745913904127769,
-      "loss": 0.7255,
-      "step": 3266
-    },
-    {
-      "epoch": 1.6049701666974228,
-      "grad_norm": 0.4128201313649831,
-      "learning_rate": 0.00011739850576956741,
-      "loss": 0.7424,
-      "step": 3267
-    },
-    {
-      "epoch": 1.6054622624100388,
-      "grad_norm": 0.4081329442940276,
-      "learning_rate": 0.00011733786590109577,
-      "loss": 0.764,
-      "step": 3268
-    },
-    {
-      "epoch": 1.6059543581226547,
-      "grad_norm": 0.4209126311839199,
-      "learning_rate": 0.00011727721945885473,
-      "loss": 0.7276,
-      "step": 3269
-    },
-    {
-      "epoch": 1.606446453835271,
-      "grad_norm": 0.39826639801393804,
-      "learning_rate": 0.00011721656646583885,
-      "loss": 0.7158,
-      "step": 3270
-    },
-    {
-      "epoch": 1.6069385495478872,
-      "grad_norm": 0.3878353242093959,
-      "learning_rate": 0.00011715590694504515,
-      "loss": 0.7083,
-      "step": 3271
-    },
-    {
-      "epoch": 1.6074306452605032,
-      "grad_norm": 0.4164322323191292,
-      "learning_rate": 0.00011709524091947304,
-      "loss": 0.7355,
-      "step": 3272
-    },
-    {
-      "epoch": 1.6079227409731192,
-      "grad_norm": 0.40521745639168855,
-      "learning_rate": 0.00011703456841212448,
-      "loss": 0.7585,
-      "step": 3273
-    },
-    {
-      "epoch": 1.6084148366857354,
-      "grad_norm": 0.42530553004267596,
-      "learning_rate": 0.00011697388944600385,
-      "loss": 0.8443,
-      "step": 3274
-    },
-    {
-      "epoch": 1.6089069323983516,
-      "grad_norm": 0.3811248910807747,
-      "learning_rate": 0.00011691320404411801,
-      "loss": 0.6883,
-      "step": 3275
-    },
-    {
-      "epoch": 1.6093990281109676,
-      "grad_norm": 0.40406970060832814,
-      "learning_rate": 0.00011685251222947621,
-      "loss": 0.7003,
-      "step": 3276
-    },
-    {
-      "epoch": 1.6098911238235836,
-      "grad_norm": 0.4323123527635906,
-      "learning_rate": 0.00011679181402509023,
-      "loss": 0.7701,
-      "step": 3277
-    },
-    {
-      "epoch": 1.6103832195361998,
-      "grad_norm": 0.4069240672567787,
-      "learning_rate": 0.00011673110945397414,
-      "loss": 0.6808,
-      "step": 3278
-    },
-    {
-      "epoch": 1.610875315248816,
-      "grad_norm": 0.42143949186525315,
-      "learning_rate": 0.0001166703985391445,
-      "loss": 0.7189,
-      "step": 3279
-    },
-    {
-      "epoch": 1.611367410961432,
-      "grad_norm": 0.4304924244856845,
-      "learning_rate": 0.00011660968130362029,
-      "loss": 0.7689,
-      "step": 3280
-    },
-    {
-      "epoch": 1.611859506674048,
-      "grad_norm": 0.46127782547072554,
-      "learning_rate": 0.00011654895777042285,
-      "loss": 0.8438,
-      "step": 3281
-    },
-    {
-      "epoch": 1.6123516023866642,
-      "grad_norm": 0.4029495862792948,
-      "learning_rate": 0.0001164882279625759,
-      "loss": 0.7118,
-      "step": 3282
-    },
-    {
-      "epoch": 1.6128436980992804,
-      "grad_norm": 0.40606752911623156,
-      "learning_rate": 0.0001164274919031056,
-      "loss": 0.7395,
-      "step": 3283
-    },
-    {
-      "epoch": 1.6133357938118964,
-      "grad_norm": 0.4299344055067732,
-      "learning_rate": 0.0001163667496150404,
-      "loss": 0.7283,
-      "step": 3284
-    },
-    {
-      "epoch": 1.6138278895245124,
-      "grad_norm": 0.41832702246026865,
-      "learning_rate": 0.00011630600112141112,
-      "loss": 0.817,
-      "step": 3285
-    },
-    {
-      "epoch": 1.6143199852371286,
-      "grad_norm": 0.3814569894630706,
-      "learning_rate": 0.00011624524644525108,
-      "loss": 0.7179,
-      "step": 3286
-    },
-    {
-      "epoch": 1.6148120809497448,
-      "grad_norm": 0.40109249265628677,
-      "learning_rate": 0.00011618448560959572,
-      "loss": 0.68,
-      "step": 3287
-    },
-    {
-      "epoch": 1.6153041766623608,
-      "grad_norm": 0.415736511427573,
-      "learning_rate": 0.00011612371863748295,
-      "loss": 0.7296,
-      "step": 3288
-    },
-    {
-      "epoch": 1.6157962723749768,
-      "grad_norm": 0.4701272415745234,
-      "learning_rate": 0.00011606294555195297,
-      "loss": 0.8022,
-      "step": 3289
-    },
-    {
-      "epoch": 1.616288368087593,
-      "grad_norm": 0.4110937527985462,
-      "learning_rate": 0.00011600216637604835,
-      "loss": 0.7109,
-      "step": 3290
-    },
-    {
-      "epoch": 1.6167804638002092,
-      "grad_norm": 0.3855358948624381,
-      "learning_rate": 0.0001159413811328139,
-      "loss": 0.6798,
-      "step": 3291
-    },
-    {
-      "epoch": 1.6172725595128252,
-      "grad_norm": 0.4043043048790532,
-      "learning_rate": 0.00011588058984529673,
-      "loss": 0.7102,
-      "step": 3292
-    },
-    {
-      "epoch": 1.6177646552254412,
-      "grad_norm": 0.41123227564716486,
-      "learning_rate": 0.00011581979253654632,
-      "loss": 0.7546,
-      "step": 3293
-    },
-    {
-      "epoch": 1.6182567509380574,
-      "grad_norm": 0.38828934945244903,
-      "learning_rate": 0.00011575898922961435,
-      "loss": 0.7667,
-      "step": 3294
-    },
-    {
-      "epoch": 1.6187488466506736,
-      "grad_norm": 0.3828094938101845,
-      "learning_rate": 0.00011569817994755481,
-      "loss": 0.6885,
-      "step": 3295
-    },
-    {
-      "epoch": 1.6192409423632896,
-      "grad_norm": 0.4440985368095448,
-      "learning_rate": 0.00011563736471342395,
-      "loss": 0.8214,
-      "step": 3296
-    },
-    {
-      "epoch": 1.6197330380759056,
-      "grad_norm": 0.408667806656832,
-      "learning_rate": 0.00011557654355028029,
-      "loss": 0.7572,
-      "step": 3297
-    },
-    {
-      "epoch": 1.6202251337885218,
-      "grad_norm": 0.4243654636971181,
-      "learning_rate": 0.00011551571648118456,
-      "loss": 0.7392,
-      "step": 3298
-    },
-    {
-      "epoch": 1.620717229501138,
-      "grad_norm": 0.38966391468078304,
-      "learning_rate": 0.0001154548835291998,
-      "loss": 0.785,
-      "step": 3299
-    },
-    {
-      "epoch": 1.621209325213754,
-      "grad_norm": 0.44030242312298823,
-      "learning_rate": 0.0001153940447173912,
-      "loss": 0.7674,
-      "step": 3300
-    },
-    {
-      "epoch": 1.62170142092637,
-      "grad_norm": 0.6004304936722245,
-      "learning_rate": 0.00011533320006882621,
-      "loss": 0.7561,
-      "step": 3301
-    },
-    {
-      "epoch": 1.6221935166389863,
-      "grad_norm": 0.4168449461682847,
-      "learning_rate": 0.00011527234960657449,
-      "loss": 0.7368,
-      "step": 3302
-    },
-    {
-      "epoch": 1.6226856123516025,
-      "grad_norm": 0.3858236200316087,
-      "learning_rate": 0.00011521149335370794,
-      "loss": 0.7126,
-      "step": 3303
-    },
-    {
-      "epoch": 1.6231777080642185,
-      "grad_norm": 0.3980531870370931,
-      "learning_rate": 0.00011515063133330057,
-      "loss": 0.7081,
-      "step": 3304
-    },
-    {
-      "epoch": 1.6236698037768345,
-      "grad_norm": 0.424134351707148,
-      "learning_rate": 0.00011508976356842867,
-      "loss": 0.7767,
-      "step": 3305
-    },
-    {
-      "epoch": 1.6241618994894507,
-      "grad_norm": 0.4138559895405696,
-      "learning_rate": 0.00011502889008217063,
-      "loss": 0.7079,
-      "step": 3306
-    },
-    {
-      "epoch": 1.6246539952020669,
-      "grad_norm": 0.4301282248555545,
-      "learning_rate": 0.00011496801089760709,
-      "loss": 0.7661,
-      "step": 3307
-    },
-    {
-      "epoch": 1.6251460909146829,
-      "grad_norm": 0.39714112078295116,
-      "learning_rate": 0.00011490712603782073,
-      "loss": 0.7219,
-      "step": 3308
-    },
-    {
-      "epoch": 1.6256381866272989,
-      "grad_norm": 0.40018756939891,
-      "learning_rate": 0.00011484623552589653,
-      "loss": 0.7125,
-      "step": 3309
-    },
-    {
-      "epoch": 1.626130282339915,
-      "grad_norm": 0.40254911696783763,
-      "learning_rate": 0.00011478533938492153,
-      "loss": 0.7989,
-      "step": 3310
-    },
-    {
-      "epoch": 1.6266223780525313,
-      "grad_norm": 0.38957602831300703,
-      "learning_rate": 0.00011472443763798486,
-      "loss": 0.7428,
-      "step": 3311
-    },
-    {
-      "epoch": 1.6271144737651473,
-      "grad_norm": 0.38243558218122387,
-      "learning_rate": 0.00011466353030817791,
-      "loss": 0.7082,
-      "step": 3312
-    },
-    {
-      "epoch": 1.6276065694777633,
-      "grad_norm": 0.42587027710463005,
-      "learning_rate": 0.00011460261741859406,
-      "loss": 0.7277,
-      "step": 3313
-    },
-    {
-      "epoch": 1.6280986651903795,
-      "grad_norm": 0.38962302245061,
-      "learning_rate": 0.00011454169899232885,
-      "loss": 0.7066,
-      "step": 3314
-    },
-    {
-      "epoch": 1.6285907609029957,
-      "grad_norm": 0.42216413584744017,
-      "learning_rate": 0.00011448077505247989,
-      "loss": 0.7611,
-      "step": 3315
-    },
-    {
-      "epoch": 1.6290828566156117,
-      "grad_norm": 0.40613777396095335,
-      "learning_rate": 0.00011441984562214693,
-      "loss": 0.6904,
-      "step": 3316
-    },
-    {
-      "epoch": 1.6295749523282277,
-      "grad_norm": 0.4074139553910867,
-      "learning_rate": 0.00011435891072443181,
-      "loss": 0.7217,
-      "step": 3317
-    },
-    {
-      "epoch": 1.630067048040844,
-      "grad_norm": 0.38582421711941,
-      "learning_rate": 0.00011429797038243836,
-      "loss": 0.7161,
-      "step": 3318
-    },
-    {
-      "epoch": 1.6305591437534601,
-      "grad_norm": 0.4309490690294988,
-      "learning_rate": 0.00011423702461927255,
-      "loss": 0.7732,
-      "step": 3319
-    },
-    {
-      "epoch": 1.631051239466076,
-      "grad_norm": 0.3907965510498859,
-      "learning_rate": 0.00011417607345804238,
-      "loss": 0.7222,
-      "step": 3320
-    },
-    {
-      "epoch": 1.631543335178692,
-      "grad_norm": 0.38648927105318526,
-      "learning_rate": 0.00011411511692185783,
-      "loss": 0.729,
-      "step": 3321
-    },
-    {
-      "epoch": 1.6320354308913083,
-      "grad_norm": 0.39061194647669145,
-      "learning_rate": 0.00011405415503383107,
-      "loss": 0.7122,
-      "step": 3322
-    },
-    {
-      "epoch": 1.6325275266039245,
-      "grad_norm": 0.4054147916082172,
-      "learning_rate": 0.0001139931878170762,
-      "loss": 0.6943,
-      "step": 3323
-    },
-    {
-      "epoch": 1.6330196223165405,
-      "grad_norm": 0.4217960383778778,
-      "learning_rate": 0.0001139322152947093,
-      "loss": 0.7283,
-      "step": 3324
-    },
-    {
-      "epoch": 1.6335117180291565,
-      "grad_norm": 0.410729409117959,
-      "learning_rate": 0.00011387123748984855,
-      "loss": 0.7121,
-      "step": 3325
-    },
-    {
-      "epoch": 1.6340038137417727,
-      "grad_norm": 0.38439346326153434,
-      "learning_rate": 0.00011381025442561415,
-      "loss": 0.6958,
-      "step": 3326
-    },
-    {
-      "epoch": 1.634495909454389,
-      "grad_norm": 0.3952011191182585,
-      "learning_rate": 0.00011374926612512814,
-      "loss": 0.6614,
-      "step": 3327
-    },
-    {
-      "epoch": 1.634988005167005,
-      "grad_norm": 0.39460101798607156,
-      "learning_rate": 0.00011368827261151473,
-      "loss": 0.7238,
-      "step": 3328
-    },
-    {
-      "epoch": 1.635480100879621,
-      "grad_norm": 0.4342295559628419,
-      "learning_rate": 0.00011362727390789998,
-      "loss": 0.7837,
-      "step": 3329
-    },
-    {
-      "epoch": 1.6359721965922371,
-      "grad_norm": 0.41323209537972694,
-      "learning_rate": 0.00011356627003741198,
-      "loss": 0.7662,
-      "step": 3330
-    },
-    {
-      "epoch": 1.6364642923048534,
-      "grad_norm": 0.461378329922537,
-      "learning_rate": 0.00011350526102318071,
-      "loss": 0.7566,
-      "step": 3331
-    },
-    {
-      "epoch": 1.6369563880174693,
-      "grad_norm": 0.5254097987571077,
-      "learning_rate": 0.00011344424688833823,
-      "loss": 0.7061,
-      "step": 3332
-    },
-    {
-      "epoch": 1.6374484837300853,
-      "grad_norm": 0.4000376526024257,
-      "learning_rate": 0.00011338322765601845,
-      "loss": 0.7018,
-      "step": 3333
-    },
-    {
-      "epoch": 1.6379405794427018,
-      "grad_norm": 0.43837849244177035,
-      "learning_rate": 0.00011332220334935715,
-      "loss": 0.6869,
-      "step": 3334
-    },
-    {
-      "epoch": 1.6384326751553178,
-      "grad_norm": 0.3881624402368388,
-      "learning_rate": 0.0001132611739914922,
-      "loss": 0.6675,
-      "step": 3335
-    },
-    {
-      "epoch": 1.6389247708679338,
-      "grad_norm": 0.39920920733537274,
-      "learning_rate": 0.00011320013960556326,
-      "loss": 0.7165,
-      "step": 3336
-    },
-    {
-      "epoch": 1.63941686658055,
-      "grad_norm": 0.39521704754594184,
-      "learning_rate": 0.00011313910021471192,
-      "loss": 0.7171,
-      "step": 3337
-    },
-    {
-      "epoch": 1.6399089622931662,
-      "grad_norm": 0.40696160903551065,
-      "learning_rate": 0.00011307805584208167,
-      "loss": 0.7165,
-      "step": 3338
-    },
-    {
-      "epoch": 1.6404010580057822,
-      "grad_norm": 0.40674244664924436,
-      "learning_rate": 0.00011301700651081795,
-      "loss": 0.713,
-      "step": 3339
-    },
-    {
-      "epoch": 1.6408931537183982,
-      "grad_norm": 0.39380716623228934,
-      "learning_rate": 0.00011295595224406796,
-      "loss": 0.7077,
-      "step": 3340
-    },
-    {
-      "epoch": 1.6413852494310144,
-      "grad_norm": 0.4170624213001939,
-      "learning_rate": 0.00011289489306498091,
-      "loss": 0.755,
-      "step": 3341
-    },
-    {
-      "epoch": 1.6418773451436306,
-      "grad_norm": 0.40302670170896127,
-      "learning_rate": 0.00011283382899670774,
-      "loss": 0.7103,
-      "step": 3342
-    },
-    {
-      "epoch": 1.6423694408562466,
-      "grad_norm": 0.40330042413991274,
-      "learning_rate": 0.00011277276006240133,
-      "loss": 0.7599,
-      "step": 3343
-    },
-    {
-      "epoch": 1.6428615365688626,
-      "grad_norm": 0.38296817165820113,
-      "learning_rate": 0.00011271168628521636,
-      "loss": 0.6903,
-      "step": 3344
-    },
-    {
-      "epoch": 1.6433536322814788,
-      "grad_norm": 0.38312626311902653,
-      "learning_rate": 0.00011265060768830942,
-      "loss": 0.6818,
-      "step": 3345
-    },
-    {
-      "epoch": 1.643845727994095,
-      "grad_norm": 0.3881725372107873,
-      "learning_rate": 0.00011258952429483882,
-      "loss": 0.7075,
-      "step": 3346
-    },
-    {
-      "epoch": 1.644337823706711,
-      "grad_norm": 0.4541579945274235,
-      "learning_rate": 0.00011252843612796475,
-      "loss": 0.7642,
-      "step": 3347
-    },
-    {
-      "epoch": 1.644829919419327,
-      "grad_norm": 0.4033663840772927,
-      "learning_rate": 0.00011246734321084925,
-      "loss": 0.7589,
-      "step": 3348
-    },
-    {
-      "epoch": 1.6453220151319432,
-      "grad_norm": 0.4523417049892899,
-      "learning_rate": 0.00011240624556665605,
-      "loss": 0.8577,
-      "step": 3349
-    },
-    {
-      "epoch": 1.6458141108445594,
-      "grad_norm": 0.41245399605827937,
-      "learning_rate": 0.00011234514321855078,
-      "loss": 0.7973,
-      "step": 3350
-    },
-    {
-      "epoch": 1.6463062065571754,
-      "grad_norm": 0.4043051870315983,
-      "learning_rate": 0.00011228403618970078,
-      "loss": 0.7732,
-      "step": 3351
-    },
-    {
-      "epoch": 1.6467983022697914,
-      "grad_norm": 0.4375866898033138,
-      "learning_rate": 0.00011222292450327523,
-      "loss": 0.7484,
-      "step": 3352
-    },
-    {
-      "epoch": 1.6472903979824076,
-      "grad_norm": 0.42413886901311715,
-      "learning_rate": 0.000112161808182445,
-      "loss": 0.6895,
-      "step": 3353
-    },
-    {
-      "epoch": 1.6477824936950238,
-      "grad_norm": 0.3601379139559961,
-      "learning_rate": 0.00011210068725038277,
-      "loss": 0.6597,
-      "step": 3354
-    },
-    {
-      "epoch": 1.6482745894076398,
-      "grad_norm": 0.41712540386337577,
-      "learning_rate": 0.00011203956173026297,
-      "loss": 0.7566,
-      "step": 3355
-    },
-    {
-      "epoch": 1.6487666851202558,
-      "grad_norm": 0.40304211283122093,
-      "learning_rate": 0.00011197843164526173,
-      "loss": 0.6943,
-      "step": 3356
-    },
-    {
-      "epoch": 1.649258780832872,
-      "grad_norm": 0.4060560773043631,
-      "learning_rate": 0.00011191729701855696,
-      "loss": 0.7507,
-      "step": 3357
-    },
-    {
-      "epoch": 1.6497508765454882,
-      "grad_norm": 0.4200077621192976,
-      "learning_rate": 0.00011185615787332826,
-      "loss": 0.7193,
-      "step": 3358
-    },
-    {
-      "epoch": 1.6502429722581042,
-      "grad_norm": 0.4135040350138502,
-      "learning_rate": 0.00011179501423275698,
-      "loss": 0.6496,
-      "step": 3359
-    },
-    {
-      "epoch": 1.6507350679707202,
-      "grad_norm": 0.4157681663944609,
-      "learning_rate": 0.00011173386612002605,
-      "loss": 0.7406,
-      "step": 3360
-    },
-    {
-      "epoch": 1.6512271636833364,
-      "grad_norm": 0.45892053218899187,
-      "learning_rate": 0.00011167271355832033,
-      "loss": 0.7491,
-      "step": 3361
-    },
-    {
-      "epoch": 1.6517192593959527,
-      "grad_norm": 0.4739170942789628,
-      "learning_rate": 0.00011161155657082611,
-      "loss": 0.716,
-      "step": 3362
-    },
-    {
-      "epoch": 1.6522113551085686,
-      "grad_norm": 0.38768998241486385,
-      "learning_rate": 0.00011155039518073156,
-      "loss": 0.6988,
-      "step": 3363
-    },
-    {
-      "epoch": 1.6527034508211846,
-      "grad_norm": 0.3778780511974089,
-      "learning_rate": 0.00011148922941122637,
-      "loss": 0.7272,
-      "step": 3364
-    },
-    {
-      "epoch": 1.6531955465338009,
-      "grad_norm": 0.4137502099541195,
-      "learning_rate": 0.000111428059285502,
-      "loss": 0.7088,
-      "step": 3365
-    },
-    {
-      "epoch": 1.653687642246417,
-      "grad_norm": 0.4033041644605808,
-      "learning_rate": 0.0001113668848267515,
-      "loss": 0.6829,
-      "step": 3366
-    },
-    {
-      "epoch": 1.654179737959033,
-      "grad_norm": 0.4126681957030245,
-      "learning_rate": 0.00011130570605816956,
-      "loss": 0.7165,
-      "step": 3367
-    },
-    {
-      "epoch": 1.654671833671649,
-      "grad_norm": 0.4190034080828862,
-      "learning_rate": 0.00011124452300295256,
-      "loss": 0.7627,
-      "step": 3368
-    },
-    {
-      "epoch": 1.6551639293842653,
-      "grad_norm": 0.428151834297233,
-      "learning_rate": 0.00011118333568429848,
-      "loss": 0.787,
-      "step": 3369
-    },
-    {
-      "epoch": 1.6556560250968815,
-      "grad_norm": 0.3905762729161534,
-      "learning_rate": 0.00011112214412540685,
-      "loss": 0.677,
-      "step": 3370
-    },
-    {
-      "epoch": 1.6561481208094975,
-      "grad_norm": 0.40359898466168803,
-      "learning_rate": 0.00011106094834947891,
-      "loss": 0.7339,
-      "step": 3371
-    },
-    {
-      "epoch": 1.6566402165221135,
-      "grad_norm": 0.4019980640244289,
-      "learning_rate": 0.00011099974837971745,
-      "loss": 0.7641,
-      "step": 3372
-    },
-    {
-      "epoch": 1.6571323122347297,
-      "grad_norm": 0.4012060101976718,
-      "learning_rate": 0.00011093854423932684,
-      "loss": 0.7098,
-      "step": 3373
-    },
-    {
-      "epoch": 1.657624407947346,
-      "grad_norm": 0.3829638774866954,
-      "learning_rate": 0.00011087733595151306,
-      "loss": 0.715,
-      "step": 3374
-    },
-    {
-      "epoch": 1.6581165036599619,
-      "grad_norm": 0.4257124339227231,
-      "learning_rate": 0.00011081612353948363,
-      "loss": 0.7411,
-      "step": 3375
-    },
-    {
-      "epoch": 1.6586085993725779,
-      "grad_norm": 0.4234409047536631,
-      "learning_rate": 0.00011075490702644765,
-      "loss": 0.7731,
-      "step": 3376
-    },
-    {
-      "epoch": 1.659100695085194,
-      "grad_norm": 0.3861076102536674,
-      "learning_rate": 0.00011069368643561578,
-      "loss": 0.6559,
-      "step": 3377
-    },
-    {
-      "epoch": 1.6595927907978103,
-      "grad_norm": 0.40596536069740446,
-      "learning_rate": 0.00011063246179020022,
-      "loss": 0.7371,
-      "step": 3378
-    },
-    {
-      "epoch": 1.6600848865104263,
-      "grad_norm": 0.41494050533596094,
-      "learning_rate": 0.00011057123311341473,
-      "loss": 0.6786,
-      "step": 3379
-    },
-    {
-      "epoch": 1.6605769822230423,
-      "grad_norm": 0.5018810497425636,
-      "learning_rate": 0.00011051000042847453,
-      "loss": 0.8102,
-      "step": 3380
-    },
-    {
-      "epoch": 1.6610690779356585,
-      "grad_norm": 0.43553815752747455,
-      "learning_rate": 0.00011044876375859647,
-      "loss": 0.752,
-      "step": 3381
-    },
-    {
-      "epoch": 1.6615611736482747,
-      "grad_norm": 0.42698662255119313,
-      "learning_rate": 0.00011038752312699883,
-      "loss": 0.6795,
-      "step": 3382
-    },
-    {
-      "epoch": 1.6620532693608907,
-      "grad_norm": 0.4173636835720084,
-      "learning_rate": 0.00011032627855690137,
-      "loss": 0.7322,
-      "step": 3383
-    },
-    {
-      "epoch": 1.6625453650735067,
-      "grad_norm": 0.3806345058818375,
-      "learning_rate": 0.00011026503007152542,
-      "loss": 0.7413,
-      "step": 3384
-    },
-    {
-      "epoch": 1.663037460786123,
-      "grad_norm": 0.4151686021688595,
-      "learning_rate": 0.00011020377769409376,
-      "loss": 0.7111,
-      "step": 3385
-    },
-    {
-      "epoch": 1.6635295564987391,
-      "grad_norm": 0.43133613729928794,
-      "learning_rate": 0.00011014252144783061,
-      "loss": 0.7359,
-      "step": 3386
-    },
-    {
-      "epoch": 1.6640216522113551,
-      "grad_norm": 0.4201512584519894,
-      "learning_rate": 0.00011008126135596175,
-      "loss": 0.6755,
-      "step": 3387
-    },
-    {
-      "epoch": 1.664513747923971,
-      "grad_norm": 0.4089224483162079,
-      "learning_rate": 0.00011001999744171431,
-      "loss": 0.7419,
-      "step": 3388
-    },
-    {
-      "epoch": 1.6650058436365873,
-      "grad_norm": 0.4260506493443712,
-      "learning_rate": 0.00010995872972831694,
-      "loss": 0.7322,
-      "step": 3389
-    },
-    {
-      "epoch": 1.6654979393492035,
-      "grad_norm": 0.3964593546037636,
-      "learning_rate": 0.00010989745823899968,
-      "loss": 0.742,
-      "step": 3390
-    },
-    {
-      "epoch": 1.6659900350618195,
-      "grad_norm": 0.4138708548288306,
-      "learning_rate": 0.00010983618299699407,
-      "loss": 0.7447,
-      "step": 3391
-    },
-    {
-      "epoch": 1.6664821307744355,
-      "grad_norm": 0.3918383348638912,
-      "learning_rate": 0.000109774904025533,
-      "loss": 0.683,
-      "step": 3392
-    },
-    {
-      "epoch": 1.6669742264870517,
-      "grad_norm": 0.42073281948385244,
-      "learning_rate": 0.00010971362134785081,
-      "loss": 0.7421,
-      "step": 3393
-    },
-    {
-      "epoch": 1.667466322199668,
-      "grad_norm": 0.46714175568654714,
-      "learning_rate": 0.0001096523349871833,
-      "loss": 0.7444,
-      "step": 3394
-    },
-    {
-      "epoch": 1.667958417912284,
-      "grad_norm": 0.4192828537110579,
-      "learning_rate": 0.00010959104496676753,
-      "loss": 0.7158,
-      "step": 3395
-    },
-    {
-      "epoch": 1.6684505136249,
-      "grad_norm": 0.43638900066541814,
-      "learning_rate": 0.00010952975130984209,
-      "loss": 0.6868,
-      "step": 3396
-    },
-    {
-      "epoch": 1.6689426093375161,
-      "grad_norm": 0.40577543224584955,
-      "learning_rate": 0.00010946845403964683,
-      "loss": 0.7062,
-      "step": 3397
-    },
-    {
-      "epoch": 1.6694347050501324,
-      "grad_norm": 0.40755905873082676,
-      "learning_rate": 0.00010940715317942308,
-      "loss": 0.7403,
-      "step": 3398
-    },
-    {
-      "epoch": 1.6699268007627484,
-      "grad_norm": 0.4089856597537415,
-      "learning_rate": 0.00010934584875241342,
-      "loss": 0.7309,
-      "step": 3399
-    },
-    {
-      "epoch": 1.6704188964753643,
-      "grad_norm": 0.41081620165600885,
-      "learning_rate": 0.00010928454078186186,
-      "loss": 0.722,
-      "step": 3400
-    },
-    {
-      "epoch": 1.6709109921879806,
-      "grad_norm": 0.4135880756035975,
-      "learning_rate": 0.00010922322929101377,
-      "loss": 0.6873,
-      "step": 3401
-    },
-    {
-      "epoch": 1.6714030879005968,
-      "grad_norm": 0.41057211036682234,
-      "learning_rate": 0.00010916191430311576,
-      "loss": 0.737,
-      "step": 3402
-    },
-    {
-      "epoch": 1.6718951836132128,
-      "grad_norm": 0.4495536689809271,
-      "learning_rate": 0.00010910059584141585,
-      "loss": 0.8352,
-      "step": 3403
-    },
-    {
-      "epoch": 1.6723872793258288,
-      "grad_norm": 0.42069116659953226,
-      "learning_rate": 0.00010903927392916335,
-      "loss": 0.6848,
-      "step": 3404
-    },
-    {
-      "epoch": 1.672879375038445,
-      "grad_norm": 0.4062013235084829,
-      "learning_rate": 0.0001089779485896089,
-      "loss": 0.7541,
-      "step": 3405
-    },
-    {
-      "epoch": 1.6733714707510612,
-      "grad_norm": 0.40290597719805843,
-      "learning_rate": 0.00010891661984600437,
-      "loss": 0.7067,
-      "step": 3406
-    },
-    {
-      "epoch": 1.6738635664636772,
-      "grad_norm": 0.3858918378569007,
-      "learning_rate": 0.00010885528772160303,
-      "loss": 0.6757,
-      "step": 3407
-    },
-    {
-      "epoch": 1.6743556621762932,
-      "grad_norm": 0.4071178562868276,
-      "learning_rate": 0.00010879395223965932,
-      "loss": 0.7594,
-      "step": 3408
-    },
-    {
-      "epoch": 1.6748477578889094,
-      "grad_norm": 0.42233551115816875,
-      "learning_rate": 0.00010873261342342902,
-      "loss": 0.7572,
-      "step": 3409
-    },
-    {
-      "epoch": 1.6753398536015256,
-      "grad_norm": 0.3697017015625988,
-      "learning_rate": 0.00010867127129616917,
-      "loss": 0.7018,
-      "step": 3410
-    },
-    {
-      "epoch": 1.6758319493141416,
-      "grad_norm": 0.38368958954840904,
-      "learning_rate": 0.00010860992588113802,
-      "loss": 0.7188,
-      "step": 3411
-    },
-    {
-      "epoch": 1.6763240450267576,
-      "grad_norm": 0.391557735397661,
-      "learning_rate": 0.00010854857720159515,
-      "loss": 0.6693,
-      "step": 3412
-    },
-    {
-      "epoch": 1.6768161407393738,
-      "grad_norm": 0.432049371194674,
-      "learning_rate": 0.00010848722528080124,
-      "loss": 0.7832,
-      "step": 3413
-    },
-    {
-      "epoch": 1.67730823645199,
-      "grad_norm": 0.39263819815769896,
-      "learning_rate": 0.0001084258701420184,
-      "loss": 0.7177,
-      "step": 3414
-    },
-    {
-      "epoch": 1.677800332164606,
-      "grad_norm": 0.4150764571185619,
-      "learning_rate": 0.00010836451180850977,
-      "loss": 0.6902,
-      "step": 3415
-    },
-    {
-      "epoch": 1.678292427877222,
-      "grad_norm": 0.4143147513945901,
-      "learning_rate": 0.00010830315030353979,
-      "loss": 0.7163,
-      "step": 3416
-    },
-    {
-      "epoch": 1.6787845235898382,
-      "grad_norm": 0.3921533202636694,
-      "learning_rate": 0.00010824178565037413,
-      "loss": 0.6945,
-      "step": 3417
-    },
-    {
-      "epoch": 1.6792766193024544,
-      "grad_norm": 0.4343821634061841,
-      "learning_rate": 0.00010818041787227957,
-      "loss": 0.7652,
-      "step": 3418
-    },
-    {
-      "epoch": 1.6797687150150704,
-      "grad_norm": 0.6397532413548314,
-      "learning_rate": 0.00010811904699252416,
-      "loss": 0.674,
-      "step": 3419
-    },
-    {
-      "epoch": 1.6802608107276864,
-      "grad_norm": 0.43034617515641505,
-      "learning_rate": 0.00010805767303437702,
-      "loss": 0.762,
-      "step": 3420
-    },
-    {
-      "epoch": 1.6807529064403026,
-      "grad_norm": 0.4197674143293288,
-      "learning_rate": 0.00010799629602110857,
-      "loss": 0.6872,
-      "step": 3421
-    },
-    {
-      "epoch": 1.6812450021529188,
-      "grad_norm": 0.40073736756931694,
-      "learning_rate": 0.00010793491597599026,
-      "loss": 0.7064,
-      "step": 3422
-    },
-    {
-      "epoch": 1.6817370978655348,
-      "grad_norm": 0.384927613805364,
-      "learning_rate": 0.00010787353292229478,
-      "loss": 0.7071,
-      "step": 3423
-    },
-    {
-      "epoch": 1.6822291935781508,
-      "grad_norm": 0.4107789427659322,
-      "learning_rate": 0.00010781214688329598,
-      "loss": 0.7093,
-      "step": 3424
-    },
-    {
-      "epoch": 1.682721289290767,
-      "grad_norm": 0.39193366996824297,
-      "learning_rate": 0.00010775075788226872,
-      "loss": 0.6758,
-      "step": 3425
-    },
-    {
-      "epoch": 1.6832133850033832,
-      "grad_norm": 0.4041036729796495,
-      "learning_rate": 0.00010768936594248904,
-      "loss": 0.7216,
-      "step": 3426
-    },
-    {
-      "epoch": 1.6837054807159992,
-      "grad_norm": 0.40704820513533807,
-      "learning_rate": 0.00010762797108723419,
-      "loss": 0.7579,
-      "step": 3427
-    },
-    {
-      "epoch": 1.6841975764286152,
-      "grad_norm": 0.40699307998332807,
-      "learning_rate": 0.00010756657333978242,
-      "loss": 0.6755,
-      "step": 3428
-    },
-    {
-      "epoch": 1.6846896721412314,
-      "grad_norm": 0.4152553427668556,
-      "learning_rate": 0.00010750517272341305,
-      "loss": 0.7296,
-      "step": 3429
-    },
-    {
-      "epoch": 1.6851817678538477,
-      "grad_norm": 0.3861353882469795,
-      "learning_rate": 0.0001074437692614066,
-      "loss": 0.69,
-      "step": 3430
-    },
-    {
-      "epoch": 1.6856738635664636,
-      "grad_norm": 0.38661979479447617,
-      "learning_rate": 0.00010738236297704458,
-      "loss": 0.7068,
-      "step": 3431
-    },
-    {
-      "epoch": 1.6861659592790796,
-      "grad_norm": 0.3830334849957862,
-      "learning_rate": 0.0001073209538936096,
-      "loss": 0.6544,
-      "step": 3432
-    },
-    {
-      "epoch": 1.6866580549916959,
-      "grad_norm": 0.4037450767030282,
-      "learning_rate": 0.00010725954203438538,
-      "loss": 0.7059,
-      "step": 3433
-    },
-    {
-      "epoch": 1.687150150704312,
-      "grad_norm": 0.42276141430010444,
-      "learning_rate": 0.00010719812742265656,
-      "loss": 0.7198,
-      "step": 3434
-    },
-    {
-      "epoch": 1.687642246416928,
-      "grad_norm": 0.41011513013201595,
-      "learning_rate": 0.00010713671008170895,
-      "loss": 0.6751,
-      "step": 3435
-    },
-    {
-      "epoch": 1.688134342129544,
-      "grad_norm": 0.42142869323810805,
-      "learning_rate": 0.00010707529003482932,
-      "loss": 0.7365,
-      "step": 3436
-    },
-    {
-      "epoch": 1.6886264378421603,
-      "grad_norm": 0.41207768141580325,
-      "learning_rate": 0.00010701386730530556,
-      "loss": 0.7399,
-      "step": 3437
-    },
-    {
-      "epoch": 1.6891185335547765,
-      "grad_norm": 0.37682752812936443,
-      "learning_rate": 0.00010695244191642648,
-      "loss": 0.6606,
-      "step": 3438
-    },
-    {
-      "epoch": 1.6896106292673925,
-      "grad_norm": 0.37990615917430404,
-      "learning_rate": 0.00010689101389148188,
-      "loss": 0.6944,
-      "step": 3439
-    },
-    {
-      "epoch": 1.6901027249800085,
-      "grad_norm": 0.3965579495629195,
-      "learning_rate": 0.00010682958325376271,
-      "loss": 0.7264,
-      "step": 3440
-    },
-    {
-      "epoch": 1.6905948206926247,
-      "grad_norm": 0.41776268306578,
-      "learning_rate": 0.00010676815002656076,
-      "loss": 0.7259,
-      "step": 3441
-    },
-    {
-      "epoch": 1.691086916405241,
-      "grad_norm": 0.3739672904934821,
-      "learning_rate": 0.00010670671423316884,
-      "loss": 0.6501,
-      "step": 3442
-    },
-    {
-      "epoch": 1.6915790121178569,
-      "grad_norm": 0.39365934144674114,
-      "learning_rate": 0.00010664527589688079,
-      "loss": 0.723,
-      "step": 3443
-    },
-    {
-      "epoch": 1.6920711078304729,
-      "grad_norm": 0.43340793664309574,
-      "learning_rate": 0.00010658383504099134,
-      "loss": 0.6439,
-      "step": 3444
-    },
-    {
-      "epoch": 1.692563203543089,
-      "grad_norm": 0.3905102487641378,
-      "learning_rate": 0.00010652239168879619,
-      "loss": 0.7088,
-      "step": 3445
-    },
-    {
-      "epoch": 1.6930552992557053,
-      "grad_norm": 0.4149420738539436,
-      "learning_rate": 0.00010646094586359203,
-      "loss": 0.765,
-      "step": 3446
-    },
-    {
-      "epoch": 1.6935473949683213,
-      "grad_norm": 0.4111554063228533,
-      "learning_rate": 0.00010639949758867649,
-      "loss": 0.7387,
-      "step": 3447
-    },
-    {
-      "epoch": 1.6940394906809373,
-      "grad_norm": 0.4051040945649876,
-      "learning_rate": 0.00010633804688734806,
-      "loss": 0.6392,
-      "step": 3448
-    },
-    {
-      "epoch": 1.6945315863935535,
-      "grad_norm": 0.42401110785586005,
-      "learning_rate": 0.00010627659378290617,
-      "loss": 0.7706,
-      "step": 3449
-    },
-    {
-      "epoch": 1.6950236821061697,
-      "grad_norm": 0.4098836282687331,
-      "learning_rate": 0.00010621513829865124,
-      "loss": 0.7824,
-      "step": 3450
-    },
-    {
-      "epoch": 1.6955157778187857,
-      "grad_norm": 0.3953800504291363,
-      "learning_rate": 0.0001061536804578845,
-      "loss": 0.7369,
-      "step": 3451
-    },
-    {
-      "epoch": 1.6960078735314017,
-      "grad_norm": 0.3879011551239534,
-      "learning_rate": 0.00010609222028390808,
-      "loss": 0.674,
-      "step": 3452
-    },
-    {
-      "epoch": 1.696499969244018,
-      "grad_norm": 0.3936995377938904,
-      "learning_rate": 0.00010603075780002507,
-      "loss": 0.7182,
-      "step": 3453
-    },
-    {
-      "epoch": 1.6969920649566341,
-      "grad_norm": 0.4124503683356637,
-      "learning_rate": 0.00010596929302953937,
-      "loss": 0.7175,
-      "step": 3454
-    },
-    {
-      "epoch": 1.6974841606692501,
-      "grad_norm": 0.4221111801284014,
-      "learning_rate": 0.00010590782599575578,
-      "loss": 0.6894,
-      "step": 3455
-    },
-    {
-      "epoch": 1.6979762563818661,
-      "grad_norm": 0.38030934929246407,
-      "learning_rate": 0.0001058463567219799,
-      "loss": 0.6647,
-      "step": 3456
-    },
-    {
-      "epoch": 1.6984683520944823,
-      "grad_norm": 0.4356937652188898,
-      "learning_rate": 0.00010578488523151829,
-      "loss": 0.734,
-      "step": 3457
-    },
-    {
-      "epoch": 1.6989604478070985,
-      "grad_norm": 0.4079031161575368,
-      "learning_rate": 0.00010572341154767817,
-      "loss": 0.6947,
-      "step": 3458
-    },
-    {
-      "epoch": 1.6994525435197145,
-      "grad_norm": 0.40368267187106555,
-      "learning_rate": 0.0001056619356937678,
-      "loss": 0.6886,
-      "step": 3459
-    },
-    {
-      "epoch": 1.6999446392323305,
-      "grad_norm": 0.39113338766393774,
-      "learning_rate": 0.00010560045769309617,
-      "loss": 0.6906,
-      "step": 3460
-    },
-    {
-      "epoch": 1.7004367349449467,
-      "grad_norm": 0.38863425958578685,
-      "learning_rate": 0.00010553897756897304,
-      "loss": 0.771,
-      "step": 3461
-    },
-    {
-      "epoch": 1.700928830657563,
-      "grad_norm": 0.39353367669803496,
-      "learning_rate": 0.00010547749534470898,
-      "loss": 0.6912,
-      "step": 3462
-    },
-    {
-      "epoch": 1.701420926370179,
-      "grad_norm": 0.382638751415916,
-      "learning_rate": 0.0001054160110436155,
-      "loss": 0.6611,
-      "step": 3463
-    },
-    {
-      "epoch": 1.7019130220827952,
-      "grad_norm": 0.4416567025751971,
-      "learning_rate": 0.00010535452468900471,
-      "loss": 0.6856,
-      "step": 3464
-    },
-    {
-      "epoch": 1.7024051177954114,
-      "grad_norm": 0.3989790595531368,
-      "learning_rate": 0.00010529303630418959,
-      "loss": 0.7047,
-      "step": 3465
-    },
-    {
-      "epoch": 1.7028972135080274,
-      "grad_norm": 0.4302125927914887,
-      "learning_rate": 0.00010523154591248387,
-      "loss": 0.6971,
-      "step": 3466
-    },
-    {
-      "epoch": 1.7033893092206434,
-      "grad_norm": 0.3897444973557634,
-      "learning_rate": 0.00010517005353720208,
-      "loss": 0.6806,
-      "step": 3467
-    },
-    {
-      "epoch": 1.7038814049332596,
-      "grad_norm": 0.3998831439045781,
-      "learning_rate": 0.00010510855920165944,
-      "loss": 0.7455,
-      "step": 3468
-    },
-    {
-      "epoch": 1.7043735006458758,
-      "grad_norm": 0.434049063496724,
-      "learning_rate": 0.00010504706292917196,
-      "loss": 0.7863,
-      "step": 3469
-    },
-    {
-      "epoch": 1.7048655963584918,
-      "grad_norm": 0.362344809714173,
-      "learning_rate": 0.00010498556474305638,
-      "loss": 0.6478,
-      "step": 3470
-    },
-    {
-      "epoch": 1.7053576920711078,
-      "grad_norm": 0.4004932249867723,
-      "learning_rate": 0.00010492406466663012,
-      "loss": 0.7207,
-      "step": 3471
-    },
-    {
-      "epoch": 1.705849787783724,
-      "grad_norm": 0.3847313072495366,
-      "learning_rate": 0.00010486256272321137,
-      "loss": 0.7137,
-      "step": 3472
-    },
-    {
-      "epoch": 1.7063418834963402,
-      "grad_norm": 0.40582667619910506,
-      "learning_rate": 0.00010480105893611902,
-      "loss": 0.6809,
-      "step": 3473
-    },
-    {
-      "epoch": 1.7068339792089562,
-      "grad_norm": 0.4081347903126302,
-      "learning_rate": 0.00010473955332867265,
-      "loss": 0.7758,
-      "step": 3474
-    },
-    {
-      "epoch": 1.7073260749215722,
-      "grad_norm": 0.39818872530915195,
-      "learning_rate": 0.00010467804592419248,
-      "loss": 0.6788,
-      "step": 3475
-    },
-    {
-      "epoch": 1.7078181706341884,
-      "grad_norm": 0.4315985147326263,
-      "learning_rate": 0.00010461653674599951,
-      "loss": 0.7231,
-      "step": 3476
-    },
-    {
-      "epoch": 1.7083102663468046,
-      "grad_norm": 0.36912180977733466,
-      "learning_rate": 0.00010455502581741536,
-      "loss": 0.6004,
-      "step": 3477
-    },
-    {
-      "epoch": 1.7088023620594206,
-      "grad_norm": 0.4252613710701617,
-      "learning_rate": 0.0001044935131617623,
-      "loss": 0.7596,
-      "step": 3478
-    },
-    {
-      "epoch": 1.7092944577720366,
-      "grad_norm": 0.411811627204728,
-      "learning_rate": 0.00010443199880236325,
-      "loss": 0.7626,
-      "step": 3479
-    },
-    {
-      "epoch": 1.7097865534846528,
-      "grad_norm": 0.4087737533711953,
-      "learning_rate": 0.00010437048276254185,
-      "loss": 0.7217,
-      "step": 3480
-    },
-    {
-      "epoch": 1.710278649197269,
-      "grad_norm": 0.3867689710603881,
-      "learning_rate": 0.00010430896506562224,
-      "loss": 0.697,
-      "step": 3481
-    },
-    {
-      "epoch": 1.710770744909885,
-      "grad_norm": 0.39001765816253,
-      "learning_rate": 0.00010424744573492937,
-      "loss": 0.7491,
-      "step": 3482
-    },
-    {
-      "epoch": 1.711262840622501,
-      "grad_norm": 0.4040324344323463,
-      "learning_rate": 0.00010418592479378863,
-      "loss": 0.6807,
-      "step": 3483
-    },
-    {
-      "epoch": 1.7117549363351172,
-      "grad_norm": 0.39205894436891514,
-      "learning_rate": 0.00010412440226552618,
-      "loss": 0.7686,
-      "step": 3484
-    },
-    {
-      "epoch": 1.7122470320477334,
-      "grad_norm": 0.39573002089340803,
-      "learning_rate": 0.0001040628781734686,
-      "loss": 0.6872,
-      "step": 3485
-    },
-    {
-      "epoch": 1.7127391277603494,
-      "grad_norm": 0.38844382559662977,
-      "learning_rate": 0.00010400135254094328,
-      "loss": 0.6386,
-      "step": 3486
-    },
-    {
-      "epoch": 1.7132312234729654,
-      "grad_norm": 0.4053088198277832,
-      "learning_rate": 0.000103939825391278,
-      "loss": 0.7674,
-      "step": 3487
-    },
-    {
-      "epoch": 1.7137233191855816,
-      "grad_norm": 0.41086119397989745,
-      "learning_rate": 0.00010387829674780123,
-      "loss": 0.7798,
-      "step": 3488
-    },
-    {
-      "epoch": 1.7142154148981978,
-      "grad_norm": 0.40896189748626177,
-      "learning_rate": 0.00010381676663384196,
-      "loss": 0.7271,
-      "step": 3489
-    },
-    {
-      "epoch": 1.7147075106108138,
-      "grad_norm": 0.38509716197342253,
-      "learning_rate": 0.00010375523507272975,
-      "loss": 0.6816,
-      "step": 3490
-    },
-    {
-      "epoch": 1.7151996063234298,
-      "grad_norm": 0.3822168242330676,
-      "learning_rate": 0.00010369370208779469,
-      "loss": 0.689,
-      "step": 3491
-    },
-    {
-      "epoch": 1.715691702036046,
-      "grad_norm": 0.39191773645131417,
-      "learning_rate": 0.0001036321677023675,
-      "loss": 0.7319,
-      "step": 3492
-    },
-    {
-      "epoch": 1.7161837977486623,
-      "grad_norm": 0.43689208804666463,
-      "learning_rate": 0.0001035706319397793,
-      "loss": 0.7913,
-      "step": 3493
-    },
-    {
-      "epoch": 1.7166758934612782,
-      "grad_norm": 0.40220947564620113,
-      "learning_rate": 0.00010350909482336176,
-      "loss": 0.7456,
-      "step": 3494
-    },
-    {
-      "epoch": 1.7171679891738942,
-      "grad_norm": 0.4026476252829011,
-      "learning_rate": 0.00010344755637644717,
-      "loss": 0.7621,
-      "step": 3495
-    },
-    {
-      "epoch": 1.7176600848865105,
-      "grad_norm": 0.4354533643323663,
-      "learning_rate": 0.00010338601662236823,
-      "loss": 0.7264,
-      "step": 3496
-    },
-    {
-      "epoch": 1.7181521805991267,
-      "grad_norm": 0.4465269393370047,
-      "learning_rate": 0.0001033244755844581,
-      "loss": 0.7096,
-      "step": 3497
-    },
-    {
-      "epoch": 1.7186442763117427,
-      "grad_norm": 0.38531188475617795,
-      "learning_rate": 0.00010326293328605052,
-      "loss": 0.7259,
-      "step": 3498
-    },
-    {
-      "epoch": 1.7191363720243586,
-      "grad_norm": 0.3955474469345509,
-      "learning_rate": 0.00010320138975047971,
-      "loss": 0.6948,
-      "step": 3499
-    },
-    {
-      "epoch": 1.7196284677369749,
-      "grad_norm": 0.372532556838744,
-      "learning_rate": 0.00010313984500108025,
-      "loss": 0.653,
-      "step": 3500
-    },
-    {
-      "epoch": 1.720120563449591,
-      "grad_norm": 0.3817651953050749,
-      "learning_rate": 0.0001030782990611873,
-      "loss": 0.7308,
-      "step": 3501
-    },
-    {
-      "epoch": 1.720612659162207,
-      "grad_norm": 0.43360979815950806,
-      "learning_rate": 0.0001030167519541364,
-      "loss": 0.7916,
-      "step": 3502
-    },
-    {
-      "epoch": 1.721104754874823,
-      "grad_norm": 0.3670332825002412,
-      "learning_rate": 0.00010295520370326355,
-      "loss": 0.734,
-      "step": 3503
-    },
-    {
-      "epoch": 1.7215968505874393,
-      "grad_norm": 0.41344384343051993,
-      "learning_rate": 0.00010289365433190514,
-      "loss": 0.7197,
-      "step": 3504
-    },
-    {
-      "epoch": 1.7220889463000555,
-      "grad_norm": 0.38374499673397267,
-      "learning_rate": 0.00010283210386339813,
-      "loss": 0.7351,
-      "step": 3505
-    },
-    {
-      "epoch": 1.7225810420126715,
-      "grad_norm": 0.3941941455094037,
-      "learning_rate": 0.00010277055232107975,
-      "loss": 0.7782,
-      "step": 3506
-    },
-    {
-      "epoch": 1.7230731377252875,
-      "grad_norm": 0.3933031831370362,
-      "learning_rate": 0.00010270899972828764,
-      "loss": 0.7382,
-      "step": 3507
-    },
-    {
-      "epoch": 1.7235652334379037,
-      "grad_norm": 0.40792635676036754,
-      "learning_rate": 0.00010264744610835995,
-      "loss": 0.7309,
-      "step": 3508
-    },
-    {
-      "epoch": 1.72405732915052,
-      "grad_norm": 0.4037259709765141,
-      "learning_rate": 0.00010258589148463513,
-      "loss": 0.7129,
-      "step": 3509
-    },
-    {
-      "epoch": 1.724549424863136,
-      "grad_norm": 0.40434860022010893,
-      "learning_rate": 0.00010252433588045203,
-      "loss": 0.7048,
-      "step": 3510
-    },
-    {
-      "epoch": 1.7250415205757519,
-      "grad_norm": 0.3930347623977529,
-      "learning_rate": 0.00010246277931914987,
-      "loss": 0.6624,
-      "step": 3511
-    },
-    {
-      "epoch": 1.725533616288368,
-      "grad_norm": 0.41578831003075,
-      "learning_rate": 0.00010240122182406824,
-      "loss": 0.7506,
-      "step": 3512
-    },
-    {
-      "epoch": 1.7260257120009843,
-      "grad_norm": 0.4079183364657208,
-      "learning_rate": 0.00010233966341854708,
-      "loss": 0.7897,
-      "step": 3513
-    },
-    {
-      "epoch": 1.7265178077136003,
-      "grad_norm": 0.4006686873888735,
-      "learning_rate": 0.00010227810412592667,
-      "loss": 0.7007,
-      "step": 3514
-    },
-    {
-      "epoch": 1.7270099034262163,
-      "grad_norm": 0.41546339244659536,
-      "learning_rate": 0.00010221654396954765,
-      "loss": 0.6996,
-      "step": 3515
-    },
-    {
-      "epoch": 1.7275019991388325,
-      "grad_norm": 0.3612698481311449,
-      "learning_rate": 0.00010215498297275095,
-      "loss": 0.6563,
-      "step": 3516
-    },
-    {
-      "epoch": 1.7279940948514487,
-      "grad_norm": 0.3788470095424007,
-      "learning_rate": 0.00010209342115887786,
-      "loss": 0.7211,
-      "step": 3517
-    },
-    {
-      "epoch": 1.7284861905640647,
-      "grad_norm": 0.38345562832385494,
-      "learning_rate": 0.00010203185855126995,
-      "loss": 0.7401,
-      "step": 3518
-    },
-    {
-      "epoch": 1.7289782862766807,
-      "grad_norm": 0.4389762480016236,
-      "learning_rate": 0.0001019702951732691,
-      "loss": 0.7652,
-      "step": 3519
-    },
-    {
-      "epoch": 1.729470381989297,
-      "grad_norm": 0.44322486526136545,
-      "learning_rate": 0.00010190873104821747,
-      "loss": 0.7873,
-      "step": 3520
-    },
-    {
-      "epoch": 1.7299624777019131,
-      "grad_norm": 0.4107555332048576,
-      "learning_rate": 0.00010184716619945753,
-      "loss": 0.7415,
-      "step": 3521
-    },
-    {
-      "epoch": 1.7304545734145291,
-      "grad_norm": 0.4175407609007106,
-      "learning_rate": 0.00010178560065033202,
-      "loss": 0.7055,
-      "step": 3522
-    },
-    {
-      "epoch": 1.7309466691271451,
-      "grad_norm": 0.3824759288659564,
-      "learning_rate": 0.00010172403442418394,
-      "loss": 0.6757,
-      "step": 3523
-    },
-    {
-      "epoch": 1.7314387648397613,
-      "grad_norm": 0.3842239327936125,
-      "learning_rate": 0.0001016624675443565,
-      "loss": 0.7057,
-      "step": 3524
-    },
-    {
-      "epoch": 1.7319308605523775,
-      "grad_norm": 0.442323995185006,
-      "learning_rate": 0.00010160090003419324,
-      "loss": 0.7685,
-      "step": 3525
-    },
-    {
-      "epoch": 1.7324229562649935,
-      "grad_norm": 0.3791806550187048,
-      "learning_rate": 0.00010153933191703789,
-      "loss": 0.6874,
-      "step": 3526
-    },
-    {
-      "epoch": 1.7329150519776095,
-      "grad_norm": 0.400614393343862,
-      "learning_rate": 0.00010147776321623437,
-      "loss": 0.7042,
-      "step": 3527
-    },
-    {
-      "epoch": 1.7334071476902257,
-      "grad_norm": 0.401074185035461,
-      "learning_rate": 0.00010141619395512694,
-      "loss": 0.6988,
-      "step": 3528
-    },
-    {
-      "epoch": 1.733899243402842,
-      "grad_norm": 0.389821659209944,
-      "learning_rate": 0.00010135462415705996,
-      "loss": 0.7341,
-      "step": 3529
-    },
-    {
-      "epoch": 1.734391339115458,
-      "grad_norm": 0.41109927618522046,
-      "learning_rate": 0.00010129305384537803,
-      "loss": 0.713,
-      "step": 3530
-    },
-    {
-      "epoch": 1.734883434828074,
-      "grad_norm": 0.4303110985782416,
-      "learning_rate": 0.000101231483043426,
-      "loss": 0.7744,
-      "step": 3531
-    },
-    {
-      "epoch": 1.7353755305406902,
-      "grad_norm": 0.3926575069143071,
-      "learning_rate": 0.00010116991177454884,
-      "loss": 0.6485,
-      "step": 3532
-    },
-    {
-      "epoch": 1.7358676262533064,
-      "grad_norm": 0.3829580495426352,
-      "learning_rate": 0.0001011083400620917,
-      "loss": 0.6795,
-      "step": 3533
-    },
-    {
-      "epoch": 1.7363597219659224,
-      "grad_norm": 0.38757407855706405,
-      "learning_rate": 0.00010104676792939991,
-      "loss": 0.6671,
-      "step": 3534
-    },
-    {
-      "epoch": 1.7368518176785384,
-      "grad_norm": 0.40191137525061993,
-      "learning_rate": 0.00010098519539981895,
-      "loss": 0.7354,
-      "step": 3535
-    },
-    {
-      "epoch": 1.7373439133911546,
-      "grad_norm": 0.4192264881292666,
-      "learning_rate": 0.00010092362249669449,
-      "loss": 0.6715,
-      "step": 3536
-    },
-    {
-      "epoch": 1.7378360091037708,
-      "grad_norm": 0.41467412279185634,
-      "learning_rate": 0.00010086204924337228,
-      "loss": 0.7166,
-      "step": 3537
-    },
-    {
-      "epoch": 1.7383281048163868,
-      "grad_norm": 0.4271854923288818,
-      "learning_rate": 0.00010080047566319828,
-      "loss": 0.7443,
-      "step": 3538
-    },
-    {
-      "epoch": 1.7388202005290028,
-      "grad_norm": 0.411345068577774,
-      "learning_rate": 0.0001007389017795185,
-      "loss": 0.7376,
-      "step": 3539
-    },
-    {
-      "epoch": 1.739312296241619,
-      "grad_norm": 0.39822703509584123,
-      "learning_rate": 0.00010067732761567909,
-      "loss": 0.7129,
-      "step": 3540
-    },
-    {
-      "epoch": 1.7398043919542352,
-      "grad_norm": 0.4109190607654912,
-      "learning_rate": 0.00010061575319502634,
-      "loss": 0.7226,
-      "step": 3541
-    },
-    {
-      "epoch": 1.7402964876668512,
-      "grad_norm": 0.3960863412071096,
-      "learning_rate": 0.00010055417854090661,
-      "loss": 0.6881,
-      "step": 3542
-    },
-    {
-      "epoch": 1.7407885833794672,
-      "grad_norm": 0.3691519032507848,
-      "learning_rate": 0.00010049260367666629,
-      "loss": 0.6281,
-      "step": 3543
-    },
-    {
-      "epoch": 1.7412806790920834,
-      "grad_norm": 0.47692918908217197,
-      "learning_rate": 0.00010043102862565197,
-      "loss": 0.7108,
-      "step": 3544
-    },
-    {
-      "epoch": 1.7417727748046996,
-      "grad_norm": 0.36731186336445404,
-      "learning_rate": 0.00010036945341121023,
-      "loss": 0.6829,
-      "step": 3545
-    },
-    {
-      "epoch": 1.7422648705173156,
-      "grad_norm": 0.41197569915970433,
-      "learning_rate": 0.00010030787805668772,
-      "loss": 0.7398,
-      "step": 3546
-    },
-    {
-      "epoch": 1.7427569662299316,
-      "grad_norm": 0.3956203084461126,
-      "learning_rate": 0.00010024630258543115,
-      "loss": 0.7136,
-      "step": 3547
-    },
-    {
-      "epoch": 1.7432490619425478,
-      "grad_norm": 0.3764832927847127,
-      "learning_rate": 0.00010018472702078731,
-      "loss": 0.6258,
-      "step": 3548
-    },
-    {
-      "epoch": 1.743741157655164,
-      "grad_norm": 0.39374042961756567,
-      "learning_rate": 0.00010012315138610296,
-      "loss": 0.6854,
-      "step": 3549
-    },
-    {
-      "epoch": 1.74423325336778,
-      "grad_norm": 0.3718920513883015,
-      "learning_rate": 0.0001000615757047249,
-      "loss": 0.7036,
-      "step": 3550
-    },
-    {
-      "epoch": 1.744725349080396,
-      "grad_norm": 0.3840694493698443,
-      "learning_rate": 0.0001,
-      "loss": 0.6737,
-      "step": 3551
-    },
-    {
-      "epoch": 1.7452174447930122,
-      "grad_norm": 0.37453480334956263,
-      "learning_rate": 9.993842429527511e-05,
-      "loss": 0.6893,
-      "step": 3552
-    },
-    {
-      "epoch": 1.7457095405056284,
-      "grad_norm": 0.3855906756896309,
-      "learning_rate": 9.98768486138971e-05,
-      "loss": 0.6951,
-      "step": 3553
-    },
-    {
-      "epoch": 1.7462016362182444,
-      "grad_norm": 0.4039952146641488,
-      "learning_rate": 9.981527297921271e-05,
-      "loss": 0.7419,
-      "step": 3554
-    },
-    {
-      "epoch": 1.7466937319308604,
-      "grad_norm": 0.374779436380444,
-      "learning_rate": 9.975369741456886e-05,
-      "loss": 0.7277,
-      "step": 3555
-    },
-    {
-      "epoch": 1.7471858276434766,
-      "grad_norm": 0.4017553648948238,
-      "learning_rate": 9.96921219433123e-05,
-      "loss": 0.7364,
-      "step": 3556
-    },
-    {
-      "epoch": 1.7476779233560928,
-      "grad_norm": 0.3874209994250295,
-      "learning_rate": 9.963054658878979e-05,
-      "loss": 0.6758,
-      "step": 3557
-    },
-    {
-      "epoch": 1.7481700190687088,
-      "grad_norm": 0.4008569565521517,
-      "learning_rate": 9.956897137434803e-05,
-      "loss": 0.6873,
-      "step": 3558
-    },
-    {
-      "epoch": 1.7486621147813248,
-      "grad_norm": 0.4084491154396771,
-      "learning_rate": 9.950739632333374e-05,
-      "loss": 0.7869,
-      "step": 3559
-    },
-    {
-      "epoch": 1.749154210493941,
-      "grad_norm": 0.3842445774848982,
-      "learning_rate": 9.944582145909342e-05,
-      "loss": 0.775,
-      "step": 3560
-    },
-    {
-      "epoch": 1.7496463062065573,
-      "grad_norm": 0.38670512594838846,
-      "learning_rate": 9.938424680497366e-05,
-      "loss": 0.7093,
-      "step": 3561
-    },
-    {
-      "epoch": 1.7501384019191732,
-      "grad_norm": 0.4232140184965088,
-      "learning_rate": 9.932267238432092e-05,
-      "loss": 0.6848,
-      "step": 3562
-    },
-    {
-      "epoch": 1.7506304976317892,
-      "grad_norm": 0.3967726417745073,
-      "learning_rate": 9.926109822048152e-05,
-      "loss": 0.6915,
-      "step": 3563
-    },
-    {
-      "epoch": 1.7511225933444055,
-      "grad_norm": 0.3992389662750598,
-      "learning_rate": 9.919952433680176e-05,
-      "loss": 0.7599,
-      "step": 3564
-    },
-    {
-      "epoch": 1.7516146890570217,
-      "grad_norm": 0.4069296621023427,
-      "learning_rate": 9.913795075662773e-05,
-      "loss": 0.7919,
-      "step": 3565
-    },
-    {
-      "epoch": 1.7521067847696377,
-      "grad_norm": 0.40169081802961726,
-      "learning_rate": 9.907637750330552e-05,
-      "loss": 0.7502,
-      "step": 3566
-    },
-    {
-      "epoch": 1.7525988804822537,
-      "grad_norm": 0.44601310963902435,
-      "learning_rate": 9.901480460018109e-05,
-      "loss": 0.6969,
-      "step": 3567
-    },
-    {
-      "epoch": 1.7530909761948699,
-      "grad_norm": 0.43521158507426405,
-      "learning_rate": 9.895323207060012e-05,
-      "loss": 0.7,
-      "step": 3568
-    },
-    {
-      "epoch": 1.753583071907486,
-      "grad_norm": 0.40027753321132364,
-      "learning_rate": 9.889165993790833e-05,
-      "loss": 0.7111,
-      "step": 3569
-    },
-    {
-      "epoch": 1.754075167620102,
-      "grad_norm": 0.40321945384277336,
-      "learning_rate": 9.883008822545118e-05,
-      "loss": 0.7506,
-      "step": 3570
-    },
-    {
-      "epoch": 1.754567263332718,
-      "grad_norm": 0.41366983129488605,
-      "learning_rate": 9.876851695657401e-05,
-      "loss": 0.7994,
-      "step": 3571
-    },
-    {
-      "epoch": 1.7550593590453343,
-      "grad_norm": 0.3958610074652765,
-      "learning_rate": 9.870694615462196e-05,
-      "loss": 0.7408,
-      "step": 3572
-    },
-    {
-      "epoch": 1.7555514547579505,
-      "grad_norm": 0.40465987135097947,
-      "learning_rate": 9.864537584294009e-05,
-      "loss": 0.7041,
-      "step": 3573
-    },
-    {
-      "epoch": 1.7560435504705665,
-      "grad_norm": 0.4131957896563108,
-      "learning_rate": 9.85838060448731e-05,
-      "loss": 0.6722,
-      "step": 3574
-    },
-    {
-      "epoch": 1.7565356461831825,
-      "grad_norm": 0.4070671999796951,
-      "learning_rate": 9.852223678376564e-05,
-      "loss": 0.6942,
-      "step": 3575
-    },
-    {
-      "epoch": 1.7570277418957987,
-      "grad_norm": 0.40961386529538774,
-      "learning_rate": 9.846066808296216e-05,
-      "loss": 0.7206,
-      "step": 3576
-    },
-    {
-      "epoch": 1.757519837608415,
-      "grad_norm": 0.3957878469143598,
-      "learning_rate": 9.839909996580678e-05,
-      "loss": 0.6698,
-      "step": 3577
-    },
-    {
-      "epoch": 1.758011933321031,
-      "grad_norm": 0.43925914277772965,
-      "learning_rate": 9.83375324556435e-05,
-      "loss": 0.7104,
-      "step": 3578
-    },
-    {
-      "epoch": 1.7585040290336469,
-      "grad_norm": 0.41004679098200747,
-      "learning_rate": 9.827596557581608e-05,
-      "loss": 0.6971,
-      "step": 3579
-    },
-    {
-      "epoch": 1.758996124746263,
-      "grad_norm": 0.40025436562277616,
-      "learning_rate": 9.821439934966799e-05,
-      "loss": 0.7186,
-      "step": 3580
-    },
-    {
-      "epoch": 1.7594882204588793,
-      "grad_norm": 0.4013237402959138,
-      "learning_rate": 9.815283380054245e-05,
-      "loss": 0.7392,
-      "step": 3581
-    },
-    {
-      "epoch": 1.7599803161714953,
-      "grad_norm": 0.39082561983900566,
-      "learning_rate": 9.809126895178255e-05,
-      "loss": 0.7119,
-      "step": 3582
-    },
-    {
-      "epoch": 1.7604724118841113,
-      "grad_norm": 0.44344168165424497,
-      "learning_rate": 9.802970482673092e-05,
-      "loss": 0.7736,
-      "step": 3583
-    },
-    {
-      "epoch": 1.7609645075967275,
-      "grad_norm": 0.4265648361327368,
-      "learning_rate": 9.796814144873006e-05,
-      "loss": 0.7046,
-      "step": 3584
-    },
-    {
-      "epoch": 1.7614566033093437,
-      "grad_norm": 0.3885664868990184,
-      "learning_rate": 9.790657884112218e-05,
-      "loss": 0.7191,
-      "step": 3585
-    },
-    {
-      "epoch": 1.7619486990219597,
-      "grad_norm": 0.43208693779897167,
-      "learning_rate": 9.784501702724906e-05,
-      "loss": 0.7569,
-      "step": 3586
-    },
-    {
-      "epoch": 1.7624407947345757,
-      "grad_norm": 6.769422811998193,
-      "learning_rate": 9.778345603045236e-05,
-      "loss": 0.9136,
-      "step": 3587
-    },
-    {
-      "epoch": 1.762932890447192,
-      "grad_norm": 0.42892805162230574,
-      "learning_rate": 9.772189587407337e-05,
-      "loss": 0.687,
-      "step": 3588
-    },
-    {
-      "epoch": 1.7634249861598081,
-      "grad_norm": 0.39870719955427314,
-      "learning_rate": 9.766033658145293e-05,
-      "loss": 0.6832,
-      "step": 3589
-    },
-    {
-      "epoch": 1.7639170818724241,
-      "grad_norm": 0.41477009627463773,
-      "learning_rate": 9.759877817593181e-05,
-      "loss": 0.7545,
-      "step": 3590
-    },
-    {
-      "epoch": 1.7644091775850403,
-      "grad_norm": 0.4338401142547945,
-      "learning_rate": 9.753722068085017e-05,
-      "loss": 0.7534,
-      "step": 3591
-    },
-    {
-      "epoch": 1.7649012732976566,
-      "grad_norm": 0.4310440685314093,
-      "learning_rate": 9.7475664119548e-05,
-      "loss": 0.7866,
-      "step": 3592
-    },
-    {
-      "epoch": 1.7653933690102726,
-      "grad_norm": 0.4006500381452929,
-      "learning_rate": 9.741410851536489e-05,
-      "loss": 0.7156,
-      "step": 3593
-    },
-    {
-      "epoch": 1.7658854647228885,
-      "grad_norm": 0.3983126074848176,
-      "learning_rate": 9.735255389164007e-05,
-      "loss": 0.7056,
-      "step": 3594
-    },
-    {
-      "epoch": 1.7663775604355048,
-      "grad_norm": 0.41122094306894436,
-      "learning_rate": 9.729100027171237e-05,
-      "loss": 0.6845,
-      "step": 3595
-    },
-    {
-      "epoch": 1.766869656148121,
-      "grad_norm": 0.43211998328105466,
-      "learning_rate": 9.72294476789203e-05,
-      "loss": 0.726,
-      "step": 3596
-    },
-    {
-      "epoch": 1.767361751860737,
-      "grad_norm": 0.4370230378403562,
-      "learning_rate": 9.716789613660188e-05,
-      "loss": 0.7546,
-      "step": 3597
-    },
-    {
-      "epoch": 1.767853847573353,
-      "grad_norm": 0.45340938619946647,
-      "learning_rate": 9.710634566809484e-05,
-      "loss": 0.7013,
-      "step": 3598
-    },
-    {
-      "epoch": 1.7683459432859692,
-      "grad_norm": 0.4086115283344419,
-      "learning_rate": 9.70447962967365e-05,
-      "loss": 0.7106,
-      "step": 3599
-    },
-    {
-      "epoch": 1.7688380389985854,
-      "grad_norm": 0.41458569303006004,
-      "learning_rate": 9.698324804586362e-05,
-      "loss": 0.6908,
-      "step": 3600
-    },
-    {
-      "epoch": 1.7693301347112014,
-      "grad_norm": 0.3746455260273381,
-      "learning_rate": 9.692170093881272e-05,
-      "loss": 0.6919,
-      "step": 3601
-    },
-    {
-      "epoch": 1.7698222304238174,
-      "grad_norm": 0.38263056436158766,
-      "learning_rate": 9.686015499891976e-05,
-      "loss": 0.7329,
-      "step": 3602
-    },
-    {
-      "epoch": 1.7703143261364336,
-      "grad_norm": 0.39690722756054714,
-      "learning_rate": 9.679861024952031e-05,
-      "loss": 0.7098,
-      "step": 3603
-    },
-    {
-      "epoch": 1.7708064218490498,
-      "grad_norm": 0.4288916707528441,
-      "learning_rate": 9.673706671394947e-05,
-      "loss": 0.7258,
-      "step": 3604
-    },
-    {
-      "epoch": 1.7712985175616658,
-      "grad_norm": 0.40693943074181127,
-      "learning_rate": 9.667552441554193e-05,
-      "loss": 0.6631,
-      "step": 3605
-    },
-    {
-      "epoch": 1.7717906132742818,
-      "grad_norm": 0.3943991617454967,
-      "learning_rate": 9.661398337763181e-05,
-      "loss": 0.6899,
-      "step": 3606
-    },
-    {
-      "epoch": 1.772282708986898,
-      "grad_norm": 0.46438425466168803,
-      "learning_rate": 9.655244362355283e-05,
-      "loss": 0.7604,
-      "step": 3607
-    },
-    {
-      "epoch": 1.7727748046995142,
-      "grad_norm": 0.39289139588856387,
-      "learning_rate": 9.649090517663825e-05,
-      "loss": 0.6647,
-      "step": 3608
-    },
-    {
-      "epoch": 1.7732669004121302,
-      "grad_norm": 0.37589813982708253,
-      "learning_rate": 9.642936806022074e-05,
-      "loss": 0.7453,
-      "step": 3609
-    },
-    {
-      "epoch": 1.7737589961247462,
-      "grad_norm": 0.43598144036229847,
-      "learning_rate": 9.63678322976325e-05,
-      "loss": 0.752,
-      "step": 3610
-    },
-    {
-      "epoch": 1.7742510918373624,
-      "grad_norm": 0.4202884364136635,
-      "learning_rate": 9.630629791220532e-05,
-      "loss": 0.7024,
-      "step": 3611
-    },
-    {
-      "epoch": 1.7747431875499786,
-      "grad_norm": 0.41047943093238265,
-      "learning_rate": 9.624476492727026e-05,
-      "loss": 0.7001,
-      "step": 3612
-    },
-    {
-      "epoch": 1.7752352832625946,
-      "grad_norm": 0.38733211562690373,
-      "learning_rate": 9.618323336615809e-05,
-      "loss": 0.6984,
-      "step": 3613
-    },
-    {
-      "epoch": 1.7757273789752106,
-      "grad_norm": 0.3915299895644845,
-      "learning_rate": 9.61217032521988e-05,
-      "loss": 0.7614,
-      "step": 3614
-    },
-    {
-      "epoch": 1.7762194746878268,
-      "grad_norm": 0.4864737172546369,
-      "learning_rate": 9.606017460872202e-05,
-      "loss": 0.702,
-      "step": 3615
-    },
-    {
-      "epoch": 1.776711570400443,
-      "grad_norm": 0.4287295052493886,
-      "learning_rate": 9.599864745905676e-05,
-      "loss": 0.7666,
-      "step": 3616
-    },
-    {
-      "epoch": 1.777203666113059,
-      "grad_norm": 0.37777073408162265,
-      "learning_rate": 9.593712182653142e-05,
-      "loss": 0.7069,
-      "step": 3617
-    },
-    {
-      "epoch": 1.777695761825675,
-      "grad_norm": 0.4041294321789745,
-      "learning_rate": 9.587559773447386e-05,
-      "loss": 0.7171,
-      "step": 3618
-    },
-    {
-      "epoch": 1.7781878575382912,
-      "grad_norm": 0.4045073679131275,
-      "learning_rate": 9.581407520621139e-05,
-      "loss": 0.7409,
-      "step": 3619
-    },
-    {
-      "epoch": 1.7786799532509074,
-      "grad_norm": 0.4203245957136642,
-      "learning_rate": 9.575255426507066e-05,
-      "loss": 0.7248,
-      "step": 3620
-    },
-    {
-      "epoch": 1.7791720489635234,
-      "grad_norm": 0.41894224530100027,
-      "learning_rate": 9.569103493437775e-05,
-      "loss": 0.7337,
-      "step": 3621
-    },
-    {
-      "epoch": 1.7796641446761394,
-      "grad_norm": 0.4010499604903825,
-      "learning_rate": 9.56295172374582e-05,
-      "loss": 0.7348,
-      "step": 3622
-    },
-    {
-      "epoch": 1.7801562403887556,
-      "grad_norm": 0.3976124715314281,
-      "learning_rate": 9.556800119763676e-05,
-      "loss": 0.7038,
-      "step": 3623
-    },
-    {
-      "epoch": 1.7806483361013719,
-      "grad_norm": 0.4316771595023527,
-      "learning_rate": 9.550648683823774e-05,
-      "loss": 0.7351,
-      "step": 3624
-    },
-    {
-      "epoch": 1.7811404318139878,
-      "grad_norm": 0.39222661297048333,
-      "learning_rate": 9.544497418258466e-05,
-      "loss": 0.7287,
-      "step": 3625
-    },
-    {
-      "epoch": 1.7816325275266038,
-      "grad_norm": 0.41841392270252,
-      "learning_rate": 9.53834632540005e-05,
-      "loss": 0.7575,
-      "step": 3626
-    },
-    {
-      "epoch": 1.78212462323922,
-      "grad_norm": 0.3927915030903514,
-      "learning_rate": 9.532195407580751e-05,
-      "loss": 0.7238,
-      "step": 3627
-    },
-    {
-      "epoch": 1.7826167189518363,
-      "grad_norm": 0.39068585990983884,
-      "learning_rate": 9.52604466713274e-05,
-      "loss": 0.7126,
-      "step": 3628
-    },
-    {
-      "epoch": 1.7831088146644523,
-      "grad_norm": 0.37218797397745434,
-      "learning_rate": 9.5198941063881e-05,
-      "loss": 0.661,
-      "step": 3629
-    },
-    {
-      "epoch": 1.7836009103770682,
-      "grad_norm": 0.4231800035280616,
-      "learning_rate": 9.513743727678862e-05,
-      "loss": 0.7152,
-      "step": 3630
-    },
-    {
-      "epoch": 1.7840930060896845,
-      "grad_norm": 0.4109899483833016,
-      "learning_rate": 9.507593533336991e-05,
-      "loss": 0.7397,
-      "step": 3631
-    },
-    {
-      "epoch": 1.7845851018023007,
-      "grad_norm": 0.4048869671348017,
-      "learning_rate": 9.501443525694364e-05,
-      "loss": 0.7574,
-      "step": 3632
-    },
-    {
-      "epoch": 1.7850771975149167,
-      "grad_norm": 0.42718661700723454,
-      "learning_rate": 9.495293707082803e-05,
-      "loss": 0.6622,
-      "step": 3633
-    },
-    {
-      "epoch": 1.7855692932275327,
-      "grad_norm": 0.40615682533588254,
-      "learning_rate": 9.489144079834057e-05,
-      "loss": 0.7172,
-      "step": 3634
-    },
-    {
-      "epoch": 1.7860613889401489,
-      "grad_norm": 0.38567327017355046,
-      "learning_rate": 9.482994646279794e-05,
-      "loss": 0.6907,
-      "step": 3635
-    },
-    {
-      "epoch": 1.786553484652765,
-      "grad_norm": 0.3898865033843525,
-      "learning_rate": 9.476845408751614e-05,
-      "loss": 0.7232,
-      "step": 3636
-    },
-    {
-      "epoch": 1.787045580365381,
-      "grad_norm": 0.38392837452996853,
-      "learning_rate": 9.470696369581043e-05,
-      "loss": 0.661,
-      "step": 3637
-    },
-    {
-      "epoch": 1.787537676077997,
-      "grad_norm": 0.3957580491721606,
-      "learning_rate": 9.464547531099531e-05,
-      "loss": 0.7026,
-      "step": 3638
-    },
-    {
-      "epoch": 1.7880297717906133,
-      "grad_norm": 0.3841754582784987,
-      "learning_rate": 9.458398895638453e-05,
-      "loss": 0.6594,
-      "step": 3639
-    },
-    {
-      "epoch": 1.7885218675032295,
-      "grad_norm": 0.381540839072192,
-      "learning_rate": 9.452250465529103e-05,
-      "loss": 0.7105,
-      "step": 3640
-    },
-    {
-      "epoch": 1.7890139632158455,
-      "grad_norm": 0.4075211291646927,
-      "learning_rate": 9.446102243102698e-05,
-      "loss": 0.7512,
-      "step": 3641
-    },
-    {
-      "epoch": 1.7895060589284615,
-      "grad_norm": 0.4110434646998583,
-      "learning_rate": 9.439954230690387e-05,
-      "loss": 0.7843,
-      "step": 3642
-    },
-    {
-      "epoch": 1.7899981546410777,
-      "grad_norm": 0.3821810793023431,
-      "learning_rate": 9.43380643062322e-05,
-      "loss": 0.7083,
-      "step": 3643
-    },
-    {
-      "epoch": 1.790490250353694,
-      "grad_norm": 0.43800897842754943,
-      "learning_rate": 9.427658845232183e-05,
-      "loss": 0.7994,
-      "step": 3644
-    },
-    {
-      "epoch": 1.79098234606631,
-      "grad_norm": 0.4060559657117266,
-      "learning_rate": 9.421511476848176e-05,
-      "loss": 0.7054,
-      "step": 3645
-    },
-    {
-      "epoch": 1.791474441778926,
-      "grad_norm": 0.4251741558666256,
-      "learning_rate": 9.41536432780201e-05,
-      "loss": 0.6839,
-      "step": 3646
-    },
-    {
-      "epoch": 1.791966537491542,
-      "grad_norm": 0.3967139786790515,
-      "learning_rate": 9.409217400424425e-05,
-      "loss": 0.6599,
-      "step": 3647
-    },
-    {
-      "epoch": 1.7924586332041583,
-      "grad_norm": 0.3844356170739326,
-      "learning_rate": 9.403070697046064e-05,
-      "loss": 0.6614,
-      "step": 3648
-    },
-    {
-      "epoch": 1.7929507289167743,
-      "grad_norm": 0.405097692915681,
-      "learning_rate": 9.396924219997495e-05,
-      "loss": 0.6524,
-      "step": 3649
-    },
-    {
-      "epoch": 1.7934428246293903,
-      "grad_norm": 0.4015999545623788,
-      "learning_rate": 9.390777971609192e-05,
-      "loss": 0.7098,
-      "step": 3650
-    },
-    {
-      "epoch": 1.7939349203420065,
-      "grad_norm": 0.39157952962502157,
-      "learning_rate": 9.384631954211556e-05,
-      "loss": 0.7294,
-      "step": 3651
-    },
-    {
-      "epoch": 1.7944270160546227,
-      "grad_norm": 0.4053712472251571,
-      "learning_rate": 9.37848617013488e-05,
-      "loss": 0.7188,
-      "step": 3652
-    },
-    {
-      "epoch": 1.7949191117672387,
-      "grad_norm": 0.42118321856598334,
-      "learning_rate": 9.372340621709384e-05,
-      "loss": 0.7789,
-      "step": 3653
-    },
-    {
-      "epoch": 1.7954112074798547,
-      "grad_norm": 0.3736988990314272,
-      "learning_rate": 9.366195311265199e-05,
-      "loss": 0.7104,
-      "step": 3654
-    },
-    {
-      "epoch": 1.795903303192471,
-      "grad_norm": 0.3800296898417442,
-      "learning_rate": 9.360050241132353e-05,
-      "loss": 0.6743,
-      "step": 3655
-    },
-    {
-      "epoch": 1.7963953989050871,
-      "grad_norm": 0.4087234152549692,
-      "learning_rate": 9.353905413640795e-05,
-      "loss": 0.716,
-      "step": 3656
-    },
-    {
-      "epoch": 1.7968874946177031,
-      "grad_norm": 0.38169405337252094,
-      "learning_rate": 9.347760831120384e-05,
-      "loss": 0.7012,
-      "step": 3657
-    },
-    {
-      "epoch": 1.7973795903303191,
-      "grad_norm": 0.39749285717983823,
-      "learning_rate": 9.34161649590087e-05,
-      "loss": 0.6815,
-      "step": 3658
-    },
-    {
-      "epoch": 1.7978716860429353,
-      "grad_norm": 0.42167945023720504,
-      "learning_rate": 9.335472410311924e-05,
-      "loss": 0.7164,
-      "step": 3659
-    },
-    {
-      "epoch": 1.7983637817555516,
-      "grad_norm": 0.3938649420809012,
-      "learning_rate": 9.329328576683117e-05,
-      "loss": 0.6846,
-      "step": 3660
-    },
-    {
-      "epoch": 1.7988558774681676,
-      "grad_norm": 0.41802174604661035,
-      "learning_rate": 9.323184997343926e-05,
-      "loss": 0.7196,
-      "step": 3661
-    },
-    {
-      "epoch": 1.7993479731807835,
-      "grad_norm": 0.39438780140295476,
-      "learning_rate": 9.317041674623731e-05,
-      "loss": 0.7902,
-      "step": 3662
-    },
-    {
-      "epoch": 1.7998400688933998,
-      "grad_norm": 0.421907736746234,
-      "learning_rate": 9.310898610851814e-05,
-      "loss": 0.6881,
-      "step": 3663
-    },
-    {
-      "epoch": 1.800332164606016,
-      "grad_norm": 0.39614091012821945,
-      "learning_rate": 9.304755808357355e-05,
-      "loss": 0.6844,
-      "step": 3664
-    },
-    {
-      "epoch": 1.800824260318632,
-      "grad_norm": 0.3733740097174482,
-      "learning_rate": 9.298613269469449e-05,
-      "loss": 0.6513,
-      "step": 3665
-    },
-    {
-      "epoch": 1.801316356031248,
-      "grad_norm": 0.3712862644921941,
-      "learning_rate": 9.292470996517069e-05,
-      "loss": 0.6777,
-      "step": 3666
-    },
-    {
-      "epoch": 1.8018084517438642,
-      "grad_norm": 0.41059073322424766,
-      "learning_rate": 9.286328991829107e-05,
-      "loss": 0.7188,
-      "step": 3667
-    },
-    {
-      "epoch": 1.8023005474564804,
-      "grad_norm": 0.3898080484982208,
-      "learning_rate": 9.280187257734349e-05,
-      "loss": 0.6804,
-      "step": 3668
-    },
-    {
-      "epoch": 1.8027926431690964,
-      "grad_norm": 0.38677746790954753,
-      "learning_rate": 9.274045796561466e-05,
-      "loss": 0.6653,
-      "step": 3669
-    },
-    {
-      "epoch": 1.8032847388817124,
-      "grad_norm": 0.4165005809660249,
-      "learning_rate": 9.26790461063904e-05,
-      "loss": 0.7119,
-      "step": 3670
-    },
-    {
-      "epoch": 1.8037768345943286,
-      "grad_norm": 0.3824033248879199,
-      "learning_rate": 9.261763702295543e-05,
-      "loss": 0.7022,
-      "step": 3671
-    },
-    {
-      "epoch": 1.8042689303069448,
-      "grad_norm": 0.4023884022227293,
-      "learning_rate": 9.255623073859343e-05,
-      "loss": 0.6696,
-      "step": 3672
-    },
-    {
-      "epoch": 1.8047610260195608,
-      "grad_norm": 0.40136840583425243,
-      "learning_rate": 9.249482727658696e-05,
-      "loss": 0.6904,
-      "step": 3673
-    },
-    {
-      "epoch": 1.8052531217321768,
-      "grad_norm": 0.3798428338020864,
-      "learning_rate": 9.243342666021764e-05,
-      "loss": 0.6783,
-      "step": 3674
-    },
-    {
-      "epoch": 1.805745217444793,
-      "grad_norm": 0.38874449867040906,
-      "learning_rate": 9.237202891276583e-05,
-      "loss": 0.6736,
-      "step": 3675
-    },
-    {
-      "epoch": 1.8062373131574092,
-      "grad_norm": 0.4108747113060928,
-      "learning_rate": 9.231063405751095e-05,
-      "loss": 0.6866,
-      "step": 3676
-    },
-    {
-      "epoch": 1.8067294088700252,
-      "grad_norm": 0.4502609393034804,
-      "learning_rate": 9.224924211773134e-05,
-      "loss": 0.7519,
-      "step": 3677
-    },
-    {
-      "epoch": 1.8072215045826412,
-      "grad_norm": 0.3796156629232663,
-      "learning_rate": 9.218785311670406e-05,
-      "loss": 0.6691,
-      "step": 3678
-    },
-    {
-      "epoch": 1.8077136002952574,
-      "grad_norm": 0.40886274368335873,
-      "learning_rate": 9.21264670777052e-05,
-      "loss": 0.7112,
-      "step": 3679
-    },
-    {
-      "epoch": 1.8082056960078736,
-      "grad_norm": 0.42355144311951637,
-      "learning_rate": 9.206508402400978e-05,
-      "loss": 0.6871,
-      "step": 3680
-    },
-    {
-      "epoch": 1.8086977917204896,
-      "grad_norm": 0.3853752762561231,
-      "learning_rate": 9.200370397889145e-05,
-      "loss": 0.7104,
-      "step": 3681
-    },
-    {
-      "epoch": 1.8091898874331056,
-      "grad_norm": 0.39264909739973025,
-      "learning_rate": 9.194232696562299e-05,
-      "loss": 0.7307,
-      "step": 3682
-    },
-    {
-      "epoch": 1.8096819831457218,
-      "grad_norm": 0.4102083225264434,
-      "learning_rate": 9.188095300747588e-05,
-      "loss": 0.7301,
-      "step": 3683
-    },
-    {
-      "epoch": 1.810174078858338,
-      "grad_norm": 0.39726478647459396,
-      "learning_rate": 9.181958212772045e-05,
-      "loss": 0.6958,
-      "step": 3684
-    },
-    {
-      "epoch": 1.810666174570954,
-      "grad_norm": 0.4021112806581545,
-      "learning_rate": 9.175821434962587e-05,
-      "loss": 0.6997,
-      "step": 3685
-    },
-    {
-      "epoch": 1.81115827028357,
-      "grad_norm": 0.41951606160623856,
-      "learning_rate": 9.169684969646022e-05,
-      "loss": 0.6895,
-      "step": 3686
-    },
-    {
-      "epoch": 1.8116503659961862,
-      "grad_norm": 0.3845688009685872,
-      "learning_rate": 9.163548819149024e-05,
-      "loss": 0.6505,
-      "step": 3687
-    },
-    {
-      "epoch": 1.8121424617088024,
-      "grad_norm": 0.37000457667998365,
-      "learning_rate": 9.157412985798164e-05,
-      "loss": 0.6849,
-      "step": 3688
-    },
-    {
-      "epoch": 1.8126345574214184,
-      "grad_norm": 0.38963423725786017,
-      "learning_rate": 9.151277471919877e-05,
-      "loss": 0.729,
-      "step": 3689
-    },
-    {
-      "epoch": 1.8131266531340344,
-      "grad_norm": 0.39149433659239297,
-      "learning_rate": 9.145142279840489e-05,
-      "loss": 0.6916,
-      "step": 3690
-    },
-    {
-      "epoch": 1.8136187488466506,
-      "grad_norm": 0.4241517404854112,
-      "learning_rate": 9.139007411886203e-05,
-      "loss": 0.7773,
-      "step": 3691
-    },
-    {
-      "epoch": 1.8141108445592669,
-      "grad_norm": 0.38811482772973754,
-      "learning_rate": 9.132872870383086e-05,
-      "loss": 0.6882,
-      "step": 3692
-    },
-    {
-      "epoch": 1.8146029402718828,
-      "grad_norm": 0.38972974305922936,
-      "learning_rate": 9.1267386576571e-05,
-      "loss": 0.7575,
-      "step": 3693
-    },
-    {
-      "epoch": 1.8150950359844988,
-      "grad_norm": 0.3707112981210964,
-      "learning_rate": 9.12060477603407e-05,
-      "loss": 0.7315,
-      "step": 3694
-    },
-    {
-      "epoch": 1.815587131697115,
-      "grad_norm": 0.37287843944915056,
-      "learning_rate": 9.114471227839701e-05,
-      "loss": 0.7211,
-      "step": 3695
-    },
-    {
-      "epoch": 1.8160792274097313,
-      "grad_norm": 0.39947783415704224,
-      "learning_rate": 9.108338015399563e-05,
-      "loss": 0.6665,
-      "step": 3696
-    },
-    {
-      "epoch": 1.8165713231223473,
-      "grad_norm": 0.36954024004603747,
-      "learning_rate": 9.102205141039115e-05,
-      "loss": 0.6781,
-      "step": 3697
-    },
-    {
-      "epoch": 1.8170634188349633,
-      "grad_norm": 0.40546968582768905,
-      "learning_rate": 9.096072607083667e-05,
-      "loss": 0.7207,
-      "step": 3698
-    },
-    {
-      "epoch": 1.8175555145475795,
-      "grad_norm": 0.42605154676674606,
-      "learning_rate": 9.089940415858416e-05,
-      "loss": 0.7944,
-      "step": 3699
-    },
-    {
-      "epoch": 1.8180476102601957,
-      "grad_norm": 0.4107813840978002,
-      "learning_rate": 9.083808569688428e-05,
-      "loss": 0.7439,
-      "step": 3700
-    },
-    {
-      "epoch": 1.8185397059728117,
-      "grad_norm": 0.4112317216098252,
-      "learning_rate": 9.077677070898627e-05,
-      "loss": 0.7473,
-      "step": 3701
-    },
-    {
-      "epoch": 1.8190318016854277,
-      "grad_norm": 0.39892042348833323,
-      "learning_rate": 9.071545921813814e-05,
-      "loss": 0.7187,
-      "step": 3702
-    },
-    {
-      "epoch": 1.8195238973980439,
-      "grad_norm": 0.39043805453135716,
-      "learning_rate": 9.065415124758663e-05,
-      "loss": 0.7001,
-      "step": 3703
-    },
-    {
-      "epoch": 1.82001599311066,
-      "grad_norm": 0.4079832837617063,
-      "learning_rate": 9.059284682057695e-05,
-      "loss": 0.6899,
-      "step": 3704
-    },
-    {
-      "epoch": 1.820508088823276,
-      "grad_norm": 0.40226960278619833,
-      "learning_rate": 9.053154596035318e-05,
-      "loss": 0.7256,
-      "step": 3705
-    },
-    {
-      "epoch": 1.821000184535892,
-      "grad_norm": 0.39845675735152963,
-      "learning_rate": 9.047024869015794e-05,
-      "loss": 0.7302,
-      "step": 3706
-    },
-    {
-      "epoch": 1.8214922802485083,
-      "grad_norm": 0.3754735430182928,
-      "learning_rate": 9.040895503323248e-05,
-      "loss": 0.7277,
-      "step": 3707
-    },
-    {
-      "epoch": 1.8219843759611245,
-      "grad_norm": 0.39474050952105155,
-      "learning_rate": 9.034766501281671e-05,
-      "loss": 0.6996,
-      "step": 3708
-    },
-    {
-      "epoch": 1.8224764716737405,
-      "grad_norm": 0.3896056496771847,
-      "learning_rate": 9.02863786521492e-05,
-      "loss": 0.7312,
-      "step": 3709
-    },
-    {
-      "epoch": 1.8229685673863565,
-      "grad_norm": 0.3696055463080941,
-      "learning_rate": 9.0225095974467e-05,
-      "loss": 0.7103,
-      "step": 3710
-    },
-    {
-      "epoch": 1.8234606630989727,
-      "grad_norm": 0.3869019189330015,
-      "learning_rate": 9.016381700300598e-05,
-      "loss": 0.793,
-      "step": 3711
-    },
-    {
-      "epoch": 1.823952758811589,
-      "grad_norm": 0.4083584908249029,
-      "learning_rate": 9.010254176100034e-05,
-      "loss": 0.7622,
-      "step": 3712
-    },
-    {
-      "epoch": 1.824444854524205,
-      "grad_norm": 0.3877063911433168,
-      "learning_rate": 9.004127027168307e-05,
-      "loss": 0.6907,
-      "step": 3713
-    },
-    {
-      "epoch": 1.824936950236821,
-      "grad_norm": 0.4029664066401883,
-      "learning_rate": 8.998000255828573e-05,
-      "loss": 0.6933,
-      "step": 3714
-    },
-    {
-      "epoch": 1.8254290459494371,
-      "grad_norm": 0.3770977332100248,
-      "learning_rate": 8.991873864403827e-05,
-      "loss": 0.6523,
-      "step": 3715
-    },
-    {
-      "epoch": 1.8259211416620533,
-      "grad_norm": 0.43050634881317557,
-      "learning_rate": 8.98574785521694e-05,
-      "loss": 0.7379,
-      "step": 3716
-    },
-    {
-      "epoch": 1.8264132373746693,
-      "grad_norm": 0.381995881375049,
-      "learning_rate": 8.979622230590627e-05,
-      "loss": 0.699,
-      "step": 3717
-    },
-    {
-      "epoch": 1.8269053330872855,
-      "grad_norm": 0.39348691077067494,
-      "learning_rate": 8.97349699284746e-05,
-      "loss": 0.7172,
-      "step": 3718
-    },
-    {
-      "epoch": 1.8273974287999017,
-      "grad_norm": 0.4081967569670485,
-      "learning_rate": 8.967372144309864e-05,
-      "loss": 0.7371,
-      "step": 3719
-    },
-    {
-      "epoch": 1.8278895245125177,
-      "grad_norm": 0.3833825353998454,
-      "learning_rate": 8.961247687300122e-05,
-      "loss": 0.7411,
-      "step": 3720
-    },
-    {
-      "epoch": 1.8283816202251337,
-      "grad_norm": 0.39388029814116815,
-      "learning_rate": 8.955123624140355e-05,
-      "loss": 0.7949,
-      "step": 3721
-    },
-    {
-      "epoch": 1.82887371593775,
-      "grad_norm": 0.4108304786282939,
-      "learning_rate": 8.948999957152547e-05,
-      "loss": 0.7347,
-      "step": 3722
-    },
-    {
-      "epoch": 1.8293658116503662,
-      "grad_norm": 0.6030504547135287,
-      "learning_rate": 8.942876688658531e-05,
-      "loss": 0.6942,
-      "step": 3723
-    },
-    {
-      "epoch": 1.8298579073629822,
-      "grad_norm": 0.4350371316300042,
-      "learning_rate": 8.936753820979981e-05,
-      "loss": 0.7535,
-      "step": 3724
-    },
-    {
-      "epoch": 1.8303500030755981,
-      "grad_norm": 0.41955100355530994,
-      "learning_rate": 8.930631356438423e-05,
-      "loss": 0.7797,
-      "step": 3725
-    },
-    {
-      "epoch": 1.8308420987882144,
-      "grad_norm": 0.4347420898352567,
-      "learning_rate": 8.924509297355239e-05,
-      "loss": 0.7939,
-      "step": 3726
-    },
-    {
-      "epoch": 1.8313341945008306,
-      "grad_norm": 0.381354374019232,
-      "learning_rate": 8.91838764605164e-05,
-      "loss": 0.6642,
-      "step": 3727
-    },
-    {
-      "epoch": 1.8318262902134466,
-      "grad_norm": 0.3718501686471073,
-      "learning_rate": 8.912266404848697e-05,
-      "loss": 0.666,
-      "step": 3728
-    },
-    {
-      "epoch": 1.8323183859260626,
-      "grad_norm": 0.3714097611358155,
-      "learning_rate": 8.906145576067319e-05,
-      "loss": 0.6676,
-      "step": 3729
-    },
-    {
-      "epoch": 1.8328104816386788,
-      "grad_norm": 0.41357898781941393,
-      "learning_rate": 8.900025162028257e-05,
-      "loss": 0.747,
-      "step": 3730
-    },
-    {
-      "epoch": 1.833302577351295,
-      "grad_norm": 0.4031385619511016,
-      "learning_rate": 8.893905165052108e-05,
-      "loss": 0.7013,
-      "step": 3731
-    },
-    {
-      "epoch": 1.833794673063911,
-      "grad_norm": 0.36808070463648157,
-      "learning_rate": 8.887785587459319e-05,
-      "loss": 0.6565,
-      "step": 3732
-    },
-    {
-      "epoch": 1.834286768776527,
-      "grad_norm": 0.40993267655499366,
-      "learning_rate": 8.881666431570155e-05,
-      "loss": 0.657,
-      "step": 3733
-    },
-    {
-      "epoch": 1.8347788644891432,
-      "grad_norm": 0.38779860392045135,
-      "learning_rate": 8.875547699704742e-05,
-      "loss": 0.7153,
-      "step": 3734
-    },
-    {
-      "epoch": 1.8352709602017594,
-      "grad_norm": 0.39033315281503617,
-      "learning_rate": 8.869429394183046e-05,
-      "loss": 0.7381,
-      "step": 3735
-    },
-    {
-      "epoch": 1.8357630559143754,
-      "grad_norm": 0.38994489830429646,
-      "learning_rate": 8.863311517324852e-05,
-      "loss": 0.7355,
-      "step": 3736
-    },
-    {
-      "epoch": 1.8362551516269914,
-      "grad_norm": 0.455438199517176,
-      "learning_rate": 8.857194071449804e-05,
-      "loss": 0.706,
-      "step": 3737
-    },
-    {
-      "epoch": 1.8367472473396076,
-      "grad_norm": 0.3922967608564001,
-      "learning_rate": 8.851077058877364e-05,
-      "loss": 0.732,
-      "step": 3738
-    },
-    {
-      "epoch": 1.8372393430522238,
-      "grad_norm": 0.38887486284354605,
-      "learning_rate": 8.844960481926847e-05,
-      "loss": 0.7134,
-      "step": 3739
-    },
-    {
-      "epoch": 1.8377314387648398,
-      "grad_norm": 0.43176106329419306,
-      "learning_rate": 8.83884434291739e-05,
-      "loss": 0.6663,
-      "step": 3740
-    },
-    {
-      "epoch": 1.8382235344774558,
-      "grad_norm": 0.3941384918487974,
-      "learning_rate": 8.83272864416797e-05,
-      "loss": 0.7481,
-      "step": 3741
-    },
-    {
-      "epoch": 1.838715630190072,
-      "grad_norm": 0.40064854172022374,
-      "learning_rate": 8.826613387997393e-05,
-      "loss": 0.6401,
-      "step": 3742
-    },
-    {
-      "epoch": 1.8392077259026882,
-      "grad_norm": 0.38157431859627394,
-      "learning_rate": 8.820498576724307e-05,
-      "loss": 0.6739,
-      "step": 3743
-    },
-    {
-      "epoch": 1.8396998216153042,
-      "grad_norm": 0.3710958900036419,
-      "learning_rate": 8.814384212667175e-05,
-      "loss": 0.6977,
-      "step": 3744
-    },
-    {
-      "epoch": 1.8401919173279202,
-      "grad_norm": 0.4127225948441869,
-      "learning_rate": 8.808270298144304e-05,
-      "loss": 0.6942,
-      "step": 3745
-    },
-    {
-      "epoch": 1.8406840130405364,
-      "grad_norm": 0.3945045820758828,
-      "learning_rate": 8.80215683547383e-05,
-      "loss": 0.7707,
-      "step": 3746
-    },
-    {
-      "epoch": 1.8411761087531526,
-      "grad_norm": 0.37529300969754065,
-      "learning_rate": 8.796043826973705e-05,
-      "loss": 0.6929,
-      "step": 3747
-    },
-    {
-      "epoch": 1.8416682044657686,
-      "grad_norm": 0.3803235173782925,
-      "learning_rate": 8.789931274961724e-05,
-      "loss": 0.6737,
-      "step": 3748
-    },
-    {
-      "epoch": 1.8421603001783846,
-      "grad_norm": 0.40410222428430514,
-      "learning_rate": 8.783819181755504e-05,
-      "loss": 0.6884,
-      "step": 3749
-    },
-    {
-      "epoch": 1.8426523958910008,
-      "grad_norm": 0.38811489242499914,
-      "learning_rate": 8.77770754967248e-05,
-      "loss": 0.7584,
-      "step": 3750
-    },
-    {
-      "epoch": 1.843144491603617,
-      "grad_norm": 0.6251943676212821,
-      "learning_rate": 8.771596381029923e-05,
-      "loss": 0.7277,
-      "step": 3751
-    },
-    {
-      "epoch": 1.843636587316233,
-      "grad_norm": 0.4208576032096547,
-      "learning_rate": 8.765485678144925e-05,
-      "loss": 0.7618,
-      "step": 3752
-    },
-    {
-      "epoch": 1.844128683028849,
-      "grad_norm": 0.4008461840171299,
-      "learning_rate": 8.759375443334396e-05,
-      "loss": 0.7259,
-      "step": 3753
-    },
-    {
-      "epoch": 1.8446207787414652,
-      "grad_norm": 0.4146321017072665,
-      "learning_rate": 8.753265678915076e-05,
-      "loss": 0.6809,
-      "step": 3754
-    },
-    {
-      "epoch": 1.8451128744540815,
-      "grad_norm": 0.42344388837986163,
-      "learning_rate": 8.747156387203528e-05,
-      "loss": 0.6932,
-      "step": 3755
-    },
-    {
-      "epoch": 1.8456049701666974,
-      "grad_norm": 0.42379754679967974,
-      "learning_rate": 8.74104757051612e-05,
-      "loss": 0.771,
-      "step": 3756
-    },
-    {
-      "epoch": 1.8460970658793134,
-      "grad_norm": 0.39649121838114987,
-      "learning_rate": 8.734939231169059e-05,
-      "loss": 0.6894,
-      "step": 3757
-    },
-    {
-      "epoch": 1.8465891615919297,
-      "grad_norm": 0.40998816993359005,
-      "learning_rate": 8.728831371478365e-05,
-      "loss": 0.6918,
-      "step": 3758
-    },
-    {
-      "epoch": 1.8470812573045459,
-      "grad_norm": 0.37188395049459383,
-      "learning_rate": 8.722723993759869e-05,
-      "loss": 0.6585,
-      "step": 3759
-    },
-    {
-      "epoch": 1.8475733530171619,
-      "grad_norm": 0.39116038860538377,
-      "learning_rate": 8.716617100329231e-05,
-      "loss": 0.7328,
-      "step": 3760
-    },
-    {
-      "epoch": 1.8480654487297778,
-      "grad_norm": 0.3932310254133637,
-      "learning_rate": 8.710510693501912e-05,
-      "loss": 0.6873,
-      "step": 3761
-    },
-    {
-      "epoch": 1.848557544442394,
-      "grad_norm": 0.3913915052836477,
-      "learning_rate": 8.704404775593205e-05,
-      "loss": 0.6605,
-      "step": 3762
-    },
-    {
-      "epoch": 1.8490496401550103,
-      "grad_norm": 0.4031210169561732,
-      "learning_rate": 8.698299348918209e-05,
-      "loss": 0.707,
-      "step": 3763
-    },
-    {
-      "epoch": 1.8495417358676263,
-      "grad_norm": 0.37107837027361323,
-      "learning_rate": 8.692194415791834e-05,
-      "loss": 0.6348,
-      "step": 3764
-    },
-    {
-      "epoch": 1.8500338315802423,
-      "grad_norm": 0.42508606082857237,
-      "learning_rate": 8.68608997852881e-05,
-      "loss": 0.8115,
-      "step": 3765
-    },
-    {
-      "epoch": 1.8505259272928585,
-      "grad_norm": 0.42053473150393383,
-      "learning_rate": 8.679986039443679e-05,
-      "loss": 0.6626,
-      "step": 3766
-    },
-    {
-      "epoch": 1.8510180230054747,
-      "grad_norm": 0.4355712014350185,
-      "learning_rate": 8.673882600850782e-05,
-      "loss": 0.6569,
-      "step": 3767
-    },
-    {
-      "epoch": 1.8515101187180907,
-      "grad_norm": 0.41190927012735623,
-      "learning_rate": 8.667779665064284e-05,
-      "loss": 0.8091,
-      "step": 3768
-    },
-    {
-      "epoch": 1.8520022144307067,
-      "grad_norm": 0.3898814652389425,
-      "learning_rate": 8.66167723439816e-05,
-      "loss": 0.7222,
-      "step": 3769
-    },
-    {
-      "epoch": 1.8524943101433229,
-      "grad_norm": 0.3614475994894846,
-      "learning_rate": 8.655575311166178e-05,
-      "loss": 0.7334,
-      "step": 3770
-    },
-    {
-      "epoch": 1.852986405855939,
-      "grad_norm": 0.4777743120470902,
-      "learning_rate": 8.649473897681927e-05,
-      "loss": 0.7197,
-      "step": 3771
-    },
-    {
-      "epoch": 1.853478501568555,
-      "grad_norm": 0.35562573345109555,
-      "learning_rate": 8.643372996258807e-05,
-      "loss": 0.6453,
-      "step": 3772
-    },
-    {
-      "epoch": 1.853970597281171,
-      "grad_norm": 0.4039928739211572,
-      "learning_rate": 8.637272609210004e-05,
-      "loss": 0.6872,
-      "step": 3773
-    },
-    {
-      "epoch": 1.8544626929937873,
-      "grad_norm": 0.38841040229625956,
-      "learning_rate": 8.63117273884853e-05,
-      "loss": 0.6808,
-      "step": 3774
-    },
-    {
-      "epoch": 1.8549547887064035,
-      "grad_norm": 0.4096237991473063,
-      "learning_rate": 8.625073387487187e-05,
-      "loss": 0.7119,
-      "step": 3775
-    },
-    {
-      "epoch": 1.8554468844190195,
-      "grad_norm": 0.38596576523341325,
-      "learning_rate": 8.618974557438588e-05,
-      "loss": 0.6666,
-      "step": 3776
-    },
-    {
-      "epoch": 1.8559389801316355,
-      "grad_norm": 0.39409470141726155,
-      "learning_rate": 8.612876251015143e-05,
-      "loss": 0.7382,
-      "step": 3777
-    },
-    {
-      "epoch": 1.8564310758442517,
-      "grad_norm": 0.38532395588137713,
-      "learning_rate": 8.606778470529072e-05,
-      "loss": 0.7215,
-      "step": 3778
-    },
-    {
-      "epoch": 1.856923171556868,
-      "grad_norm": 0.3928612260370393,
-      "learning_rate": 8.600681218292382e-05,
-      "loss": 0.7244,
-      "step": 3779
-    },
-    {
-      "epoch": 1.857415267269484,
-      "grad_norm": 0.41470702589972863,
-      "learning_rate": 8.594584496616892e-05,
-      "loss": 0.6866,
-      "step": 3780
-    },
-    {
-      "epoch": 1.8579073629821,
-      "grad_norm": 0.40288101987634306,
-      "learning_rate": 8.588488307814219e-05,
-      "loss": 0.6973,
-      "step": 3781
-    },
-    {
-      "epoch": 1.8583994586947161,
-      "grad_norm": 0.40182217552033306,
-      "learning_rate": 8.582392654195765e-05,
-      "loss": 0.7019,
-      "step": 3782
-    },
-    {
-      "epoch": 1.8588915544073323,
-      "grad_norm": 0.38297590120797004,
-      "learning_rate": 8.576297538072745e-05,
-      "loss": 0.6397,
-      "step": 3783
-    },
-    {
-      "epoch": 1.8593836501199483,
-      "grad_norm": 0.4485398175862643,
-      "learning_rate": 8.570202961756166e-05,
-      "loss": 0.7402,
-      "step": 3784
-    },
-    {
-      "epoch": 1.8598757458325643,
-      "grad_norm": 0.4066828438215999,
-      "learning_rate": 8.564108927556821e-05,
-      "loss": 0.7403,
-      "step": 3785
-    },
-    {
-      "epoch": 1.8603678415451805,
-      "grad_norm": 0.44161212245553355,
-      "learning_rate": 8.558015437785307e-05,
-      "loss": 0.7408,
-      "step": 3786
-    },
-    {
-      "epoch": 1.8608599372577967,
-      "grad_norm": 0.40158467712899054,
-      "learning_rate": 8.551922494752014e-05,
-      "loss": 0.7313,
-      "step": 3787
-    },
-    {
-      "epoch": 1.8613520329704127,
-      "grad_norm": 0.39374026714890287,
-      "learning_rate": 8.545830100767119e-05,
-      "loss": 0.8011,
-      "step": 3788
-    },
-    {
-      "epoch": 1.8618441286830287,
-      "grad_norm": 0.40035521905149524,
-      "learning_rate": 8.539738258140599e-05,
-      "loss": 0.7178,
-      "step": 3789
-    },
-    {
-      "epoch": 1.862336224395645,
-      "grad_norm": 0.385540295090378,
-      "learning_rate": 8.533646969182212e-05,
-      "loss": 0.7086,
-      "step": 3790
-    },
-    {
-      "epoch": 1.8628283201082612,
-      "grad_norm": 0.3939001451499724,
-      "learning_rate": 8.527556236201512e-05,
-      "loss": 0.7987,
-      "step": 3791
-    },
-    {
-      "epoch": 1.8633204158208772,
-      "grad_norm": 0.38453740296294187,
-      "learning_rate": 8.521466061507851e-05,
-      "loss": 0.7068,
-      "step": 3792
-    },
-    {
-      "epoch": 1.8638125115334931,
-      "grad_norm": 0.35669861262940133,
-      "learning_rate": 8.515376447410348e-05,
-      "loss": 0.652,
-      "step": 3793
-    },
-    {
-      "epoch": 1.8643046072461094,
-      "grad_norm": 0.4445744310608242,
-      "learning_rate": 8.509287396217927e-05,
-      "loss": 0.6823,
-      "step": 3794
-    },
-    {
-      "epoch": 1.8647967029587256,
-      "grad_norm": 0.4148131315069323,
-      "learning_rate": 8.503198910239296e-05,
-      "loss": 0.7272,
-      "step": 3795
-    },
-    {
-      "epoch": 1.8652887986713416,
-      "grad_norm": 0.37908432290668054,
-      "learning_rate": 8.497110991782938e-05,
-      "loss": 0.6588,
-      "step": 3796
-    },
-    {
-      "epoch": 1.8657808943839576,
-      "grad_norm": 0.36803127144774506,
-      "learning_rate": 8.491023643157135e-05,
-      "loss": 0.6546,
-      "step": 3797
-    },
-    {
-      "epoch": 1.8662729900965738,
-      "grad_norm": 0.38550432925741684,
-      "learning_rate": 8.484936866669945e-05,
-      "loss": 0.6906,
-      "step": 3798
-    },
-    {
-      "epoch": 1.86676508580919,
-      "grad_norm": 0.3773157261334979,
-      "learning_rate": 8.47885066462921e-05,
-      "loss": 0.7224,
-      "step": 3799
-    },
-    {
-      "epoch": 1.867257181521806,
-      "grad_norm": 0.3961376302003408,
-      "learning_rate": 8.472765039342551e-05,
-      "loss": 0.7273,
-      "step": 3800
-    },
-    {
-      "epoch": 1.867749277234422,
-      "grad_norm": 0.3946762702992294,
-      "learning_rate": 8.466679993117384e-05,
-      "loss": 0.7254,
-      "step": 3801
-    },
-    {
-      "epoch": 1.8682413729470382,
-      "grad_norm": 0.37808628750504814,
-      "learning_rate": 8.460595528260883e-05,
-      "loss": 0.7114,
-      "step": 3802
-    },
-    {
-      "epoch": 1.8687334686596544,
-      "grad_norm": 0.38643394808877063,
-      "learning_rate": 8.45451164708002e-05,
-      "loss": 0.7016,
-      "step": 3803
-    },
-    {
-      "epoch": 1.8692255643722704,
-      "grad_norm": 0.429300538303249,
-      "learning_rate": 8.448428351881545e-05,
-      "loss": 0.7641,
-      "step": 3804
-    },
-    {
-      "epoch": 1.8697176600848864,
-      "grad_norm": 0.43625152503785475,
-      "learning_rate": 8.442345644971972e-05,
-      "loss": 0.7163,
-      "step": 3805
-    },
-    {
-      "epoch": 1.8702097557975026,
-      "grad_norm": 0.36460530301815747,
-      "learning_rate": 8.436263528657605e-05,
-      "loss": 0.6496,
-      "step": 3806
-    },
-    {
-      "epoch": 1.8707018515101188,
-      "grad_norm": 0.38492339749305676,
-      "learning_rate": 8.430182005244521e-05,
-      "loss": 0.6765,
-      "step": 3807
-    },
-    {
-      "epoch": 1.8711939472227348,
-      "grad_norm": 0.3807975636968924,
-      "learning_rate": 8.424101077038568e-05,
-      "loss": 0.681,
-      "step": 3808
-    },
-    {
-      "epoch": 1.8716860429353508,
-      "grad_norm": 0.3732510035840216,
-      "learning_rate": 8.418020746345371e-05,
-      "loss": 0.6569,
-      "step": 3809
-    },
-    {
-      "epoch": 1.872178138647967,
-      "grad_norm": 0.40355547433002115,
-      "learning_rate": 8.411941015470329e-05,
-      "loss": 0.6905,
-      "step": 3810
-    },
-    {
-      "epoch": 1.8726702343605832,
-      "grad_norm": 0.38860278114060753,
-      "learning_rate": 8.405861886718614e-05,
-      "loss": 0.7857,
-      "step": 3811
-    },
-    {
-      "epoch": 1.8731623300731992,
-      "grad_norm": 0.3625925085463039,
-      "learning_rate": 8.399783362395169e-05,
-      "loss": 0.5675,
-      "step": 3812
-    },
-    {
-      "epoch": 1.8736544257858152,
-      "grad_norm": 0.4011507589155444,
-      "learning_rate": 8.393705444804704e-05,
-      "loss": 0.6947,
-      "step": 3813
-    },
-    {
-      "epoch": 1.8741465214984314,
-      "grad_norm": 0.3914846994760786,
-      "learning_rate": 8.387628136251707e-05,
-      "loss": 0.6686,
-      "step": 3814
-    },
-    {
-      "epoch": 1.8746386172110476,
-      "grad_norm": 0.44190174275672733,
-      "learning_rate": 8.381551439040433e-05,
-      "loss": 0.7964,
-      "step": 3815
-    },
-    {
-      "epoch": 1.8751307129236636,
-      "grad_norm": 0.3691434321689682,
-      "learning_rate": 8.375475355474895e-05,
-      "loss": 0.6825,
-      "step": 3816
-    },
-    {
-      "epoch": 1.8756228086362796,
-      "grad_norm": 0.39352120912167615,
-      "learning_rate": 8.369399887858886e-05,
-      "loss": 0.6894,
-      "step": 3817
-    },
-    {
-      "epoch": 1.8761149043488958,
-      "grad_norm": 0.378502360716238,
-      "learning_rate": 8.363325038495965e-05,
-      "loss": 0.6574,
-      "step": 3818
-    },
-    {
-      "epoch": 1.876607000061512,
-      "grad_norm": 0.40311105925092716,
-      "learning_rate": 8.357250809689442e-05,
-      "loss": 0.7525,
-      "step": 3819
-    },
-    {
-      "epoch": 1.877099095774128,
-      "grad_norm": 0.39149881189310953,
-      "learning_rate": 8.351177203742412e-05,
-      "loss": 0.7102,
-      "step": 3820
-    },
-    {
-      "epoch": 1.877591191486744,
-      "grad_norm": 0.37918226075791184,
-      "learning_rate": 8.345104222957719e-05,
-      "loss": 0.6966,
-      "step": 3821
-    },
-    {
-      "epoch": 1.8780832871993602,
-      "grad_norm": 0.4225687066624157,
-      "learning_rate": 8.339031869637974e-05,
-      "loss": 0.8542,
-      "step": 3822
-    },
-    {
-      "epoch": 1.8785753829119765,
-      "grad_norm": 0.39847575621164844,
-      "learning_rate": 8.332960146085551e-05,
-      "loss": 0.7559,
-      "step": 3823
-    },
-    {
-      "epoch": 1.8790674786245924,
-      "grad_norm": 0.385056347103929,
-      "learning_rate": 8.32688905460259e-05,
-      "loss": 0.7425,
-      "step": 3824
-    },
-    {
-      "epoch": 1.8795595743372084,
-      "grad_norm": 0.38433150729937143,
-      "learning_rate": 8.32081859749098e-05,
-      "loss": 0.7067,
-      "step": 3825
-    },
-    {
-      "epoch": 1.8800516700498247,
-      "grad_norm": 0.4001505978656989,
-      "learning_rate": 8.314748777052377e-05,
-      "loss": 0.6428,
-      "step": 3826
-    },
-    {
-      "epoch": 1.8805437657624409,
-      "grad_norm": 0.3692886739999643,
-      "learning_rate": 8.308679595588203e-05,
-      "loss": 0.6267,
-      "step": 3827
-    },
-    {
-      "epoch": 1.8810358614750569,
-      "grad_norm": 0.41017986026468556,
-      "learning_rate": 8.302611055399616e-05,
-      "loss": 0.7414,
-      "step": 3828
-    },
-    {
-      "epoch": 1.8815279571876729,
-      "grad_norm": 0.39691819101663334,
-      "learning_rate": 8.296543158787553e-05,
-      "loss": 0.6985,
-      "step": 3829
-    },
-    {
-      "epoch": 1.882020052900289,
-      "grad_norm": 0.4163231228785293,
-      "learning_rate": 8.2904759080527e-05,
-      "loss": 0.7806,
-      "step": 3830
-    },
-    {
-      "epoch": 1.8825121486129053,
-      "grad_norm": 0.3916218754190283,
-      "learning_rate": 8.284409305495489e-05,
-      "loss": 0.7054,
-      "step": 3831
-    },
-    {
-      "epoch": 1.8830042443255213,
-      "grad_norm": 0.38331071958594176,
-      "learning_rate": 8.278343353416114e-05,
-      "loss": 0.7215,
-      "step": 3832
-    },
-    {
-      "epoch": 1.8834963400381373,
-      "grad_norm": 0.39088909814304723,
-      "learning_rate": 8.27227805411453e-05,
-      "loss": 0.6823,
-      "step": 3833
-    },
-    {
-      "epoch": 1.8839884357507535,
-      "grad_norm": 0.3974172902358765,
-      "learning_rate": 8.266213409890427e-05,
-      "loss": 0.6934,
-      "step": 3834
-    },
-    {
-      "epoch": 1.8844805314633697,
-      "grad_norm": 0.3782369679192204,
-      "learning_rate": 8.260149423043263e-05,
-      "loss": 0.6891,
-      "step": 3835
-    },
-    {
-      "epoch": 1.8849726271759857,
-      "grad_norm": 0.392966638649635,
-      "learning_rate": 8.254086095872232e-05,
-      "loss": 0.7512,
-      "step": 3836
-    },
-    {
-      "epoch": 1.8854647228886017,
-      "grad_norm": 0.394558787642243,
-      "learning_rate": 8.24802343067629e-05,
-      "loss": 0.7267,
-      "step": 3837
-    },
-    {
-      "epoch": 1.8859568186012179,
-      "grad_norm": 0.3792863680994872,
-      "learning_rate": 8.241961429754144e-05,
-      "loss": 0.6789,
-      "step": 3838
-    },
-    {
-      "epoch": 1.886448914313834,
-      "grad_norm": 0.38164515637351726,
-      "learning_rate": 8.235900095404231e-05,
-      "loss": 0.7319,
-      "step": 3839
-    },
-    {
-      "epoch": 1.88694101002645,
-      "grad_norm": 0.4099016397638299,
-      "learning_rate": 8.229839429924753e-05,
-      "loss": 0.7304,
-      "step": 3840
-    },
-    {
-      "epoch": 1.887433105739066,
-      "grad_norm": 0.38415307587009967,
-      "learning_rate": 8.223779435613654e-05,
-      "loss": 0.7236,
-      "step": 3841
-    },
-    {
-      "epoch": 1.8879252014516823,
-      "grad_norm": 0.4408150407896303,
-      "learning_rate": 8.217720114768618e-05,
-      "loss": 0.7462,
-      "step": 3842
-    },
-    {
-      "epoch": 1.8884172971642985,
-      "grad_norm": 0.43569175301612517,
-      "learning_rate": 8.21166146968708e-05,
-      "loss": 0.7277,
-      "step": 3843
-    },
-    {
-      "epoch": 1.8889093928769145,
-      "grad_norm": 0.4092180401833967,
-      "learning_rate": 8.205603502666216e-05,
-      "loss": 0.7862,
-      "step": 3844
-    },
-    {
-      "epoch": 1.8894014885895305,
-      "grad_norm": 0.39599779906032656,
-      "learning_rate": 8.199546216002945e-05,
-      "loss": 0.628,
-      "step": 3845
-    },
-    {
-      "epoch": 1.889893584302147,
-      "grad_norm": 0.3979126754120914,
-      "learning_rate": 8.193489611993926e-05,
-      "loss": 0.7038,
-      "step": 3846
-    },
-    {
-      "epoch": 1.890385680014763,
-      "grad_norm": 0.4117737153450986,
-      "learning_rate": 8.18743369293557e-05,
-      "loss": 0.6989,
-      "step": 3847
-    },
-    {
-      "epoch": 1.890877775727379,
-      "grad_norm": 0.4046921419899561,
-      "learning_rate": 8.181378461124006e-05,
-      "loss": 0.6912,
-      "step": 3848
-    },
-    {
-      "epoch": 1.8913698714399951,
-      "grad_norm": 0.3711360123633831,
-      "learning_rate": 8.175323918855125e-05,
-      "loss": 0.6544,
-      "step": 3849
-    },
-    {
-      "epoch": 1.8918619671526113,
-      "grad_norm": 0.40231408623806025,
-      "learning_rate": 8.169270068424549e-05,
-      "loss": 0.6882,
-      "step": 3850
-    },
-    {
-      "epoch": 1.8923540628652273,
-      "grad_norm": 0.40330329724477676,
-      "learning_rate": 8.163216912127632e-05,
-      "loss": 0.7932,
-      "step": 3851
-    },
-    {
-      "epoch": 1.8928461585778433,
-      "grad_norm": 0.37383738566453517,
-      "learning_rate": 8.157164452259469e-05,
-      "loss": 0.718,
-      "step": 3852
-    },
-    {
-      "epoch": 1.8933382542904595,
-      "grad_norm": 0.3831247990844017,
-      "learning_rate": 8.151112691114899e-05,
-      "loss": 0.7086,
-      "step": 3853
-    },
-    {
-      "epoch": 1.8938303500030758,
-      "grad_norm": 0.35959035294624553,
-      "learning_rate": 8.145061630988479e-05,
-      "loss": 0.7009,
-      "step": 3854
-    },
-    {
-      "epoch": 1.8943224457156917,
-      "grad_norm": 0.3918304765662465,
-      "learning_rate": 8.13901127417451e-05,
-      "loss": 0.7179,
-      "step": 3855
-    },
-    {
-      "epoch": 1.8948145414283077,
-      "grad_norm": 1.0671166540530197,
-      "learning_rate": 8.132961622967035e-05,
-      "loss": 0.7583,
-      "step": 3856
-    },
-    {
-      "epoch": 1.895306637140924,
-      "grad_norm": 0.41056659421614605,
-      "learning_rate": 8.126912679659808e-05,
-      "loss": 0.7938,
-      "step": 3857
-    },
-    {
-      "epoch": 1.8957987328535402,
-      "grad_norm": 0.3940836886653735,
-      "learning_rate": 8.120864446546338e-05,
-      "loss": 0.7456,
-      "step": 3858
-    },
-    {
-      "epoch": 1.8962908285661562,
-      "grad_norm": 0.3958673026190194,
-      "learning_rate": 8.114816925919844e-05,
-      "loss": 0.6397,
-      "step": 3859
-    },
-    {
-      "epoch": 1.8967829242787722,
-      "grad_norm": 0.4154842629060033,
-      "learning_rate": 8.108770120073289e-05,
-      "loss": 0.7005,
-      "step": 3860
-    },
-    {
-      "epoch": 1.8972750199913884,
-      "grad_norm": 0.41679603606848203,
-      "learning_rate": 8.102724031299362e-05,
-      "loss": 0.6887,
-      "step": 3861
-    },
-    {
-      "epoch": 1.8977671157040046,
-      "grad_norm": 0.37960669953571574,
-      "learning_rate": 8.096678661890475e-05,
-      "loss": 0.6616,
-      "step": 3862
-    },
-    {
-      "epoch": 1.8982592114166206,
-      "grad_norm": 0.41898795118992954,
-      "learning_rate": 8.090634014138771e-05,
-      "loss": 0.7603,
-      "step": 3863
-    },
-    {
-      "epoch": 1.8987513071292366,
-      "grad_norm": 0.3901178471684261,
-      "learning_rate": 8.084590090336127e-05,
-      "loss": 0.7369,
-      "step": 3864
-    },
-    {
-      "epoch": 1.8992434028418528,
-      "grad_norm": 0.4001074664480027,
-      "learning_rate": 8.078546892774126e-05,
-      "loss": 0.7181,
-      "step": 3865
-    },
-    {
-      "epoch": 1.899735498554469,
-      "grad_norm": 0.412856381844074,
-      "learning_rate": 8.072504423744094e-05,
-      "loss": 0.7099,
-      "step": 3866
-    },
-    {
-      "epoch": 1.900227594267085,
-      "grad_norm": 0.3960452658443214,
-      "learning_rate": 8.066462685537074e-05,
-      "loss": 0.7229,
-      "step": 3867
-    },
-    {
-      "epoch": 1.900719689979701,
-      "grad_norm": 0.6320535315638518,
-      "learning_rate": 8.060421680443831e-05,
-      "loss": 0.7684,
-      "step": 3868
-    },
-    {
-      "epoch": 1.9012117856923172,
-      "grad_norm": 0.4146782766110037,
-      "learning_rate": 8.054381410754848e-05,
-      "loss": 0.7029,
-      "step": 3869
-    },
-    {
-      "epoch": 1.9017038814049334,
-      "grad_norm": 0.40409425232709,
-      "learning_rate": 8.048341878760345e-05,
-      "loss": 0.7093,
-      "step": 3870
-    },
-    {
-      "epoch": 1.9021959771175494,
-      "grad_norm": 0.3892420486298809,
-      "learning_rate": 8.042303086750241e-05,
-      "loss": 0.6532,
-      "step": 3871
-    },
-    {
-      "epoch": 1.9026880728301654,
-      "grad_norm": 0.389899415150913,
-      "learning_rate": 8.036265037014188e-05,
-      "loss": 0.7346,
-      "step": 3872
-    },
-    {
-      "epoch": 1.9031801685427816,
-      "grad_norm": 0.4252139440574835,
-      "learning_rate": 8.030227731841562e-05,
-      "loss": 0.7112,
-      "step": 3873
-    },
-    {
-      "epoch": 1.9036722642553978,
-      "grad_norm": 0.3870187866545771,
-      "learning_rate": 8.024191173521435e-05,
-      "loss": 0.7001,
-      "step": 3874
-    },
-    {
-      "epoch": 1.9041643599680138,
-      "grad_norm": 0.3896495111815748,
-      "learning_rate": 8.018155364342615e-05,
-      "loss": 0.688,
-      "step": 3875
-    },
-    {
-      "epoch": 1.9046564556806298,
-      "grad_norm": 0.37281523882922024,
-      "learning_rate": 8.012120306593623e-05,
-      "loss": 0.6826,
-      "step": 3876
-    },
-    {
-      "epoch": 1.905148551393246,
-      "grad_norm": 0.3976998963384889,
-      "learning_rate": 8.006086002562689e-05,
-      "loss": 0.749,
-      "step": 3877
-    },
-    {
-      "epoch": 1.9056406471058622,
-      "grad_norm": 0.37228780960957175,
-      "learning_rate": 8.000052454537756e-05,
-      "loss": 0.6921,
-      "step": 3878
-    },
-    {
-      "epoch": 1.9061327428184782,
-      "grad_norm": 0.4232935279839646,
-      "learning_rate": 7.994019664806495e-05,
-      "loss": 0.7563,
-      "step": 3879
-    },
-    {
-      "epoch": 1.9066248385310942,
-      "grad_norm": 0.3939288567329543,
-      "learning_rate": 7.987987635656267e-05,
-      "loss": 0.6897,
-      "step": 3880
-    },
-    {
-      "epoch": 1.9071169342437104,
-      "grad_norm": 0.39392641292806546,
-      "learning_rate": 7.981956369374164e-05,
-      "loss": 0.7314,
-      "step": 3881
-    },
-    {
-      "epoch": 1.9076090299563266,
-      "grad_norm": 0.3731836872853506,
-      "learning_rate": 7.975925868246985e-05,
-      "loss": 0.7107,
-      "step": 3882
-    },
-    {
-      "epoch": 1.9081011256689426,
-      "grad_norm": 0.39577720252997817,
-      "learning_rate": 7.969896134561226e-05,
-      "loss": 0.7301,
-      "step": 3883
-    },
-    {
-      "epoch": 1.9085932213815586,
-      "grad_norm": 0.407424447362973,
-      "learning_rate": 7.963867170603109e-05,
-      "loss": 0.7504,
-      "step": 3884
-    },
-    {
-      "epoch": 1.9090853170941748,
-      "grad_norm": 0.4036401411701683,
-      "learning_rate": 7.95783897865855e-05,
-      "loss": 0.7475,
-      "step": 3885
-    },
-    {
-      "epoch": 1.909577412806791,
-      "grad_norm": 0.4103005449659771,
-      "learning_rate": 7.951811561013181e-05,
-      "loss": 0.77,
-      "step": 3886
-    },
-    {
-      "epoch": 1.910069508519407,
-      "grad_norm": 0.38510200906900166,
-      "learning_rate": 7.945784919952345e-05,
-      "loss": 0.7327,
-      "step": 3887
-    },
-    {
-      "epoch": 1.910561604232023,
-      "grad_norm": 0.41653972275927953,
-      "learning_rate": 7.939759057761075e-05,
-      "loss": 0.7525,
-      "step": 3888
-    },
-    {
-      "epoch": 1.9110536999446393,
-      "grad_norm": 0.3832676976187471,
-      "learning_rate": 7.933733976724121e-05,
-      "loss": 0.6463,
-      "step": 3889
-    },
-    {
-      "epoch": 1.9115457956572555,
-      "grad_norm": 0.3727547330165016,
-      "learning_rate": 7.927709679125935e-05,
-      "loss": 0.6828,
-      "step": 3890
-    },
-    {
-      "epoch": 1.9120378913698715,
-      "grad_norm": 0.37988777052469785,
-      "learning_rate": 7.921686167250668e-05,
-      "loss": 0.743,
-      "step": 3891
-    },
-    {
-      "epoch": 1.9125299870824874,
-      "grad_norm": 0.39145387833034767,
-      "learning_rate": 7.915663443382173e-05,
-      "loss": 0.6608,
-      "step": 3892
-    },
-    {
-      "epoch": 1.9130220827951037,
-      "grad_norm": 0.4043528265935617,
-      "learning_rate": 7.909641509804015e-05,
-      "loss": 0.6648,
-      "step": 3893
-    },
-    {
-      "epoch": 1.9135141785077199,
-      "grad_norm": 0.4082358905888069,
-      "learning_rate": 7.903620368799439e-05,
-      "loss": 0.7488,
-      "step": 3894
-    },
-    {
-      "epoch": 1.9140062742203359,
-      "grad_norm": 0.4009678832066102,
-      "learning_rate": 7.897600022651409e-05,
-      "loss": 0.7247,
-      "step": 3895
-    },
-    {
-      "epoch": 1.9144983699329519,
-      "grad_norm": 0.38337578975659437,
-      "learning_rate": 7.891580473642582e-05,
-      "loss": 0.7509,
-      "step": 3896
-    },
-    {
-      "epoch": 1.914990465645568,
-      "grad_norm": 0.40866859945165235,
-      "learning_rate": 7.885561724055305e-05,
-      "loss": 0.69,
-      "step": 3897
-    },
-    {
-      "epoch": 1.9154825613581843,
-      "grad_norm": 0.4183788383491655,
-      "learning_rate": 7.87954377617163e-05,
-      "loss": 0.7324,
-      "step": 3898
-    },
-    {
-      "epoch": 1.9159746570708003,
-      "grad_norm": 0.37721860259448187,
-      "learning_rate": 7.873526632273304e-05,
-      "loss": 0.7288,
-      "step": 3899
-    },
-    {
-      "epoch": 1.9164667527834163,
-      "grad_norm": 0.36384014459340297,
-      "learning_rate": 7.86751029464177e-05,
-      "loss": 0.6763,
-      "step": 3900
-    },
-    {
-      "epoch": 1.9169588484960325,
-      "grad_norm": 0.3981900478021887,
-      "learning_rate": 7.861494765558153e-05,
-      "loss": 0.7455,
-      "step": 3901
-    },
-    {
-      "epoch": 1.9174509442086487,
-      "grad_norm": 0.37327464484386186,
-      "learning_rate": 7.855480047303296e-05,
-      "loss": 0.6973,
-      "step": 3902
-    },
-    {
-      "epoch": 1.9179430399212647,
-      "grad_norm": 0.3981322933258857,
-      "learning_rate": 7.849466142157708e-05,
-      "loss": 0.7177,
-      "step": 3903
-    },
-    {
-      "epoch": 1.9184351356338807,
-      "grad_norm": 0.4064397438066983,
-      "learning_rate": 7.843453052401606e-05,
-      "loss": 0.7306,
-      "step": 3904
-    },
-    {
-      "epoch": 1.918927231346497,
-      "grad_norm": 0.3880128323398334,
-      "learning_rate": 7.8374407803149e-05,
-      "loss": 0.7034,
-      "step": 3905
-    },
-    {
-      "epoch": 1.9194193270591131,
-      "grad_norm": 0.3880484616530206,
-      "learning_rate": 7.831429328177172e-05,
-      "loss": 0.6771,
-      "step": 3906
-    },
-    {
-      "epoch": 1.919911422771729,
-      "grad_norm": 0.41919304250603023,
-      "learning_rate": 7.825418698267717e-05,
-      "loss": 0.6921,
-      "step": 3907
-    },
-    {
-      "epoch": 1.920403518484345,
-      "grad_norm": 0.4092797013272745,
-      "learning_rate": 7.819408892865496e-05,
-      "loss": 0.7241,
-      "step": 3908
-    },
-    {
-      "epoch": 1.9208956141969613,
-      "grad_norm": 0.37058491846323266,
-      "learning_rate": 7.81339991424917e-05,
-      "loss": 0.606,
-      "step": 3909
-    },
-    {
-      "epoch": 1.9213877099095775,
-      "grad_norm": 0.39136055759246124,
-      "learning_rate": 7.80739176469709e-05,
-      "loss": 0.7127,
-      "step": 3910
-    },
-    {
-      "epoch": 1.9218798056221935,
-      "grad_norm": 0.37185122565731515,
-      "learning_rate": 7.801384446487278e-05,
-      "loss": 0.6649,
-      "step": 3911
-    },
-    {
-      "epoch": 1.9223719013348095,
-      "grad_norm": 0.36242614478350127,
-      "learning_rate": 7.795377961897457e-05,
-      "loss": 0.6752,
-      "step": 3912
-    },
-    {
-      "epoch": 1.9228639970474257,
-      "grad_norm": 0.38954244258229026,
-      "learning_rate": 7.78937231320502e-05,
-      "loss": 0.7113,
-      "step": 3913
-    },
-    {
-      "epoch": 1.923356092760042,
-      "grad_norm": 0.6042183187900263,
-      "learning_rate": 7.783367502687055e-05,
-      "loss": 0.6591,
-      "step": 3914
-    },
-    {
-      "epoch": 1.923848188472658,
-      "grad_norm": 0.37240936186016527,
-      "learning_rate": 7.777363532620322e-05,
-      "loss": 0.7206,
-      "step": 3915
-    },
-    {
-      "epoch": 1.924340284185274,
-      "grad_norm": 0.3908463141441568,
-      "learning_rate": 7.77136040528127e-05,
-      "loss": 0.6938,
-      "step": 3916
-    },
-    {
-      "epoch": 1.9248323798978901,
-      "grad_norm": 0.36547252043384504,
-      "learning_rate": 7.765358122946023e-05,
-      "loss": 0.681,
-      "step": 3917
-    },
-    {
-      "epoch": 1.9253244756105063,
-      "grad_norm": 0.36236724630938455,
-      "learning_rate": 7.759356687890387e-05,
-      "loss": 0.7331,
-      "step": 3918
-    },
-    {
-      "epoch": 1.9258165713231223,
-      "grad_norm": 0.3858341600511175,
-      "learning_rate": 7.753356102389851e-05,
-      "loss": 0.6595,
-      "step": 3919
-    },
-    {
-      "epoch": 1.9263086670357383,
-      "grad_norm": 0.3648928111427161,
-      "learning_rate": 7.747356368719572e-05,
-      "loss": 0.6407,
-      "step": 3920
-    },
-    {
-      "epoch": 1.9268007627483545,
-      "grad_norm": 0.35891837542860777,
-      "learning_rate": 7.741357489154393e-05,
-      "loss": 0.6502,
-      "step": 3921
-    },
-    {
-      "epoch": 1.9272928584609708,
-      "grad_norm": 0.3511748862513752,
-      "learning_rate": 7.735359465968833e-05,
-      "loss": 0.6547,
-      "step": 3922
-    },
-    {
-      "epoch": 1.9277849541735868,
-      "grad_norm": 0.38882962060119547,
-      "learning_rate": 7.729362301437076e-05,
-      "loss": 0.7529,
-      "step": 3923
-    },
-    {
-      "epoch": 1.9282770498862027,
-      "grad_norm": 0.41045132673897666,
-      "learning_rate": 7.723365997832989e-05,
-      "loss": 0.7183,
-      "step": 3924
-    },
-    {
-      "epoch": 1.928769145598819,
-      "grad_norm": 0.36737314744463456,
-      "learning_rate": 7.717370557430119e-05,
-      "loss": 0.6562,
-      "step": 3925
-    },
-    {
-      "epoch": 1.9292612413114352,
-      "grad_norm": 0.35477206335805433,
-      "learning_rate": 7.711375982501666e-05,
-      "loss": 0.6714,
-      "step": 3926
-    },
-    {
-      "epoch": 1.9297533370240512,
-      "grad_norm": 0.3644726863724075,
-      "learning_rate": 7.70538227532052e-05,
-      "loss": 0.5821,
-      "step": 3927
-    },
-    {
-      "epoch": 1.9302454327366672,
-      "grad_norm": 0.37967583596356574,
-      "learning_rate": 7.69938943815924e-05,
-      "loss": 0.6873,
-      "step": 3928
-    },
-    {
-      "epoch": 1.9307375284492834,
-      "grad_norm": 0.40669346566849485,
-      "learning_rate": 7.693397473290042e-05,
-      "loss": 0.7506,
-      "step": 3929
-    },
-    {
-      "epoch": 1.9312296241618996,
-      "grad_norm": 0.35485559877989564,
-      "learning_rate": 7.687406382984824e-05,
-      "loss": 0.7025,
-      "step": 3930
-    },
-    {
-      "epoch": 1.9317217198745156,
-      "grad_norm": 0.3874316845253912,
-      "learning_rate": 7.681416169515153e-05,
-      "loss": 0.7445,
-      "step": 3931
-    },
-    {
-      "epoch": 1.9322138155871316,
-      "grad_norm": 0.36698998710942177,
-      "learning_rate": 7.675426835152251e-05,
-      "loss": 0.6624,
-      "step": 3932
-    },
-    {
-      "epoch": 1.9327059112997478,
-      "grad_norm": 0.5317326558138038,
-      "learning_rate": 7.669438382167023e-05,
-      "loss": 0.724,
-      "step": 3933
-    },
-    {
-      "epoch": 1.933198007012364,
-      "grad_norm": 0.3952290965426452,
-      "learning_rate": 7.663450812830022e-05,
-      "loss": 0.6268,
-      "step": 3934
-    },
-    {
-      "epoch": 1.93369010272498,
-      "grad_norm": 0.3691924253171679,
-      "learning_rate": 7.657464129411482e-05,
-      "loss": 0.6754,
-      "step": 3935
-    },
-    {
-      "epoch": 1.934182198437596,
-      "grad_norm": 0.38390491352597483,
-      "learning_rate": 7.651478334181294e-05,
-      "loss": 0.6864,
-      "step": 3936
-    },
-    {
-      "epoch": 1.9346742941502122,
-      "grad_norm": 0.38709566127974504,
-      "learning_rate": 7.645493429409012e-05,
-      "loss": 0.7239,
-      "step": 3937
-    },
-    {
-      "epoch": 1.9351663898628284,
-      "grad_norm": 0.3930552781239274,
-      "learning_rate": 7.639509417363851e-05,
-      "loss": 0.6669,
-      "step": 3938
-    },
-    {
-      "epoch": 1.9356584855754444,
-      "grad_norm": 0.419724994671056,
-      "learning_rate": 7.633526300314695e-05,
-      "loss": 0.7072,
-      "step": 3939
-    },
-    {
-      "epoch": 1.9361505812880604,
-      "grad_norm": 0.3919255622772456,
-      "learning_rate": 7.627544080530077e-05,
-      "loss": 0.6986,
-      "step": 3940
-    },
-    {
-      "epoch": 1.9366426770006766,
-      "grad_norm": 0.3618611807972929,
-      "learning_rate": 7.6215627602782e-05,
-      "loss": 0.6982,
-      "step": 3941
-    },
-    {
-      "epoch": 1.9371347727132928,
-      "grad_norm": 0.384737341889195,
-      "learning_rate": 7.615582341826924e-05,
-      "loss": 0.6407,
-      "step": 3942
-    },
-    {
-      "epoch": 1.9376268684259088,
-      "grad_norm": 0.39658939521182923,
-      "learning_rate": 7.60960282744376e-05,
-      "loss": 0.6992,
-      "step": 3943
-    },
-    {
-      "epoch": 1.9381189641385248,
-      "grad_norm": 0.3862929418739502,
-      "learning_rate": 7.603624219395886e-05,
-      "loss": 0.6689,
-      "step": 3944
-    },
-    {
-      "epoch": 1.938611059851141,
-      "grad_norm": 0.358821061476227,
-      "learning_rate": 7.59764651995013e-05,
-      "loss": 0.652,
-      "step": 3945
-    },
-    {
-      "epoch": 1.9391031555637572,
-      "grad_norm": 0.38679309562512887,
-      "learning_rate": 7.591669731372977e-05,
-      "loss": 0.7125,
-      "step": 3946
-    },
-    {
-      "epoch": 1.9395952512763732,
-      "grad_norm": 0.40047442982785675,
-      "learning_rate": 7.585693855930565e-05,
-      "loss": 0.6988,
-      "step": 3947
-    },
-    {
-      "epoch": 1.9400873469889892,
-      "grad_norm": 0.40516081465275683,
-      "learning_rate": 7.579718895888693e-05,
-      "loss": 0.7243,
-      "step": 3948
-    },
-    {
-      "epoch": 1.9405794427016054,
-      "grad_norm": 0.38918640243870417,
-      "learning_rate": 7.573744853512801e-05,
-      "loss": 0.7002,
-      "step": 3949
-    },
-    {
-      "epoch": 1.9410715384142216,
-      "grad_norm": 0.4081989388987209,
-      "learning_rate": 7.56777173106799e-05,
-      "loss": 0.7325,
-      "step": 3950
-    },
-    {
-      "epoch": 1.9415636341268376,
-      "grad_norm": 0.38965796642259787,
-      "learning_rate": 7.561799530819016e-05,
-      "loss": 0.6553,
-      "step": 3951
-    },
-    {
-      "epoch": 1.9420557298394536,
-      "grad_norm": 0.4329902070473435,
-      "learning_rate": 7.555828255030269e-05,
-      "loss": 0.7718,
-      "step": 3952
-    },
-    {
-      "epoch": 1.9425478255520698,
-      "grad_norm": 0.3851929182372548,
-      "learning_rate": 7.549857905965805e-05,
-      "loss": 0.6929,
-      "step": 3953
-    },
-    {
-      "epoch": 1.943039921264686,
-      "grad_norm": 0.41348069716240204,
-      "learning_rate": 7.543888485889325e-05,
-      "loss": 0.7604,
-      "step": 3954
-    },
-    {
-      "epoch": 1.943532016977302,
-      "grad_norm": 0.3659746623034429,
-      "learning_rate": 7.537919997064166e-05,
-      "loss": 0.6784,
-      "step": 3955
-    },
-    {
-      "epoch": 1.944024112689918,
-      "grad_norm": 0.38655369915998505,
-      "learning_rate": 7.53195244175333e-05,
-      "loss": 0.7017,
-      "step": 3956
-    },
-    {
-      "epoch": 1.9445162084025343,
-      "grad_norm": 0.4381754236216023,
-      "learning_rate": 7.525985822219449e-05,
-      "loss": 0.7443,
-      "step": 3957
-    },
-    {
-      "epoch": 1.9450083041151505,
-      "grad_norm": 0.3703580602730962,
-      "learning_rate": 7.520020140724812e-05,
-      "loss": 0.6762,
-      "step": 3958
-    },
-    {
-      "epoch": 1.9455003998277665,
-      "grad_norm": 0.384051493397988,
-      "learning_rate": 7.514055399531345e-05,
-      "loss": 0.6779,
-      "step": 3959
-    },
-    {
-      "epoch": 1.9459924955403824,
-      "grad_norm": 0.42371486958296867,
-      "learning_rate": 7.508091600900622e-05,
-      "loss": 0.7155,
-      "step": 3960
-    },
-    {
-      "epoch": 1.9464845912529987,
-      "grad_norm": 0.4242852107890427,
-      "learning_rate": 7.502128747093855e-05,
-      "loss": 0.7045,
-      "step": 3961
-    },
-    {
-      "epoch": 1.9469766869656149,
-      "grad_norm": 0.3833034508840907,
-      "learning_rate": 7.496166840371905e-05,
-      "loss": 0.716,
-      "step": 3962
-    },
-    {
-      "epoch": 1.9474687826782309,
-      "grad_norm": 0.3931053734634782,
-      "learning_rate": 7.490205882995262e-05,
-      "loss": 0.693,
-      "step": 3963
-    },
-    {
-      "epoch": 1.9479608783908469,
-      "grad_norm": 0.37142856120811557,
-      "learning_rate": 7.48424587722407e-05,
-      "loss": 0.709,
-      "step": 3964
-    },
-    {
-      "epoch": 1.948452974103463,
-      "grad_norm": 0.3836266535930692,
-      "learning_rate": 7.478286825318107e-05,
-      "loss": 0.6681,
-      "step": 3965
-    },
-    {
-      "epoch": 1.9489450698160793,
-      "grad_norm": 0.38418787540701044,
-      "learning_rate": 7.472328729536778e-05,
-      "loss": 0.6837,
-      "step": 3966
-    },
-    {
-      "epoch": 1.9494371655286953,
-      "grad_norm": 0.3958444618673081,
-      "learning_rate": 7.466371592139144e-05,
-      "loss": 0.6669,
-      "step": 3967
-    },
-    {
-      "epoch": 1.9499292612413113,
-      "grad_norm": 0.4139091323368251,
-      "learning_rate": 7.460415415383892e-05,
-      "loss": 0.7679,
-      "step": 3968
-    },
-    {
-      "epoch": 1.9504213569539275,
-      "grad_norm": 0.3916373108439504,
-      "learning_rate": 7.454460201529347e-05,
-      "loss": 0.6784,
-      "step": 3969
-    },
-    {
-      "epoch": 1.9509134526665437,
-      "grad_norm": 0.3888059304432373,
-      "learning_rate": 7.448505952833467e-05,
-      "loss": 0.7274,
-      "step": 3970
-    },
-    {
-      "epoch": 1.9514055483791597,
-      "grad_norm": 0.3962591604311054,
-      "learning_rate": 7.44255267155385e-05,
-      "loss": 0.7075,
-      "step": 3971
-    },
-    {
-      "epoch": 1.9518976440917757,
-      "grad_norm": 0.3986261270565527,
-      "learning_rate": 7.436600359947716e-05,
-      "loss": 0.6868,
-      "step": 3972
-    },
-    {
-      "epoch": 1.952389739804392,
-      "grad_norm": 0.3753923584160657,
-      "learning_rate": 7.430649020271928e-05,
-      "loss": 0.6913,
-      "step": 3973
-    },
-    {
-      "epoch": 1.9528818355170081,
-      "grad_norm": 0.5451048466962539,
-      "learning_rate": 7.424698654782982e-05,
-      "loss": 0.7778,
-      "step": 3974
-    },
-    {
-      "epoch": 1.953373931229624,
-      "grad_norm": 0.38129838315942727,
-      "learning_rate": 7.418749265736988e-05,
-      "loss": 0.6884,
-      "step": 3975
-    },
-    {
-      "epoch": 1.9538660269422403,
-      "grad_norm": 0.3717298725102167,
-      "learning_rate": 7.412800855389706e-05,
-      "loss": 0.634,
-      "step": 3976
-    },
-    {
-      "epoch": 1.9543581226548565,
-      "grad_norm": 0.3753359427054715,
-      "learning_rate": 7.406853425996516e-05,
-      "loss": 0.6536,
-      "step": 3977
-    },
-    {
-      "epoch": 1.9548502183674725,
-      "grad_norm": 0.3792311434661509,
-      "learning_rate": 7.400906979812419e-05,
-      "loss": 0.6589,
-      "step": 3978
-    },
-    {
-      "epoch": 1.9553423140800885,
-      "grad_norm": 0.40045467173202365,
-      "learning_rate": 7.394961519092059e-05,
-      "loss": 0.7486,
-      "step": 3979
-    },
-    {
-      "epoch": 1.9558344097927047,
-      "grad_norm": 0.40348379740070833,
-      "learning_rate": 7.389017046089693e-05,
-      "loss": 0.7228,
-      "step": 3980
-    },
-    {
-      "epoch": 1.956326505505321,
-      "grad_norm": 0.36871799311639814,
-      "learning_rate": 7.38307356305921e-05,
-      "loss": 0.639,
-      "step": 3981
-    },
-    {
-      "epoch": 1.956818601217937,
-      "grad_norm": 0.3712312310517344,
-      "learning_rate": 7.37713107225412e-05,
-      "loss": 0.6222,
-      "step": 3982
-    },
-    {
-      "epoch": 1.957310696930553,
-      "grad_norm": 0.39149582583187675,
-      "learning_rate": 7.371189575927559e-05,
-      "loss": 0.7135,
-      "step": 3983
-    },
-    {
-      "epoch": 1.9578027926431691,
-      "grad_norm": 0.378984951027961,
-      "learning_rate": 7.365249076332286e-05,
-      "loss": 0.6857,
-      "step": 3984
-    },
-    {
-      "epoch": 1.9582948883557854,
-      "grad_norm": 0.3939353646696863,
-      "learning_rate": 7.359309575720684e-05,
-      "loss": 0.7188,
-      "step": 3985
-    },
-    {
-      "epoch": 1.9587869840684013,
-      "grad_norm": 0.37732842848980086,
-      "learning_rate": 7.35337107634475e-05,
-      "loss": 0.6829,
-      "step": 3986
-    },
-    {
-      "epoch": 1.9592790797810173,
-      "grad_norm": 0.4133237097160675,
-      "learning_rate": 7.347433580456109e-05,
-      "loss": 0.7565,
-      "step": 3987
-    },
-    {
-      "epoch": 1.9597711754936336,
-      "grad_norm": 0.41302993387558545,
-      "learning_rate": 7.341497090306007e-05,
-      "loss": 0.7545,
-      "step": 3988
-    },
-    {
-      "epoch": 1.9602632712062498,
-      "grad_norm": 0.37239730229404544,
-      "learning_rate": 7.335561608145295e-05,
-      "loss": 0.6516,
-      "step": 3989
-    },
-    {
-      "epoch": 1.9607553669188658,
-      "grad_norm": 0.3832871384393314,
-      "learning_rate": 7.329627136224459e-05,
-      "loss": 0.729,
-      "step": 3990
-    },
-    {
-      "epoch": 1.9612474626314818,
-      "grad_norm": 0.3851560377513374,
-      "learning_rate": 7.323693676793593e-05,
-      "loss": 0.7525,
-      "step": 3991
-    },
-    {
-      "epoch": 1.961739558344098,
-      "grad_norm": 0.3827623689533237,
-      "learning_rate": 7.317761232102407e-05,
-      "loss": 0.6951,
-      "step": 3992
-    },
-    {
-      "epoch": 1.9622316540567142,
-      "grad_norm": 0.37162400806550305,
-      "learning_rate": 7.311829804400225e-05,
-      "loss": 0.7051,
-      "step": 3993
-    },
-    {
-      "epoch": 1.9627237497693302,
-      "grad_norm": 0.3600410323263964,
-      "learning_rate": 7.305899395935996e-05,
-      "loss": 0.6315,
-      "step": 3994
-    },
-    {
-      "epoch": 1.9632158454819462,
-      "grad_norm": 0.38586231066607346,
-      "learning_rate": 7.299970008958263e-05,
-      "loss": 0.7077,
-      "step": 3995
-    },
-    {
-      "epoch": 1.9637079411945624,
-      "grad_norm": 0.3790292346339804,
-      "learning_rate": 7.294041645715202e-05,
-      "loss": 0.6864,
-      "step": 3996
-    },
-    {
-      "epoch": 1.9642000369071786,
-      "grad_norm": 0.39286744520815214,
-      "learning_rate": 7.288114308454594e-05,
-      "loss": 0.7155,
-      "step": 3997
-    },
-    {
-      "epoch": 1.9646921326197946,
-      "grad_norm": 0.3888345515951847,
-      "learning_rate": 7.282187999423819e-05,
-      "loss": 0.7233,
-      "step": 3998
-    },
-    {
-      "epoch": 1.9651842283324106,
-      "grad_norm": 0.3587376427302989,
-      "learning_rate": 7.276262720869886e-05,
-      "loss": 0.6629,
-      "step": 3999
-    },
-    {
-      "epoch": 1.9656763240450268,
-      "grad_norm": 0.3660232540426246,
-      "learning_rate": 7.270338475039403e-05,
-      "loss": 0.6734,
-      "step": 4000
-    },
-    {
-      "epoch": 1.966168419757643,
-      "grad_norm": 0.3699736422164832,
-      "learning_rate": 7.264415264178584e-05,
-      "loss": 0.7058,
-      "step": 4001
-    },
-    {
-      "epoch": 1.966660515470259,
-      "grad_norm": 0.3883997786079121,
-      "learning_rate": 7.258493090533258e-05,
-      "loss": 0.7026,
-      "step": 4002
-    },
-    {
-      "epoch": 1.967152611182875,
-      "grad_norm": 0.36329654656760146,
-      "learning_rate": 7.252571956348857e-05,
-      "loss": 0.6876,
-      "step": 4003
-    },
-    {
-      "epoch": 1.9676447068954912,
-      "grad_norm": 0.3934790306062663,
-      "learning_rate": 7.246651863870419e-05,
-      "loss": 0.7143,
-      "step": 4004
-    },
-    {
-      "epoch": 1.9681368026081074,
-      "grad_norm": 0.3667473853117635,
-      "learning_rate": 7.240732815342586e-05,
-      "loss": 0.698,
-      "step": 4005
-    },
-    {
-      "epoch": 1.9686288983207234,
-      "grad_norm": 0.4013537302536723,
-      "learning_rate": 7.234814813009607e-05,
-      "loss": 0.6995,
-      "step": 4006
-    },
-    {
-      "epoch": 1.9691209940333394,
-      "grad_norm": 0.41091932113589824,
-      "learning_rate": 7.228897859115328e-05,
-      "loss": 0.6696,
-      "step": 4007
-    },
-    {
-      "epoch": 1.9696130897459556,
-      "grad_norm": 0.3665492709345682,
-      "learning_rate": 7.222981955903212e-05,
-      "loss": 0.6631,
-      "step": 4008
-    },
-    {
-      "epoch": 1.9701051854585718,
-      "grad_norm": 0.39965118014858425,
-      "learning_rate": 7.217067105616303e-05,
-      "loss": 0.6935,
-      "step": 4009
-    },
-    {
-      "epoch": 1.9705972811711878,
-      "grad_norm": 0.40863817469155167,
-      "learning_rate": 7.21115331049726e-05,
-      "loss": 0.7634,
-      "step": 4010
-    },
-    {
-      "epoch": 1.9710893768838038,
-      "grad_norm": 0.3957593416304825,
-      "learning_rate": 7.205240572788347e-05,
-      "loss": 0.6962,
-      "step": 4011
-    },
-    {
-      "epoch": 1.97158147259642,
-      "grad_norm": 0.4081654308335122,
-      "learning_rate": 7.199328894731405e-05,
-      "loss": 0.6983,
-      "step": 4012
-    },
-    {
-      "epoch": 1.9720735683090362,
-      "grad_norm": 0.39040802513376455,
-      "learning_rate": 7.193418278567896e-05,
-      "loss": 0.7076,
-      "step": 4013
-    },
-    {
-      "epoch": 1.9725656640216522,
-      "grad_norm": 0.4008045733607616,
-      "learning_rate": 7.187508726538868e-05,
-      "loss": 0.7273,
-      "step": 4014
-    },
-    {
-      "epoch": 1.9730577597342682,
-      "grad_norm": 0.38090762300162073,
-      "learning_rate": 7.181600240884964e-05,
-      "loss": 0.6886,
-      "step": 4015
-    },
-    {
-      "epoch": 1.9735498554468844,
-      "grad_norm": 0.37013891486174255,
-      "learning_rate": 7.17569282384643e-05,
-      "loss": 0.6878,
-      "step": 4016
-    },
-    {
-      "epoch": 1.9740419511595007,
-      "grad_norm": 0.38334532397209775,
-      "learning_rate": 7.169786477663107e-05,
-      "loss": 0.6934,
-      "step": 4017
-    },
-    {
-      "epoch": 1.9745340468721166,
-      "grad_norm": 0.3789189230561965,
-      "learning_rate": 7.163881204574416e-05,
-      "loss": 0.6783,
-      "step": 4018
-    },
-    {
-      "epoch": 1.9750261425847326,
-      "grad_norm": 0.36729184015092375,
-      "learning_rate": 7.157977006819389e-05,
-      "loss": 0.7182,
-      "step": 4019
-    },
-    {
-      "epoch": 1.9755182382973488,
-      "grad_norm": 0.3919399646984472,
-      "learning_rate": 7.152073886636644e-05,
-      "loss": 0.6721,
-      "step": 4020
-    },
-    {
-      "epoch": 1.976010334009965,
-      "grad_norm": 0.3762141305174364,
-      "learning_rate": 7.146171846264383e-05,
-      "loss": 0.706,
-      "step": 4021
-    },
-    {
-      "epoch": 1.976502429722581,
-      "grad_norm": 0.3764993270301451,
-      "learning_rate": 7.140270887940406e-05,
-      "loss": 0.703,
-      "step": 4022
-    },
-    {
-      "epoch": 1.976994525435197,
-      "grad_norm": 0.367886036134554,
-      "learning_rate": 7.134371013902106e-05,
-      "loss": 0.6753,
-      "step": 4023
-    },
-    {
-      "epoch": 1.9774866211478133,
-      "grad_norm": 0.3848459021046618,
-      "learning_rate": 7.128472226386455e-05,
-      "loss": 0.6684,
-      "step": 4024
-    },
-    {
-      "epoch": 1.9779787168604295,
-      "grad_norm": 0.39550051877055614,
-      "learning_rate": 7.122574527630021e-05,
-      "loss": 0.6941,
-      "step": 4025
-    },
-    {
-      "epoch": 1.9784708125730455,
-      "grad_norm": 0.38016213285326045,
-      "learning_rate": 7.116677919868954e-05,
-      "loss": 0.6346,
-      "step": 4026
-    },
-    {
-      "epoch": 1.9789629082856615,
-      "grad_norm": 0.3894246527271591,
-      "learning_rate": 7.110782405338999e-05,
-      "loss": 0.6878,
-      "step": 4027
-    },
-    {
-      "epoch": 1.9794550039982777,
-      "grad_norm": 0.4133635692174051,
-      "learning_rate": 7.10488798627547e-05,
-      "loss": 0.7232,
-      "step": 4028
-    },
-    {
-      "epoch": 1.9799470997108939,
-      "grad_norm": 0.40156391540366365,
-      "learning_rate": 7.098994664913287e-05,
-      "loss": 0.696,
-      "step": 4029
-    },
-    {
-      "epoch": 1.9804391954235099,
-      "grad_norm": 0.4015682966013915,
-      "learning_rate": 7.093102443486932e-05,
-      "loss": 0.6298,
-      "step": 4030
-    },
-    {
-      "epoch": 1.9809312911361259,
-      "grad_norm": 0.42187283731355024,
-      "learning_rate": 7.087211324230492e-05,
-      "loss": 0.7534,
-      "step": 4031
-    },
-    {
-      "epoch": 1.981423386848742,
-      "grad_norm": 0.40001052591898884,
-      "learning_rate": 7.081321309377615e-05,
-      "loss": 0.7422,
-      "step": 4032
-    },
-    {
-      "epoch": 1.9819154825613583,
-      "grad_norm": 0.4018306730114361,
-      "learning_rate": 7.075432401161541e-05,
-      "loss": 0.6845,
-      "step": 4033
-    },
-    {
-      "epoch": 1.9824075782739743,
-      "grad_norm": 0.4006821583956409,
-      "learning_rate": 7.069544601815099e-05,
-      "loss": 0.6996,
-      "step": 4034
-    },
-    {
-      "epoch": 1.9828996739865903,
-      "grad_norm": 0.4330809786874021,
-      "learning_rate": 7.063657913570678e-05,
-      "loss": 0.7038,
-      "step": 4035
-    },
-    {
-      "epoch": 1.9833917696992065,
-      "grad_norm": 0.3801958588467038,
-      "learning_rate": 7.05777233866026e-05,
-      "loss": 0.6982,
-      "step": 4036
-    },
-    {
-      "epoch": 1.9838838654118227,
-      "grad_norm": 0.3947886852134686,
-      "learning_rate": 7.0518878793154e-05,
-      "loss": 0.679,
-      "step": 4037
-    },
-    {
-      "epoch": 1.9843759611244387,
-      "grad_norm": 0.38529311731196236,
-      "learning_rate": 7.04600453776723e-05,
-      "loss": 0.6694,
-      "step": 4038
-    },
-    {
-      "epoch": 1.9848680568370547,
-      "grad_norm": 0.3706573317319643,
-      "learning_rate": 7.040122316246457e-05,
-      "loss": 0.6912,
-      "step": 4039
-    },
-    {
-      "epoch": 1.985360152549671,
-      "grad_norm": 0.3844736105471381,
-      "learning_rate": 7.034241216983373e-05,
-      "loss": 0.6732,
-      "step": 4040
-    },
-    {
-      "epoch": 1.9858522482622871,
-      "grad_norm": 0.3882127364280242,
-      "learning_rate": 7.028361242207826e-05,
-      "loss": 0.749,
-      "step": 4041
-    },
-    {
-      "epoch": 1.9863443439749031,
-      "grad_norm": 0.3815591229369493,
-      "learning_rate": 7.022482394149252e-05,
-      "loss": 0.6357,
-      "step": 4042
-    },
-    {
-      "epoch": 1.986836439687519,
-      "grad_norm": 0.36098206448794196,
-      "learning_rate": 7.016604675036664e-05,
-      "loss": 0.5993,
-      "step": 4043
-    },
-    {
-      "epoch": 1.9873285354001353,
-      "grad_norm": 0.37414818025183944,
-      "learning_rate": 7.010728087098627e-05,
-      "loss": 0.6918,
-      "step": 4044
-    },
-    {
-      "epoch": 1.9878206311127515,
-      "grad_norm": 0.3925281535031267,
-      "learning_rate": 7.004852632563294e-05,
-      "loss": 0.6839,
-      "step": 4045
-    },
-    {
-      "epoch": 1.9883127268253675,
-      "grad_norm": 0.42591815350432277,
-      "learning_rate": 6.998978313658391e-05,
-      "loss": 0.7108,
-      "step": 4046
-    },
-    {
-      "epoch": 1.9888048225379835,
-      "grad_norm": 0.3888517155200917,
-      "learning_rate": 6.993105132611192e-05,
-      "loss": 0.6691,
-      "step": 4047
-    },
-    {
-      "epoch": 1.9892969182505997,
-      "grad_norm": 0.42597976309535457,
-      "learning_rate": 6.987233091648563e-05,
-      "loss": 0.6791,
-      "step": 4048
-    },
-    {
-      "epoch": 1.989789013963216,
-      "grad_norm": 0.3580610621691536,
-      "learning_rate": 6.981362192996925e-05,
-      "loss": 0.6681,
-      "step": 4049
-    },
-    {
-      "epoch": 1.990281109675832,
-      "grad_norm": 0.3862945615520119,
-      "learning_rate": 6.97549243888227e-05,
-      "loss": 0.6996,
-      "step": 4050
-    },
-    {
-      "epoch": 1.990773205388448,
-      "grad_norm": 0.3946234123343332,
-      "learning_rate": 6.969623831530153e-05,
-      "loss": 0.6948,
-      "step": 4051
-    },
-    {
-      "epoch": 1.9912653011010641,
-      "grad_norm": 0.37424551565435027,
-      "learning_rate": 6.9637563731657e-05,
-      "loss": 0.6788,
-      "step": 4052
-    },
-    {
-      "epoch": 1.9917573968136804,
-      "grad_norm": 0.40721138742140944,
-      "learning_rate": 6.957890066013591e-05,
-      "loss": 0.6949,
-      "step": 4053
-    },
-    {
-      "epoch": 1.9922494925262964,
-      "grad_norm": 0.4124145540393745,
-      "learning_rate": 6.952024912298087e-05,
-      "loss": 0.7289,
-      "step": 4054
-    },
-    {
-      "epoch": 1.9927415882389123,
-      "grad_norm": 0.37934479535897286,
-      "learning_rate": 6.946160914242987e-05,
-      "loss": 0.6843,
-      "step": 4055
-    },
-    {
-      "epoch": 1.9932336839515286,
-      "grad_norm": 0.3968682923956738,
-      "learning_rate": 6.940298074071674e-05,
-      "loss": 0.6528,
-      "step": 4056
-    },
-    {
-      "epoch": 1.9937257796641448,
-      "grad_norm": 0.39753013317271285,
-      "learning_rate": 6.934436394007088e-05,
-      "loss": 0.6873,
-      "step": 4057
-    },
-    {
-      "epoch": 1.9942178753767608,
-      "grad_norm": 0.4000299686554215,
-      "learning_rate": 6.928575876271714e-05,
-      "loss": 0.6579,
-      "step": 4058
-    },
-    {
-      "epoch": 1.9947099710893768,
-      "grad_norm": 0.4166155384433662,
-      "learning_rate": 6.922716523087613e-05,
-      "loss": 0.713,
-      "step": 4059
-    },
-    {
-      "epoch": 1.995202066801993,
-      "grad_norm": 0.3893089265316482,
-      "learning_rate": 6.916858336676399e-05,
-      "loss": 0.7017,
-      "step": 4060
-    },
-    {
-      "epoch": 1.9956941625146092,
-      "grad_norm": 0.3715116087865237,
-      "learning_rate": 6.91100131925924e-05,
-      "loss": 0.6517,
-      "step": 4061
-    },
-    {
-      "epoch": 1.9961862582272252,
-      "grad_norm": 0.40996160393797904,
-      "learning_rate": 6.905145473056866e-05,
-      "loss": 0.6874,
-      "step": 4062
-    },
-    {
-      "epoch": 1.9966783539398412,
-      "grad_norm": 0.4342963666090466,
-      "learning_rate": 6.899290800289562e-05,
-      "loss": 0.6562,
-      "step": 4063
-    },
-    {
-      "epoch": 1.9971704496524574,
-      "grad_norm": 0.3842098815912704,
-      "learning_rate": 6.893437303177162e-05,
-      "loss": 0.6818,
-      "step": 4064
-    },
-    {
-      "epoch": 1.9976625453650736,
-      "grad_norm": 0.38851779653997165,
-      "learning_rate": 6.887584983939063e-05,
-      "loss": 0.6541,
-      "step": 4065
-    },
-    {
-      "epoch": 1.9981546410776896,
-      "grad_norm": 0.4006724039006498,
-      "learning_rate": 6.881733844794213e-05,
-      "loss": 0.7806,
-      "step": 4066
-    },
-    {
-      "epoch": 1.9986467367903056,
-      "grad_norm": 0.36687676440230677,
-      "learning_rate": 6.875883887961105e-05,
-      "loss": 0.6418,
-      "step": 4067
-    },
-    {
-      "epoch": 1.9991388325029218,
-      "grad_norm": 0.3740451711557724,
-      "learning_rate": 6.870035115657795e-05,
-      "loss": 0.697,
-      "step": 4068
-    },
-    {
-      "epoch": 1.9991388325029218,
-      "eval_loss": 0.7580295205116272,
-      "eval_runtime": 6652.6164,
-      "eval_samples_per_second": 4.284,
-      "eval_steps_per_second": 2.142,
-      "step": 4068
-    },
-    {
-      "epoch": 1.999630928215538,
-      "grad_norm": 0.3894925405323842,
-      "learning_rate": 6.864187530101886e-05,
-      "loss": 0.6709,
-      "step": 4069
-    },
-    {
-      "epoch": 2.000123023928154,
-      "grad_norm": 0.3755117315889235,
-      "learning_rate": 6.858341133510524e-05,
-      "loss": 0.6201,
-      "step": 4070
-    },
-    {
-      "epoch": 2.00061511964077,
-      "grad_norm": 0.4014505915180703,
-      "learning_rate": 6.852495928100416e-05,
-      "loss": 0.7471,
-      "step": 4071
-    },
-    {
-      "epoch": 2.001107215353386,
-      "grad_norm": 0.5370893198510371,
-      "learning_rate": 6.846651916087806e-05,
-      "loss": 0.7051,
-      "step": 4072
-    },
-    {
-      "epoch": 2.0015993110660024,
-      "grad_norm": 0.4014359983059753,
-      "learning_rate": 6.840809099688493e-05,
-      "loss": 0.7049,
-      "step": 4073
-    },
-    {
-      "epoch": 2.0020914067786184,
-      "grad_norm": 0.3743263437640882,
-      "learning_rate": 6.834967481117817e-05,
-      "loss": 0.7083,
-      "step": 4074
-    },
-    {
-      "epoch": 2.000122887864823,
-      "grad_norm": 0.3815542448086518,
-      "learning_rate": 6.829127062590676e-05,
-      "loss": 0.6448,
-      "step": 4075
-    },
-    {
-      "epoch": 2.0006144393241168,
-      "grad_norm": 0.6708925998664091,
-      "learning_rate": 6.823287846321495e-05,
-      "loss": 0.4752,
-      "step": 4076
-    },
-    {
-      "epoch": 2.0011059907834103,
-      "grad_norm": 0.5173586392729002,
-      "learning_rate": 6.817449834524256e-05,
-      "loss": 0.429,
-      "step": 4077
-    },
-    {
-      "epoch": 2.0015975422427035,
-      "grad_norm": 0.4707257381318483,
-      "learning_rate": 6.811613029412485e-05,
-      "loss": 0.4924,
-      "step": 4078
-    },
-    {
-      "epoch": 2.002089093701997,
-      "grad_norm": 0.5106438786967059,
-      "learning_rate": 6.805777433199237e-05,
-      "loss": 0.4694,
-      "step": 4079
-    },
-    {
-      "epoch": 2.00258064516129,
-      "grad_norm": 0.6033965393914016,
-      "learning_rate": 6.799943048097129e-05,
-      "loss": 0.4622,
-      "step": 4080
-    },
-    {
-      "epoch": 2.003072196620584,
-      "grad_norm": 0.6244085622434824,
-      "learning_rate": 6.794109876318297e-05,
-      "loss": 0.4839,
-      "step": 4081
-    },
-    {
-      "epoch": 2.003563748079877,
-      "grad_norm": 0.4313257222073283,
-      "learning_rate": 6.788277920074433e-05,
-      "loss": 0.4298,
-      "step": 4082
-    },
-    {
-      "epoch": 2.0040552995391705,
-      "grad_norm": 0.42524213292992213,
-      "learning_rate": 6.782447181576763e-05,
-      "loss": 0.4372,
-      "step": 4083
-    },
-    {
-      "epoch": 2.004546850998464,
-      "grad_norm": 0.474606247022978,
-      "learning_rate": 6.77661766303605e-05,
-      "loss": 0.4498,
-      "step": 4084
-    },
-    {
-      "epoch": 2.005038402457757,
-      "grad_norm": 0.42730997305705487,
-      "learning_rate": 6.770789366662594e-05,
-      "loss": 0.4186,
-      "step": 4085
-    },
-    {
-      "epoch": 2.005529953917051,
-      "grad_norm": 0.4790604818251201,
-      "learning_rate": 6.764962294666237e-05,
-      "loss": 0.4768,
-      "step": 4086
-    },
-    {
-      "epoch": 2.006021505376344,
-      "grad_norm": 0.4358449485807311,
-      "learning_rate": 6.759136449256348e-05,
-      "loss": 0.4195,
-      "step": 4087
-    },
-    {
-      "epoch": 2.0065130568356375,
-      "grad_norm": 0.4194938080241657,
-      "learning_rate": 6.753311832641837e-05,
-      "loss": 0.4529,
-      "step": 4088
-    },
-    {
-      "epoch": 2.0070046082949307,
-      "grad_norm": 0.4863607881419129,
-      "learning_rate": 6.747488447031154e-05,
-      "loss": 0.4839,
-      "step": 4089
-    },
-    {
-      "epoch": 2.0074961597542242,
-      "grad_norm": 0.4160476895074089,
-      "learning_rate": 6.741666294632263e-05,
-      "loss": 0.4004,
-      "step": 4090
-    },
-    {
-      "epoch": 2.007987711213518,
-      "grad_norm": 0.4391113923613935,
-      "learning_rate": 6.735845377652679e-05,
-      "loss": 0.4228,
-      "step": 4091
-    },
-    {
-      "epoch": 2.008479262672811,
-      "grad_norm": 0.4693568766600367,
-      "learning_rate": 6.730025698299446e-05,
-      "loss": 0.4587,
-      "step": 4092
-    },
-    {
-      "epoch": 2.0089708141321045,
-      "grad_norm": 0.4813398186533998,
-      "learning_rate": 6.724207258779128e-05,
-      "loss": 0.4913,
-      "step": 4093
-    },
-    {
-      "epoch": 2.0094623655913977,
-      "grad_norm": 0.42630600115753076,
-      "learning_rate": 6.718390061297829e-05,
-      "loss": 0.4622,
-      "step": 4094
-    },
-    {
-      "epoch": 2.0099539170506913,
-      "grad_norm": 0.39257238097213143,
-      "learning_rate": 6.712574108061179e-05,
-      "loss": 0.4041,
-      "step": 4095
-    },
-    {
-      "epoch": 2.010445468509985,
-      "grad_norm": 0.4074983610993362,
-      "learning_rate": 6.706759401274334e-05,
-      "loss": 0.4451,
-      "step": 4096
-    },
-    {
-      "epoch": 2.010937019969278,
-      "grad_norm": 0.43739407714550654,
-      "learning_rate": 6.700945943141981e-05,
-      "loss": 0.4827,
-      "step": 4097
-    },
-    {
-      "epoch": 2.0114285714285716,
-      "grad_norm": 0.4085917973021729,
-      "learning_rate": 6.695133735868335e-05,
-      "loss": 0.4069,
-      "step": 4098
-    },
-    {
-      "epoch": 2.0119201228878647,
-      "grad_norm": 0.4208115463076437,
-      "learning_rate": 6.689322781657126e-05,
-      "loss": 0.4232,
-      "step": 4099
-    },
-    {
-      "epoch": 2.0124116743471583,
-      "grad_norm": 0.4510441341424042,
-      "learning_rate": 6.683513082711622e-05,
-      "loss": 0.5122,
-      "step": 4100
-    },
-    {
-      "epoch": 2.0129032258064514,
-      "grad_norm": 0.40602687554070094,
-      "learning_rate": 6.677704641234614e-05,
-      "loss": 0.4196,
-      "step": 4101
-    },
-    {
-      "epoch": 2.013394777265745,
-      "grad_norm": 0.4322923711717843,
-      "learning_rate": 6.671897459428403e-05,
-      "loss": 0.4353,
-      "step": 4102
-    },
-    {
-      "epoch": 2.0138863287250386,
-      "grad_norm": 0.4294115463406626,
-      "learning_rate": 6.666091539494828e-05,
-      "loss": 0.4143,
-      "step": 4103
-    },
-    {
-      "epoch": 2.0143778801843317,
-      "grad_norm": 0.4101529983890213,
-      "learning_rate": 6.660286883635236e-05,
-      "loss": 0.4321,
-      "step": 4104
-    },
-    {
-      "epoch": 2.0148694316436253,
-      "grad_norm": 0.4249185882895841,
-      "learning_rate": 6.65448349405051e-05,
-      "loss": 0.4331,
-      "step": 4105
-    },
-    {
-      "epoch": 2.0153609831029184,
-      "grad_norm": 0.4258412576860847,
-      "learning_rate": 6.648681372941038e-05,
-      "loss": 0.4293,
-      "step": 4106
-    },
-    {
-      "epoch": 2.015852534562212,
-      "grad_norm": 0.3912177759782342,
-      "learning_rate": 6.642880522506736e-05,
-      "loss": 0.4505,
-      "step": 4107
-    },
-    {
-      "epoch": 2.016344086021505,
-      "grad_norm": 0.41628918720304675,
-      "learning_rate": 6.637080944947036e-05,
-      "loss": 0.4119,
-      "step": 4108
-    },
-    {
-      "epoch": 2.0168356374807987,
-      "grad_norm": 0.3965045620637507,
-      "learning_rate": 6.631282642460889e-05,
-      "loss": 0.4146,
-      "step": 4109
-    },
-    {
-      "epoch": 2.0173271889400923,
-      "grad_norm": 0.42006382185977537,
-      "learning_rate": 6.625485617246756e-05,
-      "loss": 0.4523,
-      "step": 4110
-    },
-    {
-      "epoch": 2.0178187403993855,
-      "grad_norm": 0.4247212776906086,
-      "learning_rate": 6.619689871502619e-05,
-      "loss": 0.4342,
-      "step": 4111
-    },
-    {
-      "epoch": 2.018310291858679,
-      "grad_norm": 0.4275327236859406,
-      "learning_rate": 6.613895407425982e-05,
-      "loss": 0.4481,
-      "step": 4112
-    },
-    {
-      "epoch": 2.018801843317972,
-      "grad_norm": 0.3919230359262144,
-      "learning_rate": 6.608102227213843e-05,
-      "loss": 0.4319,
-      "step": 4113
-    },
-    {
-      "epoch": 2.0192933947772658,
-      "grad_norm": 0.4050965827800886,
-      "learning_rate": 6.602310333062735e-05,
-      "loss": 0.4438,
-      "step": 4114
-    },
-    {
-      "epoch": 2.0197849462365594,
-      "grad_norm": 0.4114104262486101,
-      "learning_rate": 6.596519727168692e-05,
-      "loss": 0.4382,
-      "step": 4115
-    },
-    {
-      "epoch": 2.0202764976958525,
-      "grad_norm": 0.3786829710539903,
-      "learning_rate": 6.590730411727259e-05,
-      "loss": 0.4309,
-      "step": 4116
-    },
-    {
-      "epoch": 2.020768049155146,
-      "grad_norm": 0.4706133872636162,
-      "learning_rate": 6.584942388933497e-05,
-      "loss": 0.4708,
-      "step": 4117
-    },
-    {
-      "epoch": 2.021259600614439,
-      "grad_norm": 0.3900931651516358,
-      "learning_rate": 6.579155660981973e-05,
-      "loss": 0.3856,
-      "step": 4118
-    },
-    {
-      "epoch": 2.021751152073733,
-      "grad_norm": 0.4139423552177058,
-      "learning_rate": 6.573370230066763e-05,
-      "loss": 0.4211,
-      "step": 4119
-    },
-    {
-      "epoch": 2.022242703533026,
-      "grad_norm": 0.41903910589701315,
-      "learning_rate": 6.567586098381451e-05,
-      "loss": 0.4387,
-      "step": 4120
-    },
-    {
-      "epoch": 2.0227342549923195,
-      "grad_norm": 0.427809401375755,
-      "learning_rate": 6.561803268119137e-05,
-      "loss": 0.4286,
-      "step": 4121
-    },
-    {
-      "epoch": 2.023225806451613,
-      "grad_norm": 0.3994268707226468,
-      "learning_rate": 6.55602174147241e-05,
-      "loss": 0.4321,
-      "step": 4122
-    },
-    {
-      "epoch": 2.0237173579109062,
-      "grad_norm": 0.39016984389351117,
-      "learning_rate": 6.55024152063338e-05,
-      "loss": 0.4261,
-      "step": 4123
-    },
-    {
-      "epoch": 2.0242089093702,
-      "grad_norm": 0.4411645189937886,
-      "learning_rate": 6.544462607793662e-05,
-      "loss": 0.4278,
-      "step": 4124
-    },
-    {
-      "epoch": 2.024700460829493,
-      "grad_norm": 0.4017328097731629,
-      "learning_rate": 6.538685005144361e-05,
-      "loss": 0.4039,
-      "step": 4125
-    },
-    {
-      "epoch": 2.0251920122887865,
-      "grad_norm": 0.42563631870851926,
-      "learning_rate": 6.532908714876098e-05,
-      "loss": 0.428,
-      "step": 4126
-    },
-    {
-      "epoch": 2.0256835637480797,
-      "grad_norm": 0.4075830650325921,
-      "learning_rate": 6.527133739178997e-05,
-      "loss": 0.4409,
-      "step": 4127
-    },
-    {
-      "epoch": 2.0261751152073733,
-      "grad_norm": 0.43540133618542143,
-      "learning_rate": 6.521360080242672e-05,
-      "loss": 0.456,
-      "step": 4128
-    },
-    {
-      "epoch": 2.026666666666667,
-      "grad_norm": 0.439879167469121,
-      "learning_rate": 6.515587740256249e-05,
-      "loss": 0.4744,
-      "step": 4129
-    },
-    {
-      "epoch": 2.02715821812596,
-      "grad_norm": 0.411680677061892,
-      "learning_rate": 6.509816721408349e-05,
-      "loss": 0.4405,
-      "step": 4130
-    },
-    {
-      "epoch": 2.0276497695852536,
-      "grad_norm": 0.394497213937786,
-      "learning_rate": 6.504047025887091e-05,
-      "loss": 0.4281,
-      "step": 4131
-    },
-    {
-      "epoch": 2.0281413210445467,
-      "grad_norm": 0.4120193361553592,
-      "learning_rate": 6.498278655880098e-05,
-      "loss": 0.4166,
-      "step": 4132
-    },
-    {
-      "epoch": 2.0286328725038403,
-      "grad_norm": 0.4346158339098155,
-      "learning_rate": 6.492511613574481e-05,
-      "loss": 0.4342,
-      "step": 4133
-    },
-    {
-      "epoch": 2.029124423963134,
-      "grad_norm": 0.4439111452882697,
-      "learning_rate": 6.486745901156857e-05,
-      "loss": 0.4336,
-      "step": 4134
-    },
-    {
-      "epoch": 2.029615975422427,
-      "grad_norm": 0.4208865777916461,
-      "learning_rate": 6.480981520813339e-05,
-      "loss": 0.4369,
-      "step": 4135
-    },
-    {
-      "epoch": 2.0301075268817206,
-      "grad_norm": 0.4384899856213821,
-      "learning_rate": 6.475218474729521e-05,
-      "loss": 0.4465,
-      "step": 4136
-    },
-    {
-      "epoch": 2.0305990783410137,
-      "grad_norm": 0.4295474685785935,
-      "learning_rate": 6.469456765090507e-05,
-      "loss": 0.4473,
-      "step": 4137
-    },
-    {
-      "epoch": 2.0310906298003073,
-      "grad_norm": 0.4261412429004471,
-      "learning_rate": 6.463696394080892e-05,
-      "loss": 0.4405,
-      "step": 4138
-    },
-    {
-      "epoch": 2.0315821812596004,
-      "grad_norm": 0.4498202985404661,
-      "learning_rate": 6.457937363884752e-05,
-      "loss": 0.4515,
-      "step": 4139
-    },
-    {
-      "epoch": 2.032073732718894,
-      "grad_norm": 0.43715641660298266,
-      "learning_rate": 6.452179676685666e-05,
-      "loss": 0.4435,
-      "step": 4140
-    },
-    {
-      "epoch": 2.0325652841781876,
-      "grad_norm": 0.4286604499497547,
-      "learning_rate": 6.446423334666702e-05,
-      "loss": 0.4887,
-      "step": 4141
-    },
-    {
-      "epoch": 2.0330568356374807,
-      "grad_norm": 0.420967234444815,
-      "learning_rate": 6.440668340010412e-05,
-      "loss": 0.457,
-      "step": 4142
-    },
-    {
-      "epoch": 2.0335483870967743,
-      "grad_norm": 0.415186780716354,
-      "learning_rate": 6.434914694898841e-05,
-      "loss": 0.4173,
-      "step": 4143
-    },
-    {
-      "epoch": 2.0340399385560675,
-      "grad_norm": 0.4137964320774288,
-      "learning_rate": 6.42916240151353e-05,
-      "loss": 0.4587,
-      "step": 4144
-    },
-    {
-      "epoch": 2.034531490015361,
-      "grad_norm": 0.41415684066741026,
-      "learning_rate": 6.42341146203549e-05,
-      "loss": 0.4821,
-      "step": 4145
-    },
-    {
-      "epoch": 2.035023041474654,
-      "grad_norm": 0.39270470579534494,
-      "learning_rate": 6.417661878645234e-05,
-      "loss": 0.4483,
-      "step": 4146
-    },
-    {
-      "epoch": 2.0355145929339478,
-      "grad_norm": 0.40117532015060886,
-      "learning_rate": 6.411913653522757e-05,
-      "loss": 0.4819,
-      "step": 4147
-    },
-    {
-      "epoch": 2.0360061443932413,
-      "grad_norm": 0.42198643835245975,
-      "learning_rate": 6.406166788847531e-05,
-      "loss": 0.4686,
-      "step": 4148
-    },
-    {
-      "epoch": 2.0364976958525345,
-      "grad_norm": 0.42417085484863365,
-      "learning_rate": 6.400421286798526e-05,
-      "loss": 0.4558,
-      "step": 4149
-    },
-    {
-      "epoch": 2.036989247311828,
-      "grad_norm": 0.40938232295129534,
-      "learning_rate": 6.394677149554188e-05,
-      "loss": 0.3957,
-      "step": 4150
-    },
-    {
-      "epoch": 2.037480798771121,
-      "grad_norm": 0.4081222904677301,
-      "learning_rate": 6.38893437929244e-05,
-      "loss": 0.4292,
-      "step": 4151
-    },
-    {
-      "epoch": 2.037972350230415,
-      "grad_norm": 0.4285141261745488,
-      "learning_rate": 6.383192978190695e-05,
-      "loss": 0.4418,
-      "step": 4152
-    },
-    {
-      "epoch": 2.038463901689708,
-      "grad_norm": 0.4222072895102476,
-      "learning_rate": 6.377452948425844e-05,
-      "loss": 0.4238,
-      "step": 4153
-    },
-    {
-      "epoch": 2.0389554531490015,
-      "grad_norm": 0.4457859828251274,
-      "learning_rate": 6.371714292174257e-05,
-      "loss": 0.4727,
-      "step": 4154
-    },
-    {
-      "epoch": 2.039447004608295,
-      "grad_norm": 0.4138136085922726,
-      "learning_rate": 6.36597701161179e-05,
-      "loss": 0.438,
-      "step": 4155
-    },
-    {
-      "epoch": 2.0399385560675882,
-      "grad_norm": 0.4219842851486012,
-      "learning_rate": 6.360241108913763e-05,
-      "loss": 0.4429,
-      "step": 4156
-    },
-    {
-      "epoch": 2.040430107526882,
-      "grad_norm": 0.3948076248943635,
-      "learning_rate": 6.354506586254985e-05,
-      "loss": 0.429,
-      "step": 4157
-    },
-    {
-      "epoch": 2.040921658986175,
-      "grad_norm": 0.40035883095024877,
-      "learning_rate": 6.348773445809747e-05,
-      "loss": 0.4521,
-      "step": 4158
-    },
-    {
-      "epoch": 2.0414132104454685,
-      "grad_norm": 0.4017427790558212,
-      "learning_rate": 6.343041689751798e-05,
-      "loss": 0.4331,
-      "step": 4159
-    },
-    {
-      "epoch": 2.041904761904762,
-      "grad_norm": 0.3954016603152219,
-      "learning_rate": 6.337311320254375e-05,
-      "loss": 0.3992,
-      "step": 4160
-    },
-    {
-      "epoch": 2.0423963133640552,
-      "grad_norm": 0.45375102420223135,
-      "learning_rate": 6.331582339490192e-05,
-      "loss": 0.4469,
-      "step": 4161
-    },
-    {
-      "epoch": 2.042887864823349,
-      "grad_norm": 0.4123656788480944,
-      "learning_rate": 6.325854749631423e-05,
-      "loss": 0.4169,
-      "step": 4162
-    },
-    {
-      "epoch": 2.043379416282642,
-      "grad_norm": 0.3974255647849046,
-      "learning_rate": 6.320128552849728e-05,
-      "loss": 0.4142,
-      "step": 4163
-    },
-    {
-      "epoch": 2.0438709677419356,
-      "grad_norm": 0.42228387487327623,
-      "learning_rate": 6.314403751316231e-05,
-      "loss": 0.4263,
-      "step": 4164
-    },
-    {
-      "epoch": 2.0443625192012287,
-      "grad_norm": 0.43566106954297074,
-      "learning_rate": 6.30868034720153e-05,
-      "loss": 0.4418,
-      "step": 4165
-    },
-    {
-      "epoch": 2.0448540706605223,
-      "grad_norm": 0.4506847981089266,
-      "learning_rate": 6.30295834267569e-05,
-      "loss": 0.4457,
-      "step": 4166
-    },
-    {
-      "epoch": 2.045345622119816,
-      "grad_norm": 0.368505391234191,
-      "learning_rate": 6.297237739908253e-05,
-      "loss": 0.3901,
-      "step": 4167
-    },
-    {
-      "epoch": 2.045837173579109,
-      "grad_norm": 0.39309283243786297,
-      "learning_rate": 6.291518541068217e-05,
-      "loss": 0.4152,
-      "step": 4168
-    },
-    {
-      "epoch": 2.0463287250384026,
-      "grad_norm": 0.412713494958705,
-      "learning_rate": 6.285800748324062e-05,
-      "loss": 0.4032,
-      "step": 4169
-    },
-    {
-      "epoch": 2.0468202764976957,
-      "grad_norm": 0.433805358942553,
-      "learning_rate": 6.280084363843726e-05,
-      "loss": 0.4546,
-      "step": 4170
-    },
-    {
-      "epoch": 2.0473118279569893,
-      "grad_norm": 0.4054641455222668,
-      "learning_rate": 6.274369389794612e-05,
-      "loss": 0.4543,
-      "step": 4171
-    },
-    {
-      "epoch": 2.0478033794162824,
-      "grad_norm": 0.4479040896677571,
-      "learning_rate": 6.268655828343591e-05,
-      "loss": 0.4247,
-      "step": 4172
-    },
-    {
-      "epoch": 2.048294930875576,
-      "grad_norm": 0.43764567536186005,
-      "learning_rate": 6.262943681657007e-05,
-      "loss": 0.4978,
-      "step": 4173
-    },
-    {
-      "epoch": 2.0487864823348696,
-      "grad_norm": 0.4114390085492157,
-      "learning_rate": 6.257232951900649e-05,
-      "loss": 0.4618,
-      "step": 4174
-    },
-    {
-      "epoch": 2.0492780337941627,
-      "grad_norm": 0.44117863226371595,
-      "learning_rate": 6.251523641239782e-05,
-      "loss": 0.4459,
-      "step": 4175
-    },
-    {
-      "epoch": 2.0497695852534563,
-      "grad_norm": 0.4221198902147319,
-      "learning_rate": 6.245815751839133e-05,
-      "loss": 0.4464,
-      "step": 4176
-    },
-    {
-      "epoch": 2.0502611367127495,
-      "grad_norm": 0.4102347342983765,
-      "learning_rate": 6.240109285862881e-05,
-      "loss": 0.4379,
-      "step": 4177
-    },
-    {
-      "epoch": 2.050752688172043,
-      "grad_norm": 0.4564328383403766,
-      "learning_rate": 6.23440424547468e-05,
-      "loss": 0.465,
-      "step": 4178
-    },
-    {
-      "epoch": 2.0512442396313366,
-      "grad_norm": 0.4385686966186336,
-      "learning_rate": 6.228700632837624e-05,
-      "loss": 0.4769,
-      "step": 4179
-    },
-    {
-      "epoch": 2.0517357910906298,
-      "grad_norm": 0.415179595540576,
-      "learning_rate": 6.222998450114283e-05,
-      "loss": 0.442,
-      "step": 4180
-    },
-    {
-      "epoch": 2.0522273425499233,
-      "grad_norm": 0.44994559714934196,
-      "learning_rate": 6.21729769946668e-05,
-      "loss": 0.4903,
-      "step": 4181
-    },
-    {
-      "epoch": 2.0527188940092165,
-      "grad_norm": 0.4293575390240816,
-      "learning_rate": 6.211598383056287e-05,
-      "loss": 0.4171,
-      "step": 4182
-    },
-    {
-      "epoch": 2.05321044546851,
-      "grad_norm": 0.42337786819357853,
-      "learning_rate": 6.20590050304404e-05,
-      "loss": 0.4465,
-      "step": 4183
-    },
-    {
-      "epoch": 2.053701996927803,
-      "grad_norm": 0.4287816136370031,
-      "learning_rate": 6.200204061590336e-05,
-      "loss": 0.4001,
-      "step": 4184
-    },
-    {
-      "epoch": 2.054193548387097,
-      "grad_norm": 0.44936952646255507,
-      "learning_rate": 6.19450906085501e-05,
-      "loss": 0.5181,
-      "step": 4185
-    },
-    {
-      "epoch": 2.0546850998463904,
-      "grad_norm": 0.4716491483591041,
-      "learning_rate": 6.188815502997367e-05,
-      "loss": 0.4952,
-      "step": 4186
-    },
-    {
-      "epoch": 2.0551766513056835,
-      "grad_norm": 0.41813981468267314,
-      "learning_rate": 6.183123390176154e-05,
-      "loss": 0.4456,
-      "step": 4187
-    },
-    {
-      "epoch": 2.055668202764977,
-      "grad_norm": 0.40200169789525003,
-      "learning_rate": 6.177432724549574e-05,
-      "loss": 0.4152,
-      "step": 4188
-    },
-    {
-      "epoch": 2.05615975422427,
-      "grad_norm": 0.4335209784535882,
-      "learning_rate": 6.171743508275283e-05,
-      "loss": 0.4635,
-      "step": 4189
-    },
-    {
-      "epoch": 2.056651305683564,
-      "grad_norm": 0.4479894470101658,
-      "learning_rate": 6.166055743510388e-05,
-      "loss": 0.4534,
-      "step": 4190
-    },
-    {
-      "epoch": 2.057142857142857,
-      "grad_norm": 0.39543833767960374,
-      "learning_rate": 6.160369432411438e-05,
-      "loss": 0.4368,
-      "step": 4191
-    },
-    {
-      "epoch": 2.0576344086021505,
-      "grad_norm": 0.42126283772504136,
-      "learning_rate": 6.15468457713444e-05,
-      "loss": 0.4553,
-      "step": 4192
-    },
-    {
-      "epoch": 2.058125960061444,
-      "grad_norm": 0.4179272616502525,
-      "learning_rate": 6.149001179834848e-05,
-      "loss": 0.4902,
-      "step": 4193
-    },
-    {
-      "epoch": 2.0586175115207372,
-      "grad_norm": 0.44317015319838116,
-      "learning_rate": 6.143319242667554e-05,
-      "loss": 0.4415,
-      "step": 4194
-    },
-    {
-      "epoch": 2.059109062980031,
-      "grad_norm": 0.4238443346146709,
-      "learning_rate": 6.137638767786906e-05,
-      "loss": 0.4234,
-      "step": 4195
-    },
-    {
-      "epoch": 2.059600614439324,
-      "grad_norm": 0.43151834583654336,
-      "learning_rate": 6.131959757346699e-05,
-      "loss": 0.4482,
-      "step": 4196
-    },
-    {
-      "epoch": 2.0600921658986175,
-      "grad_norm": 0.39480103202227185,
-      "learning_rate": 6.126282213500163e-05,
-      "loss": 0.4484,
-      "step": 4197
-    },
-    {
-      "epoch": 2.060583717357911,
-      "grad_norm": 0.42970638592110433,
-      "learning_rate": 6.120606138399977e-05,
-      "loss": 0.4377,
-      "step": 4198
-    },
-    {
-      "epoch": 2.0610752688172043,
-      "grad_norm": 0.47463432127306476,
-      "learning_rate": 6.114931534198268e-05,
-      "loss": 0.47,
-      "step": 4199
-    },
-    {
-      "epoch": 2.061566820276498,
-      "grad_norm": 0.4068639285001661,
-      "learning_rate": 6.109258403046593e-05,
-      "loss": 0.4311,
-      "step": 4200
-    },
-    {
-      "epoch": 2.062058371735791,
-      "grad_norm": 0.42315206102396913,
-      "learning_rate": 6.103586747095965e-05,
-      "loss": 0.505,
-      "step": 4201
-    },
-    {
-      "epoch": 2.0625499231950846,
-      "grad_norm": 0.41286570120147303,
-      "learning_rate": 6.097916568496831e-05,
-      "loss": 0.4386,
-      "step": 4202
-    },
-    {
-      "epoch": 2.0630414746543777,
-      "grad_norm": 0.4362737550070793,
-      "learning_rate": 6.092247869399073e-05,
-      "loss": 0.3975,
-      "step": 4203
-    },
-    {
-      "epoch": 2.0635330261136713,
-      "grad_norm": 0.39515381693685236,
-      "learning_rate": 6.086580651952021e-05,
-      "loss": 0.4324,
-      "step": 4204
-    },
-    {
-      "epoch": 2.064024577572965,
-      "grad_norm": 0.4517275023619289,
-      "learning_rate": 6.080914918304432e-05,
-      "loss": 0.4544,
-      "step": 4205
-    },
-    {
-      "epoch": 2.064516129032258,
-      "grad_norm": 0.42463570176493404,
-      "learning_rate": 6.0752506706045134e-05,
-      "loss": 0.449,
-      "step": 4206
-    },
-    {
-      "epoch": 2.0650076804915516,
-      "grad_norm": 0.4231749522154473,
-      "learning_rate": 6.069587910999905e-05,
-      "loss": 0.4624,
-      "step": 4207
-    },
-    {
-      "epoch": 2.0654992319508447,
-      "grad_norm": 0.40450809361098755,
-      "learning_rate": 6.063926641637674e-05,
-      "loss": 0.4774,
-      "step": 4208
-    },
-    {
-      "epoch": 2.0659907834101383,
-      "grad_norm": 0.42138351602512353,
-      "learning_rate": 6.058266864664335e-05,
-      "loss": 0.4626,
-      "step": 4209
-    },
-    {
-      "epoch": 2.0664823348694314,
-      "grad_norm": 0.4399847833526247,
-      "learning_rate": 6.052608582225827e-05,
-      "loss": 0.454,
-      "step": 4210
-    },
-    {
-      "epoch": 2.066973886328725,
-      "grad_norm": 0.3866471785643516,
-      "learning_rate": 6.0469517964675274e-05,
-      "loss": 0.3906,
-      "step": 4211
-    },
-    {
-      "epoch": 2.0674654377880186,
-      "grad_norm": 0.4091364352386822,
-      "learning_rate": 6.0412965095342425e-05,
-      "loss": 0.4211,
-      "step": 4212
-    },
-    {
-      "epoch": 2.0679569892473117,
-      "grad_norm": 0.4439567956775442,
-      "learning_rate": 6.035642723570218e-05,
-      "loss": 0.4562,
-      "step": 4213
-    },
-    {
-      "epoch": 2.0684485407066053,
-      "grad_norm": 0.42440742150518035,
-      "learning_rate": 6.02999044071912e-05,
-      "loss": 0.4332,
-      "step": 4214
-    },
-    {
-      "epoch": 2.0689400921658985,
-      "grad_norm": 0.44296274032268856,
-      "learning_rate": 6.02433966312405e-05,
-      "loss": 0.4364,
-      "step": 4215
-    },
-    {
-      "epoch": 2.069431643625192,
-      "grad_norm": 0.41404027269175936,
-      "learning_rate": 6.018690392927546e-05,
-      "loss": 0.4512,
-      "step": 4216
-    },
-    {
-      "epoch": 2.0699231950844856,
-      "grad_norm": 0.427461997274158,
-      "learning_rate": 6.013042632271556e-05,
-      "loss": 0.4112,
-      "step": 4217
-    },
-    {
-      "epoch": 2.0704147465437788,
-      "grad_norm": 0.4020576859023756,
-      "learning_rate": 6.0073963832974735e-05,
-      "loss": 0.422,
-      "step": 4218
-    },
-    {
-      "epoch": 2.0709062980030724,
-      "grad_norm": 0.4324412504665786,
-      "learning_rate": 6.001751648146115e-05,
-      "loss": 0.4869,
-      "step": 4219
-    },
-    {
-      "epoch": 2.0713978494623655,
-      "grad_norm": 0.45320776064796725,
-      "learning_rate": 5.996108428957713e-05,
-      "loss": 0.4673,
-      "step": 4220
-    },
-    {
-      "epoch": 2.071889400921659,
-      "grad_norm": 0.4075869212928584,
-      "learning_rate": 5.990466727871933e-05,
-      "loss": 0.3943,
-      "step": 4221
-    },
-    {
-      "epoch": 2.072380952380952,
-      "grad_norm": 0.43830291088572687,
-      "learning_rate": 5.984826547027871e-05,
-      "loss": 0.438,
-      "step": 4222
-    },
-    {
-      "epoch": 2.072872503840246,
-      "grad_norm": 0.4335520478543081,
-      "learning_rate": 5.9791878885640315e-05,
-      "loss": 0.4411,
-      "step": 4223
-    },
-    {
-      "epoch": 2.0733640552995394,
-      "grad_norm": 0.4044037516383188,
-      "learning_rate": 5.973550754618353e-05,
-      "loss": 0.4135,
-      "step": 4224
-    },
-    {
-      "epoch": 2.0738556067588325,
-      "grad_norm": 0.4175805564684879,
-      "learning_rate": 5.9679151473282e-05,
-      "loss": 0.4343,
-      "step": 4225
-    },
-    {
-      "epoch": 2.074347158218126,
-      "grad_norm": 0.4622345146053767,
-      "learning_rate": 5.96228106883034e-05,
-      "loss": 0.4488,
-      "step": 4226
-    },
-    {
-      "epoch": 2.0748387096774192,
-      "grad_norm": 0.4336592304310755,
-      "learning_rate": 5.956648521260979e-05,
-      "loss": 0.4369,
-      "step": 4227
-    },
-    {
-      "epoch": 2.075330261136713,
-      "grad_norm": 0.4375432816128757,
-      "learning_rate": 5.951017506755732e-05,
-      "loss": 0.4305,
-      "step": 4228
-    },
-    {
-      "epoch": 2.075821812596006,
-      "grad_norm": 0.4100065792649927,
-      "learning_rate": 5.945388027449636e-05,
-      "loss": 0.4491,
-      "step": 4229
-    },
-    {
-      "epoch": 2.0763133640552995,
-      "grad_norm": 0.4225142824113446,
-      "learning_rate": 5.939760085477155e-05,
-      "loss": 0.4647,
-      "step": 4230
-    },
-    {
-      "epoch": 2.076804915514593,
-      "grad_norm": 0.4615051490534944,
-      "learning_rate": 5.93413368297215e-05,
-      "loss": 0.4471,
-      "step": 4231
-    },
-    {
-      "epoch": 2.0772964669738863,
-      "grad_norm": 0.4386461472128062,
-      "learning_rate": 5.928508822067914e-05,
-      "loss": 0.4507,
-      "step": 4232
-    },
-    {
-      "epoch": 2.07778801843318,
-      "grad_norm": 0.43290493152027015,
-      "learning_rate": 5.922885504897153e-05,
-      "loss": 0.4214,
-      "step": 4233
-    },
-    {
-      "epoch": 2.078279569892473,
-      "grad_norm": 0.45573929418569736,
-      "learning_rate": 5.9172637335919834e-05,
-      "loss": 0.4898,
-      "step": 4234
-    },
-    {
-      "epoch": 2.0787711213517666,
-      "grad_norm": 0.40594113040743607,
-      "learning_rate": 5.911643510283937e-05,
-      "loss": 0.4013,
-      "step": 4235
-    },
-    {
-      "epoch": 2.0792626728110597,
-      "grad_norm": 0.41906807843682553,
-      "learning_rate": 5.906024837103965e-05,
-      "loss": 0.3846,
-      "step": 4236
-    },
-    {
-      "epoch": 2.0797542242703533,
-      "grad_norm": 0.40930435408629845,
-      "learning_rate": 5.900407716182418e-05,
-      "loss": 0.4135,
-      "step": 4237
-    },
-    {
-      "epoch": 2.080245775729647,
-      "grad_norm": 0.4326377182972461,
-      "learning_rate": 5.894792149649069e-05,
-      "loss": 0.4548,
-      "step": 4238
-    },
-    {
-      "epoch": 2.08073732718894,
-      "grad_norm": 0.4391786320177373,
-      "learning_rate": 5.889178139633101e-05,
-      "loss": 0.4576,
-      "step": 4239
-    },
-    {
-      "epoch": 2.0812288786482336,
-      "grad_norm": 0.44145357856165895,
-      "learning_rate": 5.883565688263099e-05,
-      "loss": 0.46,
-      "step": 4240
-    },
-    {
-      "epoch": 2.0817204301075267,
-      "grad_norm": 0.4263744073324838,
-      "learning_rate": 5.8779547976670625e-05,
-      "loss": 0.462,
-      "step": 4241
-    },
-    {
-      "epoch": 2.0822119815668203,
-      "grad_norm": 0.43403947901842116,
-      "learning_rate": 5.872345469972405e-05,
-      "loss": 0.4728,
-      "step": 4242
-    },
-    {
-      "epoch": 2.082703533026114,
-      "grad_norm": 0.40056615005581947,
-      "learning_rate": 5.866737707305935e-05,
-      "loss": 0.4019,
-      "step": 4243
-    },
-    {
-      "epoch": 2.083195084485407,
-      "grad_norm": 0.41796871662512924,
-      "learning_rate": 5.861131511793871e-05,
-      "loss": 0.4351,
-      "step": 4244
-    },
-    {
-      "epoch": 2.0836866359447006,
-      "grad_norm": 0.4335772808148304,
-      "learning_rate": 5.8555268855618504e-05,
-      "loss": 0.4506,
-      "step": 4245
-    },
-    {
-      "epoch": 2.0841781874039937,
-      "grad_norm": 0.41647793325665344,
-      "learning_rate": 5.849923830734895e-05,
-      "loss": 0.4385,
-      "step": 4246
-    },
-    {
-      "epoch": 2.0846697388632873,
-      "grad_norm": 0.44367157856839123,
-      "learning_rate": 5.844322349437443e-05,
-      "loss": 0.4189,
-      "step": 4247
-    },
-    {
-      "epoch": 2.0851612903225805,
-      "grad_norm": 0.4168511883626321,
-      "learning_rate": 5.8387224437933416e-05,
-      "loss": 0.4384,
-      "step": 4248
-    },
-    {
-      "epoch": 2.085652841781874,
-      "grad_norm": 0.4206723111717998,
-      "learning_rate": 5.8331241159258254e-05,
-      "loss": 0.4734,
-      "step": 4249
-    },
-    {
-      "epoch": 2.0861443932411676,
-      "grad_norm": 0.4906305343165687,
-      "learning_rate": 5.827527367957536e-05,
-      "loss": 0.4691,
-      "step": 4250
-    },
-    {
-      "epoch": 2.0866359447004608,
-      "grad_norm": 0.4187927938525607,
-      "learning_rate": 5.82193220201053e-05,
-      "loss": 0.4539,
-      "step": 4251
-    },
-    {
-      "epoch": 2.0871274961597543,
-      "grad_norm": 0.42085613310255154,
-      "learning_rate": 5.816338620206239e-05,
-      "loss": 0.3937,
-      "step": 4252
-    },
-    {
-      "epoch": 2.0876190476190475,
-      "grad_norm": 0.40741924993899514,
-      "learning_rate": 5.8107466246655194e-05,
-      "loss": 0.4158,
-      "step": 4253
-    },
-    {
-      "epoch": 2.088110599078341,
-      "grad_norm": 0.42475115806634434,
-      "learning_rate": 5.805156217508601e-05,
-      "loss": 0.4309,
-      "step": 4254
-    },
-    {
-      "epoch": 2.088602150537634,
-      "grad_norm": 0.44669438770800324,
-      "learning_rate": 5.799567400855136e-05,
-      "loss": 0.4448,
-      "step": 4255
-    },
-    {
-      "epoch": 2.089093701996928,
-      "grad_norm": 0.42259855388429474,
-      "learning_rate": 5.793980176824158e-05,
-      "loss": 0.437,
-      "step": 4256
-    },
-    {
-      "epoch": 2.0895852534562214,
-      "grad_norm": 0.4128000188437674,
-      "learning_rate": 5.7883945475341005e-05,
-      "loss": 0.3984,
-      "step": 4257
-    },
-    {
-      "epoch": 2.0900768049155145,
-      "grad_norm": 0.45715433167600705,
-      "learning_rate": 5.78281051510279e-05,
-      "loss": 0.4523,
-      "step": 4258
-    },
-    {
-      "epoch": 2.090568356374808,
-      "grad_norm": 0.45573679362133407,
-      "learning_rate": 5.777228081647461e-05,
-      "loss": 0.4736,
-      "step": 4259
-    },
-    {
-      "epoch": 2.0910599078341012,
-      "grad_norm": 0.4365464346634226,
-      "learning_rate": 5.771647249284715e-05,
-      "loss": 0.4547,
-      "step": 4260
-    },
-    {
-      "epoch": 2.091551459293395,
-      "grad_norm": 0.41296022755027106,
-      "learning_rate": 5.766068020130575e-05,
-      "loss": 0.4309,
-      "step": 4261
-    },
-    {
-      "epoch": 2.092043010752688,
-      "grad_norm": 0.44935687469527485,
-      "learning_rate": 5.760490396300443e-05,
-      "loss": 0.4329,
-      "step": 4262
-    },
-    {
-      "epoch": 2.0925345622119815,
-      "grad_norm": 0.3894919375689829,
-      "learning_rate": 5.754914379909102e-05,
-      "loss": 0.4411,
-      "step": 4263
-    },
-    {
-      "epoch": 2.093026113671275,
-      "grad_norm": 0.41443354525864307,
-      "learning_rate": 5.7493399730707464e-05,
-      "loss": 0.4492,
-      "step": 4264
-    },
-    {
-      "epoch": 2.0935176651305683,
-      "grad_norm": 0.41810900486078467,
-      "learning_rate": 5.743767177898948e-05,
-      "loss": 0.4372,
-      "step": 4265
-    },
-    {
-      "epoch": 2.094009216589862,
-      "grad_norm": 0.4189025385705837,
-      "learning_rate": 5.73819599650667e-05,
-      "loss": 0.4746,
-      "step": 4266
-    },
-    {
-      "epoch": 2.094500768049155,
-      "grad_norm": 0.43034018217263537,
-      "learning_rate": 5.7326264310062585e-05,
-      "loss": 0.413,
-      "step": 4267
-    },
-    {
-      "epoch": 2.0949923195084486,
-      "grad_norm": 0.39978423275874964,
-      "learning_rate": 5.727058483509463e-05,
-      "loss": 0.4372,
-      "step": 4268
-    },
-    {
-      "epoch": 2.095483870967742,
-      "grad_norm": 0.41270926386333584,
-      "learning_rate": 5.7214921561273945e-05,
-      "loss": 0.4441,
-      "step": 4269
-    },
-    {
-      "epoch": 2.0959754224270353,
-      "grad_norm": 0.4279187141373669,
-      "learning_rate": 5.715927450970577e-05,
-      "loss": 0.434,
-      "step": 4270
-    },
-    {
-      "epoch": 2.096466973886329,
-      "grad_norm": 0.4362070618606098,
-      "learning_rate": 5.7103643701488995e-05,
-      "loss": 0.4513,
-      "step": 4271
-    },
-    {
-      "epoch": 2.096958525345622,
-      "grad_norm": 0.46100877295110176,
-      "learning_rate": 5.704802915771642e-05,
-      "loss": 0.447,
-      "step": 4272
-    },
-    {
-      "epoch": 2.0974500768049156,
-      "grad_norm": 0.4176842880306042,
-      "learning_rate": 5.6992430899474684e-05,
-      "loss": 0.4289,
-      "step": 4273
-    },
-    {
-      "epoch": 2.0979416282642087,
-      "grad_norm": 0.41357265167043994,
-      "learning_rate": 5.6936848947844245e-05,
-      "loss": 0.4098,
-      "step": 4274
-    },
-    {
-      "epoch": 2.0984331797235023,
-      "grad_norm": 0.403733583839532,
-      "learning_rate": 5.688128332389937e-05,
-      "loss": 0.4539,
-      "step": 4275
-    },
-    {
-      "epoch": 2.098924731182796,
-      "grad_norm": 0.4051883313467419,
-      "learning_rate": 5.6825734048708155e-05,
-      "loss": 0.435,
-      "step": 4276
-    },
-    {
-      "epoch": 2.099416282642089,
-      "grad_norm": 0.4173185495282762,
-      "learning_rate": 5.6770201143332466e-05,
-      "loss": 0.4432,
-      "step": 4277
-    },
-    {
-      "epoch": 2.0999078341013826,
-      "grad_norm": 0.38956088264333477,
-      "learning_rate": 5.671468462882796e-05,
-      "loss": 0.4007,
-      "step": 4278
-    },
-    {
-      "epoch": 2.1003993855606757,
-      "grad_norm": 0.40783052142404524,
-      "learning_rate": 5.66591845262442e-05,
-      "loss": 0.4433,
-      "step": 4279
-    },
-    {
-      "epoch": 2.1008909370199693,
-      "grad_norm": 0.39415449939366926,
-      "learning_rate": 5.6603700856624276e-05,
-      "loss": 0.3917,
-      "step": 4280
-    },
-    {
-      "epoch": 2.1013824884792625,
-      "grad_norm": 0.4366003112326132,
-      "learning_rate": 5.654823364100532e-05,
-      "loss": 0.4607,
-      "step": 4281
-    },
-    {
-      "epoch": 2.101874039938556,
-      "grad_norm": 0.44813445324718815,
-      "learning_rate": 5.649278290041806e-05,
-      "loss": 0.4582,
-      "step": 4282
-    },
-    {
-      "epoch": 2.1023655913978496,
-      "grad_norm": 0.5816753446345644,
-      "learning_rate": 5.6437348655887014e-05,
-      "loss": 0.4849,
-      "step": 4283
-    },
-    {
-      "epoch": 2.1028571428571428,
-      "grad_norm": 0.404515910416168,
-      "learning_rate": 5.6381930928430474e-05,
-      "loss": 0.432,
-      "step": 4284
-    },
-    {
-      "epoch": 2.1033486943164363,
-      "grad_norm": 0.4720776180458684,
-      "learning_rate": 5.632652973906041e-05,
-      "loss": 0.4476,
-      "step": 4285
-    },
-    {
-      "epoch": 2.1038402457757295,
-      "grad_norm": 0.4085662152502483,
-      "learning_rate": 5.627114510878257e-05,
-      "loss": 0.4481,
-      "step": 4286
-    },
-    {
-      "epoch": 2.104331797235023,
-      "grad_norm": 0.41486082305016303,
-      "learning_rate": 5.6215777058596406e-05,
-      "loss": 0.4339,
-      "step": 4287
-    },
-    {
-      "epoch": 2.1048233486943166,
-      "grad_norm": 0.43919763438895254,
-      "learning_rate": 5.616042560949517e-05,
-      "loss": 0.4524,
-      "step": 4288
-    },
-    {
-      "epoch": 2.10531490015361,
-      "grad_norm": 0.4467773296703719,
-      "learning_rate": 5.6105090782465596e-05,
-      "loss": 0.4481,
-      "step": 4289
-    },
-    {
-      "epoch": 2.1058064516129034,
-      "grad_norm": 0.4035063773503261,
-      "learning_rate": 5.6049772598488385e-05,
-      "loss": 0.3952,
-      "step": 4290
-    },
-    {
-      "epoch": 2.1062980030721965,
-      "grad_norm": 0.40984972985964835,
-      "learning_rate": 5.5994471078537736e-05,
-      "loss": 0.3756,
-      "step": 4291
-    },
-    {
-      "epoch": 2.10678955453149,
-      "grad_norm": 0.4430332216262048,
-      "learning_rate": 5.5939186243581607e-05,
-      "loss": 0.4398,
-      "step": 4292
-    },
-    {
-      "epoch": 2.107281105990783,
-      "grad_norm": 0.7517599704205895,
-      "learning_rate": 5.5883918114581634e-05,
-      "loss": 0.4649,
-      "step": 4293
-    },
-    {
-      "epoch": 2.107772657450077,
-      "grad_norm": 0.42219048420753424,
-      "learning_rate": 5.58286667124931e-05,
-      "loss": 0.4528,
-      "step": 4294
-    },
-    {
-      "epoch": 2.1082642089093704,
-      "grad_norm": 0.41393203240996357,
-      "learning_rate": 5.577343205826492e-05,
-      "loss": 0.4224,
-      "step": 4295
-    },
-    {
-      "epoch": 2.1087557603686635,
-      "grad_norm": 0.43523521388805325,
-      "learning_rate": 5.5718214172839664e-05,
-      "loss": 0.449,
-      "step": 4296
-    },
-    {
-      "epoch": 2.109247311827957,
-      "grad_norm": 0.4588148374271314,
-      "learning_rate": 5.5663013077153705e-05,
-      "loss": 0.4492,
-      "step": 4297
-    },
-    {
-      "epoch": 2.1097388632872502,
-      "grad_norm": 0.4390774467423553,
-      "learning_rate": 5.560782879213673e-05,
-      "loss": 0.4627,
-      "step": 4298
-    },
-    {
-      "epoch": 2.110230414746544,
-      "grad_norm": 0.42403359582377603,
-      "learning_rate": 5.5552661338712356e-05,
-      "loss": 0.4335,
-      "step": 4299
-    },
-    {
-      "epoch": 2.110721966205837,
-      "grad_norm": 0.3893738533159988,
-      "learning_rate": 5.549751073779768e-05,
-      "loss": 0.3949,
-      "step": 4300
-    },
-    {
-      "epoch": 2.1112135176651305,
-      "grad_norm": 0.40235654616974886,
-      "learning_rate": 5.544237701030339e-05,
-      "loss": 0.4208,
-      "step": 4301
-    },
-    {
-      "epoch": 2.111705069124424,
-      "grad_norm": 0.40588728212682634,
-      "learning_rate": 5.538726017713385e-05,
-      "loss": 0.414,
-      "step": 4302
-    },
-    {
-      "epoch": 2.1121966205837173,
-      "grad_norm": 0.4347137345119711,
-      "learning_rate": 5.533216025918695e-05,
-      "loss": 0.4534,
-      "step": 4303
-    },
-    {
-      "epoch": 2.112688172043011,
-      "grad_norm": 0.4064793647100823,
-      "learning_rate": 5.527707727735416e-05,
-      "loss": 0.4545,
-      "step": 4304
-    },
-    {
-      "epoch": 2.113179723502304,
-      "grad_norm": 0.407632446198486,
-      "learning_rate": 5.522201125252071e-05,
-      "loss": 0.4317,
-      "step": 4305
-    },
-    {
-      "epoch": 2.1136712749615976,
-      "grad_norm": 0.37832562786643037,
-      "learning_rate": 5.5166962205565076e-05,
-      "loss": 0.4069,
-      "step": 4306
-    },
-    {
-      "epoch": 2.114162826420891,
-      "grad_norm": 0.4134656782296033,
-      "learning_rate": 5.511193015735962e-05,
-      "loss": 0.4334,
-      "step": 4307
-    },
-    {
-      "epoch": 2.1146543778801843,
-      "grad_norm": 0.4373465449969304,
-      "learning_rate": 5.505691512877007e-05,
-      "loss": 0.4182,
-      "step": 4308
-    },
-    {
-      "epoch": 2.115145929339478,
-      "grad_norm": 0.44125981660801755,
-      "learning_rate": 5.500191714065568e-05,
-      "loss": 0.4019,
-      "step": 4309
-    },
-    {
-      "epoch": 2.115637480798771,
-      "grad_norm": 0.39669666789656294,
-      "learning_rate": 5.4946936213869394e-05,
-      "loss": 0.4004,
-      "step": 4310
-    },
-    {
-      "epoch": 2.1161290322580646,
-      "grad_norm": 0.40699107590458056,
-      "learning_rate": 5.489197236925758e-05,
-      "loss": 0.4375,
-      "step": 4311
-    },
-    {
-      "epoch": 2.1166205837173577,
-      "grad_norm": 0.4426464646697577,
-      "learning_rate": 5.4837025627660154e-05,
-      "loss": 0.4613,
-      "step": 4312
-    },
-    {
-      "epoch": 2.1171121351766513,
-      "grad_norm": 0.4309200739702426,
-      "learning_rate": 5.478209600991049e-05,
-      "loss": 0.4504,
-      "step": 4313
-    },
-    {
-      "epoch": 2.117603686635945,
-      "grad_norm": 0.40561472063691545,
-      "learning_rate": 5.4727183536835645e-05,
-      "loss": 0.4159,
-      "step": 4314
-    },
-    {
-      "epoch": 2.118095238095238,
-      "grad_norm": 0.4070801765550614,
-      "learning_rate": 5.467228822925592e-05,
-      "loss": 0.4077,
-      "step": 4315
-    },
-    {
-      "epoch": 2.1185867895545316,
-      "grad_norm": 0.5851478499291302,
-      "learning_rate": 5.4617410107985334e-05,
-      "loss": 0.4621,
-      "step": 4316
-    },
-    {
-      "epoch": 2.1190783410138248,
-      "grad_norm": 0.4037222857784695,
-      "learning_rate": 5.456254919383128e-05,
-      "loss": 0.4741,
-      "step": 4317
-    },
-    {
-      "epoch": 2.1195698924731183,
-      "grad_norm": 0.454975814241343,
-      "learning_rate": 5.450770550759463e-05,
-      "loss": 0.4135,
-      "step": 4318
-    },
-    {
-      "epoch": 2.1200614439324115,
-      "grad_norm": 0.40158758582202636,
-      "learning_rate": 5.445287907006975e-05,
-      "loss": 0.446,
-      "step": 4319
-    },
-    {
-      "epoch": 2.120552995391705,
-      "grad_norm": 0.45012536558190436,
-      "learning_rate": 5.4398069902044456e-05,
-      "loss": 0.4608,
-      "step": 4320
-    },
-    {
-      "epoch": 2.1210445468509986,
-      "grad_norm": 0.43217395612265946,
-      "learning_rate": 5.434327802430002e-05,
-      "loss": 0.4506,
-      "step": 4321
-    },
-    {
-      "epoch": 2.1215360983102918,
-      "grad_norm": 0.45837058375236767,
-      "learning_rate": 5.428850345761107e-05,
-      "loss": 0.4575,
-      "step": 4322
-    },
-    {
-      "epoch": 2.1220276497695854,
-      "grad_norm": 0.43880641287499256,
-      "learning_rate": 5.423374622274594e-05,
-      "loss": 0.4221,
-      "step": 4323
-    },
-    {
-      "epoch": 2.1225192012288785,
-      "grad_norm": 0.42240569393185295,
-      "learning_rate": 5.4179006340466e-05,
-      "loss": 0.4392,
-      "step": 4324
-    },
-    {
-      "epoch": 2.123010752688172,
-      "grad_norm": 0.44762750433104964,
-      "learning_rate": 5.412428383152644e-05,
-      "loss": 0.4265,
-      "step": 4325
-    },
-    {
-      "epoch": 2.1235023041474657,
-      "grad_norm": 0.4234949194967042,
-      "learning_rate": 5.4069578716675486e-05,
-      "loss": 0.4377,
-      "step": 4326
-    },
-    {
-      "epoch": 2.123993855606759,
-      "grad_norm": 0.4049735863161532,
-      "learning_rate": 5.4014891016655065e-05,
-      "loss": 0.4474,
-      "step": 4327
-    },
-    {
-      "epoch": 2.1244854070660524,
-      "grad_norm": 0.41485118889205314,
-      "learning_rate": 5.396022075220037e-05,
-      "loss": 0.4549,
-      "step": 4328
-    },
-    {
-      "epoch": 2.1249769585253455,
-      "grad_norm": 0.42655675316538716,
-      "learning_rate": 5.390556794404e-05,
-      "loss": 0.4247,
-      "step": 4329
-    },
-    {
-      "epoch": 2.125468509984639,
-      "grad_norm": 0.4227362368948209,
-      "learning_rate": 5.385093261289594e-05,
-      "loss": 0.4455,
-      "step": 4330
-    },
-    {
-      "epoch": 2.1259600614439322,
-      "grad_norm": 0.39724971257023695,
-      "learning_rate": 5.379631477948355e-05,
-      "loss": 0.3841,
-      "step": 4331
-    },
-    {
-      "epoch": 2.126451612903226,
-      "grad_norm": 0.41390658051806384,
-      "learning_rate": 5.3741714464511526e-05,
-      "loss": 0.4298,
-      "step": 4332
-    },
-    {
-      "epoch": 2.1269431643625194,
-      "grad_norm": 0.38780469194051687,
-      "learning_rate": 5.3687131688681914e-05,
-      "loss": 0.412,
-      "step": 4333
-    },
-    {
-      "epoch": 2.1274347158218125,
-      "grad_norm": 0.435484590620052,
-      "learning_rate": 5.363256647269028e-05,
-      "loss": 0.4431,
-      "step": 4334
-    },
-    {
-      "epoch": 2.127926267281106,
-      "grad_norm": 0.42119058864037817,
-      "learning_rate": 5.3578018837225244e-05,
-      "loss": 0.4498,
-      "step": 4335
-    },
-    {
-      "epoch": 2.1284178187403993,
-      "grad_norm": 0.39102037612829293,
-      "learning_rate": 5.3523488802969e-05,
-      "loss": 0.3978,
-      "step": 4336
-    },
-    {
-      "epoch": 2.128909370199693,
-      "grad_norm": 0.4001188913866158,
-      "learning_rate": 5.346897639059696e-05,
-      "loss": 0.4239,
-      "step": 4337
-    },
-    {
-      "epoch": 2.129400921658986,
-      "grad_norm": 0.4727651469042729,
-      "learning_rate": 5.3414481620777867e-05,
-      "loss": 0.4532,
-      "step": 4338
-    },
-    {
-      "epoch": 2.1298924731182796,
-      "grad_norm": 0.41595630304601267,
-      "learning_rate": 5.33600045141738e-05,
-      "loss": 0.4467,
-      "step": 4339
-    },
-    {
-      "epoch": 2.130384024577573,
-      "grad_norm": 0.4291265838853872,
-      "learning_rate": 5.33055450914401e-05,
-      "loss": 0.4658,
-      "step": 4340
-    },
-    {
-      "epoch": 2.1308755760368663,
-      "grad_norm": 0.40854086583064586,
-      "learning_rate": 5.325110337322543e-05,
-      "loss": 0.3894,
-      "step": 4341
-    },
-    {
-      "epoch": 2.13136712749616,
-      "grad_norm": 0.4438761789467226,
-      "learning_rate": 5.31966793801717e-05,
-      "loss": 0.4918,
-      "step": 4342
-    },
-    {
-      "epoch": 2.131858678955453,
-      "grad_norm": 0.41391216359375127,
-      "learning_rate": 5.314227313291427e-05,
-      "loss": 0.4445,
-      "step": 4343
-    },
-    {
-      "epoch": 2.1323502304147466,
-      "grad_norm": 0.45122684773947314,
-      "learning_rate": 5.308788465208146e-05,
-      "loss": 0.4901,
-      "step": 4344
-    },
-    {
-      "epoch": 2.13284178187404,
-      "grad_norm": 0.416500040730317,
-      "learning_rate": 5.3033513958295145e-05,
-      "loss": 0.457,
-      "step": 4345
-    },
-    {
-      "epoch": 2.1333333333333333,
-      "grad_norm": 0.4349749020563684,
-      "learning_rate": 5.297916107217033e-05,
-      "loss": 0.4102,
-      "step": 4346
-    },
-    {
-      "epoch": 2.133824884792627,
-      "grad_norm": 0.4016279125462083,
-      "learning_rate": 5.2924826014315246e-05,
-      "loss": 0.4174,
-      "step": 4347
-    },
-    {
-      "epoch": 2.13431643625192,
-      "grad_norm": 0.4410167232425387,
-      "learning_rate": 5.287050880533138e-05,
-      "loss": 0.467,
-      "step": 4348
-    },
-    {
-      "epoch": 2.1348079877112136,
-      "grad_norm": 0.4795203199587206,
-      "learning_rate": 5.2816209465813595e-05,
-      "loss": 0.395,
-      "step": 4349
-    },
-    {
-      "epoch": 2.1352995391705067,
-      "grad_norm": 0.45030586232055475,
-      "learning_rate": 5.276192801634967e-05,
-      "loss": 0.4265,
-      "step": 4350
-    },
-    {
-      "epoch": 2.1357910906298003,
-      "grad_norm": 0.43040429273516784,
-      "learning_rate": 5.270766447752097e-05,
-      "loss": 0.3789,
-      "step": 4351
-    },
-    {
-      "epoch": 2.136282642089094,
-      "grad_norm": 0.4405157905881135,
-      "learning_rate": 5.2653418869901714e-05,
-      "loss": 0.4402,
-      "step": 4352
-    },
-    {
-      "epoch": 2.136774193548387,
-      "grad_norm": 0.41373487724637437,
-      "learning_rate": 5.259919121405961e-05,
-      "loss": 0.4064,
-      "step": 4353
-    },
-    {
-      "epoch": 2.1372657450076806,
-      "grad_norm": 0.4649562006226589,
-      "learning_rate": 5.2544981530555425e-05,
-      "loss": 0.4808,
-      "step": 4354
-    },
-    {
-      "epoch": 2.1377572964669738,
-      "grad_norm": 0.43818021032391435,
-      "learning_rate": 5.2490789839943036e-05,
-      "loss": 0.4029,
-      "step": 4355
-    },
-    {
-      "epoch": 2.1382488479262673,
-      "grad_norm": 0.4176186544370395,
-      "learning_rate": 5.243661616276969e-05,
-      "loss": 0.4538,
-      "step": 4356
-    },
-    {
-      "epoch": 2.1387403993855605,
-      "grad_norm": 0.4250411574079382,
-      "learning_rate": 5.2382460519575674e-05,
-      "loss": 0.452,
-      "step": 4357
-    },
-    {
-      "epoch": 2.139231950844854,
-      "grad_norm": 0.4255661163081559,
-      "learning_rate": 5.2328322930894466e-05,
-      "loss": 0.4673,
-      "step": 4358
-    },
-    {
-      "epoch": 2.1397235023041477,
-      "grad_norm": 0.4712447049756326,
-      "learning_rate": 5.227420341725267e-05,
-      "loss": 0.4585,
-      "step": 4359
-    },
-    {
-      "epoch": 2.140215053763441,
-      "grad_norm": 0.41687159849900807,
-      "learning_rate": 5.2220101999170156e-05,
-      "loss": 0.42,
-      "step": 4360
-    },
-    {
-      "epoch": 2.1407066052227344,
-      "grad_norm": 0.3876350752448147,
-      "learning_rate": 5.2166018697159714e-05,
-      "loss": 0.4025,
-      "step": 4361
-    },
-    {
-      "epoch": 2.1411981566820275,
-      "grad_norm": 0.4398360394424511,
-      "learning_rate": 5.2111953531727484e-05,
-      "loss": 0.5072,
-      "step": 4362
-    },
-    {
-      "epoch": 2.141689708141321,
-      "grad_norm": 0.4475877916395095,
-      "learning_rate": 5.205790652337264e-05,
-      "loss": 0.4411,
-      "step": 4363
-    },
-    {
-      "epoch": 2.1421812596006147,
-      "grad_norm": 0.42813264116827077,
-      "learning_rate": 5.2003877692587435e-05,
-      "loss": 0.4383,
-      "step": 4364
-    },
-    {
-      "epoch": 2.142672811059908,
-      "grad_norm": 0.4210238187824671,
-      "learning_rate": 5.1949867059857296e-05,
-      "loss": 0.4158,
-      "step": 4365
-    },
-    {
-      "epoch": 2.1431643625192014,
-      "grad_norm": 0.42162936077012486,
-      "learning_rate": 5.189587464566069e-05,
-      "loss": 0.4331,
-      "step": 4366
-    },
-    {
-      "epoch": 2.1436559139784945,
-      "grad_norm": 0.43042589148166177,
-      "learning_rate": 5.184190047046923e-05,
-      "loss": 0.4211,
-      "step": 4367
-    },
-    {
-      "epoch": 2.144147465437788,
-      "grad_norm": 0.45073364278941475,
-      "learning_rate": 5.178794455474754e-05,
-      "loss": 0.4512,
-      "step": 4368
-    },
-    {
-      "epoch": 2.1446390168970813,
-      "grad_norm": 0.4315029400695387,
-      "learning_rate": 5.173400691895349e-05,
-      "loss": 0.457,
-      "step": 4369
-    },
-    {
-      "epoch": 2.145130568356375,
-      "grad_norm": 0.411411302822659,
-      "learning_rate": 5.168008758353775e-05,
-      "loss": 0.416,
-      "step": 4370
-    },
-    {
-      "epoch": 2.145622119815668,
-      "grad_norm": 0.73511778409704,
-      "learning_rate": 5.162618656894431e-05,
-      "loss": 0.4794,
-      "step": 4371
-    },
-    {
-      "epoch": 2.1461136712749616,
-      "grad_norm": 0.4162527447396153,
-      "learning_rate": 5.1572303895610086e-05,
-      "loss": 0.4059,
-      "step": 4372
-    },
-    {
-      "epoch": 2.146605222734255,
-      "grad_norm": 0.44990882816501676,
-      "learning_rate": 5.151843958396503e-05,
-      "loss": 0.4622,
-      "step": 4373
-    },
-    {
-      "epoch": 2.1470967741935483,
-      "grad_norm": 0.4031846943438368,
-      "learning_rate": 5.146459365443217e-05,
-      "loss": 0.3995,
-      "step": 4374
-    },
-    {
-      "epoch": 2.147588325652842,
-      "grad_norm": 0.43375475174688627,
-      "learning_rate": 5.141076612742757e-05,
-      "loss": 0.4593,
-      "step": 4375
-    },
-    {
-      "epoch": 2.148079877112135,
-      "grad_norm": 0.4396761341089915,
-      "learning_rate": 5.1356957023360287e-05,
-      "loss": 0.4808,
-      "step": 4376
-    },
-    {
-      "epoch": 2.1485714285714286,
-      "grad_norm": 0.4409281915702271,
-      "learning_rate": 5.13031663626324e-05,
-      "loss": 0.4978,
-      "step": 4377
-    },
-    {
-      "epoch": 2.149062980030722,
-      "grad_norm": 0.44717845168424386,
-      "learning_rate": 5.124939416563903e-05,
-      "loss": 0.4443,
-      "step": 4378
-    },
-    {
-      "epoch": 2.1495545314900153,
-      "grad_norm": 0.3813846296539454,
-      "learning_rate": 5.119564045276821e-05,
-      "loss": 0.4195,
-      "step": 4379
-    },
-    {
-      "epoch": 2.150046082949309,
-      "grad_norm": 0.3996798873020046,
-      "learning_rate": 5.1141905244401144e-05,
-      "loss": 0.4108,
-      "step": 4380
-    },
-    {
-      "epoch": 2.150537634408602,
-      "grad_norm": 0.4299087809995918,
-      "learning_rate": 5.1088188560911733e-05,
-      "loss": 0.4342,
-      "step": 4381
-    },
-    {
-      "epoch": 2.1510291858678956,
-      "grad_norm": 0.4217456028047463,
-      "learning_rate": 5.103449042266717e-05,
-      "loss": 0.4045,
-      "step": 4382
-    },
-    {
-      "epoch": 2.1515207373271887,
-      "grad_norm": 0.41749867581150746,
-      "learning_rate": 5.0980810850027404e-05,
-      "loss": 0.4201,
-      "step": 4383
-    },
-    {
-      "epoch": 2.1520122887864823,
-      "grad_norm": 0.4286667184185848,
-      "learning_rate": 5.0927149863345416e-05,
-      "loss": 0.4422,
-      "step": 4384
-    },
-    {
-      "epoch": 2.152503840245776,
-      "grad_norm": 0.4219826069627986,
-      "learning_rate": 5.087350748296714e-05,
-      "loss": 0.4471,
-      "step": 4385
-    },
-    {
-      "epoch": 2.152995391705069,
-      "grad_norm": 0.41988649075766477,
-      "learning_rate": 5.081988372923143e-05,
-      "loss": 0.4347,
-      "step": 4386
-    },
-    {
-      "epoch": 2.1534869431643626,
-      "grad_norm": 0.41939660893342745,
-      "learning_rate": 5.076627862247012e-05,
-      "loss": 0.4185,
-      "step": 4387
-    },
-    {
-      "epoch": 2.1539784946236558,
-      "grad_norm": 0.41779965967946603,
-      "learning_rate": 5.071269218300789e-05,
-      "loss": 0.448,
-      "step": 4388
-    },
-    {
-      "epoch": 2.1544700460829493,
-      "grad_norm": 0.44787804981382706,
-      "learning_rate": 5.065912443116252e-05,
-      "loss": 0.4544,
-      "step": 4389
-    },
-    {
-      "epoch": 2.1549615975422425,
-      "grad_norm": 0.42412804495993484,
-      "learning_rate": 5.060557538724444e-05,
-      "loss": 0.4275,
-      "step": 4390
-    },
-    {
-      "epoch": 2.155453149001536,
-      "grad_norm": 0.4738682238040403,
-      "learning_rate": 5.0552045071557245e-05,
-      "loss": 0.4901,
-      "step": 4391
-    },
-    {
-      "epoch": 2.1559447004608296,
-      "grad_norm": 0.39120760867964455,
-      "learning_rate": 5.0498533504397286e-05,
-      "loss": 0.3948,
-      "step": 4392
-    },
-    {
-      "epoch": 2.156436251920123,
-      "grad_norm": 0.42834411621168483,
-      "learning_rate": 5.044504070605381e-05,
-      "loss": 0.4241,
-      "step": 4393
-    },
-    {
-      "epoch": 2.1569278033794164,
-      "grad_norm": 0.4292925707764463,
-      "learning_rate": 5.039156669680898e-05,
-      "loss": 0.4626,
-      "step": 4394
-    },
-    {
-      "epoch": 2.1574193548387095,
-      "grad_norm": 0.42117150287635424,
-      "learning_rate": 5.03381114969379e-05,
-      "loss": 0.4164,
-      "step": 4395
-    },
-    {
-      "epoch": 2.157910906298003,
-      "grad_norm": 0.4556973258772338,
-      "learning_rate": 5.028467512670834e-05,
-      "loss": 0.4401,
-      "step": 4396
-    },
-    {
-      "epoch": 2.1584024577572967,
-      "grad_norm": 0.4005896738237106,
-      "learning_rate": 5.0231257606381174e-05,
-      "loss": 0.4168,
-      "step": 4397
-    },
-    {
-      "epoch": 2.15889400921659,
-      "grad_norm": 0.41955820433982366,
-      "learning_rate": 5.017785895620999e-05,
-      "loss": 0.4531,
-      "step": 4398
-    },
-    {
-      "epoch": 2.1593855606758834,
-      "grad_norm": 0.44005067322569746,
-      "learning_rate": 5.012447919644122e-05,
-      "loss": 0.405,
-      "step": 4399
-    },
-    {
-      "epoch": 2.1598771121351765,
-      "grad_norm": 0.43738368249977705,
-      "learning_rate": 5.007111834731422e-05,
-      "loss": 0.4579,
-      "step": 4400
-    },
-    {
-      "epoch": 2.16036866359447,
-      "grad_norm": 0.4460818048531665,
-      "learning_rate": 5.0017776429061004e-05,
-      "loss": 0.4408,
-      "step": 4401
-    },
-    {
-      "epoch": 2.1608602150537632,
-      "grad_norm": 0.4147449048415123,
-      "learning_rate": 4.9964453461906626e-05,
-      "loss": 0.3757,
-      "step": 4402
-    },
-    {
-      "epoch": 2.161351766513057,
-      "grad_norm": 0.45692420521761984,
-      "learning_rate": 4.991114946606882e-05,
-      "loss": 0.4608,
-      "step": 4403
-    },
-    {
-      "epoch": 2.1618433179723504,
-      "grad_norm": 0.39778311288667356,
-      "learning_rate": 4.985786446175815e-05,
-      "loss": 0.3995,
-      "step": 4404
-    },
-    {
-      "epoch": 2.1623348694316435,
-      "grad_norm": 0.41081249623918414,
-      "learning_rate": 4.980459846917797e-05,
-      "loss": 0.4023,
-      "step": 4405
-    },
-    {
-      "epoch": 2.162826420890937,
-      "grad_norm": 0.4440057390082081,
-      "learning_rate": 4.975135150852452e-05,
-      "loss": 0.446,
-      "step": 4406
-    },
-    {
-      "epoch": 2.1633179723502303,
-      "grad_norm": 0.4238231167375056,
-      "learning_rate": 4.969812359998662e-05,
-      "loss": 0.4423,
-      "step": 4407
-    },
-    {
-      "epoch": 2.163809523809524,
-      "grad_norm": 0.4068182614428961,
-      "learning_rate": 4.964491476374611e-05,
-      "loss": 0.4271,
-      "step": 4408
-    },
-    {
-      "epoch": 2.164301075268817,
-      "grad_norm": 0.4110332195034115,
-      "learning_rate": 4.959172501997742e-05,
-      "loss": 0.444,
-      "step": 4409
-    },
-    {
-      "epoch": 2.1647926267281106,
-      "grad_norm": 0.4261451735273132,
-      "learning_rate": 4.953855438884782e-05,
-      "loss": 0.4235,
-      "step": 4410
-    },
-    {
-      "epoch": 2.165284178187404,
-      "grad_norm": 0.43062714070941965,
-      "learning_rate": 4.9485402890517306e-05,
-      "loss": 0.4402,
-      "step": 4411
-    },
-    {
-      "epoch": 2.1657757296466973,
-      "grad_norm": 0.4043275262104169,
-      "learning_rate": 4.943227054513864e-05,
-      "loss": 0.4418,
-      "step": 4412
-    },
-    {
-      "epoch": 2.166267281105991,
-      "grad_norm": 0.40076859148893473,
-      "learning_rate": 4.9379157372857295e-05,
-      "loss": 0.4246,
-      "step": 4413
-    },
-    {
-      "epoch": 2.166758832565284,
-      "grad_norm": 0.3996514965672806,
-      "learning_rate": 4.932606339381146e-05,
-      "loss": 0.3997,
-      "step": 4414
-    },
-    {
-      "epoch": 2.1672503840245776,
-      "grad_norm": 0.41932968565298395,
-      "learning_rate": 4.92729886281322e-05,
-      "loss": 0.4488,
-      "step": 4415
-    },
-    {
-      "epoch": 2.167741935483871,
-      "grad_norm": 0.41813494970217113,
-      "learning_rate": 4.9219933095943005e-05,
-      "loss": 0.4262,
-      "step": 4416
-    },
-    {
-      "epoch": 2.1682334869431643,
-      "grad_norm": 0.4515181929220674,
-      "learning_rate": 4.916689681736036e-05,
-      "loss": 0.4611,
-      "step": 4417
-    },
-    {
-      "epoch": 2.168725038402458,
-      "grad_norm": 0.43052490479356864,
-      "learning_rate": 4.911387981249329e-05,
-      "loss": 0.4889,
-      "step": 4418
-    },
-    {
-      "epoch": 2.169216589861751,
-      "grad_norm": 0.4144350250575182,
-      "learning_rate": 4.906088210144356e-05,
-      "loss": 0.4445,
-      "step": 4419
-    },
-    {
-      "epoch": 2.1697081413210446,
-      "grad_norm": 0.43388096264822634,
-      "learning_rate": 4.90079037043056e-05,
-      "loss": 0.4693,
-      "step": 4420
-    },
-    {
-      "epoch": 2.1701996927803378,
-      "grad_norm": 0.45783415204779754,
-      "learning_rate": 4.895494464116653e-05,
-      "loss": 0.4563,
-      "step": 4421
-    },
-    {
-      "epoch": 2.1706912442396313,
-      "grad_norm": 0.43757618312677327,
-      "learning_rate": 4.890200493210615e-05,
-      "loss": 0.4557,
-      "step": 4422
-    },
-    {
-      "epoch": 2.171182795698925,
-      "grad_norm": 0.4086225003044922,
-      "learning_rate": 4.8849084597196896e-05,
-      "loss": 0.4077,
-      "step": 4423
-    },
-    {
-      "epoch": 2.171674347158218,
-      "grad_norm": 0.42413792999109734,
-      "learning_rate": 4.879618365650387e-05,
-      "loss": 0.4107,
-      "step": 4424
-    },
-    {
-      "epoch": 2.1721658986175116,
-      "grad_norm": 0.3957085764936461,
-      "learning_rate": 4.87433021300848e-05,
-      "loss": 0.3994,
-      "step": 4425
-    },
-    {
-      "epoch": 2.1726574500768048,
-      "grad_norm": 0.4146731465061143,
-      "learning_rate": 4.869044003799017e-05,
-      "loss": 0.4397,
-      "step": 4426
-    },
-    {
-      "epoch": 2.1731490015360984,
-      "grad_norm": 0.629670447790838,
-      "learning_rate": 4.863759740026286e-05,
-      "loss": 0.5132,
-      "step": 4427
-    },
-    {
-      "epoch": 2.1736405529953915,
-      "grad_norm": 0.41846108849849256,
-      "learning_rate": 4.858477423693862e-05,
-      "loss": 0.4465,
-      "step": 4428
-    },
-    {
-      "epoch": 2.174132104454685,
-      "grad_norm": 0.4119282564612937,
-      "learning_rate": 4.853197056804569e-05,
-      "loss": 0.4068,
-      "step": 4429
-    },
-    {
-      "epoch": 2.1746236559139787,
-      "grad_norm": 0.4153153738135856,
-      "learning_rate": 4.8479186413604924e-05,
-      "loss": 0.4316,
-      "step": 4430
-    },
-    {
-      "epoch": 2.175115207373272,
-      "grad_norm": 0.4541094117283424,
-      "learning_rate": 4.8426421793629795e-05,
-      "loss": 0.3986,
-      "step": 4431
-    },
-    {
-      "epoch": 2.1756067588325654,
-      "grad_norm": 0.4350741526371811,
-      "learning_rate": 4.837367672812636e-05,
-      "loss": 0.44,
-      "step": 4432
-    },
-    {
-      "epoch": 2.1760983102918585,
-      "grad_norm": 0.4367682504678159,
-      "learning_rate": 4.832095123709328e-05,
-      "loss": 0.421,
-      "step": 4433
-    },
-    {
-      "epoch": 2.176589861751152,
-      "grad_norm": 0.42496427203676856,
-      "learning_rate": 4.826824534052174e-05,
-      "loss": 0.4454,
-      "step": 4434
-    },
-    {
-      "epoch": 2.1770814132104457,
-      "grad_norm": 0.4724552533650487,
-      "learning_rate": 4.8215559058395645e-05,
-      "loss": 0.4492,
-      "step": 4435
-    },
-    {
-      "epoch": 2.177572964669739,
-      "grad_norm": 0.4225075059301563,
-      "learning_rate": 4.816289241069122e-05,
-      "loss": 0.4374,
-      "step": 4436
-    },
-    {
-      "epoch": 2.1780645161290324,
-      "grad_norm": 0.46626577619958326,
-      "learning_rate": 4.811024541737747e-05,
-      "loss": 0.4531,
-      "step": 4437
-    },
-    {
-      "epoch": 2.1785560675883255,
-      "grad_norm": 0.4388792550230304,
-      "learning_rate": 4.8057618098415845e-05,
-      "loss": 0.4799,
-      "step": 4438
-    },
-    {
-      "epoch": 2.179047619047619,
-      "grad_norm": 0.46133312440643043,
-      "learning_rate": 4.800501047376034e-05,
-      "loss": 0.439,
-      "step": 4439
-    },
-    {
-      "epoch": 2.1795391705069123,
-      "grad_norm": 0.4578713126408955,
-      "learning_rate": 4.7952422563357444e-05,
-      "loss": 0.4662,
-      "step": 4440
-    },
-    {
-      "epoch": 2.180030721966206,
-      "grad_norm": 0.44059885691155315,
-      "learning_rate": 4.789985438714636e-05,
-      "loss": 0.4439,
-      "step": 4441
-    },
-    {
-      "epoch": 2.1805222734254994,
-      "grad_norm": 0.40775031092238395,
-      "learning_rate": 4.7847305965058497e-05,
-      "loss": 0.4289,
-      "step": 4442
-    },
-    {
-      "epoch": 2.1810138248847926,
-      "grad_norm": 0.41936920870705424,
-      "learning_rate": 4.779477731701806e-05,
-      "loss": 0.4472,
-      "step": 4443
-    },
-    {
-      "epoch": 2.181505376344086,
-      "grad_norm": 0.44219260372325364,
-      "learning_rate": 4.7742268462941606e-05,
-      "loss": 0.476,
-      "step": 4444
-    },
-    {
-      "epoch": 2.1819969278033793,
-      "grad_norm": 0.40176090187932306,
-      "learning_rate": 4.768977942273822e-05,
-      "loss": 0.4392,
-      "step": 4445
-    },
-    {
-      "epoch": 2.182488479262673,
-      "grad_norm": 0.43136033668182955,
-      "learning_rate": 4.763731021630949e-05,
-      "loss": 0.4627,
-      "step": 4446
-    },
-    {
-      "epoch": 2.182980030721966,
-      "grad_norm": 0.45410435241183505,
-      "learning_rate": 4.758486086354946e-05,
-      "loss": 0.4492,
-      "step": 4447
-    },
-    {
-      "epoch": 2.1834715821812596,
-      "grad_norm": 0.3803836135972755,
-      "learning_rate": 4.7532431384344666e-05,
-      "loss": 0.4006,
-      "step": 4448
-    },
-    {
-      "epoch": 2.183963133640553,
-      "grad_norm": 0.4361763004935449,
-      "learning_rate": 4.7480021798574094e-05,
-      "loss": 0.4412,
-      "step": 4449
-    },
-    {
-      "epoch": 2.1844546850998463,
-      "grad_norm": 0.3917833693828462,
-      "learning_rate": 4.7427632126109186e-05,
-      "loss": 0.4373,
-      "step": 4450
-    },
-    {
-      "epoch": 2.18494623655914,
-      "grad_norm": 0.40969285412603657,
-      "learning_rate": 4.7375262386813814e-05,
-      "loss": 0.4344,
-      "step": 4451
-    },
-    {
-      "epoch": 2.185437788018433,
-      "grad_norm": 0.4453853483380478,
-      "learning_rate": 4.7322912600544435e-05,
-      "loss": 0.492,
-      "step": 4452
-    },
-    {
-      "epoch": 2.1859293394777266,
-      "grad_norm": 0.4104727458845761,
-      "learning_rate": 4.727058278714966e-05,
-      "loss": 0.4271,
-      "step": 4453
-    },
-    {
-      "epoch": 2.18642089093702,
-      "grad_norm": 0.4277137440758201,
-      "learning_rate": 4.721827296647083e-05,
-      "loss": 0.4349,
-      "step": 4454
-    },
-    {
-      "epoch": 2.1869124423963133,
-      "grad_norm": 0.44356378878174674,
-      "learning_rate": 4.716598315834151e-05,
-      "loss": 0.4357,
-      "step": 4455
-    },
-    {
-      "epoch": 2.187403993855607,
-      "grad_norm": 0.4308645590911913,
-      "learning_rate": 4.7113713382587745e-05,
-      "loss": 0.4405,
-      "step": 4456
-    },
-    {
-      "epoch": 2.1878955453149,
-      "grad_norm": 0.4307110305176749,
-      "learning_rate": 4.706146365902796e-05,
-      "loss": 0.4205,
-      "step": 4457
-    },
-    {
-      "epoch": 2.1883870967741936,
-      "grad_norm": 0.44090083753213666,
-      "learning_rate": 4.7009234007473016e-05,
-      "loss": 0.4509,
-      "step": 4458
-    },
-    {
-      "epoch": 2.1888786482334868,
-      "grad_norm": 0.42396812896536706,
-      "learning_rate": 4.695702444772611e-05,
-      "loss": 0.4522,
-      "step": 4459
-    },
-    {
-      "epoch": 2.1893701996927803,
-      "grad_norm": 0.4390162544767842,
-      "learning_rate": 4.6904834999582834e-05,
-      "loss": 0.4613,
-      "step": 4460
-    },
-    {
-      "epoch": 2.189861751152074,
-      "grad_norm": 0.40877584824911817,
-      "learning_rate": 4.6852665682831284e-05,
-      "loss": 0.4188,
-      "step": 4461
-    },
-    {
-      "epoch": 2.190353302611367,
-      "grad_norm": 0.4261914855380424,
-      "learning_rate": 4.6800516517251644e-05,
-      "loss": 0.4718,
-      "step": 4462
-    },
-    {
-      "epoch": 2.1908448540706607,
-      "grad_norm": 0.43261030424633407,
-      "learning_rate": 4.674838752261675e-05,
-      "loss": 0.4257,
-      "step": 4463
-    },
-    {
-      "epoch": 2.191336405529954,
-      "grad_norm": 0.4082558536404945,
-      "learning_rate": 4.6696278718691635e-05,
-      "loss": 0.4445,
-      "step": 4464
-    },
-    {
-      "epoch": 2.1918279569892474,
-      "grad_norm": 0.41713268918343116,
-      "learning_rate": 4.6644190125233675e-05,
-      "loss": 0.4367,
-      "step": 4465
-    },
-    {
-      "epoch": 2.1923195084485405,
-      "grad_norm": 0.4205254545193984,
-      "learning_rate": 4.659212176199264e-05,
-      "loss": 0.4579,
-      "step": 4466
-    },
-    {
-      "epoch": 2.192811059907834,
-      "grad_norm": 0.4294159493533712,
-      "learning_rate": 4.65400736487106e-05,
-      "loss": 0.4867,
-      "step": 4467
-    },
-    {
-      "epoch": 2.1933026113671277,
-      "grad_norm": 0.4096270514696592,
-      "learning_rate": 4.6488045805121936e-05,
-      "loss": 0.4472,
-      "step": 4468
-    },
-    {
-      "epoch": 2.193794162826421,
-      "grad_norm": 0.4222234775429786,
-      "learning_rate": 4.643603825095333e-05,
-      "loss": 0.4012,
-      "step": 4469
-    },
-    {
-      "epoch": 2.1942857142857144,
-      "grad_norm": 0.43404196111820115,
-      "learning_rate": 4.63840510059239e-05,
-      "loss": 0.4539,
-      "step": 4470
-    },
-    {
-      "epoch": 2.1947772657450075,
-      "grad_norm": 0.4225395948168961,
-      "learning_rate": 4.6332084089744834e-05,
-      "loss": 0.451,
-      "step": 4471
-    },
-    {
-      "epoch": 2.195268817204301,
-      "grad_norm": 0.47872341738559837,
-      "learning_rate": 4.628013752211987e-05,
-      "loss": 0.4689,
-      "step": 4472
-    },
-    {
-      "epoch": 2.1957603686635947,
-      "grad_norm": 0.40990643697997525,
-      "learning_rate": 4.622821132274475e-05,
-      "loss": 0.4241,
-      "step": 4473
-    },
-    {
-      "epoch": 2.196251920122888,
-      "grad_norm": 0.4166483040482699,
-      "learning_rate": 4.617630551130778e-05,
-      "loss": 0.4133,
-      "step": 4474
-    },
-    {
-      "epoch": 2.1967434715821814,
-      "grad_norm": 0.430535564954758,
-      "learning_rate": 4.612442010748934e-05,
-      "loss": 0.4434,
-      "step": 4475
-    },
-    {
-      "epoch": 2.1972350230414746,
-      "grad_norm": 0.3999902275332258,
-      "learning_rate": 4.607255513096215e-05,
-      "loss": 0.4503,
-      "step": 4476
-    },
-    {
-      "epoch": 2.197726574500768,
-      "grad_norm": 0.41627649583279697,
-      "learning_rate": 4.602071060139115e-05,
-      "loss": 0.4418,
-      "step": 4477
-    },
-    {
-      "epoch": 2.1982181259600613,
-      "grad_norm": 0.39641632122986575,
-      "learning_rate": 4.596888653843354e-05,
-      "loss": 0.403,
-      "step": 4478
-    },
-    {
-      "epoch": 2.198709677419355,
-      "grad_norm": 0.4161578906013969,
-      "learning_rate": 4.59170829617388e-05,
-      "loss": 0.4831,
-      "step": 4479
-    },
-    {
-      "epoch": 2.1992012288786484,
-      "grad_norm": 0.45074209732039683,
-      "learning_rate": 4.586529989094853e-05,
-      "loss": 0.4994,
-      "step": 4480
-    },
-    {
-      "epoch": 2.1996927803379416,
-      "grad_norm": 0.4195719471519759,
-      "learning_rate": 4.581353734569678e-05,
-      "loss": 0.4556,
-      "step": 4481
-    },
-    {
-      "epoch": 2.200184331797235,
-      "grad_norm": 0.4364300571102807,
-      "learning_rate": 4.576179534560948e-05,
-      "loss": 0.408,
-      "step": 4482
-    },
-    {
-      "epoch": 2.2006758832565283,
-      "grad_norm": 0.4102148959058698,
-      "learning_rate": 4.571007391030511e-05,
-      "loss": 0.4293,
-      "step": 4483
-    },
-    {
-      "epoch": 2.201167434715822,
-      "grad_norm": 0.4379611186924836,
-      "learning_rate": 4.565837305939414e-05,
-      "loss": 0.4479,
-      "step": 4484
-    },
-    {
-      "epoch": 2.201658986175115,
-      "grad_norm": 0.44502080054836296,
-      "learning_rate": 4.560669281247931e-05,
-      "loss": 0.4442,
-      "step": 4485
-    },
-    {
-      "epoch": 2.2021505376344086,
-      "grad_norm": 0.44082396196571616,
-      "learning_rate": 4.5555033189155505e-05,
-      "loss": 0.4042,
-      "step": 4486
-    },
-    {
-      "epoch": 2.202642089093702,
-      "grad_norm": 0.5436717499586232,
-      "learning_rate": 4.550339420900992e-05,
-      "loss": 0.4569,
-      "step": 4487
-    },
-    {
-      "epoch": 2.2031336405529953,
-      "grad_norm": 0.43565266629182936,
-      "learning_rate": 4.545177589162167e-05,
-      "loss": 0.4196,
-      "step": 4488
-    },
-    {
-      "epoch": 2.203625192012289,
-      "grad_norm": 0.45680818945809476,
-      "learning_rate": 4.540017825656232e-05,
-      "loss": 0.4289,
-      "step": 4489
-    },
-    {
-      "epoch": 2.204116743471582,
-      "grad_norm": 0.4216188706854949,
-      "learning_rate": 4.5348601323395415e-05,
-      "loss": 0.4583,
-      "step": 4490
-    },
-    {
-      "epoch": 2.2046082949308756,
-      "grad_norm": 0.47074766549661506,
-      "learning_rate": 4.529704511167669e-05,
-      "loss": 0.4386,
-      "step": 4491
-    },
-    {
-      "epoch": 2.205099846390169,
-      "grad_norm": 0.4044953118483657,
-      "learning_rate": 4.5245509640954057e-05,
-      "loss": 0.3919,
-      "step": 4492
-    },
-    {
-      "epoch": 2.2055913978494623,
-      "grad_norm": 0.41423338895930417,
-      "learning_rate": 4.5193994930767526e-05,
-      "loss": 0.4137,
-      "step": 4493
-    },
-    {
-      "epoch": 2.206082949308756,
-      "grad_norm": 0.4021240623311567,
-      "learning_rate": 4.514250100064924e-05,
-      "loss": 0.4459,
-      "step": 4494
-    },
-    {
-      "epoch": 2.206574500768049,
-      "grad_norm": 0.4487918225338265,
-      "learning_rate": 4.509102787012344e-05,
-      "loss": 0.4128,
-      "step": 4495
-    },
-    {
-      "epoch": 2.2070660522273426,
-      "grad_norm": 0.4104860899519224,
-      "learning_rate": 4.5039575558706625e-05,
-      "loss": 0.4045,
-      "step": 4496
-    },
-    {
-      "epoch": 2.207557603686636,
-      "grad_norm": 0.40373145049891,
-      "learning_rate": 4.4988144085907136e-05,
-      "loss": 0.4309,
-      "step": 4497
-    },
-    {
-      "epoch": 2.2080491551459294,
-      "grad_norm": 0.4135194906388253,
-      "learning_rate": 4.493673347122572e-05,
-      "loss": 0.4417,
-      "step": 4498
-    },
-    {
-      "epoch": 2.2085407066052225,
-      "grad_norm": 0.4165902757659008,
-      "learning_rate": 4.488534373415492e-05,
-      "loss": 0.4521,
-      "step": 4499
-    },
-    {
-      "epoch": 2.209032258064516,
-      "grad_norm": 0.38993335048743943,
-      "learning_rate": 4.483397489417959e-05,
-      "loss": 0.4061,
-      "step": 4500
-    },
-    {
-      "epoch": 2.2095238095238097,
-      "grad_norm": 0.4524942637057196,
-      "learning_rate": 4.4782626970776544e-05,
-      "loss": 0.4763,
-      "step": 4501
-    },
-    {
-      "epoch": 2.210015360983103,
-      "grad_norm": 0.4190968947250465,
-      "learning_rate": 4.47312999834147e-05,
-      "loss": 0.4632,
-      "step": 4502
-    },
-    {
-      "epoch": 2.2105069124423964,
-      "grad_norm": 0.4170480670203764,
-      "learning_rate": 4.4679993951555044e-05,
-      "loss": 0.4108,
-      "step": 4503
-    },
-    {
-      "epoch": 2.2109984639016895,
-      "grad_norm": 0.41696081364198906,
-      "learning_rate": 4.462870889465058e-05,
-      "loss": 0.4285,
-      "step": 4504
-    },
-    {
-      "epoch": 2.211490015360983,
-      "grad_norm": 0.4686791565256886,
-      "learning_rate": 4.457744483214641e-05,
-      "loss": 0.4826,
-      "step": 4505
-    },
-    {
-      "epoch": 2.2119815668202767,
-      "grad_norm": 0.42478544683028374,
-      "learning_rate": 4.45262017834796e-05,
-      "loss": 0.4707,
-      "step": 4506
-    },
-    {
-      "epoch": 2.21247311827957,
-      "grad_norm": 0.40480018628254527,
-      "learning_rate": 4.447497976807942e-05,
-      "loss": 0.3889,
-      "step": 4507
-    },
-    {
-      "epoch": 2.2129646697388634,
-      "grad_norm": 0.441810988481185,
-      "learning_rate": 4.442377880536689e-05,
-      "loss": 0.4446,
-      "step": 4508
-    },
-    {
-      "epoch": 2.2134562211981565,
-      "grad_norm": 0.4255315141753949,
-      "learning_rate": 4.43725989147553e-05,
-      "loss": 0.4394,
-      "step": 4509
-    },
-    {
-      "epoch": 2.21394777265745,
-      "grad_norm": 0.3998457670599458,
-      "learning_rate": 4.4321440115649835e-05,
-      "loss": 0.4022,
-      "step": 4510
-    },
-    {
-      "epoch": 2.2144393241167433,
-      "grad_norm": 0.4440745234989787,
-      "learning_rate": 4.42703024274477e-05,
-      "loss": 0.4275,
-      "step": 4511
-    },
-    {
-      "epoch": 2.214930875576037,
-      "grad_norm": 0.44696329121917866,
-      "learning_rate": 4.421918586953808e-05,
-      "loss": 0.4257,
-      "step": 4512
-    },
-    {
-      "epoch": 2.2154224270353304,
-      "grad_norm": 0.8390629282171552,
-      "learning_rate": 4.416809046130218e-05,
-      "loss": 0.4722,
-      "step": 4513
-    },
-    {
-      "epoch": 2.2159139784946236,
-      "grad_norm": 0.4333858632285727,
-      "learning_rate": 4.411701622211316e-05,
-      "loss": 0.4548,
-      "step": 4514
-    },
-    {
-      "epoch": 2.216405529953917,
-      "grad_norm": 0.4540794328456241,
-      "learning_rate": 4.4065963171336144e-05,
-      "loss": 0.4199,
-      "step": 4515
-    },
-    {
-      "epoch": 2.2168970814132103,
-      "grad_norm": 0.4151006477780833,
-      "learning_rate": 4.401493132832832e-05,
-      "loss": 0.41,
-      "step": 4516
-    },
-    {
-      "epoch": 2.217388632872504,
-      "grad_norm": 0.4539538777684501,
-      "learning_rate": 4.396392071243864e-05,
-      "loss": 0.4461,
-      "step": 4517
-    },
-    {
-      "epoch": 2.217880184331797,
-      "grad_norm": 0.4232795307120172,
-      "learning_rate": 4.391293134300824e-05,
-      "loss": 0.4165,
-      "step": 4518
-    },
-    {
-      "epoch": 2.2183717357910906,
-      "grad_norm": 0.39441453701000495,
-      "learning_rate": 4.3861963239370007e-05,
-      "loss": 0.3831,
-      "step": 4519
-    },
-    {
-      "epoch": 2.218863287250384,
-      "grad_norm": 0.4153403752586208,
-      "learning_rate": 4.3811016420848884e-05,
-      "loss": 0.382,
-      "step": 4520
-    },
-    {
-      "epoch": 2.2193548387096773,
-      "grad_norm": 0.4330494209968192,
-      "learning_rate": 4.3760090906761686e-05,
-      "loss": 0.4707,
-      "step": 4521
-    },
-    {
-      "epoch": 2.219846390168971,
-      "grad_norm": 0.39644422590373857,
-      "learning_rate": 4.370918671641716e-05,
-      "loss": 0.4211,
-      "step": 4522
-    },
-    {
-      "epoch": 2.220337941628264,
-      "grad_norm": 0.4515300199036469,
-      "learning_rate": 4.3658303869115994e-05,
-      "loss": 0.474,
-      "step": 4523
-    },
-    {
-      "epoch": 2.2208294930875576,
-      "grad_norm": 0.4153569127675756,
-      "learning_rate": 4.360744238415075e-05,
-      "loss": 0.446,
-      "step": 4524
-    },
-    {
-      "epoch": 2.221321044546851,
-      "grad_norm": 0.40705322607325733,
-      "learning_rate": 4.3556602280805904e-05,
-      "loss": 0.3911,
-      "step": 4525
-    },
-    {
-      "epoch": 2.2218125960061443,
-      "grad_norm": 0.47516826610389534,
-      "learning_rate": 4.350578357835781e-05,
-      "loss": 0.4324,
-      "step": 4526
-    },
-    {
-      "epoch": 2.222304147465438,
-      "grad_norm": 0.4192287978383726,
-      "learning_rate": 4.3454986296074795e-05,
-      "loss": 0.411,
-      "step": 4527
-    },
-    {
-      "epoch": 2.222795698924731,
-      "grad_norm": 0.419728918031921,
-      "learning_rate": 4.340421045321688e-05,
-      "loss": 0.4338,
-      "step": 4528
-    },
-    {
-      "epoch": 2.2232872503840246,
-      "grad_norm": 0.5017966085307272,
-      "learning_rate": 4.335345606903616e-05,
-      "loss": 0.4972,
-      "step": 4529
-    },
-    {
-      "epoch": 2.2237788018433178,
-      "grad_norm": 0.3932223866674703,
-      "learning_rate": 4.330272316277648e-05,
-      "loss": 0.3854,
-      "step": 4530
-    },
-    {
-      "epoch": 2.2242703533026114,
-      "grad_norm": 0.4223532152873605,
-      "learning_rate": 4.325201175367356e-05,
-      "loss": 0.4434,
-      "step": 4531
-    },
-    {
-      "epoch": 2.224761904761905,
-      "grad_norm": 0.45051744969778434,
-      "learning_rate": 4.3201321860954943e-05,
-      "loss": 0.455,
-      "step": 4532
-    },
-    {
-      "epoch": 2.225253456221198,
-      "grad_norm": 0.43922627478970927,
-      "learning_rate": 4.3150653503840145e-05,
-      "loss": 0.4594,
-      "step": 4533
-    },
-    {
-      "epoch": 2.2257450076804917,
-      "grad_norm": 0.40800420038643515,
-      "learning_rate": 4.3100006701540274e-05,
-      "loss": 0.449,
-      "step": 4534
-    },
-    {
-      "epoch": 2.226236559139785,
-      "grad_norm": 0.4156611109687932,
-      "learning_rate": 4.3049381473258534e-05,
-      "loss": 0.4262,
-      "step": 4535
-    },
-    {
-      "epoch": 2.2267281105990784,
-      "grad_norm": 0.44412597623468436,
-      "learning_rate": 4.299877783818975e-05,
-      "loss": 0.4556,
-      "step": 4536
-    },
-    {
-      "epoch": 2.2272196620583715,
-      "grad_norm": 0.43052028167325,
-      "learning_rate": 4.2948195815520675e-05,
-      "loss": 0.4365,
-      "step": 4537
-    },
-    {
-      "epoch": 2.227711213517665,
-      "grad_norm": 0.44273384583205344,
-      "learning_rate": 4.2897635424429795e-05,
-      "loss": 0.4237,
-      "step": 4538
-    },
-    {
-      "epoch": 2.2282027649769587,
-      "grad_norm": 0.42175756925343066,
-      "learning_rate": 4.284709668408744e-05,
-      "loss": 0.4175,
-      "step": 4539
-    },
-    {
-      "epoch": 2.228694316436252,
-      "grad_norm": 0.4291815039613073,
-      "learning_rate": 4.279657961365572e-05,
-      "loss": 0.4359,
-      "step": 4540
-    },
-    {
-      "epoch": 2.2291858678955454,
-      "grad_norm": 0.43043977435781366,
-      "learning_rate": 4.2746084232288476e-05,
-      "loss": 0.4639,
-      "step": 4541
-    },
-    {
-      "epoch": 2.2296774193548385,
-      "grad_norm": 0.4153525016413108,
-      "learning_rate": 4.269561055913148e-05,
-      "loss": 0.4399,
-      "step": 4542
-    },
-    {
-      "epoch": 2.230168970814132,
-      "grad_norm": 0.41924135802962675,
-      "learning_rate": 4.2645158613322044e-05,
-      "loss": 0.4431,
-      "step": 4543
-    },
-    {
-      "epoch": 2.2306605222734257,
-      "grad_norm": 0.4356805792830745,
-      "learning_rate": 4.259472841398945e-05,
-      "loss": 0.4298,
-      "step": 4544
-    },
-    {
-      "epoch": 2.231152073732719,
-      "grad_norm": 0.4191414001168072,
-      "learning_rate": 4.254431998025462e-05,
-      "loss": 0.4408,
-      "step": 4545
-    },
-    {
-      "epoch": 2.2316436251920124,
-      "grad_norm": 0.425152767662572,
-      "learning_rate": 4.249393333123026e-05,
-      "loss": 0.4503,
-      "step": 4546
-    },
-    {
-      "epoch": 2.2321351766513056,
-      "grad_norm": 0.4140228911065242,
-      "learning_rate": 4.244356848602081e-05,
-      "loss": 0.4404,
-      "step": 4547
-    },
-    {
-      "epoch": 2.232626728110599,
-      "grad_norm": 0.42846747020990583,
-      "learning_rate": 4.239322546372244e-05,
-      "loss": 0.4842,
-      "step": 4548
-    },
-    {
-      "epoch": 2.2331182795698923,
-      "grad_norm": 0.41412832765079294,
-      "learning_rate": 4.234290428342305e-05,
-      "loss": 0.4338,
-      "step": 4549
-    },
-    {
-      "epoch": 2.233609831029186,
-      "grad_norm": 0.42718140381970093,
-      "learning_rate": 4.229260496420224e-05,
-      "loss": 0.4394,
-      "step": 4550
-    },
-    {
-      "epoch": 2.2341013824884794,
-      "grad_norm": 0.4269230817888805,
-      "learning_rate": 4.2242327525131365e-05,
-      "loss": 0.4593,
-      "step": 4551
-    },
-    {
-      "epoch": 2.2345929339477726,
-      "grad_norm": 0.42697239288245986,
-      "learning_rate": 4.219207198527339e-05,
-      "loss": 0.4386,
-      "step": 4552
-    },
-    {
-      "epoch": 2.235084485407066,
-      "grad_norm": 0.4297789439398567,
-      "learning_rate": 4.2141838363683185e-05,
-      "loss": 0.4379,
-      "step": 4553
-    },
-    {
-      "epoch": 2.2355760368663593,
-      "grad_norm": 0.3892118154155358,
-      "learning_rate": 4.2091626679407004e-05,
-      "loss": 0.3629,
-      "step": 4554
-    },
-    {
-      "epoch": 2.236067588325653,
-      "grad_norm": 0.4343031562397274,
-      "learning_rate": 4.2041436951483045e-05,
-      "loss": 0.4398,
-      "step": 4555
-    },
-    {
-      "epoch": 2.236559139784946,
-      "grad_norm": 0.43960752875867515,
-      "learning_rate": 4.1991269198941084e-05,
-      "loss": 0.4193,
-      "step": 4556
-    },
-    {
-      "epoch": 2.2370506912442396,
-      "grad_norm": 0.42335764116690994,
-      "learning_rate": 4.194112344080252e-05,
-      "loss": 0.4044,
-      "step": 4557
-    },
-    {
-      "epoch": 2.237542242703533,
-      "grad_norm": 0.4237042651183192,
-      "learning_rate": 4.189099969608049e-05,
-      "loss": 0.4187,
-      "step": 4558
-    },
-    {
-      "epoch": 2.2380337941628263,
-      "grad_norm": 0.45150880876275945,
-      "learning_rate": 4.184089798377975e-05,
-      "loss": 0.4465,
-      "step": 4559
-    },
-    {
-      "epoch": 2.23852534562212,
-      "grad_norm": 0.41438305293174793,
-      "learning_rate": 4.179081832289667e-05,
-      "loss": 0.4002,
-      "step": 4560
-    },
-    {
-      "epoch": 2.239016897081413,
-      "grad_norm": 0.4437423339084064,
-      "learning_rate": 4.17407607324193e-05,
-      "loss": 0.4248,
-      "step": 4561
-    },
-    {
-      "epoch": 2.2395084485407066,
-      "grad_norm": 0.44924414981628963,
-      "learning_rate": 4.16907252313274e-05,
-      "loss": 0.4421,
-      "step": 4562
-    },
-    {
-      "epoch": 2.24,
-      "grad_norm": 0.42788494613359257,
-      "learning_rate": 4.1640711838592114e-05,
-      "loss": 0.3978,
-      "step": 4563
-    },
-    {
-      "epoch": 2.2404915514592934,
-      "grad_norm": 0.4149591545549934,
-      "learning_rate": 4.15907205731765e-05,
-      "loss": 0.4634,
-      "step": 4564
-    },
-    {
-      "epoch": 2.240983102918587,
-      "grad_norm": 0.41417067877207575,
-      "learning_rate": 4.154075145403502e-05,
-      "loss": 0.4568,
-      "step": 4565
-    },
-    {
-      "epoch": 2.24147465437788,
-      "grad_norm": 0.4418227367043439,
-      "learning_rate": 4.149080450011382e-05,
-      "loss": 0.4825,
-      "step": 4566
-    },
-    {
-      "epoch": 2.2419662058371737,
-      "grad_norm": 0.40096107162784733,
-      "learning_rate": 4.144087973035062e-05,
-      "loss": 0.4264,
-      "step": 4567
-    },
-    {
-      "epoch": 2.242457757296467,
-      "grad_norm": 0.40050554535120325,
-      "learning_rate": 4.139097716367474e-05,
-      "loss": 0.452,
-      "step": 4568
-    },
-    {
-      "epoch": 2.2429493087557604,
-      "grad_norm": 0.4265216565740003,
-      "learning_rate": 4.1341096819007065e-05,
-      "loss": 0.4575,
-      "step": 4569
-    },
-    {
-      "epoch": 2.243440860215054,
-      "grad_norm": 0.4301429842592259,
-      "learning_rate": 4.129123871526007e-05,
-      "loss": 0.4515,
-      "step": 4570
-    },
-    {
-      "epoch": 2.243932411674347,
-      "grad_norm": 0.4098690448745217,
-      "learning_rate": 4.1241402871337806e-05,
-      "loss": 0.4304,
-      "step": 4571
-    },
-    {
-      "epoch": 2.2444239631336407,
-      "grad_norm": 0.4331944219648293,
-      "learning_rate": 4.1191589306135824e-05,
-      "loss": 0.42,
-      "step": 4572
-    },
-    {
-      "epoch": 2.244915514592934,
-      "grad_norm": 0.41091299019535316,
-      "learning_rate": 4.114179803854138e-05,
-      "loss": 0.4545,
-      "step": 4573
-    },
-    {
-      "epoch": 2.2454070660522274,
-      "grad_norm": 0.4053814514597819,
-      "learning_rate": 4.109202908743303e-05,
-      "loss": 0.4123,
-      "step": 4574
-    },
-    {
-      "epoch": 2.2458986175115205,
-      "grad_norm": 0.45923192428403736,
-      "learning_rate": 4.1042282471681114e-05,
-      "loss": 0.4697,
-      "step": 4575
-    },
-    {
-      "epoch": 2.246390168970814,
-      "grad_norm": 0.4430307828698487,
-      "learning_rate": 4.0992558210147366e-05,
-      "loss": 0.4332,
-      "step": 4576
-    },
-    {
-      "epoch": 2.2468817204301077,
-      "grad_norm": 0.4425697406910303,
-      "learning_rate": 4.094285632168507e-05,
-      "loss": 0.4686,
-      "step": 4577
-    },
-    {
-      "epoch": 2.247373271889401,
-      "grad_norm": 0.4217068811747446,
-      "learning_rate": 4.089317682513902e-05,
-      "loss": 0.4476,
-      "step": 4578
-    },
-    {
-      "epoch": 2.2478648233486944,
-      "grad_norm": 0.41911384986879874,
-      "learning_rate": 4.084351973934561e-05,
-      "loss": 0.422,
-      "step": 4579
-    },
-    {
-      "epoch": 2.2483563748079876,
-      "grad_norm": 0.4283120421304942,
-      "learning_rate": 4.079388508313255e-05,
-      "loss": 0.4526,
-      "step": 4580
-    },
-    {
-      "epoch": 2.248847926267281,
-      "grad_norm": 0.4345917705047502,
-      "learning_rate": 4.0744272875319246e-05,
-      "loss": 0.4741,
-      "step": 4581
-    },
-    {
-      "epoch": 2.2493394777265747,
-      "grad_norm": 0.43016562011256765,
-      "learning_rate": 4.069468313471646e-05,
-      "loss": 0.4079,
-      "step": 4582
-    },
-    {
-      "epoch": 2.249831029185868,
-      "grad_norm": 0.4346654484961386,
-      "learning_rate": 4.0645115880126486e-05,
-      "loss": 0.4169,
-      "step": 4583
-    },
-    {
-      "epoch": 2.2503225806451614,
-      "grad_norm": 0.445379207826208,
-      "learning_rate": 4.059557113034308e-05,
-      "loss": 0.4491,
-      "step": 4584
-    },
-    {
-      "epoch": 2.2508141321044546,
-      "grad_norm": 0.41233474037321344,
-      "learning_rate": 4.054604890415148e-05,
-      "loss": 0.441,
-      "step": 4585
-    },
-    {
-      "epoch": 2.251305683563748,
-      "grad_norm": 0.42230985031111795,
-      "learning_rate": 4.0496549220328376e-05,
-      "loss": 0.4399,
-      "step": 4586
-    },
-    {
-      "epoch": 2.2517972350230413,
-      "grad_norm": 0.8896807787548994,
-      "learning_rate": 4.044707209764185e-05,
-      "loss": 0.4918,
-      "step": 4587
-    },
-    {
-      "epoch": 2.252288786482335,
-      "grad_norm": 0.38927116952415103,
-      "learning_rate": 4.0397617554851616e-05,
-      "loss": 0.4003,
-      "step": 4588
-    },
-    {
-      "epoch": 2.252780337941628,
-      "grad_norm": 0.43685660049787034,
-      "learning_rate": 4.034818561070855e-05,
-      "loss": 0.4221,
-      "step": 4589
-    },
-    {
-      "epoch": 2.2532718894009216,
-      "grad_norm": 0.4354103191565989,
-      "learning_rate": 4.029877628395522e-05,
-      "loss": 0.4806,
-      "step": 4590
-    },
-    {
-      "epoch": 2.253763440860215,
-      "grad_norm": 0.41956558625499557,
-      "learning_rate": 4.024938959332546e-05,
-      "loss": 0.3949,
-      "step": 4591
-    },
-    {
-      "epoch": 2.2542549923195083,
-      "grad_norm": 0.39096443425428223,
-      "learning_rate": 4.020002555754459e-05,
-      "loss": 0.418,
-      "step": 4592
-    },
-    {
-      "epoch": 2.254746543778802,
-      "grad_norm": 0.41839045046187673,
-      "learning_rate": 4.015068419532929e-05,
-      "loss": 0.397,
-      "step": 4593
-    },
-    {
-      "epoch": 2.255238095238095,
-      "grad_norm": 0.4754808995685527,
-      "learning_rate": 4.010136552538769e-05,
-      "loss": 0.3927,
-      "step": 4594
-    },
-    {
-      "epoch": 2.2557296466973886,
-      "grad_norm": 0.4460660744439945,
-      "learning_rate": 4.0052069566419305e-05,
-      "loss": 0.4739,
-      "step": 4595
-    },
-    {
-      "epoch": 2.256221198156682,
-      "grad_norm": 0.39726997380970774,
-      "learning_rate": 4.000279633711501e-05,
-      "loss": 0.4115,
-      "step": 4596
-    },
-    {
-      "epoch": 2.2567127496159753,
-      "grad_norm": 0.40592278154823924,
-      "learning_rate": 3.9953545856157104e-05,
-      "loss": 0.435,
-      "step": 4597
-    },
-    {
-      "epoch": 2.257204301075269,
-      "grad_norm": 0.4009355216838383,
-      "learning_rate": 3.990431814221919e-05,
-      "loss": 0.454,
-      "step": 4598
-    },
-    {
-      "epoch": 2.257695852534562,
-      "grad_norm": 0.44290535146492604,
-      "learning_rate": 3.98551132139664e-05,
-      "loss": 0.4698,
-      "step": 4599
-    },
-    {
-      "epoch": 2.2581874039938556,
-      "grad_norm": 0.4460439017698404,
-      "learning_rate": 3.980593109005498e-05,
-      "loss": 0.4639,
-      "step": 4600
-    },
-    {
-      "epoch": 2.2586789554531492,
-      "grad_norm": 0.43143466802333363,
-      "learning_rate": 3.9756771789132776e-05,
-      "loss": 0.3946,
-      "step": 4601
-    },
-    {
-      "epoch": 2.2591705069124424,
-      "grad_norm": 0.41429758273657125,
-      "learning_rate": 3.970763532983882e-05,
-      "loss": 0.3789,
-      "step": 4602
-    },
-    {
-      "epoch": 2.259662058371736,
-      "grad_norm": 0.4508457716819504,
-      "learning_rate": 3.9658521730803544e-05,
-      "loss": 0.4037,
-      "step": 4603
-    },
-    {
-      "epoch": 2.260153609831029,
-      "grad_norm": 0.4139978709261018,
-      "learning_rate": 3.960943101064869e-05,
-      "loss": 0.4341,
-      "step": 4604
-    },
-    {
-      "epoch": 2.2606451612903227,
-      "grad_norm": 0.4027056905353451,
-      "learning_rate": 3.956036318798736e-05,
-      "loss": 0.4269,
-      "step": 4605
-    },
-    {
-      "epoch": 2.261136712749616,
-      "grad_norm": 0.40922513331030813,
-      "learning_rate": 3.9511318281423923e-05,
-      "loss": 0.4359,
-      "step": 4606
-    },
-    {
-      "epoch": 2.2616282642089094,
-      "grad_norm": 0.4020937804835981,
-      "learning_rate": 3.946229630955407e-05,
-      "loss": 0.4513,
-      "step": 4607
-    },
-    {
-      "epoch": 2.2621198156682025,
-      "grad_norm": 0.4611820833640951,
-      "learning_rate": 3.941329729096492e-05,
-      "loss": 0.4137,
-      "step": 4608
-    },
-    {
-      "epoch": 2.262611367127496,
-      "grad_norm": 0.4820967140485992,
-      "learning_rate": 3.9364321244234635e-05,
-      "loss": 0.4502,
-      "step": 4609
-    },
-    {
-      "epoch": 2.2631029185867897,
-      "grad_norm": 0.43654407121472033,
-      "learning_rate": 3.9315368187932934e-05,
-      "loss": 0.4428,
-      "step": 4610
-    },
-    {
-      "epoch": 2.263594470046083,
-      "grad_norm": 0.39993384927913134,
-      "learning_rate": 3.926643814062064e-05,
-      "loss": 0.407,
-      "step": 4611
-    },
-    {
-      "epoch": 2.2640860215053764,
-      "grad_norm": 0.4384874094190952,
-      "learning_rate": 3.921753112084995e-05,
-      "loss": 0.4215,
-      "step": 4612
-    },
-    {
-      "epoch": 2.2645775729646695,
-      "grad_norm": 0.41828219618852935,
-      "learning_rate": 3.916864714716425e-05,
-      "loss": 0.4317,
-      "step": 4613
-    },
-    {
-      "epoch": 2.265069124423963,
-      "grad_norm": 0.4053890493789178,
-      "learning_rate": 3.911978623809826e-05,
-      "loss": 0.4272,
-      "step": 4614
-    },
-    {
-      "epoch": 2.2655606758832567,
-      "grad_norm": 0.41603414503411545,
-      "learning_rate": 3.90709484121779e-05,
-      "loss": 0.4342,
-      "step": 4615
-    },
-    {
-      "epoch": 2.26605222734255,
-      "grad_norm": 0.40679673512646974,
-      "learning_rate": 3.9022133687920346e-05,
-      "loss": 0.4235,
-      "step": 4616
-    },
-    {
-      "epoch": 2.2665437788018434,
-      "grad_norm": 0.4114182523644404,
-      "learning_rate": 3.897334208383413e-05,
-      "loss": 0.4611,
-      "step": 4617
-    },
-    {
-      "epoch": 2.2670353302611366,
-      "grad_norm": 0.3994714572521716,
-      "learning_rate": 3.892457361841879e-05,
-      "loss": 0.3932,
-      "step": 4618
-    },
-    {
-      "epoch": 2.26752688172043,
-      "grad_norm": 0.40916287775468124,
-      "learning_rate": 3.8875828310165354e-05,
-      "loss": 0.4548,
-      "step": 4619
-    },
-    {
-      "epoch": 2.2680184331797237,
-      "grad_norm": 0.4171924973429695,
-      "learning_rate": 3.882710617755578e-05,
-      "loss": 0.4244,
-      "step": 4620
-    },
-    {
-      "epoch": 2.268509984639017,
-      "grad_norm": 0.40797104003939494,
-      "learning_rate": 3.8778407239063517e-05,
-      "loss": 0.3729,
-      "step": 4621
-    },
-    {
-      "epoch": 2.2690015360983105,
-      "grad_norm": 0.42958891391811654,
-      "learning_rate": 3.8729731513153065e-05,
-      "loss": 0.4199,
-      "step": 4622
-    },
-    {
-      "epoch": 2.2694930875576036,
-      "grad_norm": 0.40084594217473807,
-      "learning_rate": 3.8681079018280144e-05,
-      "loss": 0.4236,
-      "step": 4623
-    },
-    {
-      "epoch": 2.269984639016897,
-      "grad_norm": 0.4120488217245618,
-      "learning_rate": 3.863244977289165e-05,
-      "loss": 0.4156,
-      "step": 4624
-    },
-    {
-      "epoch": 2.2704761904761903,
-      "grad_norm": 0.4907505186219989,
-      "learning_rate": 3.85838437954258e-05,
-      "loss": 0.4791,
-      "step": 4625
-    },
-    {
-      "epoch": 2.270967741935484,
-      "grad_norm": 0.42749915402943883,
-      "learning_rate": 3.8535261104311725e-05,
-      "loss": 0.4193,
-      "step": 4626
-    },
-    {
-      "epoch": 2.271459293394777,
-      "grad_norm": 0.42235208614465153,
-      "learning_rate": 3.8486701717969996e-05,
-      "loss": 0.4127,
-      "step": 4627
-    },
-    {
-      "epoch": 2.2719508448540706,
-      "grad_norm": 0.4005869369497063,
-      "learning_rate": 3.8438165654812194e-05,
-      "loss": 0.3838,
-      "step": 4628
-    },
-    {
-      "epoch": 2.272442396313364,
-      "grad_norm": 0.433845300926815,
-      "learning_rate": 3.8389652933241106e-05,
-      "loss": 0.4477,
-      "step": 4629
-    },
-    {
-      "epoch": 2.2729339477726573,
-      "grad_norm": 0.5183877041218403,
-      "learning_rate": 3.834116357165064e-05,
-      "loss": 0.4818,
-      "step": 4630
-    },
-    {
-      "epoch": 2.273425499231951,
-      "grad_norm": 0.4270202814672601,
-      "learning_rate": 3.82926975884259e-05,
-      "loss": 0.4319,
-      "step": 4631
-    },
-    {
-      "epoch": 2.273917050691244,
-      "grad_norm": 0.565245370179597,
-      "learning_rate": 3.824425500194305e-05,
-      "loss": 0.4465,
-      "step": 4632
-    },
-    {
-      "epoch": 2.2744086021505376,
-      "grad_norm": 0.4039578902394242,
-      "learning_rate": 3.819583583056941e-05,
-      "loss": 0.4241,
-      "step": 4633
-    },
-    {
-      "epoch": 2.274900153609831,
-      "grad_norm": 0.39639954565666735,
-      "learning_rate": 3.814744009266356e-05,
-      "loss": 0.4233,
-      "step": 4634
-    },
-    {
-      "epoch": 2.2753917050691244,
-      "grad_norm": 0.3800405835264967,
-      "learning_rate": 3.8099067806574905e-05,
-      "loss": 0.3608,
-      "step": 4635
-    },
-    {
-      "epoch": 2.275883256528418,
-      "grad_norm": 0.41372590167111695,
-      "learning_rate": 3.805071899064424e-05,
-      "loss": 0.4371,
-      "step": 4636
-    },
-    {
-      "epoch": 2.276374807987711,
-      "grad_norm": 0.43731242578098184,
-      "learning_rate": 3.800239366320332e-05,
-      "loss": 0.4414,
-      "step": 4637
-    },
-    {
-      "epoch": 2.2768663594470047,
-      "grad_norm": 0.42546293065024243,
-      "learning_rate": 3.7954091842575004e-05,
-      "loss": 0.4605,
-      "step": 4638
-    },
-    {
-      "epoch": 2.2773579109062982,
-      "grad_norm": 0.41181112298545686,
-      "learning_rate": 3.790581354707325e-05,
-      "loss": 0.4259,
-      "step": 4639
-    },
-    {
-      "epoch": 2.2778494623655914,
-      "grad_norm": 0.4304807424305234,
-      "learning_rate": 3.785755879500312e-05,
-      "loss": 0.3906,
-      "step": 4640
-    },
-    {
-      "epoch": 2.278341013824885,
-      "grad_norm": 0.43232025613389385,
-      "learning_rate": 3.7809327604660735e-05,
-      "loss": 0.4393,
-      "step": 4641
-    },
-    {
-      "epoch": 2.278832565284178,
-      "grad_norm": 0.45054294034429826,
-      "learning_rate": 3.7761119994333215e-05,
-      "loss": 0.4152,
-      "step": 4642
-    },
-    {
-      "epoch": 2.2793241167434717,
-      "grad_norm": 0.4465477853112635,
-      "learning_rate": 3.771293598229894e-05,
-      "loss": 0.4478,
-      "step": 4643
-    },
-    {
-      "epoch": 2.279815668202765,
-      "grad_norm": 0.3948504344265393,
-      "learning_rate": 3.766477558682704e-05,
-      "loss": 0.4047,
-      "step": 4644
-    },
-    {
-      "epoch": 2.2803072196620584,
-      "grad_norm": 0.41226392387825544,
-      "learning_rate": 3.7616638826178e-05,
-      "loss": 0.4278,
-      "step": 4645
-    },
-    {
-      "epoch": 2.2807987711213515,
-      "grad_norm": 0.4280468876965564,
-      "learning_rate": 3.756852571860307e-05,
-      "loss": 0.4811,
-      "step": 4646
-    },
-    {
-      "epoch": 2.281290322580645,
-      "grad_norm": 0.42466977270288475,
-      "learning_rate": 3.752043628234474e-05,
-      "loss": 0.4437,
-      "step": 4647
-    },
-    {
-      "epoch": 2.2817818740399387,
-      "grad_norm": 0.44802496332099906,
-      "learning_rate": 3.7472370535636445e-05,
-      "loss": 0.4484,
-      "step": 4648
-    },
-    {
-      "epoch": 2.282273425499232,
-      "grad_norm": 0.404386504733684,
-      "learning_rate": 3.742432849670261e-05,
-      "loss": 0.4295,
-      "step": 4649
-    },
-    {
-      "epoch": 2.2827649769585254,
-      "grad_norm": 0.4353369101584708,
-      "learning_rate": 3.737631018375872e-05,
-      "loss": 0.3975,
-      "step": 4650
-    },
-    {
-      "epoch": 2.2832565284178186,
-      "grad_norm": 0.43022041895287205,
-      "learning_rate": 3.732831561501123e-05,
-      "loss": 0.426,
-      "step": 4651
-    },
-    {
-      "epoch": 2.283748079877112,
-      "grad_norm": 0.477706570306163,
-      "learning_rate": 3.728034480865763e-05,
-      "loss": 0.4867,
-      "step": 4652
-    },
-    {
-      "epoch": 2.2842396313364057,
-      "grad_norm": 0.4583047865256892,
-      "learning_rate": 3.7232397782886307e-05,
-      "loss": 0.4172,
-      "step": 4653
-    },
-    {
-      "epoch": 2.284731182795699,
-      "grad_norm": 0.404847525461319,
-      "learning_rate": 3.718447455587682e-05,
-      "loss": 0.3846,
-      "step": 4654
-    },
-    {
-      "epoch": 2.2852227342549924,
-      "grad_norm": 0.4226098979981869,
-      "learning_rate": 3.7136575145799454e-05,
-      "loss": 0.3868,
-      "step": 4655
-    },
-    {
-      "epoch": 2.2857142857142856,
-      "grad_norm": 0.4296273361161548,
-      "learning_rate": 3.7088699570815686e-05,
-      "loss": 0.46,
-      "step": 4656
-    },
-    {
-      "epoch": 2.286205837173579,
-      "grad_norm": 0.4179785530279156,
-      "learning_rate": 3.704084784907784e-05,
-      "loss": 0.4354,
-      "step": 4657
-    },
-    {
-      "epoch": 2.2866973886328728,
-      "grad_norm": 0.41141204803552384,
-      "learning_rate": 3.699301999872922e-05,
-      "loss": 0.4006,
-      "step": 4658
-    },
-    {
-      "epoch": 2.287188940092166,
-      "grad_norm": 0.44950649673149085,
-      "learning_rate": 3.694521603790407e-05,
-      "loss": 0.3993,
-      "step": 4659
-    },
-    {
-      "epoch": 2.2876804915514595,
-      "grad_norm": 0.4210103788988504,
-      "learning_rate": 3.6897435984727605e-05,
-      "loss": 0.4149,
-      "step": 4660
-    },
-    {
-      "epoch": 2.2881720430107526,
-      "grad_norm": 0.4186445614979278,
-      "learning_rate": 3.684967985731593e-05,
-      "loss": 0.3967,
-      "step": 4661
-    },
-    {
-      "epoch": 2.288663594470046,
-      "grad_norm": 0.4220170813609645,
-      "learning_rate": 3.680194767377609e-05,
-      "loss": 0.4291,
-      "step": 4662
-    },
-    {
-      "epoch": 2.2891551459293393,
-      "grad_norm": 0.4153687874063938,
-      "learning_rate": 3.675423945220617e-05,
-      "loss": 0.3952,
-      "step": 4663
-    },
-    {
-      "epoch": 2.289646697388633,
-      "grad_norm": 0.40545911262391143,
-      "learning_rate": 3.6706555210694914e-05,
-      "loss": 0.4456,
-      "step": 4664
-    },
-    {
-      "epoch": 2.290138248847926,
-      "grad_norm": 0.40532664376354205,
-      "learning_rate": 3.665889496732223e-05,
-      "loss": 0.4161,
-      "step": 4665
-    },
-    {
-      "epoch": 2.2906298003072196,
-      "grad_norm": 0.40857490132512503,
-      "learning_rate": 3.661125874015881e-05,
-      "loss": 0.4277,
-      "step": 4666
-    },
-    {
-      "epoch": 2.291121351766513,
-      "grad_norm": 0.4510073318625711,
-      "learning_rate": 3.6563646547266214e-05,
-      "loss": 0.4523,
-      "step": 4667
-    },
-    {
-      "epoch": 2.2916129032258064,
-      "grad_norm": 0.4471361763741866,
-      "learning_rate": 3.651605840669695e-05,
-      "loss": 0.4391,
-      "step": 4668
-    },
-    {
-      "epoch": 2.2921044546851,
-      "grad_norm": 0.4363126110022241,
-      "learning_rate": 3.64684943364944e-05,
-      "loss": 0.4731,
-      "step": 4669
-    },
-    {
-      "epoch": 2.292596006144393,
-      "grad_norm": 0.39810227969431883,
-      "learning_rate": 3.642095435469274e-05,
-      "loss": 0.4485,
-      "step": 4670
-    },
-    {
-      "epoch": 2.2930875576036867,
-      "grad_norm": 0.4000949470384779,
-      "learning_rate": 3.637343847931719e-05,
-      "loss": 0.4105,
-      "step": 4671
-    },
-    {
-      "epoch": 2.2935791090629802,
-      "grad_norm": 0.39800246987335286,
-      "learning_rate": 3.6325946728383584e-05,
-      "loss": 0.4192,
-      "step": 4672
-    },
-    {
-      "epoch": 2.2940706605222734,
-      "grad_norm": 0.4602780711104513,
-      "learning_rate": 3.6278479119898846e-05,
-      "loss": 0.4248,
-      "step": 4673
-    },
-    {
-      "epoch": 2.294562211981567,
-      "grad_norm": 0.39877452127112756,
-      "learning_rate": 3.62310356718606e-05,
-      "loss": 0.398,
-      "step": 4674
-    },
-    {
-      "epoch": 2.29505376344086,
-      "grad_norm": 0.40087525919646655,
-      "learning_rate": 3.618361640225735e-05,
-      "loss": 0.4321,
-      "step": 4675
-    },
-    {
-      "epoch": 2.2955453149001537,
-      "grad_norm": 0.436675040585018,
-      "learning_rate": 3.613622132906843e-05,
-      "loss": 0.4336,
-      "step": 4676
-    },
-    {
-      "epoch": 2.296036866359447,
-      "grad_norm": 0.42876675485031573,
-      "learning_rate": 3.6088850470264015e-05,
-      "loss": 0.4349,
-      "step": 4677
-    },
-    {
-      "epoch": 2.2965284178187404,
-      "grad_norm": 0.4111778213451194,
-      "learning_rate": 3.604150384380508e-05,
-      "loss": 0.4389,
-      "step": 4678
-    },
-    {
-      "epoch": 2.297019969278034,
-      "grad_norm": 0.46268139426787824,
-      "learning_rate": 3.59941814676434e-05,
-      "loss": 0.4418,
-      "step": 4679
-    },
-    {
-      "epoch": 2.297511520737327,
-      "grad_norm": 0.4609072912667764,
-      "learning_rate": 3.594688335972164e-05,
-      "loss": 0.4382,
-      "step": 4680
-    },
-    {
-      "epoch": 2.2980030721966207,
-      "grad_norm": 0.42369563482353995,
-      "learning_rate": 3.58996095379731e-05,
-      "loss": 0.4335,
-      "step": 4681
-    },
-    {
-      "epoch": 2.298494623655914,
-      "grad_norm": 0.42286306020186715,
-      "learning_rate": 3.585236002032205e-05,
-      "loss": 0.4543,
-      "step": 4682
-    },
-    {
-      "epoch": 2.2989861751152074,
-      "grad_norm": 0.42080638209148447,
-      "learning_rate": 3.580513482468344e-05,
-      "loss": 0.4184,
-      "step": 4683
-    },
-    {
-      "epoch": 2.2994777265745006,
-      "grad_norm": 0.42167806986936546,
-      "learning_rate": 3.575793396896303e-05,
-      "loss": 0.4501,
-      "step": 4684
-    },
-    {
-      "epoch": 2.299969278033794,
-      "grad_norm": 0.3912828211039465,
-      "learning_rate": 3.571075747105732e-05,
-      "loss": 0.4071,
-      "step": 4685
-    },
-    {
-      "epoch": 2.3004608294930877,
-      "grad_norm": 0.4286467302118104,
-      "learning_rate": 3.5663605348853625e-05,
-      "loss": 0.4538,
-      "step": 4686
-    },
-    {
-      "epoch": 2.300952380952381,
-      "grad_norm": 0.44100720276496774,
-      "learning_rate": 3.561647762022998e-05,
-      "loss": 0.4564,
-      "step": 4687
-    },
-    {
-      "epoch": 2.3014439324116744,
-      "grad_norm": 0.43837868754997794,
-      "learning_rate": 3.556937430305515e-05,
-      "loss": 0.4276,
-      "step": 4688
-    },
-    {
-      "epoch": 2.3019354838709676,
-      "grad_norm": 0.4203693226345875,
-      "learning_rate": 3.5522295415188774e-05,
-      "loss": 0.3944,
-      "step": 4689
-    },
-    {
-      "epoch": 2.302427035330261,
-      "grad_norm": 0.4319344765012458,
-      "learning_rate": 3.5475240974481006e-05,
-      "loss": 0.4006,
-      "step": 4690
-    },
-    {
-      "epoch": 2.3029185867895547,
-      "grad_norm": 0.404823184954575,
-      "learning_rate": 3.542821099877295e-05,
-      "loss": 0.384,
-      "step": 4691
-    },
-    {
-      "epoch": 2.303410138248848,
-      "grad_norm": 0.42483465979867385,
-      "learning_rate": 3.538120550589631e-05,
-      "loss": 0.4313,
-      "step": 4692
-    },
-    {
-      "epoch": 2.3039016897081415,
-      "grad_norm": 0.41230286891851775,
-      "learning_rate": 3.533422451367353e-05,
-      "loss": 0.4464,
-      "step": 4693
-    },
-    {
-      "epoch": 2.3043932411674346,
-      "grad_norm": 0.4325844945211286,
-      "learning_rate": 3.5287268039917785e-05,
-      "loss": 0.4464,
-      "step": 4694
-    },
-    {
-      "epoch": 2.304884792626728,
-      "grad_norm": 0.42635586169823536,
-      "learning_rate": 3.524033610243293e-05,
-      "loss": 0.4111,
-      "step": 4695
-    },
-    {
-      "epoch": 2.3053763440860213,
-      "grad_norm": 0.4301726105167208,
-      "learning_rate": 3.5193428719013523e-05,
-      "loss": 0.4086,
-      "step": 4696
-    },
-    {
-      "epoch": 2.305867895545315,
-      "grad_norm": 0.4198929305556593,
-      "learning_rate": 3.514654590744483e-05,
-      "loss": 0.4632,
-      "step": 4697
-    },
-    {
-      "epoch": 2.3063594470046085,
-      "grad_norm": 0.39637988836601756,
-      "learning_rate": 3.509968768550278e-05,
-      "loss": 0.3943,
-      "step": 4698
-    },
-    {
-      "epoch": 2.3068509984639016,
-      "grad_norm": 0.40120828939709524,
-      "learning_rate": 3.505285407095394e-05,
-      "loss": 0.3931,
-      "step": 4699
-    },
-    {
-      "epoch": 2.307342549923195,
-      "grad_norm": 0.42584379039334697,
-      "learning_rate": 3.500604508155571e-05,
-      "loss": 0.4136,
-      "step": 4700
-    },
-    {
-      "epoch": 2.3078341013824883,
-      "grad_norm": 0.3999601796476035,
-      "learning_rate": 3.4959260735055896e-05,
-      "loss": 0.4432,
-      "step": 4701
-    },
-    {
-      "epoch": 2.308325652841782,
-      "grad_norm": 0.41602318062853544,
-      "learning_rate": 3.491250104919321e-05,
-      "loss": 0.445,
-      "step": 4702
-    },
-    {
-      "epoch": 2.308817204301075,
-      "grad_norm": 0.4030289261290131,
-      "learning_rate": 3.486576604169686e-05,
-      "loss": 0.4214,
-      "step": 4703
-    },
-    {
-      "epoch": 2.3093087557603686,
-      "grad_norm": 0.41813375817643905,
-      "learning_rate": 3.481905573028673e-05,
-      "loss": 0.4051,
-      "step": 4704
-    },
-    {
-      "epoch": 2.3098003072196622,
-      "grad_norm": 0.4376203451293558,
-      "learning_rate": 3.477237013267337e-05,
-      "loss": 0.436,
-      "step": 4705
-    },
-    {
-      "epoch": 2.3102918586789554,
-      "grad_norm": 0.4035687214978179,
-      "learning_rate": 3.4725709266557924e-05,
-      "loss": 0.4085,
-      "step": 4706
-    },
-    {
-      "epoch": 2.310783410138249,
-      "grad_norm": 0.4228343437108084,
-      "learning_rate": 3.467907314963219e-05,
-      "loss": 0.428,
-      "step": 4707
-    },
-    {
-      "epoch": 2.311274961597542,
-      "grad_norm": 0.4458722455069885,
-      "learning_rate": 3.4632461799578534e-05,
-      "loss": 0.4808,
-      "step": 4708
-    },
-    {
-      "epoch": 2.3117665130568357,
-      "grad_norm": 0.42559318325555223,
-      "learning_rate": 3.4585875234070056e-05,
-      "loss": 0.4096,
-      "step": 4709
-    },
-    {
-      "epoch": 2.3122580645161293,
-      "grad_norm": 0.41919658507677926,
-      "learning_rate": 3.453931347077024e-05,
-      "loss": 0.4092,
-      "step": 4710
-    },
-    {
-      "epoch": 2.3127496159754224,
-      "grad_norm": 0.4241294540352367,
-      "learning_rate": 3.44927765273334e-05,
-      "loss": 0.4104,
-      "step": 4711
-    },
-    {
-      "epoch": 2.313241167434716,
-      "grad_norm": 0.42778857671006904,
-      "learning_rate": 3.444626442140428e-05,
-      "loss": 0.4387,
-      "step": 4712
-    },
-    {
-      "epoch": 2.313732718894009,
-      "grad_norm": 0.43551370942469136,
-      "learning_rate": 3.4399777170618286e-05,
-      "loss": 0.4587,
-      "step": 4713
-    },
-    {
-      "epoch": 2.3142242703533027,
-      "grad_norm": 0.4477057519800819,
-      "learning_rate": 3.435331479260133e-05,
-      "loss": 0.4485,
-      "step": 4714
-    },
-    {
-      "epoch": 2.314715821812596,
-      "grad_norm": 0.4277263298977427,
-      "learning_rate": 3.430687730497003e-05,
-      "loss": 0.4641,
-      "step": 4715
-    },
-    {
-      "epoch": 2.3152073732718894,
-      "grad_norm": 0.45950082107122175,
-      "learning_rate": 3.4260464725331345e-05,
-      "loss": 0.4639,
-      "step": 4716
-    },
-    {
-      "epoch": 2.3156989247311826,
-      "grad_norm": 0.4204903418651704,
-      "learning_rate": 3.421407707128306e-05,
-      "loss": 0.4543,
-      "step": 4717
-    },
-    {
-      "epoch": 2.316190476190476,
-      "grad_norm": 0.4227449257009278,
-      "learning_rate": 3.416771436041323e-05,
-      "loss": 0.4302,
-      "step": 4718
-    },
-    {
-      "epoch": 2.3166820276497697,
-      "grad_norm": 0.41669218211322623,
-      "learning_rate": 3.412137661030068e-05,
-      "loss": 0.3881,
-      "step": 4719
-    },
-    {
-      "epoch": 2.317173579109063,
-      "grad_norm": 0.41996009125661177,
-      "learning_rate": 3.4075063838514676e-05,
-      "loss": 0.4236,
-      "step": 4720
-    },
-    {
-      "epoch": 2.3176651305683564,
-      "grad_norm": 0.39659493085845643,
-      "learning_rate": 3.402877606261499e-05,
-      "loss": 0.3903,
-      "step": 4721
-    },
-    {
-      "epoch": 2.3181566820276496,
-      "grad_norm": 0.4036941455892041,
-      "learning_rate": 3.3982513300151964e-05,
-      "loss": 0.4006,
-      "step": 4722
-    },
-    {
-      "epoch": 2.318648233486943,
-      "grad_norm": 0.4123896872395089,
-      "learning_rate": 3.393627556866643e-05,
-      "loss": 0.4246,
-      "step": 4723
-    },
-    {
-      "epoch": 2.3191397849462367,
-      "grad_norm": 0.4096088359337234,
-      "learning_rate": 3.389006288568974e-05,
-      "loss": 0.4157,
-      "step": 4724
-    },
-    {
-      "epoch": 2.31963133640553,
-      "grad_norm": 0.4326586651819651,
-      "learning_rate": 3.384387526874371e-05,
-      "loss": 0.4508,
-      "step": 4725
-    },
-    {
-      "epoch": 2.3201228878648235,
-      "grad_norm": 0.40072572740826556,
-      "learning_rate": 3.3797712735340794e-05,
-      "loss": 0.4002,
-      "step": 4726
-    },
-    {
-      "epoch": 2.3206144393241166,
-      "grad_norm": 0.40810647346308376,
-      "learning_rate": 3.3751575302983694e-05,
-      "loss": 0.4288,
-      "step": 4727
-    },
-    {
-      "epoch": 2.32110599078341,
-      "grad_norm": 0.47708630588981893,
-      "learning_rate": 3.370546298916583e-05,
-      "loss": 0.4604,
-      "step": 4728
-    },
-    {
-      "epoch": 2.3215975422427038,
-      "grad_norm": 0.4502626136190057,
-      "learning_rate": 3.365937581137095e-05,
-      "loss": 0.4312,
-      "step": 4729
-    },
-    {
-      "epoch": 2.322089093701997,
-      "grad_norm": 0.40699201528340645,
-      "learning_rate": 3.361331378707334e-05,
-      "loss": 0.4288,
-      "step": 4730
-    },
-    {
-      "epoch": 2.3225806451612905,
-      "grad_norm": 0.4165295196834001,
-      "learning_rate": 3.3567276933737734e-05,
-      "loss": 0.4499,
-      "step": 4731
-    },
-    {
-      "epoch": 2.3230721966205836,
-      "grad_norm": 0.396397589918511,
-      "learning_rate": 3.3521265268819294e-05,
-      "loss": 0.4364,
-      "step": 4732
-    },
-    {
-      "epoch": 2.323563748079877,
-      "grad_norm": 0.458204883704109,
-      "learning_rate": 3.347527880976367e-05,
-      "loss": 0.4586,
-      "step": 4733
-    },
-    {
-      "epoch": 2.3240552995391703,
-      "grad_norm": 0.42073207194970696,
-      "learning_rate": 3.342931757400689e-05,
-      "loss": 0.4087,
-      "step": 4734
-    },
-    {
-      "epoch": 2.324546850998464,
-      "grad_norm": 0.4647240458644532,
-      "learning_rate": 3.338338157897559e-05,
-      "loss": 0.4255,
-      "step": 4735
-    },
-    {
-      "epoch": 2.325038402457757,
-      "grad_norm": 0.43744954731761865,
-      "learning_rate": 3.333747084208657e-05,
-      "loss": 0.4913,
-      "step": 4736
-    },
-    {
-      "epoch": 2.3255299539170506,
-      "grad_norm": 0.3982797540445961,
-      "learning_rate": 3.329158538074729e-05,
-      "loss": 0.4168,
-      "step": 4737
-    },
-    {
-      "epoch": 2.3260215053763442,
-      "grad_norm": 0.41405732467800316,
-      "learning_rate": 3.324572521235552e-05,
-      "loss": 0.4225,
-      "step": 4738
-    },
-    {
-      "epoch": 2.3265130568356374,
-      "grad_norm": 0.43401156821621784,
-      "learning_rate": 3.3199890354299435e-05,
-      "loss": 0.4473,
-      "step": 4739
-    },
-    {
-      "epoch": 2.327004608294931,
-      "grad_norm": 0.4106466316435039,
-      "learning_rate": 3.3154080823957635e-05,
-      "loss": 0.4377,
-      "step": 4740
-    },
-    {
-      "epoch": 2.327496159754224,
-      "grad_norm": 0.41976505694764593,
-      "learning_rate": 3.310829663869912e-05,
-      "loss": 0.4355,
-      "step": 4741
-    },
-    {
-      "epoch": 2.3279877112135177,
-      "grad_norm": 0.42346622278551865,
-      "learning_rate": 3.306253781588327e-05,
-      "loss": 0.428,
-      "step": 4742
-    },
-    {
-      "epoch": 2.3284792626728112,
-      "grad_norm": 0.4306519527183008,
-      "learning_rate": 3.3016804372859854e-05,
-      "loss": 0.4358,
-      "step": 4743
-    },
-    {
-      "epoch": 2.3289708141321044,
-      "grad_norm": 0.42950444364385965,
-      "learning_rate": 3.297109632696902e-05,
-      "loss": 0.4283,
-      "step": 4744
-    },
-    {
-      "epoch": 2.329462365591398,
-      "grad_norm": 0.43906553733635073,
-      "learning_rate": 3.2925413695541244e-05,
-      "loss": 0.4674,
-      "step": 4745
-    },
-    {
-      "epoch": 2.329953917050691,
-      "grad_norm": 0.4255153153818455,
-      "learning_rate": 3.287975649589751e-05,
-      "loss": 0.4244,
-      "step": 4746
-    },
-    {
-      "epoch": 2.3304454685099847,
-      "grad_norm": 0.43026277125446477,
-      "learning_rate": 3.283412474534893e-05,
-      "loss": 0.4237,
-      "step": 4747
-    },
-    {
-      "epoch": 2.3309370199692783,
-      "grad_norm": 0.4093513150854702,
-      "learning_rate": 3.2788518461197157e-05,
-      "loss": 0.433,
-      "step": 4748
-    },
-    {
-      "epoch": 2.3314285714285714,
-      "grad_norm": 0.43143713432610686,
-      "learning_rate": 3.274293766073413e-05,
-      "loss": 0.4422,
-      "step": 4749
-    },
-    {
-      "epoch": 2.331920122887865,
-      "grad_norm": 0.4196447328123866,
-      "learning_rate": 3.2697382361242104e-05,
-      "loss": 0.4036,
-      "step": 4750
-    },
-    {
-      "epoch": 2.332411674347158,
-      "grad_norm": 0.4204694343348252,
-      "learning_rate": 3.265185257999367e-05,
-      "loss": 0.459,
-      "step": 4751
-    },
-    {
-      "epoch": 2.3329032258064517,
-      "grad_norm": 0.44006084728563805,
-      "learning_rate": 3.2606348334251755e-05,
-      "loss": 0.434,
-      "step": 4752
-    },
-    {
-      "epoch": 2.333394777265745,
-      "grad_norm": 0.39821864349299974,
-      "learning_rate": 3.256086964126962e-05,
-      "loss": 0.4071,
-      "step": 4753
-    },
-    {
-      "epoch": 2.3338863287250384,
-      "grad_norm": 0.3972280648289152,
-      "learning_rate": 3.2515416518290774e-05,
-      "loss": 0.4173,
-      "step": 4754
-    },
-    {
-      "epoch": 2.3343778801843316,
-      "grad_norm": 0.41062672806796124,
-      "learning_rate": 3.246998898254917e-05,
-      "loss": 0.4248,
-      "step": 4755
-    },
-    {
-      "epoch": 2.334869431643625,
-      "grad_norm": 0.46399725882181564,
-      "learning_rate": 3.2424587051268837e-05,
-      "loss": 0.4566,
-      "step": 4756
-    },
-    {
-      "epoch": 2.3353609831029187,
-      "grad_norm": 0.4211860928938646,
-      "learning_rate": 3.2379210741664336e-05,
-      "loss": 0.3884,
-      "step": 4757
-    },
-    {
-      "epoch": 2.335852534562212,
-      "grad_norm": 0.4552154019131506,
-      "learning_rate": 3.233386007094036e-05,
-      "loss": 0.4691,
-      "step": 4758
-    },
-    {
-      "epoch": 2.3363440860215055,
-      "grad_norm": 0.42889915354629227,
-      "learning_rate": 3.228853505629192e-05,
-      "loss": 0.4218,
-      "step": 4759
-    },
-    {
-      "epoch": 2.3368356374807986,
-      "grad_norm": 0.39334507688502596,
-      "learning_rate": 3.224323571490428e-05,
-      "loss": 0.3941,
-      "step": 4760
-    },
-    {
-      "epoch": 2.337327188940092,
-      "grad_norm": 0.38969390775568347,
-      "learning_rate": 3.219796206395307e-05,
-      "loss": 0.4331,
-      "step": 4761
-    },
-    {
-      "epoch": 2.3378187403993858,
-      "grad_norm": 0.4079829619829583,
-      "learning_rate": 3.2152714120603986e-05,
-      "loss": 0.4394,
-      "step": 4762
-    },
-    {
-      "epoch": 2.338310291858679,
-      "grad_norm": 0.40411940197509744,
-      "learning_rate": 3.21074919020132e-05,
-      "loss": 0.4302,
-      "step": 4763
-    },
-    {
-      "epoch": 2.3388018433179725,
-      "grad_norm": 0.43858726812015925,
-      "learning_rate": 3.206229542532697e-05,
-      "loss": 0.4084,
-      "step": 4764
-    },
-    {
-      "epoch": 2.3392933947772656,
-      "grad_norm": 0.4156139932197252,
-      "learning_rate": 3.201712470768185e-05,
-      "loss": 0.4037,
-      "step": 4765
-    },
-    {
-      "epoch": 2.339784946236559,
-      "grad_norm": 0.43863161229391,
-      "learning_rate": 3.197197976620469e-05,
-      "loss": 0.4255,
-      "step": 4766
-    },
-    {
-      "epoch": 2.3402764976958528,
-      "grad_norm": 0.4480292358381235,
-      "learning_rate": 3.1926860618012346e-05,
-      "loss": 0.4522,
-      "step": 4767
-    },
-    {
-      "epoch": 2.340768049155146,
-      "grad_norm": 0.3890287573111709,
-      "learning_rate": 3.188176728021218e-05,
-      "loss": 0.4002,
-      "step": 4768
-    },
-    {
-      "epoch": 2.3412596006144395,
-      "grad_norm": 0.4184492254526786,
-      "learning_rate": 3.18366997699016e-05,
-      "loss": 0.4123,
-      "step": 4769
-    },
-    {
-      "epoch": 2.3417511520737326,
-      "grad_norm": 0.42506116602466937,
-      "learning_rate": 3.179165810416826e-05,
-      "loss": 0.455,
-      "step": 4770
-    },
-    {
-      "epoch": 2.342242703533026,
-      "grad_norm": 0.4122783842189617,
-      "learning_rate": 3.174664230008998e-05,
-      "loss": 0.4315,
-      "step": 4771
-    },
-    {
-      "epoch": 2.3427342549923194,
-      "grad_norm": 0.4148489845101186,
-      "learning_rate": 3.17016523747349e-05,
-      "loss": 0.422,
-      "step": 4772
-    },
-    {
-      "epoch": 2.343225806451613,
-      "grad_norm": 0.42936894470285625,
-      "learning_rate": 3.165668834516112e-05,
-      "loss": 0.4284,
-      "step": 4773
-    },
-    {
-      "epoch": 2.343717357910906,
-      "grad_norm": 0.4159236673561618,
-      "learning_rate": 3.161175022841717e-05,
-      "loss": 0.4283,
-      "step": 4774
-    },
-    {
-      "epoch": 2.3442089093701997,
-      "grad_norm": 0.42347653630964766,
-      "learning_rate": 3.15668380415416e-05,
-      "loss": 0.4114,
-      "step": 4775
-    },
-    {
-      "epoch": 2.3447004608294932,
-      "grad_norm": 0.4651548232142126,
-      "learning_rate": 3.152195180156317e-05,
-      "loss": 0.4897,
-      "step": 4776
-    },
-    {
-      "epoch": 2.3451920122887864,
-      "grad_norm": 0.40571874403564384,
-      "learning_rate": 3.14770915255008e-05,
-      "loss": 0.4521,
-      "step": 4777
-    },
-    {
-      "epoch": 2.34568356374808,
-      "grad_norm": 0.4436730221680044,
-      "learning_rate": 3.143225723036357e-05,
-      "loss": 0.4196,
-      "step": 4778
-    },
-    {
-      "epoch": 2.346175115207373,
-      "grad_norm": 0.42543159822317894,
-      "learning_rate": 3.13874489331507e-05,
-      "loss": 0.4238,
-      "step": 4779
-    },
-    {
-      "epoch": 2.3466666666666667,
-      "grad_norm": 0.4083009504906106,
-      "learning_rate": 3.134266665085154e-05,
-      "loss": 0.4262,
-      "step": 4780
-    },
-    {
-      "epoch": 2.3471582181259603,
-      "grad_norm": 0.4283724148865335,
-      "learning_rate": 3.1297910400445686e-05,
-      "loss": 0.4164,
-      "step": 4781
-    },
-    {
-      "epoch": 2.3476497695852534,
-      "grad_norm": 0.4198485592266913,
-      "learning_rate": 3.1253180198902655e-05,
-      "loss": 0.4225,
-      "step": 4782
-    },
-    {
-      "epoch": 2.348141321044547,
-      "grad_norm": 0.41970964022106316,
-      "learning_rate": 3.120847606318228e-05,
-      "loss": 0.461,
-      "step": 4783
-    },
-    {
-      "epoch": 2.34863287250384,
-      "grad_norm": 0.43024997503838885,
-      "learning_rate": 3.1163798010234424e-05,
-      "loss": 0.4452,
-      "step": 4784
-    },
-    {
-      "epoch": 2.3491244239631337,
-      "grad_norm": 0.4410145878631437,
-      "learning_rate": 3.111914605699906e-05,
-      "loss": 0.4551,
-      "step": 4785
-    },
-    {
-      "epoch": 2.3496159754224273,
-      "grad_norm": 0.4892838955368589,
-      "learning_rate": 3.10745202204063e-05,
-      "loss": 0.4335,
-      "step": 4786
-    },
-    {
-      "epoch": 2.3501075268817204,
-      "grad_norm": 0.38238948191787914,
-      "learning_rate": 3.102992051737631e-05,
-      "loss": 0.3736,
-      "step": 4787
-    },
-    {
-      "epoch": 2.350599078341014,
-      "grad_norm": 0.38302858860671996,
-      "learning_rate": 3.098534696481937e-05,
-      "loss": 0.411,
-      "step": 4788
-    },
-    {
-      "epoch": 2.351090629800307,
-      "grad_norm": 0.4084756345531803,
-      "learning_rate": 3.094079957963584e-05,
-      "loss": 0.4268,
-      "step": 4789
-    },
-    {
-      "epoch": 2.3515821812596007,
-      "grad_norm": 0.4165753600155168,
-      "learning_rate": 3.089627837871623e-05,
-      "loss": 0.4602,
-      "step": 4790
-    },
-    {
-      "epoch": 2.352073732718894,
-      "grad_norm": 0.42206661017082336,
-      "learning_rate": 3.085178337894093e-05,
-      "loss": 0.4471,
-      "step": 4791
-    },
-    {
-      "epoch": 2.3525652841781874,
-      "grad_norm": 0.3800883691971558,
-      "learning_rate": 3.080731459718067e-05,
-      "loss": 0.4186,
-      "step": 4792
-    },
-    {
-      "epoch": 2.3530568356374806,
-      "grad_norm": 0.41444170721371476,
-      "learning_rate": 3.0762872050295935e-05,
-      "loss": 0.4204,
-      "step": 4793
-    },
-    {
-      "epoch": 2.353548387096774,
-      "grad_norm": 0.40085496217796385,
-      "learning_rate": 3.0718455755137534e-05,
-      "loss": 0.3856,
-      "step": 4794
-    },
-    {
-      "epoch": 2.3540399385560677,
-      "grad_norm": 0.4189243886227632,
-      "learning_rate": 3.0674065728546166e-05,
-      "loss": 0.4074,
-      "step": 4795
-    },
-    {
-      "epoch": 2.354531490015361,
-      "grad_norm": 0.41448934031974805,
-      "learning_rate": 3.06297019873526e-05,
-      "loss": 0.4704,
-      "step": 4796
-    },
-    {
-      "epoch": 2.3550230414746545,
-      "grad_norm": 0.42468459699633776,
-      "learning_rate": 3.058536454837767e-05,
-      "loss": 0.4089,
-      "step": 4797
-    },
-    {
-      "epoch": 2.3555145929339476,
-      "grad_norm": 0.40372523180114434,
-      "learning_rate": 3.054105342843221e-05,
-      "loss": 0.38,
-      "step": 4798
-    },
-    {
-      "epoch": 2.356006144393241,
-      "grad_norm": 0.4099057107716398,
-      "learning_rate": 3.0496768644317077e-05,
-      "loss": 0.3934,
-      "step": 4799
-    },
-    {
-      "epoch": 2.3564976958525348,
-      "grad_norm": 0.41255023543990066,
-      "learning_rate": 3.0452510212823104e-05,
-      "loss": 0.3916,
-      "step": 4800
-    },
-    {
-      "epoch": 2.356989247311828,
-      "grad_norm": 0.43795605125926107,
-      "learning_rate": 3.0408278150731297e-05,
-      "loss": 0.4368,
-      "step": 4801
-    },
-    {
-      "epoch": 2.3574807987711215,
-      "grad_norm": 0.4251178192904342,
-      "learning_rate": 3.03640724748124e-05,
-      "loss": 0.4965,
-      "step": 4802
-    },
-    {
-      "epoch": 2.3579723502304146,
-      "grad_norm": 0.42830745049729313,
-      "learning_rate": 3.0319893201827387e-05,
-      "loss": 0.4128,
-      "step": 4803
-    },
-    {
-      "epoch": 2.358463901689708,
-      "grad_norm": 0.42090651381117394,
-      "learning_rate": 3.0275740348527093e-05,
-      "loss": 0.4478,
-      "step": 4804
-    },
-    {
-      "epoch": 2.3589554531490013,
-      "grad_norm": 0.3973716768544322,
-      "learning_rate": 3.0231613931652392e-05,
-      "loss": 0.3933,
-      "step": 4805
-    },
-    {
-      "epoch": 2.359447004608295,
-      "grad_norm": 0.4243238260429334,
-      "learning_rate": 3.0187513967934067e-05,
-      "loss": 0.4549,
-      "step": 4806
-    },
-    {
-      "epoch": 2.3599385560675885,
-      "grad_norm": 0.4343255673653745,
-      "learning_rate": 3.014344047409301e-05,
-      "loss": 0.438,
-      "step": 4807
-    },
-    {
-      "epoch": 2.3604301075268816,
-      "grad_norm": 0.41149175869544385,
-      "learning_rate": 3.0099393466839864e-05,
-      "loss": 0.4129,
-      "step": 4808
-    },
-    {
-      "epoch": 2.3609216589861752,
-      "grad_norm": 0.42358128335906053,
-      "learning_rate": 3.005537296287546e-05,
-      "loss": 0.4609,
-      "step": 4809
-    },
-    {
-      "epoch": 2.3614132104454684,
-      "grad_norm": 0.6738010749080955,
-      "learning_rate": 3.0011378978890416e-05,
-      "loss": 0.465,
-      "step": 4810
-    },
-    {
-      "epoch": 2.361904761904762,
-      "grad_norm": 0.41749798248955805,
-      "learning_rate": 2.996741153156535e-05,
-      "loss": 0.4288,
-      "step": 4811
-    },
-    {
-      "epoch": 2.362396313364055,
-      "grad_norm": 0.42248666624639847,
-      "learning_rate": 2.9923470637570827e-05,
-      "loss": 0.4292,
-      "step": 4812
-    },
-    {
-      "epoch": 2.3628878648233487,
-      "grad_norm": 0.4134649845494821,
-      "learning_rate": 2.9879556313567335e-05,
-      "loss": 0.42,
-      "step": 4813
-    },
-    {
-      "epoch": 2.3633794162826423,
-      "grad_norm": 0.4346405485385153,
-      "learning_rate": 2.9835668576205288e-05,
-      "loss": 0.4261,
-      "step": 4814
-    },
-    {
-      "epoch": 2.3638709677419354,
-      "grad_norm": 0.4355669902937064,
-      "learning_rate": 2.979180744212502e-05,
-      "loss": 0.4543,
-      "step": 4815
-    },
-    {
-      "epoch": 2.364362519201229,
-      "grad_norm": 0.4063349307832069,
-      "learning_rate": 2.9747972927956768e-05,
-      "loss": 0.4085,
-      "step": 4816
-    },
-    {
-      "epoch": 2.364854070660522,
-      "grad_norm": 0.4007487000617257,
-      "learning_rate": 2.9704165050320652e-05,
-      "loss": 0.4044,
-      "step": 4817
-    },
-    {
-      "epoch": 2.3653456221198157,
-      "grad_norm": 0.40464878367350376,
-      "learning_rate": 2.9660383825826842e-05,
-      "loss": 0.3812,
-      "step": 4818
-    },
-    {
-      "epoch": 2.3658371735791093,
-      "grad_norm": 0.39128615180039056,
-      "learning_rate": 2.9616629271075137e-05,
-      "loss": 0.4422,
-      "step": 4819
-    },
-    {
-      "epoch": 2.3663287250384024,
-      "grad_norm": 0.4184770545329243,
-      "learning_rate": 2.9572901402655484e-05,
-      "loss": 0.4307,
-      "step": 4820
-    },
-    {
-      "epoch": 2.366820276497696,
-      "grad_norm": 0.419322774265694,
-      "learning_rate": 2.9529200237147546e-05,
-      "loss": 0.4534,
-      "step": 4821
-    },
-    {
-      "epoch": 2.367311827956989,
-      "grad_norm": 0.48334917449392434,
-      "learning_rate": 2.948552579112095e-05,
-      "loss": 0.4712,
-      "step": 4822
-    },
-    {
-      "epoch": 2.3678033794162827,
-      "grad_norm": 0.43277034543550946,
-      "learning_rate": 2.944187808113512e-05,
-      "loss": 0.4331,
-      "step": 4823
-    },
-    {
-      "epoch": 2.368294930875576,
-      "grad_norm": 0.42527901485206754,
-      "learning_rate": 2.9398257123739413e-05,
-      "loss": 0.3937,
-      "step": 4824
-    },
-    {
-      "epoch": 2.3687864823348694,
-      "grad_norm": 0.4281167188782922,
-      "learning_rate": 2.9354662935472997e-05,
-      "loss": 0.4838,
-      "step": 4825
-    },
-    {
-      "epoch": 2.369278033794163,
-      "grad_norm": 0.4135117960722692,
-      "learning_rate": 2.9311095532864874e-05,
-      "loss": 0.4136,
-      "step": 4826
-    },
-    {
-      "epoch": 2.369769585253456,
-      "grad_norm": 0.39062068942264067,
-      "learning_rate": 2.9267554932434006e-05,
-      "loss": 0.3675,
-      "step": 4827
-    },
-    {
-      "epoch": 2.3702611367127497,
-      "grad_norm": 0.4081901507608112,
-      "learning_rate": 2.9224041150688997e-05,
-      "loss": 0.3844,
-      "step": 4828
-    },
-    {
-      "epoch": 2.370752688172043,
-      "grad_norm": 0.4312197072803686,
-      "learning_rate": 2.9180554204128473e-05,
-      "loss": 0.4536,
-      "step": 4829
-    },
-    {
-      "epoch": 2.3712442396313365,
-      "grad_norm": 0.4116497918018741,
-      "learning_rate": 2.913709410924078e-05,
-      "loss": 0.4114,
-      "step": 4830
-    },
-    {
-      "epoch": 2.3717357910906296,
-      "grad_norm": 0.422336565762676,
-      "learning_rate": 2.9093660882504105e-05,
-      "loss": 0.4422,
-      "step": 4831
-    },
-    {
-      "epoch": 2.372227342549923,
-      "grad_norm": 0.4685740962927458,
-      "learning_rate": 2.9050254540386457e-05,
-      "loss": 0.4537,
-      "step": 4832
-    },
-    {
-      "epoch": 2.3727188940092168,
-      "grad_norm": 0.4411007892285831,
-      "learning_rate": 2.900687509934563e-05,
-      "loss": 0.4403,
-      "step": 4833
-    },
-    {
-      "epoch": 2.37321044546851,
-      "grad_norm": 0.4145793060508992,
-      "learning_rate": 2.896352257582925e-05,
-      "loss": 0.438,
-      "step": 4834
-    },
-    {
-      "epoch": 2.3737019969278035,
-      "grad_norm": 0.39228570624315656,
-      "learning_rate": 2.892019698627467e-05,
-      "loss": 0.4041,
-      "step": 4835
-    },
-    {
-      "epoch": 2.3741935483870966,
-      "grad_norm": 0.43856759654947197,
-      "learning_rate": 2.8876898347109195e-05,
-      "loss": 0.4585,
-      "step": 4836
-    },
-    {
-      "epoch": 2.37468509984639,
-      "grad_norm": 0.39721503017618237,
-      "learning_rate": 2.883362667474967e-05,
-      "loss": 0.4156,
-      "step": 4837
-    },
-    {
-      "epoch": 2.375176651305684,
-      "grad_norm": 0.40519165834452425,
-      "learning_rate": 2.8790381985602922e-05,
-      "loss": 0.4371,
-      "step": 4838
-    },
-    {
-      "epoch": 2.375668202764977,
-      "grad_norm": 0.4155310705391595,
-      "learning_rate": 2.874716429606543e-05,
-      "loss": 0.4015,
-      "step": 4839
-    },
-    {
-      "epoch": 2.3761597542242705,
-      "grad_norm": 0.43033233439082563,
-      "learning_rate": 2.8703973622523505e-05,
-      "loss": 0.4733,
-      "step": 4840
-    },
-    {
-      "epoch": 2.3766513056835636,
-      "grad_norm": 0.4464480701023259,
-      "learning_rate": 2.8660809981353165e-05,
-      "loss": 0.4662,
-      "step": 4841
-    },
-    {
-      "epoch": 2.3771428571428572,
-      "grad_norm": 0.44938378423038566,
-      "learning_rate": 2.8617673388920197e-05,
-      "loss": 0.4815,
-      "step": 4842
-    },
-    {
-      "epoch": 2.3776344086021504,
-      "grad_norm": 0.4463081462483464,
-      "learning_rate": 2.857456386158014e-05,
-      "loss": 0.4135,
-      "step": 4843
-    },
-    {
-      "epoch": 2.378125960061444,
-      "grad_norm": 0.44035002552097285,
-      "learning_rate": 2.853148141567824e-05,
-      "loss": 0.4938,
-      "step": 4844
-    },
-    {
-      "epoch": 2.378617511520737,
-      "grad_norm": 0.41162086220383853,
-      "learning_rate": 2.8488426067549533e-05,
-      "loss": 0.4236,
-      "step": 4845
-    },
-    {
-      "epoch": 2.3791090629800307,
-      "grad_norm": 0.4379675645948824,
-      "learning_rate": 2.8445397833518695e-05,
-      "loss": 0.3664,
-      "step": 4846
-    },
-    {
-      "epoch": 2.3796006144393242,
-      "grad_norm": 0.4214342115439349,
-      "learning_rate": 2.840239672990026e-05,
-      "loss": 0.4378,
-      "step": 4847
-    },
-    {
-      "epoch": 2.3800921658986174,
-      "grad_norm": 0.42574262166261506,
-      "learning_rate": 2.8359422772998278e-05,
-      "loss": 0.4603,
-      "step": 4848
-    },
-    {
-      "epoch": 2.380583717357911,
-      "grad_norm": 0.4633004664942074,
-      "learning_rate": 2.8316475979106706e-05,
-      "loss": 0.4625,
-      "step": 4849
-    },
-    {
-      "epoch": 2.381075268817204,
-      "grad_norm": 0.41098575800817566,
-      "learning_rate": 2.8273556364509078e-05,
-      "loss": 0.4274,
-      "step": 4850
-    },
-    {
-      "epoch": 2.3815668202764977,
-      "grad_norm": 0.40466716665847524,
-      "learning_rate": 2.8230663945478663e-05,
-      "loss": 0.3734,
-      "step": 4851
-    },
-    {
-      "epoch": 2.3820583717357913,
-      "grad_norm": 0.4229969232897735,
-      "learning_rate": 2.8187798738278358e-05,
-      "loss": 0.4428,
-      "step": 4852
-    },
-    {
-      "epoch": 2.3825499231950844,
-      "grad_norm": 0.429880737144034,
-      "learning_rate": 2.8144960759160922e-05,
-      "loss": 0.4675,
-      "step": 4853
-    },
-    {
-      "epoch": 2.383041474654378,
-      "grad_norm": 0.38905584639360397,
-      "learning_rate": 2.8102150024368525e-05,
-      "loss": 0.4004,
-      "step": 4854
-    },
-    {
-      "epoch": 2.383533026113671,
-      "grad_norm": 0.4126523752904647,
-      "learning_rate": 2.8059366550133237e-05,
-      "loss": 0.414,
-      "step": 4855
-    },
-    {
-      "epoch": 2.3840245775729647,
-      "grad_norm": 0.4303438862748314,
-      "learning_rate": 2.8016610352676675e-05,
-      "loss": 0.4628,
-      "step": 4856
-    },
-    {
-      "epoch": 2.3845161290322583,
-      "grad_norm": 0.4148592383125127,
-      "learning_rate": 2.797388144821015e-05,
-      "loss": 0.3957,
-      "step": 4857
-    },
-    {
-      "epoch": 2.3850076804915514,
-      "grad_norm": 0.41798689874148054,
-      "learning_rate": 2.7931179852934596e-05,
-      "loss": 0.4635,
-      "step": 4858
-    },
-    {
-      "epoch": 2.385499231950845,
-      "grad_norm": 0.4044476611099365,
-      "learning_rate": 2.7888505583040638e-05,
-      "loss": 0.4393,
-      "step": 4859
-    },
-    {
-      "epoch": 2.385990783410138,
-      "grad_norm": 0.41043263527171825,
-      "learning_rate": 2.784585865470849e-05,
-      "loss": 0.4243,
-      "step": 4860
-    },
-    {
-      "epoch": 2.3864823348694317,
-      "grad_norm": 0.4677727877314856,
-      "learning_rate": 2.7803239084108002e-05,
-      "loss": 0.4973,
-      "step": 4861
-    },
-    {
-      "epoch": 2.386973886328725,
-      "grad_norm": 0.5059274749460108,
-      "learning_rate": 2.776064688739878e-05,
-      "loss": 0.5678,
-      "step": 4862
-    },
-    {
-      "epoch": 2.3874654377880185,
-      "grad_norm": 0.4197777784887138,
-      "learning_rate": 2.771808208072979e-05,
-      "loss": 0.4243,
-      "step": 4863
-    },
-    {
-      "epoch": 2.3879569892473116,
-      "grad_norm": 0.5906791477620013,
-      "learning_rate": 2.767554468023992e-05,
-      "loss": 0.4261,
-      "step": 4864
-    },
-    {
-      "epoch": 2.388448540706605,
-      "grad_norm": 0.43051146996828155,
-      "learning_rate": 2.763303470205738e-05,
-      "loss": 0.4663,
-      "step": 4865
-    },
-    {
-      "epoch": 2.3889400921658988,
-      "grad_norm": 0.43273203777980085,
-      "learning_rate": 2.7590552162300198e-05,
-      "loss": 0.3899,
-      "step": 4866
-    },
-    {
-      "epoch": 2.389431643625192,
-      "grad_norm": 0.423888044247044,
-      "learning_rate": 2.754809707707591e-05,
-      "loss": 0.4366,
-      "step": 4867
-    },
-    {
-      "epoch": 2.3899231950844855,
-      "grad_norm": 0.41197697162827823,
-      "learning_rate": 2.750566946248162e-05,
-      "loss": 0.428,
-      "step": 4868
-    },
-    {
-      "epoch": 2.3904147465437786,
-      "grad_norm": 0.48095076415614696,
-      "learning_rate": 2.746326933460406e-05,
-      "loss": 0.415,
-      "step": 4869
-    },
-    {
-      "epoch": 2.390906298003072,
-      "grad_norm": 0.4138766262651402,
-      "learning_rate": 2.742089670951954e-05,
-      "loss": 0.4291,
-      "step": 4870
-    },
-    {
-      "epoch": 2.3913978494623658,
-      "grad_norm": 0.44173556551645354,
-      "learning_rate": 2.73785516032939e-05,
-      "loss": 0.427,
-      "step": 4871
-    },
-    {
-      "epoch": 2.391889400921659,
-      "grad_norm": 0.46368609749867074,
-      "learning_rate": 2.7336234031982565e-05,
-      "loss": 0.5256,
-      "step": 4872
-    },
-    {
-      "epoch": 2.3923809523809525,
-      "grad_norm": 0.4290467557078167,
-      "learning_rate": 2.7293944011630613e-05,
-      "loss": 0.3894,
-      "step": 4873
-    },
-    {
-      "epoch": 2.3928725038402456,
-      "grad_norm": 0.5258175337993901,
-      "learning_rate": 2.7251681558272456e-05,
-      "loss": 0.4305,
-      "step": 4874
-    },
-    {
-      "epoch": 2.393364055299539,
-      "grad_norm": 0.43304886320230745,
-      "learning_rate": 2.7209446687932316e-05,
-      "loss": 0.452,
-      "step": 4875
-    },
-    {
-      "epoch": 2.393855606758833,
-      "grad_norm": 0.4961713659640867,
-      "learning_rate": 2.716723941662377e-05,
-      "loss": 0.4404,
-      "step": 4876
-    },
-    {
-      "epoch": 2.394347158218126,
-      "grad_norm": 0.4205551150015031,
-      "learning_rate": 2.712505976034999e-05,
-      "loss": 0.4314,
-      "step": 4877
-    },
-    {
-      "epoch": 2.3948387096774195,
-      "grad_norm": 0.40112376247949666,
-      "learning_rate": 2.7082907735103703e-05,
-      "loss": 0.4302,
-      "step": 4878
-    },
-    {
-      "epoch": 2.3953302611367127,
-      "grad_norm": 0.43438082954604723,
-      "learning_rate": 2.7040783356867104e-05,
-      "loss": 0.4065,
-      "step": 4879
-    },
-    {
-      "epoch": 2.3958218125960062,
-      "grad_norm": 0.44615159590308784,
-      "learning_rate": 2.6998686641611968e-05,
-      "loss": 0.4705,
-      "step": 4880
-    },
-    {
-      "epoch": 2.3963133640552994,
-      "grad_norm": 0.4319043997238457,
-      "learning_rate": 2.6956617605299507e-05,
-      "loss": 0.4343,
-      "step": 4881
-    },
-    {
-      "epoch": 2.396804915514593,
-      "grad_norm": 0.409673154227293,
-      "learning_rate": 2.6914576263880554e-05,
-      "loss": 0.4208,
-      "step": 4882
-    },
-    {
-      "epoch": 2.397296466973886,
-      "grad_norm": 0.42954057203393664,
-      "learning_rate": 2.6872562633295273e-05,
-      "loss": 0.4217,
-      "step": 4883
-    },
-    {
-      "epoch": 2.3977880184331797,
-      "grad_norm": 0.39676202093293017,
-      "learning_rate": 2.683057672947349e-05,
-      "loss": 0.4014,
-      "step": 4884
-    },
-    {
-      "epoch": 2.3982795698924733,
-      "grad_norm": 0.45852711552494085,
-      "learning_rate": 2.6788618568334434e-05,
-      "loss": 0.4446,
-      "step": 4885
-    },
-    {
-      "epoch": 2.3987711213517664,
-      "grad_norm": 0.39368541536487933,
-      "learning_rate": 2.6746688165786804e-05,
-      "loss": 0.4047,
-      "step": 4886
-    },
-    {
-      "epoch": 2.39926267281106,
-      "grad_norm": 0.4046290203797557,
-      "learning_rate": 2.670478553772878e-05,
-      "loss": 0.3784,
-      "step": 4887
-    },
-    {
-      "epoch": 2.399754224270353,
-      "grad_norm": 0.3947818780504029,
-      "learning_rate": 2.6662910700048115e-05,
-      "loss": 0.3927,
-      "step": 4888
-    },
-    {
-      "epoch": 2.4002457757296467,
-      "grad_norm": 0.3860916354336387,
-      "learning_rate": 2.662106366862184e-05,
-      "loss": 0.3936,
-      "step": 4889
-    },
-    {
-      "epoch": 2.4007373271889403,
-      "grad_norm": 0.4098875617715916,
-      "learning_rate": 2.6579244459316587e-05,
-      "loss": 0.4115,
-      "step": 4890
-    },
-    {
-      "epoch": 2.4012288786482334,
-      "grad_norm": 0.43718389732951696,
-      "learning_rate": 2.6537453087988372e-05,
-      "loss": 0.4125,
-      "step": 4891
-    },
-    {
-      "epoch": 2.401720430107527,
-      "grad_norm": 0.4886500766381444,
-      "learning_rate": 2.649568957048266e-05,
-      "loss": 0.51,
-      "step": 4892
-    },
-    {
-      "epoch": 2.40221198156682,
-      "grad_norm": 0.4578957117175013,
-      "learning_rate": 2.6453953922634466e-05,
-      "loss": 0.3931,
-      "step": 4893
-    },
-    {
-      "epoch": 2.4027035330261137,
-      "grad_norm": 0.4163101826419494,
-      "learning_rate": 2.6412246160268016e-05,
-      "loss": 0.3926,
-      "step": 4894
-    },
-    {
-      "epoch": 2.4031950844854073,
-      "grad_norm": 0.42597069878871974,
-      "learning_rate": 2.6370566299197175e-05,
-      "loss": 0.4475,
-      "step": 4895
-    },
-    {
-      "epoch": 2.4036866359447004,
-      "grad_norm": 0.39540056045529925,
-      "learning_rate": 2.6328914355225133e-05,
-      "loss": 0.3912,
-      "step": 4896
-    },
-    {
-      "epoch": 2.404178187403994,
-      "grad_norm": 0.43385914069692255,
-      "learning_rate": 2.6287290344144476e-05,
-      "loss": 0.3984,
-      "step": 4897
-    },
-    {
-      "epoch": 2.404669738863287,
-      "grad_norm": 0.42772163840712074,
-      "learning_rate": 2.624569428173723e-05,
-      "loss": 0.4776,
-      "step": 4898
-    },
-    {
-      "epoch": 2.4051612903225807,
-      "grad_norm": 0.40961723983935033,
-      "learning_rate": 2.6204126183774914e-05,
-      "loss": 0.4087,
-      "step": 4899
-    },
-    {
-      "epoch": 2.405652841781874,
-      "grad_norm": 0.40408596146818493,
-      "learning_rate": 2.6162586066018213e-05,
-      "loss": 0.4104,
-      "step": 4900
-    },
-    {
-      "epoch": 2.4061443932411675,
-      "grad_norm": 0.4038564775408092,
-      "learning_rate": 2.6121073944217434e-05,
-      "loss": 0.4047,
-      "step": 4901
-    },
-    {
-      "epoch": 2.4066359447004606,
-      "grad_norm": 0.43283620280296115,
-      "learning_rate": 2.607958983411217e-05,
-      "loss": 0.4211,
-      "step": 4902
-    },
-    {
-      "epoch": 2.407127496159754,
-      "grad_norm": 0.40441583245398277,
-      "learning_rate": 2.6038133751431392e-05,
-      "loss": 0.3765,
-      "step": 4903
-    },
-    {
-      "epoch": 2.4076190476190478,
-      "grad_norm": 0.40547721938521414,
-      "learning_rate": 2.5996705711893453e-05,
-      "loss": 0.4157,
-      "step": 4904
-    },
-    {
-      "epoch": 2.408110599078341,
-      "grad_norm": 0.41947227165717454,
-      "learning_rate": 2.595530573120608e-05,
-      "loss": 0.4461,
-      "step": 4905
-    },
-    {
-      "epoch": 2.4086021505376345,
-      "grad_norm": 0.443470010362414,
-      "learning_rate": 2.591393382506635e-05,
-      "loss": 0.4154,
-      "step": 4906
-    },
-    {
-      "epoch": 2.4090937019969276,
-      "grad_norm": 0.41808356762443494,
-      "learning_rate": 2.587259000916068e-05,
-      "loss": 0.4542,
-      "step": 4907
-    },
-    {
-      "epoch": 2.409585253456221,
-      "grad_norm": 0.3980796046816956,
-      "learning_rate": 2.5831274299164932e-05,
-      "loss": 0.4078,
-      "step": 4908
-    },
-    {
-      "epoch": 2.410076804915515,
-      "grad_norm": 0.4144579759623887,
-      "learning_rate": 2.578998671074414e-05,
-      "loss": 0.4291,
-      "step": 4909
-    },
-    {
-      "epoch": 2.410568356374808,
-      "grad_norm": 0.4202373811110107,
-      "learning_rate": 2.574872725955284e-05,
-      "loss": 0.4373,
-      "step": 4910
-    },
-    {
-      "epoch": 2.4110599078341015,
-      "grad_norm": 0.5234339238283734,
-      "learning_rate": 2.570749596123482e-05,
-      "loss": 0.5084,
-      "step": 4911
-    },
-    {
-      "epoch": 2.4115514592933947,
-      "grad_norm": 0.4141148789563729,
-      "learning_rate": 2.56662928314232e-05,
-      "loss": 0.4108,
-      "step": 4912
-    },
-    {
-      "epoch": 2.4120430107526882,
-      "grad_norm": 0.40809541573997604,
-      "learning_rate": 2.5625117885740425e-05,
-      "loss": 0.401,
-      "step": 4913
-    },
-    {
-      "epoch": 2.412534562211982,
-      "grad_norm": 0.4202059515704432,
-      "learning_rate": 2.5583971139798246e-05,
-      "loss": 0.4238,
-      "step": 4914
-    },
-    {
-      "epoch": 2.413026113671275,
-      "grad_norm": 0.41970170514797916,
-      "learning_rate": 2.5542852609197755e-05,
-      "loss": 0.4534,
-      "step": 4915
-    },
-    {
-      "epoch": 2.4135176651305685,
-      "grad_norm": 0.404516100355974,
-      "learning_rate": 2.5501762309529298e-05,
-      "loss": 0.4114,
-      "step": 4916
-    },
-    {
-      "epoch": 2.4140092165898617,
-      "grad_norm": 0.3975486909589504,
-      "learning_rate": 2.5460700256372548e-05,
-      "loss": 0.4004,
-      "step": 4917
-    },
-    {
-      "epoch": 2.4145007680491553,
-      "grad_norm": 0.41601915799679556,
-      "learning_rate": 2.541966646529643e-05,
-      "loss": 0.4211,
-      "step": 4918
-    },
-    {
-      "epoch": 2.4149923195084484,
-      "grad_norm": 0.41902679270124227,
-      "learning_rate": 2.537866095185929e-05,
-      "loss": 0.4484,
-      "step": 4919
-    },
-    {
-      "epoch": 2.415483870967742,
-      "grad_norm": 0.43915333885554253,
-      "learning_rate": 2.5337683731608496e-05,
-      "loss": 0.4654,
-      "step": 4920
-    },
-    {
-      "epoch": 2.415975422427035,
-      "grad_norm": 0.4228083435989576,
-      "learning_rate": 2.529673482008096e-05,
-      "loss": 0.4482,
-      "step": 4921
-    },
-    {
-      "epoch": 2.4164669738863287,
-      "grad_norm": 0.4086327788975475,
-      "learning_rate": 2.5255814232802677e-05,
-      "loss": 0.4104,
-      "step": 4922
-    },
-    {
-      "epoch": 2.4169585253456223,
-      "grad_norm": 0.45485819020667084,
-      "learning_rate": 2.5214921985288998e-05,
-      "loss": 0.4239,
-      "step": 4923
-    },
-    {
-      "epoch": 2.4174500768049154,
-      "grad_norm": 0.4019267618333613,
-      "learning_rate": 2.517405809304446e-05,
-      "loss": 0.4363,
-      "step": 4924
-    },
-    {
-      "epoch": 2.417941628264209,
-      "grad_norm": 0.414968583317999,
-      "learning_rate": 2.513322257156292e-05,
-      "loss": 0.4244,
-      "step": 4925
-    },
-    {
-      "epoch": 2.418433179723502,
-      "grad_norm": 0.4116238153943015,
-      "learning_rate": 2.5092415436327432e-05,
-      "loss": 0.4268,
-      "step": 4926
-    },
-    {
-      "epoch": 2.4189247311827957,
-      "grad_norm": 0.4327787732980884,
-      "learning_rate": 2.5051636702810254e-05,
-      "loss": 0.3985,
-      "step": 4927
-    },
-    {
-      "epoch": 2.4194162826420893,
-      "grad_norm": 0.42628721591466123,
-      "learning_rate": 2.501088638647302e-05,
-      "loss": 0.4237,
-      "step": 4928
-    },
-    {
-      "epoch": 2.4199078341013824,
-      "grad_norm": 0.4200807107563955,
-      "learning_rate": 2.497016450276637e-05,
-      "loss": 0.4167,
-      "step": 4929
-    },
-    {
-      "epoch": 2.420399385560676,
-      "grad_norm": 0.42466717652922603,
-      "learning_rate": 2.4929471067130363e-05,
-      "loss": 0.4053,
-      "step": 4930
-    },
-    {
-      "epoch": 2.420890937019969,
-      "grad_norm": 0.43364568944423604,
-      "learning_rate": 2.4888806094994167e-05,
-      "loss": 0.4379,
-      "step": 4931
-    },
-    {
-      "epoch": 2.4213824884792627,
-      "grad_norm": 0.4607203868890819,
-      "learning_rate": 2.4848169601776183e-05,
-      "loss": 0.4309,
-      "step": 4932
-    },
-    {
-      "epoch": 2.421874039938556,
-      "grad_norm": 0.38461586947658116,
-      "learning_rate": 2.480756160288402e-05,
-      "loss": 0.3772,
-      "step": 4933
-    },
-    {
-      "epoch": 2.4223655913978495,
-      "grad_norm": 0.461284179658862,
-      "learning_rate": 2.4766982113714454e-05,
-      "loss": 0.4538,
-      "step": 4934
-    },
-    {
-      "epoch": 2.422857142857143,
-      "grad_norm": 0.4336982695994764,
-      "learning_rate": 2.4726431149653496e-05,
-      "loss": 0.3941,
-      "step": 4935
-    },
-    {
-      "epoch": 2.423348694316436,
-      "grad_norm": 0.4253973098327933,
-      "learning_rate": 2.468590872607628e-05,
-      "loss": 0.4254,
-      "step": 4936
-    },
-    {
-      "epoch": 2.4238402457757298,
-      "grad_norm": 0.46035205327708106,
-      "learning_rate": 2.464541485834725e-05,
-      "loss": 0.5255,
-      "step": 4937
-    },
-    {
-      "epoch": 2.424331797235023,
-      "grad_norm": 0.4448176931120402,
-      "learning_rate": 2.4604949561819803e-05,
-      "loss": 0.3926,
-      "step": 4938
-    },
-    {
-      "epoch": 2.4248233486943165,
-      "grad_norm": 0.43535803154222147,
-      "learning_rate": 2.456451285183675e-05,
-      "loss": 0.4444,
-      "step": 4939
-    },
-    {
-      "epoch": 2.4253149001536096,
-      "grad_norm": 0.39337543079748555,
-      "learning_rate": 2.4524104743729838e-05,
-      "loss": 0.373,
-      "step": 4940
-    },
-    {
-      "epoch": 2.425806451612903,
-      "grad_norm": 0.41262199352475254,
-      "learning_rate": 2.4483725252820157e-05,
-      "loss": 0.3906,
-      "step": 4941
-    },
-    {
-      "epoch": 2.426298003072197,
-      "grad_norm": 0.5067843197376108,
-      "learning_rate": 2.444337439441783e-05,
-      "loss": 0.4257,
-      "step": 4942
-    },
-    {
-      "epoch": 2.42678955453149,
-      "grad_norm": 0.42402781111853366,
-      "learning_rate": 2.440305218382216e-05,
-      "loss": 0.4183,
-      "step": 4943
-    },
-    {
-      "epoch": 2.4272811059907835,
-      "grad_norm": 0.3952064600763445,
-      "learning_rate": 2.436275863632156e-05,
-      "loss": 0.3582,
-      "step": 4944
-    },
-    {
-      "epoch": 2.4277726574500766,
-      "grad_norm": 0.4141091231354883,
-      "learning_rate": 2.432249376719369e-05,
-      "loss": 0.4621,
-      "step": 4945
-    },
-    {
-      "epoch": 2.4282642089093702,
-      "grad_norm": 0.4170927380952518,
-      "learning_rate": 2.4282257591705127e-05,
-      "loss": 0.418,
-      "step": 4946
-    },
-    {
-      "epoch": 2.428755760368664,
-      "grad_norm": 0.43790661875108805,
-      "learning_rate": 2.424205012511178e-05,
-      "loss": 0.4095,
-      "step": 4947
-    },
-    {
-      "epoch": 2.429247311827957,
-      "grad_norm": 0.46353406479874043,
-      "learning_rate": 2.420187138265856e-05,
-      "loss": 0.4313,
-      "step": 4948
-    },
-    {
-      "epoch": 2.4297388632872505,
-      "grad_norm": 0.4070085035504352,
-      "learning_rate": 2.4161721379579493e-05,
-      "loss": 0.4408,
-      "step": 4949
-    },
-    {
-      "epoch": 2.4302304147465437,
-      "grad_norm": 0.4550178445002215,
-      "learning_rate": 2.412160013109773e-05,
-      "loss": 0.4178,
-      "step": 4950
-    },
-    {
-      "epoch": 2.4307219662058372,
-      "grad_norm": 0.3944676701943834,
-      "learning_rate": 2.4081507652425537e-05,
-      "loss": 0.4008,
-      "step": 4951
-    },
-    {
-      "epoch": 2.4312135176651304,
-      "grad_norm": 0.3870608732317602,
-      "learning_rate": 2.404144395876422e-05,
-      "loss": 0.3754,
-      "step": 4952
-    },
-    {
-      "epoch": 2.431705069124424,
-      "grad_norm": 0.4109331386351621,
-      "learning_rate": 2.4001409065304182e-05,
-      "loss": 0.4039,
-      "step": 4953
-    },
-    {
-      "epoch": 2.4321966205837176,
-      "grad_norm": 0.38205813098274866,
-      "learning_rate": 2.396140298722501e-05,
-      "loss": 0.424,
-      "step": 4954
-    },
-    {
-      "epoch": 2.4326881720430107,
-      "grad_norm": 0.46074408777096326,
-      "learning_rate": 2.392142573969516e-05,
-      "loss": 0.4483,
-      "step": 4955
-    },
-    {
-      "epoch": 2.4331797235023043,
-      "grad_norm": 0.45619132072538193,
-      "learning_rate": 2.388147733787237e-05,
-      "loss": 0.4124,
-      "step": 4956
-    },
-    {
-      "epoch": 2.4336712749615974,
-      "grad_norm": 0.43533052904614056,
-      "learning_rate": 2.3841557796903323e-05,
-      "loss": 0.4377,
-      "step": 4957
-    },
-    {
-      "epoch": 2.434162826420891,
-      "grad_norm": 0.45549689420536416,
-      "learning_rate": 2.3801667131923778e-05,
-      "loss": 0.4424,
-      "step": 4958
-    },
-    {
-      "epoch": 2.434654377880184,
-      "grad_norm": 0.41206127919664676,
-      "learning_rate": 2.376180535805854e-05,
-      "loss": 0.4631,
-      "step": 4959
-    },
-    {
-      "epoch": 2.4351459293394777,
-      "grad_norm": 0.4358014767790808,
-      "learning_rate": 2.3721972490421486e-05,
-      "loss": 0.4162,
-      "step": 4960
-    },
-    {
-      "epoch": 2.4356374807987713,
-      "grad_norm": 0.43791299923490723,
-      "learning_rate": 2.3682168544115514e-05,
-      "loss": 0.449,
-      "step": 4961
-    },
-    {
-      "epoch": 2.4361290322580644,
-      "grad_norm": 0.40285694468060407,
-      "learning_rate": 2.3642393534232543e-05,
-      "loss": 0.436,
-      "step": 4962
-    },
-    {
-      "epoch": 2.436620583717358,
-      "grad_norm": 0.41769793235393443,
-      "learning_rate": 2.3602647475853567e-05,
-      "loss": 0.442,
-      "step": 4963
-    },
-    {
-      "epoch": 2.437112135176651,
-      "grad_norm": 0.4412574100483954,
-      "learning_rate": 2.3562930384048533e-05,
-      "loss": 0.4162,
-      "step": 4964
-    },
-    {
-      "epoch": 2.4376036866359447,
-      "grad_norm": 0.40350597371130936,
-      "learning_rate": 2.352324227387651e-05,
-      "loss": 0.4368,
-      "step": 4965
-    },
-    {
-      "epoch": 2.4380952380952383,
-      "grad_norm": 0.43548997015869667,
-      "learning_rate": 2.3483583160385435e-05,
-      "loss": 0.4225,
-      "step": 4966
-    },
-    {
-      "epoch": 2.4385867895545315,
-      "grad_norm": 0.38334190191802703,
-      "learning_rate": 2.34439530586124e-05,
-      "loss": 0.3845,
-      "step": 4967
-    },
-    {
-      "epoch": 2.439078341013825,
-      "grad_norm": 0.41171341418336127,
-      "learning_rate": 2.3404351983583385e-05,
-      "loss": 0.4179,
-      "step": 4968
-    },
-    {
-      "epoch": 2.439569892473118,
-      "grad_norm": 0.4196890334736673,
-      "learning_rate": 2.336477995031342e-05,
-      "loss": 0.4383,
-      "step": 4969
-    },
-    {
-      "epoch": 2.4400614439324118,
-      "grad_norm": 0.45342298310994344,
-      "learning_rate": 2.332523697380652e-05,
-      "loss": 0.3981,
-      "step": 4970
-    },
-    {
-      "epoch": 2.440552995391705,
-      "grad_norm": 0.41924147643358856,
-      "learning_rate": 2.3285723069055644e-05,
-      "loss": 0.4414,
-      "step": 4971
-    },
-    {
-      "epoch": 2.4410445468509985,
-      "grad_norm": 0.4494783755277245,
-      "learning_rate": 2.3246238251042783e-05,
-      "loss": 0.4593,
-      "step": 4972
-    },
-    {
-      "epoch": 2.4415360983102916,
-      "grad_norm": 0.43869194137597284,
-      "learning_rate": 2.3206782534738825e-05,
-      "loss": 0.4298,
-      "step": 4973
-    },
-    {
-      "epoch": 2.442027649769585,
-      "grad_norm": 0.4052401911466543,
-      "learning_rate": 2.3167355935103775e-05,
-      "loss": 0.4286,
-      "step": 4974
-    },
-    {
-      "epoch": 2.4425192012288788,
-      "grad_norm": 0.418322005014125,
-      "learning_rate": 2.3127958467086376e-05,
-      "loss": 0.4126,
-      "step": 4975
-    },
-    {
-      "epoch": 2.443010752688172,
-      "grad_norm": 0.4400683935885187,
-      "learning_rate": 2.3088590145624524e-05,
-      "loss": 0.4453,
-      "step": 4976
-    },
-    {
-      "epoch": 2.4435023041474655,
-      "grad_norm": 0.44435139050927097,
-      "learning_rate": 2.3049250985644956e-05,
-      "loss": 0.4142,
-      "step": 4977
-    },
-    {
-      "epoch": 2.4439938556067586,
-      "grad_norm": 0.40584398439709796,
-      "learning_rate": 2.3009941002063386e-05,
-      "loss": 0.3947,
-      "step": 4978
-    },
-    {
-      "epoch": 2.444485407066052,
-      "grad_norm": 0.41334247645487726,
-      "learning_rate": 2.2970660209784468e-05,
-      "loss": 0.4072,
-      "step": 4979
-    },
-    {
-      "epoch": 2.444976958525346,
-      "grad_norm": 0.41956178081404166,
-      "learning_rate": 2.2931408623701768e-05,
-      "loss": 0.4147,
-      "step": 4980
-    },
-    {
-      "epoch": 2.445468509984639,
-      "grad_norm": 0.4500041655607618,
-      "learning_rate": 2.289218625869779e-05,
-      "loss": 0.4524,
-      "step": 4981
-    },
-    {
-      "epoch": 2.4459600614439325,
-      "grad_norm": 0.42882187483367806,
-      "learning_rate": 2.2852993129643953e-05,
-      "loss": 0.4342,
-      "step": 4982
-    },
-    {
-      "epoch": 2.4464516129032257,
-      "grad_norm": 0.39609811982163823,
-      "learning_rate": 2.2813829251400654e-05,
-      "loss": 0.4188,
-      "step": 4983
-    },
-    {
-      "epoch": 2.4469431643625192,
-      "grad_norm": 0.39380472899425994,
-      "learning_rate": 2.277469463881704e-05,
-      "loss": 0.3837,
-      "step": 4984
-    },
-    {
-      "epoch": 2.447434715821813,
-      "grad_norm": 0.39798016820271653,
-      "learning_rate": 2.273558930673135e-05,
-      "loss": 0.4118,
-      "step": 4985
-    },
-    {
-      "epoch": 2.447926267281106,
-      "grad_norm": 0.41092609830192445,
-      "learning_rate": 2.269651326997062e-05,
-      "loss": 0.4136,
-      "step": 4986
-    },
-    {
-      "epoch": 2.4484178187403995,
-      "grad_norm": 0.4307680149634275,
-      "learning_rate": 2.265746654335078e-05,
-      "loss": 0.4662,
-      "step": 4987
-    },
-    {
-      "epoch": 2.4489093701996927,
-      "grad_norm": 0.42927589692719964,
-      "learning_rate": 2.2618449141676666e-05,
-      "loss": 0.4253,
-      "step": 4988
-    },
-    {
-      "epoch": 2.4494009216589863,
-      "grad_norm": 0.40368383133013636,
-      "learning_rate": 2.2579461079741982e-05,
-      "loss": 0.427,
-      "step": 4989
-    },
-    {
-      "epoch": 2.4498924731182794,
-      "grad_norm": 0.41741848226983125,
-      "learning_rate": 2.2540502372329298e-05,
-      "loss": 0.4367,
-      "step": 4990
-    },
-    {
-      "epoch": 2.450384024577573,
-      "grad_norm": 0.42486087819110996,
-      "learning_rate": 2.2501573034210154e-05,
-      "loss": 0.4553,
-      "step": 4991
-    },
-    {
-      "epoch": 2.450875576036866,
-      "grad_norm": 0.40558882107282096,
-      "learning_rate": 2.2462673080144747e-05,
-      "loss": 0.3978,
-      "step": 4992
-    },
-    {
-      "epoch": 2.4513671274961597,
-      "grad_norm": 0.4294212555649955,
-      "learning_rate": 2.2423802524882366e-05,
-      "loss": 0.4097,
-      "step": 4993
-    },
-    {
-      "epoch": 2.4518586789554533,
-      "grad_norm": 0.42746809790960366,
-      "learning_rate": 2.238496138316101e-05,
-      "loss": 0.4217,
-      "step": 4994
-    },
-    {
-      "epoch": 2.4523502304147464,
-      "grad_norm": 0.4060989520545923,
-      "learning_rate": 2.234614966970754e-05,
-      "loss": 0.4027,
-      "step": 4995
-    },
-    {
-      "epoch": 2.45284178187404,
-      "grad_norm": 0.45248182842262413,
-      "learning_rate": 2.230736739923771e-05,
-      "loss": 0.4473,
-      "step": 4996
-    },
-    {
-      "epoch": 2.453333333333333,
-      "grad_norm": 0.3980380260404687,
-      "learning_rate": 2.2268614586456062e-05,
-      "loss": 0.4282,
-      "step": 4997
-    },
-    {
-      "epoch": 2.4538248847926267,
-      "grad_norm": 0.4112502253597491,
-      "learning_rate": 2.222989124605599e-05,
-      "loss": 0.4188,
-      "step": 4998
-    },
-    {
-      "epoch": 2.4543164362519203,
-      "grad_norm": 0.41725701212807187,
-      "learning_rate": 2.2191197392719688e-05,
-      "loss": 0.4205,
-      "step": 4999
-    },
-    {
-      "epoch": 2.4548079877112134,
-      "grad_norm": 0.40087848677582605,
-      "learning_rate": 2.2152533041118275e-05,
-      "loss": 0.3713,
-      "step": 5000
-    },
-    {
-      "epoch": 2.455299539170507,
-      "grad_norm": 0.4170737879851873,
-      "learning_rate": 2.2113898205911487e-05,
-      "loss": 0.4103,
-      "step": 5001
-    },
-    {
-      "epoch": 2.4557910906298,
-      "grad_norm": 0.3938070235299884,
-      "learning_rate": 2.207529290174808e-05,
-      "loss": 0.3661,
-      "step": 5002
-    },
-    {
-      "epoch": 2.4562826420890937,
-      "grad_norm": 0.39246320315860966,
-      "learning_rate": 2.2036717143265474e-05,
-      "loss": 0.3649,
-      "step": 5003
-    },
-    {
-      "epoch": 2.4567741935483873,
-      "grad_norm": 0.4113108642558232,
-      "learning_rate": 2.1998170945089923e-05,
-      "loss": 0.4207,
-      "step": 5004
-    },
-    {
-      "epoch": 2.4572657450076805,
-      "grad_norm": 0.44280160720408823,
-      "learning_rate": 2.1959654321836497e-05,
-      "loss": 0.4618,
-      "step": 5005
-    },
-    {
-      "epoch": 2.457757296466974,
-      "grad_norm": 0.4350799927654252,
-      "learning_rate": 2.1921167288109034e-05,
-      "loss": 0.4383,
-      "step": 5006
-    },
-    {
-      "epoch": 2.458248847926267,
-      "grad_norm": 0.4096536750530943,
-      "learning_rate": 2.188270985850015e-05,
-      "loss": 0.426,
-      "step": 5007
-    },
-    {
-      "epoch": 2.4587403993855608,
-      "grad_norm": 0.47453302718344237,
-      "learning_rate": 2.18442820475912e-05,
-      "loss": 0.4694,
-      "step": 5008
-    },
-    {
-      "epoch": 2.459231950844854,
-      "grad_norm": 0.4447458622489744,
-      "learning_rate": 2.1805883869952447e-05,
-      "loss": 0.4309,
-      "step": 5009
-    },
-    {
-      "epoch": 2.4597235023041475,
-      "grad_norm": 0.41288963387212224,
-      "learning_rate": 2.1767515340142708e-05,
-      "loss": 0.4322,
-      "step": 5010
-    },
-    {
-      "epoch": 2.4602150537634406,
-      "grad_norm": 0.3970000208134351,
-      "learning_rate": 2.172917647270977e-05,
-      "loss": 0.3888,
-      "step": 5011
-    },
-    {
-      "epoch": 2.460706605222734,
-      "grad_norm": 0.3975775174626306,
-      "learning_rate": 2.1690867282189974e-05,
-      "loss": 0.3799,
-      "step": 5012
-    },
-    {
-      "epoch": 2.461198156682028,
-      "grad_norm": 0.43998952557391197,
-      "learning_rate": 2.165258778310859e-05,
-      "loss": 0.4556,
-      "step": 5013
-    },
-    {
-      "epoch": 2.461689708141321,
-      "grad_norm": 0.4266765045152995,
-      "learning_rate": 2.1614337989979527e-05,
-      "loss": 0.482,
-      "step": 5014
-    },
-    {
-      "epoch": 2.4621812596006145,
-      "grad_norm": 0.42075334461583025,
-      "learning_rate": 2.1576117917305448e-05,
-      "loss": 0.4073,
-      "step": 5015
-    },
-    {
-      "epoch": 2.4626728110599077,
-      "grad_norm": 0.8023675217745048,
-      "learning_rate": 2.1537927579577754e-05,
-      "loss": 0.4906,
-      "step": 5016
-    },
-    {
-      "epoch": 2.4631643625192012,
-      "grad_norm": 0.4306678896059732,
-      "learning_rate": 2.149976699127657e-05,
-      "loss": 0.4357,
-      "step": 5017
-    },
-    {
-      "epoch": 2.463655913978495,
-      "grad_norm": 0.4093322198679003,
-      "learning_rate": 2.1461636166870735e-05,
-      "loss": 0.3864,
-      "step": 5018
-    },
-    {
-      "epoch": 2.464147465437788,
-      "grad_norm": 0.4519684061952061,
-      "learning_rate": 2.1423535120817796e-05,
-      "loss": 0.438,
-      "step": 5019
-    },
-    {
-      "epoch": 2.4646390168970815,
-      "grad_norm": 0.430634076064964,
-      "learning_rate": 2.1385463867564093e-05,
-      "loss": 0.4399,
-      "step": 5020
-    },
-    {
-      "epoch": 2.4651305683563747,
-      "grad_norm": 0.3930196163060272,
-      "learning_rate": 2.1347422421544495e-05,
-      "loss": 0.3756,
-      "step": 5021
-    },
-    {
-      "epoch": 2.4656221198156683,
-      "grad_norm": 0.4506395296409869,
-      "learning_rate": 2.1309410797182748e-05,
-      "loss": 0.5219,
-      "step": 5022
-    },
-    {
-      "epoch": 2.466113671274962,
-      "grad_norm": 0.4139398596955898,
-      "learning_rate": 2.1271429008891207e-05,
-      "loss": 0.4328,
-      "step": 5023
-    },
-    {
-      "epoch": 2.466605222734255,
-      "grad_norm": 0.43207514947020714,
-      "learning_rate": 2.123347707107092e-05,
-      "loss": 0.4358,
-      "step": 5024
-    },
-    {
-      "epoch": 2.4670967741935486,
-      "grad_norm": 0.4162246279419766,
-      "learning_rate": 2.11955549981116e-05,
-      "loss": 0.424,
-      "step": 5025
-    },
-    {
-      "epoch": 2.4675883256528417,
-      "grad_norm": 0.41748950764840775,
-      "learning_rate": 2.1157662804391688e-05,
-      "loss": 0.4327,
-      "step": 5026
-    },
-    {
-      "epoch": 2.4680798771121353,
-      "grad_norm": 0.43412189455031686,
-      "learning_rate": 2.1119800504278243e-05,
-      "loss": 0.4454,
-      "step": 5027
-    },
-    {
-      "epoch": 2.4685714285714284,
-      "grad_norm": 0.4022725907507483,
-      "learning_rate": 2.1081968112127004e-05,
-      "loss": 0.4442,
-      "step": 5028
-    },
-    {
-      "epoch": 2.469062980030722,
-      "grad_norm": 0.4098521047997809,
-      "learning_rate": 2.1044165642282443e-05,
-      "loss": 0.4076,
-      "step": 5029
-    },
-    {
-      "epoch": 2.469554531490015,
-      "grad_norm": 0.4404444318518348,
-      "learning_rate": 2.1006393109077525e-05,
-      "loss": 0.4215,
-      "step": 5030
-    },
-    {
-      "epoch": 2.4700460829493087,
-      "grad_norm": 0.3985616449976521,
-      "learning_rate": 2.0968650526834044e-05,
-      "loss": 0.4096,
-      "step": 5031
-    },
-    {
-      "epoch": 2.4705376344086023,
-      "grad_norm": 0.42915733479069945,
-      "learning_rate": 2.0930937909862315e-05,
-      "loss": 0.43,
-      "step": 5032
-    },
-    {
-      "epoch": 2.4710291858678954,
-      "grad_norm": 0.42843053935322273,
-      "learning_rate": 2.0893255272461353e-05,
-      "loss": 0.4468,
-      "step": 5033
-    },
-    {
-      "epoch": 2.471520737327189,
-      "grad_norm": 0.416855017295221,
-      "learning_rate": 2.0855602628918747e-05,
-      "loss": 0.402,
-      "step": 5034
-    },
-    {
-      "epoch": 2.472012288786482,
-      "grad_norm": 0.4420058552315824,
-      "learning_rate": 2.0817979993510828e-05,
-      "loss": 0.4171,
-      "step": 5035
-    },
-    {
-      "epoch": 2.4725038402457757,
-      "grad_norm": 0.4178801901237165,
-      "learning_rate": 2.0780387380502375e-05,
-      "loss": 0.4329,
-      "step": 5036
-    },
-    {
-      "epoch": 2.4729953917050693,
-      "grad_norm": 0.411539190274043,
-      "learning_rate": 2.0742824804146978e-05,
-      "loss": 0.4401,
-      "step": 5037
-    },
-    {
-      "epoch": 2.4734869431643625,
-      "grad_norm": 0.44661260544536424,
-      "learning_rate": 2.070529227868665e-05,
-      "loss": 0.4192,
-      "step": 5038
-    },
-    {
-      "epoch": 2.473978494623656,
-      "grad_norm": 0.41256879609161523,
-      "learning_rate": 2.066778981835218e-05,
-      "loss": 0.4402,
-      "step": 5039
-    },
-    {
-      "epoch": 2.474470046082949,
-      "grad_norm": 0.42217334590518796,
-      "learning_rate": 2.0630317437362834e-05,
-      "loss": 0.4154,
-      "step": 5040
-    },
-    {
-      "epoch": 2.4749615975422428,
-      "grad_norm": 0.43713094034517735,
-      "learning_rate": 2.059287514992655e-05,
-      "loss": 0.4334,
-      "step": 5041
-    },
-    {
-      "epoch": 2.4754531490015363,
-      "grad_norm": 0.4104594007405608,
-      "learning_rate": 2.05554629702398e-05,
-      "loss": 0.4299,
-      "step": 5042
-    },
-    {
-      "epoch": 2.4759447004608295,
-      "grad_norm": 0.4178253665500614,
-      "learning_rate": 2.051808091248768e-05,
-      "loss": 0.4353,
-      "step": 5043
-    },
-    {
-      "epoch": 2.476436251920123,
-      "grad_norm": 0.3866877583558031,
-      "learning_rate": 2.0480728990843833e-05,
-      "loss": 0.4092,
-      "step": 5044
-    },
-    {
-      "epoch": 2.476927803379416,
-      "grad_norm": 0.42417960596452214,
-      "learning_rate": 2.044340721947049e-05,
-      "loss": 0.4131,
-      "step": 5045
-    },
-    {
-      "epoch": 2.47741935483871,
-      "grad_norm": 0.4258239623041988,
-      "learning_rate": 2.040611561251853e-05,
-      "loss": 0.4147,
-      "step": 5046
-    },
-    {
-      "epoch": 2.477910906298003,
-      "grad_norm": 0.3956867903698109,
-      "learning_rate": 2.0368854184127183e-05,
-      "loss": 0.386,
-      "step": 5047
-    },
-    {
-      "epoch": 2.4784024577572965,
-      "grad_norm": 0.43681721120071376,
-      "learning_rate": 2.033162294842449e-05,
-      "loss": 0.4384,
-      "step": 5048
-    },
-    {
-      "epoch": 2.4788940092165896,
-      "grad_norm": 0.4070485063242609,
-      "learning_rate": 2.0294421919526873e-05,
-      "loss": 0.4249,
-      "step": 5049
-    },
-    {
-      "epoch": 2.4793855606758832,
-      "grad_norm": 0.3981539220893719,
-      "learning_rate": 2.0257251111539365e-05,
-      "loss": 0.4029,
-      "step": 5050
-    },
-    {
-      "epoch": 2.479877112135177,
-      "grad_norm": 0.4231654410800758,
-      "learning_rate": 2.0220110538555536e-05,
-      "loss": 0.4184,
-      "step": 5051
-    },
-    {
-      "epoch": 2.48036866359447,
-      "grad_norm": 0.41648044030596626,
-      "learning_rate": 2.0183000214657478e-05,
-      "loss": 0.4409,
-      "step": 5052
-    },
-    {
-      "epoch": 2.4808602150537635,
-      "grad_norm": 0.45193386894941884,
-      "learning_rate": 2.0145920153915808e-05,
-      "loss": 0.438,
-      "step": 5053
-    },
-    {
-      "epoch": 2.4813517665130567,
-      "grad_norm": 0.4392599785531329,
-      "learning_rate": 2.0108870370389687e-05,
-      "loss": 0.4267,
-      "step": 5054
-    },
-    {
-      "epoch": 2.4818433179723502,
-      "grad_norm": 0.45213477948595066,
-      "learning_rate": 2.0071850878126842e-05,
-      "loss": 0.4416,
-      "step": 5055
-    },
-    {
-      "epoch": 2.482334869431644,
-      "grad_norm": 0.40661863824003847,
-      "learning_rate": 2.0034861691163374e-05,
-      "loss": 0.4043,
-      "step": 5056
-    },
-    {
-      "epoch": 2.482826420890937,
-      "grad_norm": 0.40849375501249074,
-      "learning_rate": 1.9997902823524072e-05,
-      "loss": 0.4292,
-      "step": 5057
-    },
-    {
-      "epoch": 2.4833179723502306,
-      "grad_norm": 0.4119444960928005,
-      "learning_rate": 1.996097428922209e-05,
-      "loss": 0.3844,
-      "step": 5058
-    },
-    {
-      "epoch": 2.4838095238095237,
-      "grad_norm": 0.41662555624285125,
-      "learning_rate": 1.992407610225915e-05,
-      "loss": 0.4067,
-      "step": 5059
-    },
-    {
-      "epoch": 2.4843010752688173,
-      "grad_norm": 0.40819255656770076,
-      "learning_rate": 1.9887208276625446e-05,
-      "loss": 0.4199,
-      "step": 5060
-    },
-    {
-      "epoch": 2.4847926267281104,
-      "grad_norm": 0.41579911013234894,
-      "learning_rate": 1.9850370826299658e-05,
-      "loss": 0.4456,
-      "step": 5061
-    },
-    {
-      "epoch": 2.485284178187404,
-      "grad_norm": 0.38752784168055804,
-      "learning_rate": 1.9813563765248953e-05,
-      "loss": 0.3971,
-      "step": 5062
-    },
-    {
-      "epoch": 2.4857757296466976,
-      "grad_norm": 0.5381697267195698,
-      "learning_rate": 1.9776787107428997e-05,
-      "loss": 0.4279,
-      "step": 5063
-    },
-    {
-      "epoch": 2.4862672811059907,
-      "grad_norm": 0.41004036314598546,
-      "learning_rate": 1.9740040866783892e-05,
-      "loss": 0.3947,
-      "step": 5064
-    },
-    {
-      "epoch": 2.4867588325652843,
-      "grad_norm": 0.4254939276978934,
-      "learning_rate": 1.9703325057246203e-05,
-      "loss": 0.4036,
-      "step": 5065
-    },
-    {
-      "epoch": 2.4872503840245774,
-      "grad_norm": 0.4488112548607266,
-      "learning_rate": 1.9666639692737055e-05,
-      "loss": 0.436,
-      "step": 5066
-    },
-    {
-      "epoch": 2.487741935483871,
-      "grad_norm": 0.3918922658874846,
-      "learning_rate": 1.9629984787165844e-05,
-      "loss": 0.4034,
-      "step": 5067
-    },
-    {
-      "epoch": 2.488233486943164,
-      "grad_norm": 0.4032443888329688,
-      "learning_rate": 1.959336035443061e-05,
-      "loss": 0.4216,
-      "step": 5068
-    },
-    {
-      "epoch": 2.4887250384024577,
-      "grad_norm": 0.42438477756228404,
-      "learning_rate": 1.9556766408417738e-05,
-      "loss": 0.3937,
-      "step": 5069
-    },
-    {
-      "epoch": 2.4892165898617513,
-      "grad_norm": 0.39924274987500186,
-      "learning_rate": 1.9520202963002043e-05,
-      "loss": 0.3551,
-      "step": 5070
-    },
-    {
-      "epoch": 2.4897081413210445,
-      "grad_norm": 0.4039276954336843,
-      "learning_rate": 1.948367003204683e-05,
-      "loss": 0.3552,
-      "step": 5071
-    },
-    {
-      "epoch": 2.490199692780338,
-      "grad_norm": 0.41615304776373657,
-      "learning_rate": 1.9447167629403805e-05,
-      "loss": 0.4487,
-      "step": 5072
-    },
-    {
-      "epoch": 2.490691244239631,
-      "grad_norm": 0.40021005414883415,
-      "learning_rate": 1.9410695768913077e-05,
-      "loss": 0.3788,
-      "step": 5073
-    },
-    {
-      "epoch": 2.4911827956989248,
-      "grad_norm": 0.444686626726618,
-      "learning_rate": 1.9374254464403207e-05,
-      "loss": 0.4463,
-      "step": 5074
-    },
-    {
-      "epoch": 2.4916743471582183,
-      "grad_norm": 0.42130227798357583,
-      "learning_rate": 1.9337843729691217e-05,
-      "loss": 0.4238,
-      "step": 5075
-    },
-    {
-      "epoch": 2.4921658986175115,
-      "grad_norm": 0.3853644670730329,
-      "learning_rate": 1.930146357858238e-05,
-      "loss": 0.379,
-      "step": 5076
-    },
-    {
-      "epoch": 2.492657450076805,
-      "grad_norm": 0.42086831424076065,
-      "learning_rate": 1.926511402487057e-05,
-      "loss": 0.4184,
-      "step": 5077
-    },
-    {
-      "epoch": 2.493149001536098,
-      "grad_norm": 0.4214061439423063,
-      "learning_rate": 1.9228795082337926e-05,
-      "loss": 0.433,
-      "step": 5078
-    },
-    {
-      "epoch": 2.493640552995392,
-      "grad_norm": 0.3994475477371375,
-      "learning_rate": 1.9192506764755036e-05,
-      "loss": 0.4115,
-      "step": 5079
-    },
-    {
-      "epoch": 2.494132104454685,
-      "grad_norm": 0.42562824089908347,
-      "learning_rate": 1.9156249085880818e-05,
-      "loss": 0.4017,
-      "step": 5080
-    },
-    {
-      "epoch": 2.4946236559139785,
-      "grad_norm": 0.39344823805678997,
-      "learning_rate": 1.9120022059462706e-05,
-      "loss": 0.4285,
-      "step": 5081
-    },
-    {
-      "epoch": 2.495115207373272,
-      "grad_norm": 0.39159133762471077,
-      "learning_rate": 1.9083825699236323e-05,
-      "loss": 0.3667,
-      "step": 5082
-    },
-    {
-      "epoch": 2.495606758832565,
-      "grad_norm": 0.42521385079766616,
-      "learning_rate": 1.9047660018925828e-05,
-      "loss": 0.4225,
-      "step": 5083
-    },
-    {
-      "epoch": 2.496098310291859,
-      "grad_norm": 0.4159750219904371,
-      "learning_rate": 1.901152503224366e-05,
-      "loss": 0.4106,
-      "step": 5084
-    },
-    {
-      "epoch": 2.496589861751152,
-      "grad_norm": 0.38528449506553747,
-      "learning_rate": 1.897542075289066e-05,
-      "loss": 0.3769,
-      "step": 5085
-    },
-    {
-      "epoch": 2.496589861751152,
-      "eval_loss": 0.8140223026275635,
-      "eval_runtime": 6658.8517,
-      "eval_samples_per_second": 4.28,
-      "eval_steps_per_second": 2.14,
-      "step": 5085
-    },
-    {
-      "epoch": 2.4970814132104455,
-      "grad_norm": 0.4426154703832519,
-      "learning_rate": 1.893934719455599e-05,
-      "loss": 0.4107,
-      "step": 5086
-    },
-    {
-      "epoch": 2.4975729646697387,
-      "grad_norm": 0.39328617200473986,
-      "learning_rate": 1.8903304370917208e-05,
-      "loss": 0.422,
-      "step": 5087
-    },
-    {
-      "epoch": 2.4980645161290322,
-      "grad_norm": 0.39941817138053703,
-      "learning_rate": 1.886729229564016e-05,
-      "loss": 0.3807,
-      "step": 5088
-    },
-    {
-      "epoch": 2.498556067588326,
-      "grad_norm": 0.4038335550501196,
-      "learning_rate": 1.8831310982379103e-05,
-      "loss": 0.3992,
-      "step": 5089
-    },
-    {
-      "epoch": 2.499047619047619,
-      "grad_norm": 0.401618039340355,
-      "learning_rate": 1.8795360444776577e-05,
-      "loss": 0.4176,
-      "step": 5090
-    },
-    {
-      "epoch": 2.4995391705069125,
-      "grad_norm": 0.427198751051607,
-      "learning_rate": 1.875944069646346e-05,
-      "loss": 0.4183,
-      "step": 5091
-    },
-    {
-      "epoch": 2.5000307219662057,
-      "grad_norm": 0.40657213362173045,
-      "learning_rate": 1.872355175105902e-05,
-      "loss": 0.4765,
-      "step": 5092
-    },
-    {
-      "epoch": 2.5005222734254993,
-      "grad_norm": 0.40959209087807935,
-      "learning_rate": 1.8687693622170723e-05,
-      "loss": 0.4245,
-      "step": 5093
-    },
-    {
-      "epoch": 2.501013824884793,
-      "grad_norm": 0.4528650314771083,
-      "learning_rate": 1.8651866323394473e-05,
-      "loss": 0.4205,
-      "step": 5094
-    },
-    {
-      "epoch": 2.501505376344086,
-      "grad_norm": 0.45053998340422075,
-      "learning_rate": 1.861606986831441e-05,
-      "loss": 0.4499,
-      "step": 5095
-    },
-    {
-      "epoch": 2.5019969278033796,
-      "grad_norm": 0.3973370539956065,
-      "learning_rate": 1.8580304270503024e-05,
-      "loss": 0.3997,
-      "step": 5096
-    },
-    {
-      "epoch": 2.5024884792626727,
-      "grad_norm": 0.387607528425279,
-      "learning_rate": 1.8544569543521063e-05,
-      "loss": 0.3998,
-      "step": 5097
-    },
-    {
-      "epoch": 2.5029800307219663,
-      "grad_norm": 0.4430245555765779,
-      "learning_rate": 1.8508865700917598e-05,
-      "loss": 0.4339,
-      "step": 5098
-    },
-    {
-      "epoch": 2.50347158218126,
-      "grad_norm": 0.4334586784959611,
-      "learning_rate": 1.8473192756229985e-05,
-      "loss": 0.4558,
-      "step": 5099
-    },
-    {
-      "epoch": 2.503963133640553,
-      "grad_norm": 0.42461552764421445,
-      "learning_rate": 1.843755072298383e-05,
-      "loss": 0.3974,
-      "step": 5100
-    },
-    {
-      "epoch": 2.504454685099846,
-      "grad_norm": 0.42644282145434476,
-      "learning_rate": 1.8401939614693142e-05,
-      "loss": 0.4605,
-      "step": 5101
-    },
-    {
-      "epoch": 2.5049462365591397,
-      "grad_norm": 0.4014263088473555,
-      "learning_rate": 1.836635944485999e-05,
-      "loss": 0.3903,
-      "step": 5102
-    },
-    {
-      "epoch": 2.5054377880184333,
-      "grad_norm": 0.4595881033592045,
-      "learning_rate": 1.8330810226974928e-05,
-      "loss": 0.4364,
-      "step": 5103
-    },
-    {
-      "epoch": 2.5059293394777264,
-      "grad_norm": 0.46305780211307357,
-      "learning_rate": 1.829529197451666e-05,
-      "loss": 0.4295,
-      "step": 5104
-    },
-    {
-      "epoch": 2.50642089093702,
-      "grad_norm": 0.46629544746215856,
-      "learning_rate": 1.8259804700952166e-05,
-      "loss": 0.4377,
-      "step": 5105
-    },
-    {
-      "epoch": 2.506912442396313,
-      "grad_norm": 0.4095067468753335,
-      "learning_rate": 1.8224348419736693e-05,
-      "loss": 0.4142,
-      "step": 5106
-    },
-    {
-      "epoch": 2.5074039938556067,
-      "grad_norm": 0.4115643732643572,
-      "learning_rate": 1.8188923144313707e-05,
-      "loss": 0.4298,
-      "step": 5107
-    },
-    {
-      "epoch": 2.5078955453149003,
-      "grad_norm": 0.41417800237491703,
-      "learning_rate": 1.815352888811498e-05,
-      "loss": 0.4053,
-      "step": 5108
-    },
-    {
-      "epoch": 2.5083870967741935,
-      "grad_norm": 0.4420769583010539,
-      "learning_rate": 1.811816566456045e-05,
-      "loss": 0.4688,
-      "step": 5109
-    },
-    {
-      "epoch": 2.508878648233487,
-      "grad_norm": 0.4145580955031008,
-      "learning_rate": 1.8082833487058337e-05,
-      "loss": 0.4236,
-      "step": 5110
-    },
-    {
-      "epoch": 2.50937019969278,
-      "grad_norm": 0.4149205274435076,
-      "learning_rate": 1.804753236900505e-05,
-      "loss": 0.3725,
-      "step": 5111
-    },
-    {
-      "epoch": 2.5098617511520738,
-      "grad_norm": 0.4424062922484215,
-      "learning_rate": 1.8012262323785323e-05,
-      "loss": 0.4401,
-      "step": 5112
-    },
-    {
-      "epoch": 2.5103533026113674,
-      "grad_norm": 0.4014599813892684,
-      "learning_rate": 1.7977023364771937e-05,
-      "loss": 0.39,
-      "step": 5113
-    },
-    {
-      "epoch": 2.5108448540706605,
-      "grad_norm": 0.4065142469082297,
-      "learning_rate": 1.7941815505326042e-05,
-      "loss": 0.3848,
-      "step": 5114
-    },
-    {
-      "epoch": 2.511336405529954,
-      "grad_norm": 0.42958077209054774,
-      "learning_rate": 1.790663875879692e-05,
-      "loss": 0.4489,
-      "step": 5115
-    },
-    {
-      "epoch": 2.511827956989247,
-      "grad_norm": 0.4434811452856223,
-      "learning_rate": 1.787149313852209e-05,
-      "loss": 0.4297,
-      "step": 5116
-    },
-    {
-      "epoch": 2.512319508448541,
-      "grad_norm": 0.4195995036202588,
-      "learning_rate": 1.7836378657827224e-05,
-      "loss": 0.4092,
-      "step": 5117
-    },
-    {
-      "epoch": 2.5128110599078344,
-      "grad_norm": 0.4243217501066077,
-      "learning_rate": 1.7801295330026223e-05,
-      "loss": 0.4727,
-      "step": 5118
-    },
-    {
-      "epoch": 2.5133026113671275,
-      "grad_norm": 0.4300611395351345,
-      "learning_rate": 1.7766243168421182e-05,
-      "loss": 0.4136,
-      "step": 5119
-    },
-    {
-      "epoch": 2.5137941628264207,
-      "grad_norm": 0.3823968322916278,
-      "learning_rate": 1.7731222186302323e-05,
-      "loss": 0.3961,
-      "step": 5120
-    },
-    {
-      "epoch": 2.5142857142857142,
-      "grad_norm": 0.40991509197708587,
-      "learning_rate": 1.7696232396948177e-05,
-      "loss": 0.4273,
-      "step": 5121
-    },
-    {
-      "epoch": 2.514777265745008,
-      "grad_norm": 0.4091697241015664,
-      "learning_rate": 1.7661273813625256e-05,
-      "loss": 0.4159,
-      "step": 5122
-    },
-    {
-      "epoch": 2.515268817204301,
-      "grad_norm": 0.43010247842834093,
-      "learning_rate": 1.7626346449588416e-05,
-      "loss": 0.4428,
-      "step": 5123
-    },
-    {
-      "epoch": 2.5157603686635945,
-      "grad_norm": 0.3754503027673675,
-      "learning_rate": 1.7591450318080573e-05,
-      "loss": 0.3731,
-      "step": 5124
-    },
-    {
-      "epoch": 2.5162519201228877,
-      "grad_norm": 0.42428740891599254,
-      "learning_rate": 1.755658543233285e-05,
-      "loss": 0.4144,
-      "step": 5125
-    },
-    {
-      "epoch": 2.5167434715821813,
-      "grad_norm": 0.43077476712147833,
-      "learning_rate": 1.752175180556447e-05,
-      "loss": 0.4206,
-      "step": 5126
-    },
-    {
-      "epoch": 2.517235023041475,
-      "grad_norm": 0.38311260300187444,
-      "learning_rate": 1.7486949450982904e-05,
-      "loss": 0.4567,
-      "step": 5127
-    },
-    {
-      "epoch": 2.517726574500768,
-      "grad_norm": 0.3701679906246134,
-      "learning_rate": 1.7452178381783613e-05,
-      "loss": 0.3704,
-      "step": 5128
-    },
-    {
-      "epoch": 2.5182181259600616,
-      "grad_norm": 0.4388550280755766,
-      "learning_rate": 1.741743861115035e-05,
-      "loss": 0.4535,
-      "step": 5129
-    },
-    {
-      "epoch": 2.5187096774193547,
-      "grad_norm": 0.44016639905895155,
-      "learning_rate": 1.7382730152254922e-05,
-      "loss": 0.3892,
-      "step": 5130
-    },
-    {
-      "epoch": 2.5192012288786483,
-      "grad_norm": 0.40077274726571227,
-      "learning_rate": 1.734805301825726e-05,
-      "loss": 0.4011,
-      "step": 5131
-    },
-    {
-      "epoch": 2.519692780337942,
-      "grad_norm": 0.4216679186261871,
-      "learning_rate": 1.7313407222305456e-05,
-      "loss": 0.4013,
-      "step": 5132
-    },
-    {
-      "epoch": 2.520184331797235,
-      "grad_norm": 0.4281045163434053,
-      "learning_rate": 1.727879277753568e-05,
-      "loss": 0.4058,
-      "step": 5133
-    },
-    {
-      "epoch": 2.5206758832565286,
-      "grad_norm": 0.42294901793632517,
-      "learning_rate": 1.7244209697072233e-05,
-      "loss": 0.436,
-      "step": 5134
-    },
-    {
-      "epoch": 2.5211674347158217,
-      "grad_norm": 0.41768516071747835,
-      "learning_rate": 1.7209657994027527e-05,
-      "loss": 0.3979,
-      "step": 5135
-    },
-    {
-      "epoch": 2.5216589861751153,
-      "grad_norm": 0.42351779230794256,
-      "learning_rate": 1.7175137681502085e-05,
-      "loss": 0.3784,
-      "step": 5136
-    },
-    {
-      "epoch": 2.5221505376344084,
-      "grad_norm": 0.41401189736057803,
-      "learning_rate": 1.714064877258449e-05,
-      "loss": 0.3834,
-      "step": 5137
-    },
-    {
-      "epoch": 2.522642089093702,
-      "grad_norm": 0.4239662779891976,
-      "learning_rate": 1.71061912803515e-05,
-      "loss": 0.4137,
-      "step": 5138
-    },
-    {
-      "epoch": 2.523133640552995,
-      "grad_norm": 0.4493118796360464,
-      "learning_rate": 1.707176521786784e-05,
-      "loss": 0.4286,
-      "step": 5139
-    },
-    {
-      "epoch": 2.5236251920122887,
-      "grad_norm": 0.42753180575411415,
-      "learning_rate": 1.703737059818643e-05,
-      "loss": 0.4124,
-      "step": 5140
-    },
-    {
-      "epoch": 2.5241167434715823,
-      "grad_norm": 0.41350002774613603,
-      "learning_rate": 1.7003007434348218e-05,
-      "loss": 0.4259,
-      "step": 5141
-    },
-    {
-      "epoch": 2.5246082949308755,
-      "grad_norm": 0.4728153585992248,
-      "learning_rate": 1.6968675739382213e-05,
-      "loss": 0.4309,
-      "step": 5142
-    },
-    {
-      "epoch": 2.525099846390169,
-      "grad_norm": 0.4620370104264429,
-      "learning_rate": 1.6934375526305524e-05,
-      "loss": 0.473,
-      "step": 5143
-    },
-    {
-      "epoch": 2.525591397849462,
-      "grad_norm": 0.39439966311719304,
-      "learning_rate": 1.6900106808123307e-05,
-      "loss": 0.3884,
-      "step": 5144
-    },
-    {
-      "epoch": 2.5260829493087558,
-      "grad_norm": 0.42179306170473924,
-      "learning_rate": 1.6865869597828765e-05,
-      "loss": 0.4374,
-      "step": 5145
-    },
-    {
-      "epoch": 2.5265745007680493,
-      "grad_norm": 0.4164924671983075,
-      "learning_rate": 1.6831663908403173e-05,
-      "loss": 0.4553,
-      "step": 5146
-    },
-    {
-      "epoch": 2.5270660522273425,
-      "grad_norm": 0.4184462447707974,
-      "learning_rate": 1.6797489752815886e-05,
-      "loss": 0.4085,
-      "step": 5147
-    },
-    {
-      "epoch": 2.527557603686636,
-      "grad_norm": 0.43716333565598975,
-      "learning_rate": 1.6763347144024198e-05,
-      "loss": 0.4415,
-      "step": 5148
-    },
-    {
-      "epoch": 2.528049155145929,
-      "grad_norm": 0.420920991688011,
-      "learning_rate": 1.6729236094973577e-05,
-      "loss": 0.3847,
-      "step": 5149
-    },
-    {
-      "epoch": 2.528540706605223,
-      "grad_norm": 0.37705013432666884,
-      "learning_rate": 1.669515661859743e-05,
-      "loss": 0.3617,
-      "step": 5150
-    },
-    {
-      "epoch": 2.5290322580645164,
-      "grad_norm": 0.3941126515595328,
-      "learning_rate": 1.666110872781722e-05,
-      "loss": 0.4069,
-      "step": 5151
-    },
-    {
-      "epoch": 2.5295238095238095,
-      "grad_norm": 0.42429417080600657,
-      "learning_rate": 1.6627092435542448e-05,
-      "loss": 0.4088,
-      "step": 5152
-    },
-    {
-      "epoch": 2.5300153609831026,
-      "grad_norm": 0.4852191182017364,
-      "learning_rate": 1.65931077546706e-05,
-      "loss": 0.4312,
-      "step": 5153
-    },
-    {
-      "epoch": 2.5305069124423962,
-      "grad_norm": 0.4146656415487168,
-      "learning_rate": 1.655915469808722e-05,
-      "loss": 0.4109,
-      "step": 5154
-    },
-    {
-      "epoch": 2.53099846390169,
-      "grad_norm": 0.42159691766921686,
-      "learning_rate": 1.6525233278665795e-05,
-      "loss": 0.3944,
-      "step": 5155
-    },
-    {
-      "epoch": 2.531490015360983,
-      "grad_norm": 0.4270272139299303,
-      "learning_rate": 1.6491343509267955e-05,
-      "loss": 0.3809,
-      "step": 5156
-    },
-    {
-      "epoch": 2.5319815668202765,
-      "grad_norm": 0.41857493800827444,
-      "learning_rate": 1.645748540274311e-05,
-      "loss": 0.4207,
-      "step": 5157
-    },
-    {
-      "epoch": 2.5324731182795697,
-      "grad_norm": 0.4234857212342732,
-      "learning_rate": 1.6423658971928923e-05,
-      "loss": 0.4273,
-      "step": 5158
-    },
-    {
-      "epoch": 2.5329646697388633,
-      "grad_norm": 0.39772500065506156,
-      "learning_rate": 1.6389864229650797e-05,
-      "loss": 0.4083,
-      "step": 5159
-    },
-    {
-      "epoch": 2.533456221198157,
-      "grad_norm": 0.4009462548155493,
-      "learning_rate": 1.6356101188722305e-05,
-      "loss": 0.3938,
-      "step": 5160
-    },
-    {
-      "epoch": 2.53394777265745,
-      "grad_norm": 0.4381124873214228,
-      "learning_rate": 1.6322369861944918e-05,
-      "loss": 0.4384,
-      "step": 5161
-    },
-    {
-      "epoch": 2.5344393241167436,
-      "grad_norm": 0.4166788073611714,
-      "learning_rate": 1.6288670262108108e-05,
-      "loss": 0.416,
-      "step": 5162
-    },
-    {
-      "epoch": 2.5349308755760367,
-      "grad_norm": 0.3943217111666468,
-      "learning_rate": 1.6255002401989282e-05,
-      "loss": 0.4026,
-      "step": 5163
-    },
-    {
-      "epoch": 2.5354224270353303,
-      "grad_norm": 0.4152791886301715,
-      "learning_rate": 1.6221366294353857e-05,
-      "loss": 0.4231,
-      "step": 5164
-    },
-    {
-      "epoch": 2.535913978494624,
-      "grad_norm": 0.38759262098189234,
-      "learning_rate": 1.6187761951955184e-05,
-      "loss": 0.4072,
-      "step": 5165
-    },
-    {
-      "epoch": 2.536405529953917,
-      "grad_norm": 0.43096175102999695,
-      "learning_rate": 1.6154189387534546e-05,
-      "loss": 0.4025,
-      "step": 5166
-    },
-    {
-      "epoch": 2.5368970814132106,
-      "grad_norm": 0.3877248213476849,
-      "learning_rate": 1.6120648613821298e-05,
-      "loss": 0.4014,
-      "step": 5167
-    },
-    {
-      "epoch": 2.5373886328725037,
-      "grad_norm": 0.4108620546866996,
-      "learning_rate": 1.6087139643532533e-05,
-      "loss": 0.4094,
-      "step": 5168
-    },
-    {
-      "epoch": 2.5378801843317973,
-      "grad_norm": 0.425865320899337,
-      "learning_rate": 1.60536624893735e-05,
-      "loss": 0.443,
-      "step": 5169
-    },
-    {
-      "epoch": 2.538371735791091,
-      "grad_norm": 0.4155784774856912,
-      "learning_rate": 1.6020217164037244e-05,
-      "loss": 0.4039,
-      "step": 5170
-    },
-    {
-      "epoch": 2.538863287250384,
-      "grad_norm": 0.4376177289011271,
-      "learning_rate": 1.59868036802048e-05,
-      "loss": 0.4841,
-      "step": 5171
-    },
-    {
-      "epoch": 2.539354838709677,
-      "grad_norm": 0.45147907378585744,
-      "learning_rate": 1.5953422050545098e-05,
-      "loss": 0.4361,
-      "step": 5172
-    },
-    {
-      "epoch": 2.5398463901689707,
-      "grad_norm": 0.4453501546122892,
-      "learning_rate": 1.592007228771506e-05,
-      "loss": 0.4474,
-      "step": 5173
-    },
-    {
-      "epoch": 2.5403379416282643,
-      "grad_norm": 0.4429132250084318,
-      "learning_rate": 1.5886754404359393e-05,
-      "loss": 0.4276,
-      "step": 5174
-    },
-    {
-      "epoch": 2.5408294930875575,
-      "grad_norm": 0.4201883815292871,
-      "learning_rate": 1.5853468413110873e-05,
-      "loss": 0.3932,
-      "step": 5175
-    },
-    {
-      "epoch": 2.541321044546851,
-      "grad_norm": 0.42536375263496035,
-      "learning_rate": 1.5820214326590066e-05,
-      "loss": 0.4122,
-      "step": 5176
-    },
-    {
-      "epoch": 2.541812596006144,
-      "grad_norm": 0.4376619698641515,
-      "learning_rate": 1.5786992157405513e-05,
-      "loss": 0.4108,
-      "step": 5177
-    },
-    {
-      "epoch": 2.5423041474654378,
-      "grad_norm": 0.4311526090004511,
-      "learning_rate": 1.575380191815361e-05,
-      "loss": 0.4069,
-      "step": 5178
-    },
-    {
-      "epoch": 2.5427956989247313,
-      "grad_norm": 0.46529232878365767,
-      "learning_rate": 1.572064362141865e-05,
-      "loss": 0.4275,
-      "step": 5179
-    },
-    {
-      "epoch": 2.5432872503840245,
-      "grad_norm": 0.4234866701254169,
-      "learning_rate": 1.5687517279772846e-05,
-      "loss": 0.4068,
-      "step": 5180
-    },
-    {
-      "epoch": 2.543778801843318,
-      "grad_norm": 0.43433667062631454,
-      "learning_rate": 1.5654422905776235e-05,
-      "loss": 0.4171,
-      "step": 5181
-    },
-    {
-      "epoch": 2.544270353302611,
-      "grad_norm": 0.40308818593140394,
-      "learning_rate": 1.5621360511976857e-05,
-      "loss": 0.4259,
-      "step": 5182
-    },
-    {
-      "epoch": 2.544761904761905,
-      "grad_norm": 0.42058027825717587,
-      "learning_rate": 1.5588330110910443e-05,
-      "loss": 0.4171,
-      "step": 5183
-    },
-    {
-      "epoch": 2.5452534562211984,
-      "grad_norm": 0.41003745626655774,
-      "learning_rate": 1.5555331715100784e-05,
-      "loss": 0.4023,
-      "step": 5184
-    },
-    {
-      "epoch": 2.5457450076804915,
-      "grad_norm": 0.42199906371832907,
-      "learning_rate": 1.5522365337059363e-05,
-      "loss": 0.4218,
-      "step": 5185
-    },
-    {
-      "epoch": 2.546236559139785,
-      "grad_norm": 0.43297104336064485,
-      "learning_rate": 1.548943098928566e-05,
-      "loss": 0.4525,
-      "step": 5186
-    },
-    {
-      "epoch": 2.546728110599078,
-      "grad_norm": 0.4168197407504609,
-      "learning_rate": 1.5456528684266937e-05,
-      "loss": 0.4019,
-      "step": 5187
-    },
-    {
-      "epoch": 2.547219662058372,
-      "grad_norm": 0.418580175053328,
-      "learning_rate": 1.5423658434478338e-05,
-      "loss": 0.4249,
-      "step": 5188
-    },
-    {
-      "epoch": 2.5477112135176654,
-      "grad_norm": 0.4036913461879293,
-      "learning_rate": 1.539082025238283e-05,
-      "loss": 0.4255,
-      "step": 5189
-    },
-    {
-      "epoch": 2.5482027649769585,
-      "grad_norm": 0.43030777175460977,
-      "learning_rate": 1.535801415043123e-05,
-      "loss": 0.4126,
-      "step": 5190
-    },
-    {
-      "epoch": 2.5486943164362517,
-      "grad_norm": 0.42928372693046496,
-      "learning_rate": 1.5325240141062204e-05,
-      "loss": 0.3849,
-      "step": 5191
-    },
-    {
-      "epoch": 2.5491858678955452,
-      "grad_norm": 0.4127866494129654,
-      "learning_rate": 1.5292498236702213e-05,
-      "loss": 0.4488,
-      "step": 5192
-    },
-    {
-      "epoch": 2.549677419354839,
-      "grad_norm": 0.4273024908598793,
-      "learning_rate": 1.525978844976563e-05,
-      "loss": 0.4333,
-      "step": 5193
-    },
-    {
-      "epoch": 2.550168970814132,
-      "grad_norm": 0.40558111778023015,
-      "learning_rate": 1.522711079265452e-05,
-      "loss": 0.4198,
-      "step": 5194
-    },
-    {
-      "epoch": 2.5506605222734255,
-      "grad_norm": 0.40250482152085676,
-      "learning_rate": 1.5194465277758884e-05,
-      "loss": 0.4353,
-      "step": 5195
-    },
-    {
-      "epoch": 2.5511520737327187,
-      "grad_norm": 0.42983730915198437,
-      "learning_rate": 1.5161851917456494e-05,
-      "loss": 0.4107,
-      "step": 5196
-    },
-    {
-      "epoch": 2.5516436251920123,
-      "grad_norm": 0.4216749696468753,
-      "learning_rate": 1.5129270724112909e-05,
-      "loss": 0.4098,
-      "step": 5197
-    },
-    {
-      "epoch": 2.552135176651306,
-      "grad_norm": 0.4101392008164522,
-      "learning_rate": 1.5096721710081507e-05,
-      "loss": 0.4181,
-      "step": 5198
-    },
-    {
-      "epoch": 2.552626728110599,
-      "grad_norm": 0.4289937808734976,
-      "learning_rate": 1.5064204887703481e-05,
-      "loss": 0.4545,
-      "step": 5199
-    },
-    {
-      "epoch": 2.5531182795698926,
-      "grad_norm": 0.4081670093298884,
-      "learning_rate": 1.5031720269307792e-05,
-      "loss": 0.4087,
-      "step": 5200
-    },
-    {
-      "epoch": 2.5536098310291857,
-      "grad_norm": 0.43613123936449977,
-      "learning_rate": 1.4999267867211175e-05,
-      "loss": 0.4203,
-      "step": 5201
-    },
-    {
-      "epoch": 2.5541013824884793,
-      "grad_norm": 0.4234365538141309,
-      "learning_rate": 1.4966847693718279e-05,
-      "loss": 0.3969,
-      "step": 5202
-    },
-    {
-      "epoch": 2.554592933947773,
-      "grad_norm": 0.41756166466185746,
-      "learning_rate": 1.49344597611213e-05,
-      "loss": 0.4127,
-      "step": 5203
-    },
-    {
-      "epoch": 2.555084485407066,
-      "grad_norm": 0.4208935025456197,
-      "learning_rate": 1.4902104081700441e-05,
-      "loss": 0.3804,
-      "step": 5204
-    },
-    {
-      "epoch": 2.5555760368663596,
-      "grad_norm": 0.44268488906719217,
-      "learning_rate": 1.4869780667723531e-05,
-      "loss": 0.3979,
-      "step": 5205
-    },
-    {
-      "epoch": 2.5560675883256527,
-      "grad_norm": 0.4524267489306665,
-      "learning_rate": 1.4837489531446237e-05,
-      "loss": 0.4605,
-      "step": 5206
-    },
-    {
-      "epoch": 2.5565591397849463,
-      "grad_norm": 0.4329443386600355,
-      "learning_rate": 1.4805230685111937e-05,
-      "loss": 0.4415,
-      "step": 5207
-    },
-    {
-      "epoch": 2.55705069124424,
-      "grad_norm": 0.43016940710358653,
-      "learning_rate": 1.4773004140951807e-05,
-      "loss": 0.4188,
-      "step": 5208
-    },
-    {
-      "epoch": 2.557542242703533,
-      "grad_norm": 0.4352897381845784,
-      "learning_rate": 1.474080991118476e-05,
-      "loss": 0.4332,
-      "step": 5209
-    },
-    {
-      "epoch": 2.558033794162826,
-      "grad_norm": 0.3924524832111949,
-      "learning_rate": 1.4708648008017444e-05,
-      "loss": 0.4202,
-      "step": 5210
-    },
-    {
-      "epoch": 2.5585253456221198,
-      "grad_norm": 0.4198360475188516,
-      "learning_rate": 1.4676518443644282e-05,
-      "loss": 0.4123,
-      "step": 5211
-    },
-    {
-      "epoch": 2.5590168970814133,
-      "grad_norm": 0.4131811141251363,
-      "learning_rate": 1.4644421230247374e-05,
-      "loss": 0.4319,
-      "step": 5212
-    },
-    {
-      "epoch": 2.5595084485407065,
-      "grad_norm": 0.4197501650993057,
-      "learning_rate": 1.4612356379996672e-05,
-      "loss": 0.3766,
-      "step": 5213
-    },
-    {
-      "epoch": 2.56,
-      "grad_norm": 0.41959695002969705,
-      "learning_rate": 1.4580323905049686e-05,
-      "loss": 0.4737,
-      "step": 5214
-    },
-    {
-      "epoch": 2.560491551459293,
-      "grad_norm": 0.3871973737477092,
-      "learning_rate": 1.4548323817551824e-05,
-      "loss": 0.3772,
-      "step": 5215
-    },
-    {
-      "epoch": 2.5609831029185868,
-      "grad_norm": 0.3806914992210328,
-      "learning_rate": 1.451635612963611e-05,
-      "loss": 0.3624,
-      "step": 5216
-    },
-    {
-      "epoch": 2.5614746543778804,
-      "grad_norm": 0.4299269379033961,
-      "learning_rate": 1.4484420853423297e-05,
-      "loss": 0.4356,
-      "step": 5217
-    },
-    {
-      "epoch": 2.5619662058371735,
-      "grad_norm": 0.40142642244715,
-      "learning_rate": 1.4452518001021864e-05,
-      "loss": 0.376,
-      "step": 5218
-    },
-    {
-      "epoch": 2.562457757296467,
-      "grad_norm": 0.42096708275060635,
-      "learning_rate": 1.4420647584528036e-05,
-      "loss": 0.4344,
-      "step": 5219
-    },
-    {
-      "epoch": 2.56294930875576,
-      "grad_norm": 0.4935593284023261,
-      "learning_rate": 1.4388809616025622e-05,
-      "loss": 0.441,
-      "step": 5220
-    },
-    {
-      "epoch": 2.563440860215054,
-      "grad_norm": 0.4501725033871413,
-      "learning_rate": 1.4357004107586258e-05,
-      "loss": 0.3772,
-      "step": 5221
-    },
-    {
-      "epoch": 2.5639324116743474,
-      "grad_norm": 0.4230627057043371,
-      "learning_rate": 1.4325231071269218e-05,
-      "loss": 0.442,
-      "step": 5222
-    },
-    {
-      "epoch": 2.5644239631336405,
-      "grad_norm": 0.41904655321415124,
-      "learning_rate": 1.4293490519121434e-05,
-      "loss": 0.4264,
-      "step": 5223
-    },
-    {
-      "epoch": 2.564915514592934,
-      "grad_norm": 0.4198199995562333,
-      "learning_rate": 1.4261782463177587e-05,
-      "loss": 0.4176,
-      "step": 5224
-    },
-    {
-      "epoch": 2.5654070660522272,
-      "grad_norm": 0.43115357569969237,
-      "learning_rate": 1.4230106915459963e-05,
-      "loss": 0.4152,
-      "step": 5225
-    },
-    {
-      "epoch": 2.565898617511521,
-      "grad_norm": 0.41160266998534767,
-      "learning_rate": 1.4198463887978596e-05,
-      "loss": 0.3711,
-      "step": 5226
-    },
-    {
-      "epoch": 2.5663901689708144,
-      "grad_norm": 0.4158553892016337,
-      "learning_rate": 1.4166853392731105e-05,
-      "loss": 0.4141,
-      "step": 5227
-    },
-    {
-      "epoch": 2.5668817204301075,
-      "grad_norm": 0.3973635607235415,
-      "learning_rate": 1.41352754417029e-05,
-      "loss": 0.3913,
-      "step": 5228
-    },
-    {
-      "epoch": 2.5673732718894007,
-      "grad_norm": 0.41063752774565754,
-      "learning_rate": 1.4103730046866893e-05,
-      "loss": 0.4164,
-      "step": 5229
-    },
-    {
-      "epoch": 2.5678648233486943,
-      "grad_norm": 0.42980867328518335,
-      "learning_rate": 1.4072217220183793e-05,
-      "loss": 0.4241,
-      "step": 5230
-    },
-    {
-      "epoch": 2.568356374807988,
-      "grad_norm": 0.42278273610439054,
-      "learning_rate": 1.4040736973601887e-05,
-      "loss": 0.436,
-      "step": 5231
-    },
-    {
-      "epoch": 2.568847926267281,
-      "grad_norm": 0.4017290939967276,
-      "learning_rate": 1.4009289319057118e-05,
-      "loss": 0.4332,
-      "step": 5232
-    },
-    {
-      "epoch": 2.5693394777265746,
-      "grad_norm": 0.37897490577983095,
-      "learning_rate": 1.3977874268473069e-05,
-      "loss": 0.3911,
-      "step": 5233
-    },
-    {
-      "epoch": 2.5698310291858677,
-      "grad_norm": 0.437632664900981,
-      "learning_rate": 1.3946491833760988e-05,
-      "loss": 0.4721,
-      "step": 5234
-    },
-    {
-      "epoch": 2.5703225806451613,
-      "grad_norm": 0.4281669222895695,
-      "learning_rate": 1.3915142026819716e-05,
-      "loss": 0.4556,
-      "step": 5235
-    },
-    {
-      "epoch": 2.570814132104455,
-      "grad_norm": 0.47070890803857696,
-      "learning_rate": 1.3883824859535765e-05,
-      "loss": 0.4295,
-      "step": 5236
-    },
-    {
-      "epoch": 2.571305683563748,
-      "grad_norm": 0.43931489919317734,
-      "learning_rate": 1.3852540343783249e-05,
-      "loss": 0.4012,
-      "step": 5237
-    },
-    {
-      "epoch": 2.5717972350230416,
-      "grad_norm": 0.4031329712999231,
-      "learning_rate": 1.3821288491423867e-05,
-      "loss": 0.4152,
-      "step": 5238
-    },
-    {
-      "epoch": 2.5722887864823347,
-      "grad_norm": 0.4160293874214144,
-      "learning_rate": 1.3790069314307053e-05,
-      "loss": 0.4616,
-      "step": 5239
-    },
-    {
-      "epoch": 2.5727803379416283,
-      "grad_norm": 0.4516909753487584,
-      "learning_rate": 1.3758882824269659e-05,
-      "loss": 0.4716,
-      "step": 5240
-    },
-    {
-      "epoch": 2.573271889400922,
-      "grad_norm": 0.4286934490832245,
-      "learning_rate": 1.3727729033136349e-05,
-      "loss": 0.478,
-      "step": 5241
-    },
-    {
-      "epoch": 2.573763440860215,
-      "grad_norm": 0.4624019227545422,
-      "learning_rate": 1.3696607952719253e-05,
-      "loss": 0.4237,
-      "step": 5242
-    },
-    {
-      "epoch": 2.5742549923195086,
-      "grad_norm": 0.45014270805912443,
-      "learning_rate": 1.3665519594818155e-05,
-      "loss": 0.4216,
-      "step": 5243
-    },
-    {
-      "epoch": 2.5747465437788017,
-      "grad_norm": 0.4295199509350521,
-      "learning_rate": 1.3634463971220402e-05,
-      "loss": 0.3833,
-      "step": 5244
-    },
-    {
-      "epoch": 2.5752380952380953,
-      "grad_norm": 0.41309087112776827,
-      "learning_rate": 1.3603441093700941e-05,
-      "loss": 0.3903,
-      "step": 5245
-    },
-    {
-      "epoch": 2.575729646697389,
-      "grad_norm": 0.4132812626586564,
-      "learning_rate": 1.3572450974022321e-05,
-      "loss": 0.425,
-      "step": 5246
-    },
-    {
-      "epoch": 2.576221198156682,
-      "grad_norm": 0.37905119216522815,
-      "learning_rate": 1.354149362393462e-05,
-      "loss": 0.3938,
-      "step": 5247
-    },
-    {
-      "epoch": 2.576712749615975,
-      "grad_norm": 0.4044672010327601,
-      "learning_rate": 1.3510569055175603e-05,
-      "loss": 0.397,
-      "step": 5248
-    },
-    {
-      "epoch": 2.5772043010752688,
-      "grad_norm": 0.41093249446353486,
-      "learning_rate": 1.3479677279470448e-05,
-      "loss": 0.4425,
-      "step": 5249
-    },
-    {
-      "epoch": 2.5776958525345623,
-      "grad_norm": 0.400832755999386,
-      "learning_rate": 1.3448818308532029e-05,
-      "loss": 0.4071,
-      "step": 5250
-    },
-    {
-      "epoch": 2.5781874039938555,
-      "grad_norm": 0.4267841036431294,
-      "learning_rate": 1.3417992154060721e-05,
-      "loss": 0.4117,
-      "step": 5251
-    },
-    {
-      "epoch": 2.578678955453149,
-      "grad_norm": 0.41382688696643344,
-      "learning_rate": 1.3387198827744473e-05,
-      "loss": 0.3933,
-      "step": 5252
-    },
-    {
-      "epoch": 2.579170506912442,
-      "grad_norm": 0.45030195704791587,
-      "learning_rate": 1.335643834125876e-05,
-      "loss": 0.3901,
-      "step": 5253
-    },
-    {
-      "epoch": 2.579662058371736,
-      "grad_norm": 0.41456953369304156,
-      "learning_rate": 1.3325710706266692e-05,
-      "loss": 0.4414,
-      "step": 5254
-    },
-    {
-      "epoch": 2.5801536098310294,
-      "grad_norm": 0.4206699474990611,
-      "learning_rate": 1.3295015934418787e-05,
-      "loss": 0.412,
-      "step": 5255
-    },
-    {
-      "epoch": 2.5806451612903225,
-      "grad_norm": 0.4020092415941204,
-      "learning_rate": 1.326435403735321e-05,
-      "loss": 0.4212,
-      "step": 5256
-    },
-    {
-      "epoch": 2.581136712749616,
-      "grad_norm": 0.42552309596455123,
-      "learning_rate": 1.323372502669562e-05,
-      "loss": 0.4346,
-      "step": 5257
-    },
-    {
-      "epoch": 2.5816282642089092,
-      "grad_norm": 0.45233053570079085,
-      "learning_rate": 1.3203128914059194e-05,
-      "loss": 0.3929,
-      "step": 5258
-    },
-    {
-      "epoch": 2.582119815668203,
-      "grad_norm": 0.4038714180820211,
-      "learning_rate": 1.3172565711044727e-05,
-      "loss": 0.4142,
-      "step": 5259
-    },
-    {
-      "epoch": 2.5826113671274964,
-      "grad_norm": 0.4286245868920105,
-      "learning_rate": 1.3142035429240373e-05,
-      "loss": 0.4394,
-      "step": 5260
-    },
-    {
-      "epoch": 2.5831029185867895,
-      "grad_norm": 0.3947956652013941,
-      "learning_rate": 1.3111538080221952e-05,
-      "loss": 0.3856,
-      "step": 5261
-    },
-    {
-      "epoch": 2.583594470046083,
-      "grad_norm": 0.40575206356814003,
-      "learning_rate": 1.3081073675552735e-05,
-      "loss": 0.3895,
-      "step": 5262
-    },
-    {
-      "epoch": 2.5840860215053763,
-      "grad_norm": 0.3783440197370085,
-      "learning_rate": 1.3050642226783493e-05,
-      "loss": 0.3846,
-      "step": 5263
-    },
-    {
-      "epoch": 2.58457757296467,
-      "grad_norm": 0.4230362394713855,
-      "learning_rate": 1.3020243745452498e-05,
-      "loss": 0.4619,
-      "step": 5264
-    },
-    {
-      "epoch": 2.585069124423963,
-      "grad_norm": 0.4519557923289012,
-      "learning_rate": 1.2989878243085619e-05,
-      "loss": 0.411,
-      "step": 5265
-    },
-    {
-      "epoch": 2.5855606758832566,
-      "grad_norm": 0.40034258223343333,
-      "learning_rate": 1.2959545731196032e-05,
-      "loss": 0.3981,
-      "step": 5266
-    },
-    {
-      "epoch": 2.5860522273425497,
-      "grad_norm": 0.4065504050587561,
-      "learning_rate": 1.2929246221284597e-05,
-      "loss": 0.3863,
-      "step": 5267
-    },
-    {
-      "epoch": 2.5865437788018433,
-      "grad_norm": 0.42539197693503783,
-      "learning_rate": 1.2898979724839545e-05,
-      "loss": 0.4525,
-      "step": 5268
-    },
-    {
-      "epoch": 2.587035330261137,
-      "grad_norm": 0.4268125402751333,
-      "learning_rate": 1.2868746253336638e-05,
-      "loss": 0.4587,
-      "step": 5269
-    },
-    {
-      "epoch": 2.58752688172043,
-      "grad_norm": 0.4450892053251256,
-      "learning_rate": 1.283854581823909e-05,
-      "loss": 0.4734,
-      "step": 5270
-    },
-    {
-      "epoch": 2.5880184331797236,
-      "grad_norm": 0.4096266303695234,
-      "learning_rate": 1.2808378430997591e-05,
-      "loss": 0.4482,
-      "step": 5271
-    },
-    {
-      "epoch": 2.5885099846390167,
-      "grad_norm": 0.4073859428706417,
-      "learning_rate": 1.2778244103050341e-05,
-      "loss": 0.4046,
-      "step": 5272
-    },
-    {
-      "epoch": 2.5890015360983103,
-      "grad_norm": 0.3893176076411962,
-      "learning_rate": 1.2748142845822931e-05,
-      "loss": 0.4106,
-      "step": 5273
-    },
-    {
-      "epoch": 2.589493087557604,
-      "grad_norm": 0.40142306934620814,
-      "learning_rate": 1.271807467072852e-05,
-      "loss": 0.4285,
-      "step": 5274
-    },
-    {
-      "epoch": 2.589984639016897,
-      "grad_norm": 0.40061330517437627,
-      "learning_rate": 1.2688039589167578e-05,
-      "loss": 0.397,
-      "step": 5275
-    },
-    {
-      "epoch": 2.5904761904761906,
-      "grad_norm": 0.42711729902195467,
-      "learning_rate": 1.2658037612528184e-05,
-      "loss": 0.4175,
-      "step": 5276
-    },
-    {
-      "epoch": 2.5909677419354837,
-      "grad_norm": 0.4384587559454611,
-      "learning_rate": 1.2628068752185751e-05,
-      "loss": 0.4478,
-      "step": 5277
-    },
-    {
-      "epoch": 2.5914592933947773,
-      "grad_norm": 0.3995933598796077,
-      "learning_rate": 1.2598133019503189e-05,
-      "loss": 0.3969,
-      "step": 5278
-    },
-    {
-      "epoch": 2.591950844854071,
-      "grad_norm": 0.4150669053758655,
-      "learning_rate": 1.256823042583084e-05,
-      "loss": 0.4211,
-      "step": 5279
-    },
-    {
-      "epoch": 2.592442396313364,
-      "grad_norm": 0.3941671749373715,
-      "learning_rate": 1.2538360982506459e-05,
-      "loss": 0.3898,
-      "step": 5280
-    },
-    {
-      "epoch": 2.592933947772657,
-      "grad_norm": 0.4408742353833633,
-      "learning_rate": 1.2508524700855252e-05,
-      "loss": 0.4762,
-      "step": 5281
-    },
-    {
-      "epoch": 2.5934254992319508,
-      "grad_norm": 0.4049296429399314,
-      "learning_rate": 1.2478721592189846e-05,
-      "loss": 0.4181,
-      "step": 5282
-    },
-    {
-      "epoch": 2.5939170506912443,
-      "grad_norm": 0.4419985604059515,
-      "learning_rate": 1.2448951667810305e-05,
-      "loss": 0.398,
-      "step": 5283
-    },
-    {
-      "epoch": 2.5944086021505375,
-      "grad_norm": 0.4304218206479302,
-      "learning_rate": 1.2419214939004065e-05,
-      "loss": 0.4549,
-      "step": 5284
-    },
-    {
-      "epoch": 2.594900153609831,
-      "grad_norm": 0.4683803136024223,
-      "learning_rate": 1.2389511417046074e-05,
-      "loss": 0.4366,
-      "step": 5285
-    },
-    {
-      "epoch": 2.595391705069124,
-      "grad_norm": 0.44115516139270294,
-      "learning_rate": 1.2359841113198534e-05,
-      "loss": 0.3851,
-      "step": 5286
-    },
-    {
-      "epoch": 2.595883256528418,
-      "grad_norm": 0.41651016750665665,
-      "learning_rate": 1.2330204038711213e-05,
-      "loss": 0.4077,
-      "step": 5287
-    },
-    {
-      "epoch": 2.5963748079877114,
-      "grad_norm": 0.4222474066577131,
-      "learning_rate": 1.2300600204821178e-05,
-      "loss": 0.4108,
-      "step": 5288
-    },
-    {
-      "epoch": 2.5968663594470045,
-      "grad_norm": 0.43932138030503176,
-      "learning_rate": 1.2271029622752939e-05,
-      "loss": 0.4221,
-      "step": 5289
-    },
-    {
-      "epoch": 2.597357910906298,
-      "grad_norm": 0.41988399311680097,
-      "learning_rate": 1.2241492303718372e-05,
-      "loss": 0.4106,
-      "step": 5290
-    },
-    {
-      "epoch": 2.5978494623655912,
-      "grad_norm": 0.42322642269398886,
-      "learning_rate": 1.2211988258916752e-05,
-      "loss": 0.4235,
-      "step": 5291
-    },
-    {
-      "epoch": 2.598341013824885,
-      "grad_norm": 0.41203733168819306,
-      "learning_rate": 1.2182517499534728e-05,
-      "loss": 0.4133,
-      "step": 5292
-    },
-    {
-      "epoch": 2.5988325652841784,
-      "grad_norm": 0.3983076817788457,
-      "learning_rate": 1.2153080036746345e-05,
-      "loss": 0.4233,
-      "step": 5293
-    },
-    {
-      "epoch": 2.5993241167434715,
-      "grad_norm": 0.37847333226681684,
-      "learning_rate": 1.2123675881713048e-05,
-      "loss": 0.3893,
-      "step": 5294
-    },
-    {
-      "epoch": 2.599815668202765,
-      "grad_norm": 0.389824916268444,
-      "learning_rate": 1.2094305045583565e-05,
-      "loss": 0.4209,
-      "step": 5295
-    },
-    {
-      "epoch": 2.6003072196620582,
-      "grad_norm": 0.4017719347973111,
-      "learning_rate": 1.2064967539494087e-05,
-      "loss": 0.4043,
-      "step": 5296
-    },
-    {
-      "epoch": 2.600798771121352,
-      "grad_norm": 0.4113509117924611,
-      "learning_rate": 1.2035663374568118e-05,
-      "loss": 0.3814,
-      "step": 5297
-    },
-    {
-      "epoch": 2.6012903225806454,
-      "grad_norm": 0.4165745386087526,
-      "learning_rate": 1.2006392561916535e-05,
-      "loss": 0.4531,
-      "step": 5298
-    },
-    {
-      "epoch": 2.6017818740399385,
-      "grad_norm": 0.43949575300973437,
-      "learning_rate": 1.197715511263754e-05,
-      "loss": 0.4329,
-      "step": 5299
-    },
-    {
-      "epoch": 2.6022734254992317,
-      "grad_norm": 0.4431857617960204,
-      "learning_rate": 1.1947951037816762e-05,
-      "loss": 0.4386,
-      "step": 5300
-    },
-    {
-      "epoch": 2.6027649769585253,
-      "grad_norm": 0.423727275273249,
-      "learning_rate": 1.191878034852708e-05,
-      "loss": 0.4069,
-      "step": 5301
-    },
-    {
-      "epoch": 2.603256528417819,
-      "grad_norm": 0.432004334240044,
-      "learning_rate": 1.1889643055828758e-05,
-      "loss": 0.4249,
-      "step": 5302
-    },
-    {
-      "epoch": 2.603748079877112,
-      "grad_norm": 0.4083443107901837,
-      "learning_rate": 1.1860539170769436e-05,
-      "loss": 0.4193,
-      "step": 5303
-    },
-    {
-      "epoch": 2.6042396313364056,
-      "grad_norm": 0.4138588548351003,
-      "learning_rate": 1.1831468704383997e-05,
-      "loss": 0.3929,
-      "step": 5304
-    },
-    {
-      "epoch": 2.6047311827956987,
-      "grad_norm": 0.4070966211280334,
-      "learning_rate": 1.1802431667694768e-05,
-      "loss": 0.4134,
-      "step": 5305
-    },
-    {
-      "epoch": 2.6052227342549923,
-      "grad_norm": 0.4249990693964303,
-      "learning_rate": 1.1773428071711256e-05,
-      "loss": 0.4435,
-      "step": 5306
-    },
-    {
-      "epoch": 2.605714285714286,
-      "grad_norm": 0.4395434828072506,
-      "learning_rate": 1.1744457927430441e-05,
-      "loss": 0.4256,
-      "step": 5307
-    },
-    {
-      "epoch": 2.606205837173579,
-      "grad_norm": 0.4297304564073295,
-      "learning_rate": 1.1715521245836524e-05,
-      "loss": 0.4125,
-      "step": 5308
-    },
-    {
-      "epoch": 2.6066973886328726,
-      "grad_norm": 0.409718107651262,
-      "learning_rate": 1.1686618037901053e-05,
-      "loss": 0.4493,
-      "step": 5309
-    },
-    {
-      "epoch": 2.6071889400921657,
-      "grad_norm": 0.38519808583489984,
-      "learning_rate": 1.1657748314582851e-05,
-      "loss": 0.3788,
-      "step": 5310
-    },
-    {
-      "epoch": 2.6076804915514593,
-      "grad_norm": 0.419851656874679,
-      "learning_rate": 1.1628912086828115e-05,
-      "loss": 0.4297,
-      "step": 5311
-    },
-    {
-      "epoch": 2.608172043010753,
-      "grad_norm": 0.4565295533193233,
-      "learning_rate": 1.1600109365570234e-05,
-      "loss": 0.4016,
-      "step": 5312
-    },
-    {
-      "epoch": 2.608663594470046,
-      "grad_norm": 0.4031212214363553,
-      "learning_rate": 1.1571340161730016e-05,
-      "loss": 0.4145,
-      "step": 5313
-    },
-    {
-      "epoch": 2.6091551459293396,
-      "grad_norm": 0.39337275037809627,
-      "learning_rate": 1.1542604486215458e-05,
-      "loss": 0.4352,
-      "step": 5314
-    },
-    {
-      "epoch": 2.6096466973886328,
-      "grad_norm": 0.3978657181019612,
-      "learning_rate": 1.1513902349921913e-05,
-      "loss": 0.3963,
-      "step": 5315
-    },
-    {
-      "epoch": 2.6101382488479263,
-      "grad_norm": 0.42965276979044525,
-      "learning_rate": 1.1485233763731961e-05,
-      "loss": 0.4448,
-      "step": 5316
-    },
-    {
-      "epoch": 2.61062980030722,
-      "grad_norm": 0.4461801093207912,
-      "learning_rate": 1.1456598738515522e-05,
-      "loss": 0.4925,
-      "step": 5317
-    },
-    {
-      "epoch": 2.611121351766513,
-      "grad_norm": 0.4091718963844529,
-      "learning_rate": 1.1427997285129743e-05,
-      "loss": 0.4008,
-      "step": 5318
-    },
-    {
-      "epoch": 2.611612903225806,
-      "grad_norm": 0.41745303764794084,
-      "learning_rate": 1.1399429414419039e-05,
-      "loss": 0.3986,
-      "step": 5319
-    },
-    {
-      "epoch": 2.6121044546850998,
-      "grad_norm": 0.45964238945245806,
-      "learning_rate": 1.1370895137215176e-05,
-      "loss": 0.4585,
-      "step": 5320
-    },
-    {
-      "epoch": 2.6125960061443934,
-      "grad_norm": 0.3821186882089041,
-      "learning_rate": 1.1342394464337036e-05,
-      "loss": 0.4113,
-      "step": 5321
-    },
-    {
-      "epoch": 2.6130875576036865,
-      "grad_norm": 0.4301965137833599,
-      "learning_rate": 1.1313927406590908e-05,
-      "loss": 0.4159,
-      "step": 5322
-    },
-    {
-      "epoch": 2.61357910906298,
-      "grad_norm": 0.4150937380830955,
-      "learning_rate": 1.1285493974770245e-05,
-      "loss": 0.4138,
-      "step": 5323
-    },
-    {
-      "epoch": 2.614070660522273,
-      "grad_norm": 0.44047787380907416,
-      "learning_rate": 1.1257094179655769e-05,
-      "loss": 0.4505,
-      "step": 5324
-    },
-    {
-      "epoch": 2.614562211981567,
-      "grad_norm": 0.3941074837180245,
-      "learning_rate": 1.1228728032015468e-05,
-      "loss": 0.419,
-      "step": 5325
-    },
-    {
-      "epoch": 2.6150537634408604,
-      "grad_norm": 0.4267820860409458,
-      "learning_rate": 1.1200395542604547e-05,
-      "loss": 0.4607,
-      "step": 5326
-    },
-    {
-      "epoch": 2.6155453149001535,
-      "grad_norm": 0.445705231325303,
-      "learning_rate": 1.117209672216546e-05,
-      "loss": 0.4758,
-      "step": 5327
-    },
-    {
-      "epoch": 2.616036866359447,
-      "grad_norm": 0.4137045225724416,
-      "learning_rate": 1.1143831581427889e-05,
-      "loss": 0.4261,
-      "step": 5328
-    },
-    {
-      "epoch": 2.6165284178187402,
-      "grad_norm": 0.41750063648094804,
-      "learning_rate": 1.1115600131108817e-05,
-      "loss": 0.4103,
-      "step": 5329
-    },
-    {
-      "epoch": 2.617019969278034,
-      "grad_norm": 0.4165904590596343,
-      "learning_rate": 1.1087402381912293e-05,
-      "loss": 0.405,
-      "step": 5330
-    },
-    {
-      "epoch": 2.6175115207373274,
-      "grad_norm": 0.4058081989695218,
-      "learning_rate": 1.1059238344529765e-05,
-      "loss": 0.4402,
-      "step": 5331
-    },
-    {
-      "epoch": 2.6180030721966205,
-      "grad_norm": 0.4109268389232956,
-      "learning_rate": 1.1031108029639758e-05,
-      "loss": 0.4918,
-      "step": 5332
-    },
-    {
-      "epoch": 2.618494623655914,
-      "grad_norm": 0.40958818773959266,
-      "learning_rate": 1.1003011447908107e-05,
-      "loss": 0.4237,
-      "step": 5333
-    },
-    {
-      "epoch": 2.6189861751152073,
-      "grad_norm": 0.4325584243493363,
-      "learning_rate": 1.0974948609987823e-05,
-      "loss": 0.4012,
-      "step": 5334
-    },
-    {
-      "epoch": 2.619477726574501,
-      "grad_norm": 0.40835986383831824,
-      "learning_rate": 1.0946919526519118e-05,
-      "loss": 0.3711,
-      "step": 5335
-    },
-    {
-      "epoch": 2.6199692780337944,
-      "grad_norm": 0.4017252475489325,
-      "learning_rate": 1.0918924208129389e-05,
-      "loss": 0.3932,
-      "step": 5336
-    },
-    {
-      "epoch": 2.6204608294930876,
-      "grad_norm": 0.4439418722984641,
-      "learning_rate": 1.0890962665433268e-05,
-      "loss": 0.4567,
-      "step": 5337
-    },
-    {
-      "epoch": 2.6209523809523807,
-      "grad_norm": 0.3921230212544778,
-      "learning_rate": 1.0863034909032566e-05,
-      "loss": 0.4009,
-      "step": 5338
-    },
-    {
-      "epoch": 2.6214439324116743,
-      "grad_norm": 0.42325750347872054,
-      "learning_rate": 1.0835140949516253e-05,
-      "loss": 0.417,
-      "step": 5339
-    },
-    {
-      "epoch": 2.621935483870968,
-      "grad_norm": 0.389111937867253,
-      "learning_rate": 1.0807280797460572e-05,
-      "loss": 0.3746,
-      "step": 5340
-    },
-    {
-      "epoch": 2.622427035330261,
-      "grad_norm": 0.3978330318344825,
-      "learning_rate": 1.0779454463428817e-05,
-      "loss": 0.3961,
-      "step": 5341
-    },
-    {
-      "epoch": 2.6229185867895546,
-      "grad_norm": 0.42183186628517433,
-      "learning_rate": 1.0751661957971582e-05,
-      "loss": 0.3936,
-      "step": 5342
-    },
-    {
-      "epoch": 2.6234101382488477,
-      "grad_norm": 0.4302000139158192,
-      "learning_rate": 1.072390329162657e-05,
-      "loss": 0.4403,
-      "step": 5343
-    },
-    {
-      "epoch": 2.6239016897081413,
-      "grad_norm": 0.38113505252068064,
-      "learning_rate": 1.069617847491866e-05,
-      "loss": 0.4018,
-      "step": 5344
-    },
-    {
-      "epoch": 2.624393241167435,
-      "grad_norm": 0.41235578152989044,
-      "learning_rate": 1.0668487518359894e-05,
-      "loss": 0.4342,
-      "step": 5345
-    },
-    {
-      "epoch": 2.624884792626728,
-      "grad_norm": 0.428612445641925,
-      "learning_rate": 1.0640830432449534e-05,
-      "loss": 0.4016,
-      "step": 5346
-    },
-    {
-      "epoch": 2.6253763440860216,
-      "grad_norm": 0.39589738917147255,
-      "learning_rate": 1.0613207227673904e-05,
-      "loss": 0.4335,
-      "step": 5347
-    },
-    {
-      "epoch": 2.6258678955453147,
-      "grad_norm": 0.3918735791818909,
-      "learning_rate": 1.0585617914506529e-05,
-      "loss": 0.3989,
-      "step": 5348
-    },
-    {
-      "epoch": 2.6263594470046083,
-      "grad_norm": 0.4656303513460634,
-      "learning_rate": 1.0558062503408127e-05,
-      "loss": 0.436,
-      "step": 5349
-    },
-    {
-      "epoch": 2.626850998463902,
-      "grad_norm": 0.44756144086005695,
-      "learning_rate": 1.0530541004826455e-05,
-      "loss": 0.4687,
-      "step": 5350
-    },
-    {
-      "epoch": 2.627342549923195,
-      "grad_norm": 0.41186455865336674,
-      "learning_rate": 1.050305342919653e-05,
-      "loss": 0.3764,
-      "step": 5351
-    },
-    {
-      "epoch": 2.6278341013824886,
-      "grad_norm": 0.4305150431206504,
-      "learning_rate": 1.0475599786940438e-05,
-      "loss": 0.446,
-      "step": 5352
-    },
-    {
-      "epoch": 2.6283256528417818,
-      "grad_norm": 0.4141689143924113,
-      "learning_rate": 1.0448180088467407e-05,
-      "loss": 0.4131,
-      "step": 5353
-    },
-    {
-      "epoch": 2.6288172043010753,
-      "grad_norm": 0.43361065141307503,
-      "learning_rate": 1.0420794344173812e-05,
-      "loss": 0.4557,
-      "step": 5354
-    },
-    {
-      "epoch": 2.629308755760369,
-      "grad_norm": 0.4006286832518753,
-      "learning_rate": 1.0393442564443123e-05,
-      "loss": 0.4159,
-      "step": 5355
-    },
-    {
-      "epoch": 2.629800307219662,
-      "grad_norm": 0.40716025365309905,
-      "learning_rate": 1.0366124759645957e-05,
-      "loss": 0.4055,
-      "step": 5356
-    },
-    {
-      "epoch": 2.630291858678955,
-      "grad_norm": 0.4000233448184684,
-      "learning_rate": 1.0338840940140082e-05,
-      "loss": 0.4113,
-      "step": 5357
-    },
-    {
-      "epoch": 2.630783410138249,
-      "grad_norm": 0.42355353723854106,
-      "learning_rate": 1.031159111627028e-05,
-      "loss": 0.4623,
-      "step": 5358
-    },
-    {
-      "epoch": 2.6312749615975424,
-      "grad_norm": 0.46332201152106145,
-      "learning_rate": 1.0284375298368554e-05,
-      "loss": 0.4184,
-      "step": 5359
-    },
-    {
-      "epoch": 2.6317665130568355,
-      "grad_norm": 0.39945161000758245,
-      "learning_rate": 1.025719349675396e-05,
-      "loss": 0.3862,
-      "step": 5360
-    },
-    {
-      "epoch": 2.632258064516129,
-      "grad_norm": 0.4670259379687622,
-      "learning_rate": 1.0230045721732651e-05,
-      "loss": 0.3921,
-      "step": 5361
-    },
-    {
-      "epoch": 2.6327496159754222,
-      "grad_norm": 0.41491852944904284,
-      "learning_rate": 1.0202931983597896e-05,
-      "loss": 0.4386,
-      "step": 5362
-    },
-    {
-      "epoch": 2.633241167434716,
-      "grad_norm": 0.4458484761912659,
-      "learning_rate": 1.017585229263005e-05,
-      "loss": 0.3777,
-      "step": 5363
-    },
-    {
-      "epoch": 2.6337327188940094,
-      "grad_norm": 0.40792236037122076,
-      "learning_rate": 1.0148806659096555e-05,
-      "loss": 0.4108,
-      "step": 5364
-    },
-    {
-      "epoch": 2.6342242703533025,
-      "grad_norm": 0.4065104483513571,
-      "learning_rate": 1.012179509325194e-05,
-      "loss": 0.4498,
-      "step": 5365
-    },
-    {
-      "epoch": 2.634715821812596,
-      "grad_norm": 0.4323095333267805,
-      "learning_rate": 1.009481760533787e-05,
-      "loss": 0.4201,
-      "step": 5366
-    },
-    {
-      "epoch": 2.6352073732718893,
-      "grad_norm": 0.43581406930950106,
-      "learning_rate": 1.0067874205582971e-05,
-      "loss": 0.4406,
-      "step": 5367
-    },
-    {
-      "epoch": 2.635698924731183,
-      "grad_norm": 0.397598261902562,
-      "learning_rate": 1.0040964904203076e-05,
-      "loss": 0.4098,
-      "step": 5368
-    },
-    {
-      "epoch": 2.6361904761904764,
-      "grad_norm": 0.4264061380044999,
-      "learning_rate": 1.0014089711400998e-05,
-      "loss": 0.4197,
-      "step": 5369
-    },
-    {
-      "epoch": 2.6366820276497696,
-      "grad_norm": 0.4100340717425536,
-      "learning_rate": 9.987248637366664e-06,
-      "loss": 0.4339,
-      "step": 5370
-    },
-    {
-      "epoch": 2.637173579109063,
-      "grad_norm": 0.4169090206120225,
-      "learning_rate": 9.960441692277034e-06,
-      "loss": 0.4263,
-      "step": 5371
-    },
-    {
-      "epoch": 2.6376651305683563,
-      "grad_norm": 0.42544068109575156,
-      "learning_rate": 9.933668886296155e-06,
-      "loss": 0.4248,
-      "step": 5372
-    },
-    {
-      "epoch": 2.63815668202765,
-      "grad_norm": 0.4226205801571498,
-      "learning_rate": 9.906930229575118e-06,
-      "loss": 0.4481,
-      "step": 5373
-    },
-    {
-      "epoch": 2.6386482334869434,
-      "grad_norm": 0.41718955703675054,
-      "learning_rate": 9.880225732252035e-06,
-      "loss": 0.3969,
-      "step": 5374
-    },
-    {
-      "epoch": 2.6391397849462366,
-      "grad_norm": 0.39303734634685444,
-      "learning_rate": 9.853555404452164e-06,
-      "loss": 0.3754,
-      "step": 5375
-    },
-    {
-      "epoch": 2.6396313364055297,
-      "grad_norm": 0.44728570212843055,
-      "learning_rate": 9.82691925628766e-06,
-      "loss": 0.46,
-      "step": 5376
-    },
-    {
-      "epoch": 2.6401228878648233,
-      "grad_norm": 0.45493053165477176,
-      "learning_rate": 9.800317297857863e-06,
-      "loss": 0.4147,
-      "step": 5377
-    },
-    {
-      "epoch": 2.640614439324117,
-      "grad_norm": 0.41485918930308463,
-      "learning_rate": 9.773749539249055e-06,
-      "loss": 0.3874,
-      "step": 5378
-    },
-    {
-      "epoch": 2.64110599078341,
-      "grad_norm": 0.41486991053144257,
-      "learning_rate": 9.747215990534586e-06,
-      "loss": 0.4066,
-      "step": 5379
-    },
-    {
-      "epoch": 2.6415975422427036,
-      "grad_norm": 0.3776377283859583,
-      "learning_rate": 9.720716661774843e-06,
-      "loss": 0.3979,
-      "step": 5380
-    },
-    {
-      "epoch": 2.6420890937019967,
-      "grad_norm": 0.3989356508422083,
-      "learning_rate": 9.694251563017209e-06,
-      "loss": 0.3963,
-      "step": 5381
-    },
-    {
-      "epoch": 2.6425806451612903,
-      "grad_norm": 0.4190610328489551,
-      "learning_rate": 9.667820704296116e-06,
-      "loss": 0.4151,
-      "step": 5382
-    },
-    {
-      "epoch": 2.643072196620584,
-      "grad_norm": 0.4239017165612543,
-      "learning_rate": 9.641424095633e-06,
-      "loss": 0.4124,
-      "step": 5383
-    },
-    {
-      "epoch": 2.643563748079877,
-      "grad_norm": 0.3923533858174038,
-      "learning_rate": 9.615061747036314e-06,
-      "loss": 0.393,
-      "step": 5384
-    },
-    {
-      "epoch": 2.6440552995391706,
-      "grad_norm": 0.4289306006867652,
-      "learning_rate": 9.588733668501504e-06,
-      "loss": 0.4267,
-      "step": 5385
-    },
-    {
-      "epoch": 2.6445468509984638,
-      "grad_norm": 0.4260560989284525,
-      "learning_rate": 9.562439870011098e-06,
-      "loss": 0.4487,
-      "step": 5386
-    },
-    {
-      "epoch": 2.6450384024577573,
-      "grad_norm": 0.40920955468712666,
-      "learning_rate": 9.536180361534474e-06,
-      "loss": 0.3819,
-      "step": 5387
-    },
-    {
-      "epoch": 2.645529953917051,
-      "grad_norm": 0.4231466985521743,
-      "learning_rate": 9.509955153028193e-06,
-      "loss": 0.4047,
-      "step": 5388
-    },
-    {
-      "epoch": 2.646021505376344,
-      "grad_norm": 0.42796937170567456,
-      "learning_rate": 9.483764254435667e-06,
-      "loss": 0.4287,
-      "step": 5389
-    },
-    {
-      "epoch": 2.6465130568356376,
-      "grad_norm": 0.41928972491223726,
-      "learning_rate": 9.457607675687374e-06,
-      "loss": 0.4411,
-      "step": 5390
-    },
-    {
-      "epoch": 2.647004608294931,
-      "grad_norm": 0.40376676752035046,
-      "learning_rate": 9.431485426700736e-06,
-      "loss": 0.4051,
-      "step": 5391
-    },
-    {
-      "epoch": 2.6474961597542244,
-      "grad_norm": 0.4150879963716973,
-      "learning_rate": 9.405397517380233e-06,
-      "loss": 0.4147,
-      "step": 5392
-    },
-    {
-      "epoch": 2.6479877112135175,
-      "grad_norm": 0.42185871069050235,
-      "learning_rate": 9.379343957617226e-06,
-      "loss": 0.394,
-      "step": 5393
-    },
-    {
-      "epoch": 2.648479262672811,
-      "grad_norm": 0.4145208597484563,
-      "learning_rate": 9.353324757290082e-06,
-      "loss": 0.4214,
-      "step": 5394
-    },
-    {
-      "epoch": 2.6489708141321042,
-      "grad_norm": 0.4405632271679106,
-      "learning_rate": 9.327339926264222e-06,
-      "loss": 0.4362,
-      "step": 5395
-    },
-    {
-      "epoch": 2.649462365591398,
-      "grad_norm": 0.4092628443925841,
-      "learning_rate": 9.301389474391897e-06,
-      "loss": 0.4599,
-      "step": 5396
-    },
-    {
-      "epoch": 2.6499539170506914,
-      "grad_norm": 0.4864430244853022,
-      "learning_rate": 9.275473411512447e-06,
-      "loss": 0.5269,
-      "step": 5397
-    },
-    {
-      "epoch": 2.6504454685099845,
-      "grad_norm": 0.4299412011518461,
-      "learning_rate": 9.249591747452124e-06,
-      "loss": 0.3975,
-      "step": 5398
-    },
-    {
-      "epoch": 2.650937019969278,
-      "grad_norm": 0.3930617188361656,
-      "learning_rate": 9.223744492024112e-06,
-      "loss": 0.4113,
-      "step": 5399
-    },
-    {
-      "epoch": 2.6514285714285712,
-      "grad_norm": 0.40919081756109565,
-      "learning_rate": 9.197931655028558e-06,
-      "loss": 0.4134,
-      "step": 5400
-    },
-    {
-      "epoch": 2.651920122887865,
-      "grad_norm": 0.4307777252146958,
-      "learning_rate": 9.172153246252645e-06,
-      "loss": 0.4225,
-      "step": 5401
-    },
-    {
-      "epoch": 2.6524116743471584,
-      "grad_norm": 0.42298449468261196,
-      "learning_rate": 9.146409275470346e-06,
-      "loss": 0.4257,
-      "step": 5402
-    },
-    {
-      "epoch": 2.6529032258064515,
-      "grad_norm": 0.4318203258890334,
-      "learning_rate": 9.120699752442741e-06,
-      "loss": 0.3941,
-      "step": 5403
-    },
-    {
-      "epoch": 2.653394777265745,
-      "grad_norm": 0.3826337240679449,
-      "learning_rate": 9.095024686917687e-06,
-      "loss": 0.3645,
-      "step": 5404
-    },
-    {
-      "epoch": 2.6538863287250383,
-      "grad_norm": 0.39433634232551706,
-      "learning_rate": 9.069384088630117e-06,
-      "loss": 0.4512,
-      "step": 5405
-    },
-    {
-      "epoch": 2.654377880184332,
-      "grad_norm": 0.4121891187252731,
-      "learning_rate": 9.043777967301836e-06,
-      "loss": 0.45,
-      "step": 5406
-    },
-    {
-      "epoch": 2.6548694316436254,
-      "grad_norm": 0.4252387662041164,
-      "learning_rate": 9.018206332641554e-06,
-      "loss": 0.461,
-      "step": 5407
-    },
-    {
-      "epoch": 2.6553609831029186,
-      "grad_norm": 0.4111014762637659,
-      "learning_rate": 8.99266919434495e-06,
-      "loss": 0.3983,
-      "step": 5408
-    },
-    {
-      "epoch": 2.6558525345622117,
-      "grad_norm": 0.452111474133612,
-      "learning_rate": 8.967166562094587e-06,
-      "loss": 0.4309,
-      "step": 5409
-    },
-    {
-      "epoch": 2.6563440860215053,
-      "grad_norm": 0.4125007190544734,
-      "learning_rate": 8.941698445559965e-06,
-      "loss": 0.4159,
-      "step": 5410
-    },
-    {
-      "epoch": 2.656835637480799,
-      "grad_norm": 0.515996668294049,
-      "learning_rate": 8.916264854397483e-06,
-      "loss": 0.4522,
-      "step": 5411
-    },
-    {
-      "epoch": 2.657327188940092,
-      "grad_norm": 0.41386177213146924,
-      "learning_rate": 8.890865798250503e-06,
-      "loss": 0.4031,
-      "step": 5412
-    },
-    {
-      "epoch": 2.6578187403993856,
-      "grad_norm": 0.39125648600094315,
-      "learning_rate": 8.865501286749189e-06,
-      "loss": 0.396,
-      "step": 5413
-    },
-    {
-      "epoch": 2.6583102918586787,
-      "grad_norm": 0.4163082831106158,
-      "learning_rate": 8.840171329510705e-06,
-      "loss": 0.4085,
-      "step": 5414
-    },
-    {
-      "epoch": 2.6588018433179723,
-      "grad_norm": 0.42440505791526517,
-      "learning_rate": 8.814875936139078e-06,
-      "loss": 0.4416,
-      "step": 5415
-    },
-    {
-      "epoch": 2.659293394777266,
-      "grad_norm": 0.42560092976027736,
-      "learning_rate": 8.789615116225213e-06,
-      "loss": 0.4278,
-      "step": 5416
-    },
-    {
-      "epoch": 2.659784946236559,
-      "grad_norm": 0.6423512002747916,
-      "learning_rate": 8.76438887934693e-06,
-      "loss": 0.4591,
-      "step": 5417
-    },
-    {
-      "epoch": 2.6602764976958526,
-      "grad_norm": 0.4332157443761783,
-      "learning_rate": 8.739197235068918e-06,
-      "loss": 0.4607,
-      "step": 5418
-    },
-    {
-      "epoch": 2.6607680491551458,
-      "grad_norm": 0.409489123795808,
-      "learning_rate": 8.714040192942763e-06,
-      "loss": 0.3945,
-      "step": 5419
-    },
-    {
-      "epoch": 2.6612596006144393,
-      "grad_norm": 0.4303309799959123,
-      "learning_rate": 8.68891776250691e-06,
-      "loss": 0.4256,
-      "step": 5420
-    },
-    {
-      "epoch": 2.661751152073733,
-      "grad_norm": 0.4267798390985559,
-      "learning_rate": 8.663829953286762e-06,
-      "loss": 0.4233,
-      "step": 5421
-    },
-    {
-      "epoch": 2.662242703533026,
-      "grad_norm": 0.4810385904442036,
-      "learning_rate": 8.638776774794454e-06,
-      "loss": 0.4245,
-      "step": 5422
-    },
-    {
-      "epoch": 2.6627342549923196,
-      "grad_norm": 0.42308100332427523,
-      "learning_rate": 8.613758236529113e-06,
-      "loss": 0.4154,
-      "step": 5423
-    },
-    {
-      "epoch": 2.6632258064516128,
-      "grad_norm": 0.414959213266651,
-      "learning_rate": 8.58877434797668e-06,
-      "loss": 0.4528,
-      "step": 5424
-    },
-    {
-      "epoch": 2.6637173579109064,
-      "grad_norm": 0.37985402740042473,
-      "learning_rate": 8.563825118609969e-06,
-      "loss": 0.3577,
-      "step": 5425
-    },
-    {
-      "epoch": 2.6642089093702,
-      "grad_norm": 0.39331900486038013,
-      "learning_rate": 8.538910557888635e-06,
-      "loss": 0.3632,
-      "step": 5426
-    },
-    {
-      "epoch": 2.664700460829493,
-      "grad_norm": 0.408271435452002,
-      "learning_rate": 8.514030675259221e-06,
-      "loss": 0.4388,
-      "step": 5427
-    },
-    {
-      "epoch": 2.665192012288786,
-      "grad_norm": 0.40828619896231644,
-      "learning_rate": 8.489185480155082e-06,
-      "loss": 0.4219,
-      "step": 5428
-    },
-    {
-      "epoch": 2.66568356374808,
-      "grad_norm": 0.4338039540406892,
-      "learning_rate": 8.464374981996458e-06,
-      "loss": 0.4274,
-      "step": 5429
-    },
-    {
-      "epoch": 2.6661751152073734,
-      "grad_norm": 0.4766909380065768,
-      "learning_rate": 8.439599190190417e-06,
-      "loss": 0.4438,
-      "step": 5430
-    },
-    {
-      "epoch": 2.6666666666666665,
-      "grad_norm": 0.39862044241868255,
-      "learning_rate": 8.414858114130842e-06,
-      "loss": 0.4331,
-      "step": 5431
-    },
-    {
-      "epoch": 2.66715821812596,
-      "grad_norm": 0.3988996663069451,
-      "learning_rate": 8.390151763198528e-06,
-      "loss": 0.4306,
-      "step": 5432
-    },
-    {
-      "epoch": 2.6676497695852532,
-      "grad_norm": 0.4079597306392764,
-      "learning_rate": 8.365480146761006e-06,
-      "loss": 0.4177,
-      "step": 5433
-    },
-    {
-      "epoch": 2.668141321044547,
-      "grad_norm": 0.42331957670189707,
-      "learning_rate": 8.340843274172728e-06,
-      "loss": 0.4543,
-      "step": 5434
-    },
-    {
-      "epoch": 2.6686328725038404,
-      "grad_norm": 0.44553664600185905,
-      "learning_rate": 8.316241154774906e-06,
-      "loss": 0.4243,
-      "step": 5435
-    },
-    {
-      "epoch": 2.6691244239631335,
-      "grad_norm": 0.41428620785650344,
-      "learning_rate": 8.291673797895616e-06,
-      "loss": 0.4049,
-      "step": 5436
-    },
-    {
-      "epoch": 2.669615975422427,
-      "grad_norm": 0.41890171133499243,
-      "learning_rate": 8.267141212849704e-06,
-      "loss": 0.4415,
-      "step": 5437
-    },
-    {
-      "epoch": 2.6701075268817203,
-      "grad_norm": 0.4208486063219193,
-      "learning_rate": 8.242643408938922e-06,
-      "loss": 0.4313,
-      "step": 5438
-    },
-    {
-      "epoch": 2.670599078341014,
-      "grad_norm": 0.42531973876252605,
-      "learning_rate": 8.218180395451735e-06,
-      "loss": 0.4477,
-      "step": 5439
-    },
-    {
-      "epoch": 2.6710906298003074,
-      "grad_norm": 0.3930556158310125,
-      "learning_rate": 8.193752181663461e-06,
-      "loss": 0.4162,
-      "step": 5440
-    },
-    {
-      "epoch": 2.6715821812596006,
-      "grad_norm": 0.41512643601245525,
-      "learning_rate": 8.169358776836266e-06,
-      "loss": 0.3853,
-      "step": 5441
-    },
-    {
-      "epoch": 2.672073732718894,
-      "grad_norm": 0.4365926230250442,
-      "learning_rate": 8.145000190219e-06,
-      "loss": 0.4257,
-      "step": 5442
-    },
-    {
-      "epoch": 2.6725652841781873,
-      "grad_norm": 0.42705422252950365,
-      "learning_rate": 8.120676431047458e-06,
-      "loss": 0.4246,
-      "step": 5443
-    },
-    {
-      "epoch": 2.673056835637481,
-      "grad_norm": 0.41396914387957756,
-      "learning_rate": 8.096387508544123e-06,
-      "loss": 0.4348,
-      "step": 5444
-    },
-    {
-      "epoch": 2.6735483870967744,
-      "grad_norm": 0.410243392264974,
-      "learning_rate": 8.072133431918316e-06,
-      "loss": 0.4209,
-      "step": 5445
-    },
-    {
-      "epoch": 2.6740399385560676,
-      "grad_norm": 0.4049885626060416,
-      "learning_rate": 8.047914210366104e-06,
-      "loss": 0.3968,
-      "step": 5446
-    },
-    {
-      "epoch": 2.6745314900153607,
-      "grad_norm": 0.37332485096170237,
-      "learning_rate": 8.023729853070439e-06,
-      "loss": 0.3585,
-      "step": 5447
-    },
-    {
-      "epoch": 2.6750230414746543,
-      "grad_norm": 0.42614093446772877,
-      "learning_rate": 7.999580369200898e-06,
-      "loss": 0.3783,
-      "step": 5448
-    },
-    {
-      "epoch": 2.675514592933948,
-      "grad_norm": 0.4162593184506544,
-      "learning_rate": 7.975465767913981e-06,
-      "loss": 0.4034,
-      "step": 5449
-    },
-    {
-      "epoch": 2.676006144393241,
-      "grad_norm": 0.42459606567290314,
-      "learning_rate": 7.951386058352894e-06,
-      "loss": 0.4178,
-      "step": 5450
-    },
-    {
-      "epoch": 2.6764976958525346,
-      "grad_norm": 0.3980127332139539,
-      "learning_rate": 7.927341249647601e-06,
-      "loss": 0.3451,
-      "step": 5451
-    },
-    {
-      "epoch": 2.6769892473118277,
-      "grad_norm": 0.4256421613569688,
-      "learning_rate": 7.903331350914867e-06,
-      "loss": 0.441,
-      "step": 5452
-    },
-    {
-      "epoch": 2.6774807987711213,
-      "grad_norm": 0.44487912604201296,
-      "learning_rate": 7.879356371258218e-06,
-      "loss": 0.4597,
-      "step": 5453
-    },
-    {
-      "epoch": 2.677972350230415,
-      "grad_norm": 0.4075066995046214,
-      "learning_rate": 7.855416319767905e-06,
-      "loss": 0.3722,
-      "step": 5454
-    },
-    {
-      "epoch": 2.678463901689708,
-      "grad_norm": 0.4095127828141176,
-      "learning_rate": 7.83151120552098e-06,
-      "loss": 0.3888,
-      "step": 5455
-    },
-    {
-      "epoch": 2.6789554531490016,
-      "grad_norm": 0.3845272925313367,
-      "learning_rate": 7.807641037581226e-06,
-      "loss": 0.4136,
-      "step": 5456
-    },
-    {
-      "epoch": 2.6794470046082948,
-      "grad_norm": 0.4257334648775762,
-      "learning_rate": 7.783805824999157e-06,
-      "loss": 0.4207,
-      "step": 5457
-    },
-    {
-      "epoch": 2.6799385560675884,
-      "grad_norm": 0.40704957813240444,
-      "learning_rate": 7.760005576812112e-06,
-      "loss": 0.3943,
-      "step": 5458
-    },
-    {
-      "epoch": 2.680430107526882,
-      "grad_norm": 0.4303423979860556,
-      "learning_rate": 7.73624030204404e-06,
-      "loss": 0.4059,
-      "step": 5459
-    },
-    {
-      "epoch": 2.680921658986175,
-      "grad_norm": 0.4092923842280974,
-      "learning_rate": 7.712510009705764e-06,
-      "loss": 0.4251,
-      "step": 5460
-    },
-    {
-      "epoch": 2.6814132104454687,
-      "grad_norm": 0.44414524974282177,
-      "learning_rate": 7.688814708794766e-06,
-      "loss": 0.4186,
-      "step": 5461
-    },
-    {
-      "epoch": 2.681904761904762,
-      "grad_norm": 0.4565612987079781,
-      "learning_rate": 7.665154408295283e-06,
-      "loss": 0.4025,
-      "step": 5462
-    },
-    {
-      "epoch": 2.6823963133640554,
-      "grad_norm": 0.4064968763263214,
-      "learning_rate": 7.64152911717827e-06,
-      "loss": 0.4588,
-      "step": 5463
-    },
-    {
-      "epoch": 2.682887864823349,
-      "grad_norm": 0.4230784853496628,
-      "learning_rate": 7.617938844401429e-06,
-      "loss": 0.4135,
-      "step": 5464
-    },
-    {
-      "epoch": 2.683379416282642,
-      "grad_norm": 0.39190676123324375,
-      "learning_rate": 7.594383598909161e-06,
-      "loss": 0.3934,
-      "step": 5465
-    },
-    {
-      "epoch": 2.6838709677419352,
-      "grad_norm": 0.37307168866269835,
-      "learning_rate": 7.570863389632588e-06,
-      "loss": 0.3794,
-      "step": 5466
-    },
-    {
-      "epoch": 2.684362519201229,
-      "grad_norm": 0.41743804704716,
-      "learning_rate": 7.547378225489599e-06,
-      "loss": 0.4001,
-      "step": 5467
-    },
-    {
-      "epoch": 2.6848540706605224,
-      "grad_norm": 0.41777787964864926,
-      "learning_rate": 7.523928115384682e-06,
-      "loss": 0.4015,
-      "step": 5468
-    },
-    {
-      "epoch": 2.6853456221198155,
-      "grad_norm": 0.41587500762223234,
-      "learning_rate": 7.500513068209181e-06,
-      "loss": 0.421,
-      "step": 5469
-    },
-    {
-      "epoch": 2.685837173579109,
-      "grad_norm": 0.3878773263647131,
-      "learning_rate": 7.477133092841027e-06,
-      "loss": 0.3643,
-      "step": 5470
-    },
-    {
-      "epoch": 2.6863287250384023,
-      "grad_norm": 0.39004440967641163,
-      "learning_rate": 7.4537881981448954e-06,
-      "loss": 0.3995,
-      "step": 5471
-    },
-    {
-      "epoch": 2.686820276497696,
-      "grad_norm": 0.43869456975339577,
-      "learning_rate": 7.430478392972184e-06,
-      "loss": 0.4649,
-      "step": 5472
-    },
-    {
-      "epoch": 2.6873118279569894,
-      "grad_norm": 0.4115065209219869,
-      "learning_rate": 7.4072036861609464e-06,
-      "loss": 0.3871,
-      "step": 5473
-    },
-    {
-      "epoch": 2.6878033794162826,
-      "grad_norm": 0.4314534206231697,
-      "learning_rate": 7.3839640865359435e-06,
-      "loss": 0.423,
-      "step": 5474
-    },
-    {
-      "epoch": 2.688294930875576,
-      "grad_norm": 0.47178153428138386,
-      "learning_rate": 7.360759602908618e-06,
-      "loss": 0.3804,
-      "step": 5475
-    },
-    {
-      "epoch": 2.6887864823348693,
-      "grad_norm": 0.4235575725733525,
-      "learning_rate": 7.3375902440771635e-06,
-      "loss": 0.4365,
-      "step": 5476
-    },
-    {
-      "epoch": 2.689278033794163,
-      "grad_norm": 0.45252503817051837,
-      "learning_rate": 7.314456018826321e-06,
-      "loss": 0.4495,
-      "step": 5477
-    },
-    {
-      "epoch": 2.6897695852534564,
-      "grad_norm": 0.39021995944891386,
-      "learning_rate": 7.291356935927651e-06,
-      "loss": 0.3991,
-      "step": 5478
-    },
-    {
-      "epoch": 2.6902611367127496,
-      "grad_norm": 0.425949661897845,
-      "learning_rate": 7.268293004139271e-06,
-      "loss": 0.4086,
-      "step": 5479
-    },
-    {
-      "epoch": 2.690752688172043,
-      "grad_norm": 0.4401504805673719,
-      "learning_rate": 7.245264232206073e-06,
-      "loss": 0.4081,
-      "step": 5480
-    },
-    {
-      "epoch": 2.6912442396313363,
-      "grad_norm": 0.4464911473396545,
-      "learning_rate": 7.222270628859562e-06,
-      "loss": 0.4434,
-      "step": 5481
-    },
-    {
-      "epoch": 2.69173579109063,
-      "grad_norm": 0.4008652142682673,
-      "learning_rate": 7.199312202817909e-06,
-      "loss": 0.4039,
-      "step": 5482
-    },
-    {
-      "epoch": 2.6922273425499235,
-      "grad_norm": 0.46932450787482,
-      "learning_rate": 7.176388962785952e-06,
-      "loss": 0.4652,
-      "step": 5483
-    },
-    {
-      "epoch": 2.6927188940092166,
-      "grad_norm": 0.4209872429872233,
-      "learning_rate": 7.153500917455225e-06,
-      "loss": 0.3941,
-      "step": 5484
-    },
-    {
-      "epoch": 2.6932104454685097,
-      "grad_norm": 0.4638061944951321,
-      "learning_rate": 7.130648075503843e-06,
-      "loss": 0.4816,
-      "step": 5485
-    },
-    {
-      "epoch": 2.6937019969278033,
-      "grad_norm": 0.41302638234078337,
-      "learning_rate": 7.107830445596631e-06,
-      "loss": 0.4156,
-      "step": 5486
-    },
-    {
-      "epoch": 2.694193548387097,
-      "grad_norm": 0.40260901197167054,
-      "learning_rate": 7.085048036385078e-06,
-      "loss": 0.4305,
-      "step": 5487
-    },
-    {
-      "epoch": 2.69468509984639,
-      "grad_norm": 0.43152109742967315,
-      "learning_rate": 7.062300856507231e-06,
-      "loss": 0.4385,
-      "step": 5488
-    },
-    {
-      "epoch": 2.6951766513056836,
-      "grad_norm": 0.41329099828376487,
-      "learning_rate": 7.039588914587891e-06,
-      "loss": 0.4449,
-      "step": 5489
-    },
-    {
-      "epoch": 2.6956682027649768,
-      "grad_norm": 0.4279799231581329,
-      "learning_rate": 7.0169122192384144e-06,
-      "loss": 0.434,
-      "step": 5490
-    },
-    {
-      "epoch": 2.6961597542242703,
-      "grad_norm": 0.3929479731471267,
-      "learning_rate": 6.994270779056833e-06,
-      "loss": 0.3593,
-      "step": 5491
-    },
-    {
-      "epoch": 2.696651305683564,
-      "grad_norm": 0.41398419861491503,
-      "learning_rate": 6.971664602627792e-06,
-      "loss": 0.4526,
-      "step": 5492
-    },
-    {
-      "epoch": 2.697142857142857,
-      "grad_norm": 0.4497409827849861,
-      "learning_rate": 6.949093698522613e-06,
-      "loss": 0.5157,
-      "step": 5493
-    },
-    {
-      "epoch": 2.6976344086021506,
-      "grad_norm": 0.4097101827921944,
-      "learning_rate": 6.9265580752991495e-06,
-      "loss": 0.3983,
-      "step": 5494
-    },
-    {
-      "epoch": 2.698125960061444,
-      "grad_norm": 0.42514745491146183,
-      "learning_rate": 6.90405774150199e-06,
-      "loss": 0.4237,
-      "step": 5495
-    },
-    {
-      "epoch": 2.6986175115207374,
-      "grad_norm": 0.3824039272095052,
-      "learning_rate": 6.881592705662265e-06,
-      "loss": 0.3631,
-      "step": 5496
-    },
-    {
-      "epoch": 2.699109062980031,
-      "grad_norm": 0.42045176301648407,
-      "learning_rate": 6.8591629762977396e-06,
-      "loss": 0.4034,
-      "step": 5497
-    },
-    {
-      "epoch": 2.699600614439324,
-      "grad_norm": 0.4568161670097656,
-      "learning_rate": 6.836768561912798e-06,
-      "loss": 0.4174,
-      "step": 5498
-    },
-    {
-      "epoch": 2.7000921658986177,
-      "grad_norm": 0.4156900179818879,
-      "learning_rate": 6.8144094709984504e-06,
-      "loss": 0.4234,
-      "step": 5499
-    },
-    {
-      "epoch": 2.700583717357911,
-      "grad_norm": 0.39477858041860353,
-      "learning_rate": 6.792085712032281e-06,
-      "loss": 0.3605,
-      "step": 5500
-    },
-    {
-      "epoch": 2.7010752688172044,
-      "grad_norm": 0.46185864360589834,
-      "learning_rate": 6.769797293478486e-06,
-      "loss": 0.4075,
-      "step": 5501
-    },
-    {
-      "epoch": 2.701566820276498,
-      "grad_norm": 0.40614384264872005,
-      "learning_rate": 6.747544223787916e-06,
-      "loss": 0.4011,
-      "step": 5502
-    },
-    {
-      "epoch": 2.702058371735791,
-      "grad_norm": 0.4040622279379188,
-      "learning_rate": 6.7253265113979225e-06,
-      "loss": 0.4121,
-      "step": 5503
-    },
-    {
-      "epoch": 2.7025499231950842,
-      "grad_norm": 0.3932175902213402,
-      "learning_rate": 6.7031441647325335e-06,
-      "loss": 0.4,
-      "step": 5504
-    },
-    {
-      "epoch": 2.703041474654378,
-      "grad_norm": 0.4033607995851353,
-      "learning_rate": 6.680997192202299e-06,
-      "loss": 0.4296,
-      "step": 5505
-    },
-    {
-      "epoch": 2.7035330261136714,
-      "grad_norm": 0.40796652445487475,
-      "learning_rate": 6.658885602204446e-06,
-      "loss": 0.4381,
-      "step": 5506
-    },
-    {
-      "epoch": 2.7040245775729645,
-      "grad_norm": 0.39909841834260235,
-      "learning_rate": 6.63680940312269e-06,
-      "loss": 0.4056,
-      "step": 5507
-    },
-    {
-      "epoch": 2.704516129032258,
-      "grad_norm": 0.4209460608124384,
-      "learning_rate": 6.614768603327393e-06,
-      "loss": 0.3929,
-      "step": 5508
-    },
-    {
-      "epoch": 2.7050076804915513,
-      "grad_norm": 0.43719805515392945,
-      "learning_rate": 6.592763211175467e-06,
-      "loss": 0.4152,
-      "step": 5509
-    },
-    {
-      "epoch": 2.705499231950845,
-      "grad_norm": 0.3920812340842254,
-      "learning_rate": 6.570793235010408e-06,
-      "loss": 0.4159,
-      "step": 5510
-    },
-    {
-      "epoch": 2.7059907834101384,
-      "grad_norm": 0.44013974108592474,
-      "learning_rate": 6.548858683162284e-06,
-      "loss": 0.4124,
-      "step": 5511
-    },
-    {
-      "epoch": 2.7064823348694316,
-      "grad_norm": 0.4252420172153551,
-      "learning_rate": 6.52695956394771e-06,
-      "loss": 0.4456,
-      "step": 5512
-    },
-    {
-      "epoch": 2.706973886328725,
-      "grad_norm": 0.4064544627889784,
-      "learning_rate": 6.505095885669921e-06,
-      "loss": 0.4055,
-      "step": 5513
-    },
-    {
-      "epoch": 2.7074654377880183,
-      "grad_norm": 0.4136392213161951,
-      "learning_rate": 6.483267656618641e-06,
-      "loss": 0.3946,
-      "step": 5514
-    },
-    {
-      "epoch": 2.707956989247312,
-      "grad_norm": 0.42928931669236947,
-      "learning_rate": 6.4614748850702175e-06,
-      "loss": 0.404,
-      "step": 5515
-    },
-    {
-      "epoch": 2.7084485407066055,
-      "grad_norm": 0.42949999014574364,
-      "learning_rate": 6.439717579287518e-06,
-      "loss": 0.4155,
-      "step": 5516
-    },
-    {
-      "epoch": 2.7089400921658986,
-      "grad_norm": 0.4015711691051889,
-      "learning_rate": 6.417995747519967e-06,
-      "loss": 0.4021,
-      "step": 5517
-    },
-    {
-      "epoch": 2.709431643625192,
-      "grad_norm": 0.3892934049595953,
-      "learning_rate": 6.396309398003564e-06,
-      "loss": 0.3795,
-      "step": 5518
-    },
-    {
-      "epoch": 2.7099231950844853,
-      "grad_norm": 0.41761366564574076,
-      "learning_rate": 6.374658538960809e-06,
-      "loss": 0.4209,
-      "step": 5519
-    },
-    {
-      "epoch": 2.710414746543779,
-      "grad_norm": 0.39633714846566165,
-      "learning_rate": 6.3530431786007926e-06,
-      "loss": 0.4211,
-      "step": 5520
-    },
-    {
-      "epoch": 2.710906298003072,
-      "grad_norm": 0.432710141786852,
-      "learning_rate": 6.331463325119091e-06,
-      "loss": 0.4421,
-      "step": 5521
-    },
-    {
-      "epoch": 2.7113978494623656,
-      "grad_norm": 0.3789983734393442,
-      "learning_rate": 6.309918986697916e-06,
-      "loss": 0.3539,
-      "step": 5522
-    },
-    {
-      "epoch": 2.7118894009216588,
-      "grad_norm": 0.40750132326610417,
-      "learning_rate": 6.288410171505887e-06,
-      "loss": 0.4179,
-      "step": 5523
-    },
-    {
-      "epoch": 2.7123809523809523,
-      "grad_norm": 0.3977210793440475,
-      "learning_rate": 6.26693688769825e-06,
-      "loss": 0.3676,
-      "step": 5524
-    },
-    {
-      "epoch": 2.712872503840246,
-      "grad_norm": 0.4172629691498353,
-      "learning_rate": 6.245499143416733e-06,
-      "loss": 0.45,
-      "step": 5525
-    },
-    {
-      "epoch": 2.713364055299539,
-      "grad_norm": 0.45504849644046713,
-      "learning_rate": 6.224096946789615e-06,
-      "loss": 0.4375,
-      "step": 5526
-    },
-    {
-      "epoch": 2.7138556067588326,
-      "grad_norm": 0.43003664665807395,
-      "learning_rate": 6.20273030593167e-06,
-      "loss": 0.4195,
-      "step": 5527
-    },
-    {
-      "epoch": 2.7143471582181258,
-      "grad_norm": 0.4531115642820368,
-      "learning_rate": 6.181399228944218e-06,
-      "loss": 0.4194,
-      "step": 5528
-    },
-    {
-      "epoch": 2.7148387096774194,
-      "grad_norm": 0.4494584646762748,
-      "learning_rate": 6.160103723915067e-06,
-      "loss": 0.4388,
-      "step": 5529
-    },
-    {
-      "epoch": 2.715330261136713,
-      "grad_norm": 0.40059391291025565,
-      "learning_rate": 6.138843798918558e-06,
-      "loss": 0.3915,
-      "step": 5530
-    },
-    {
-      "epoch": 2.715821812596006,
-      "grad_norm": 0.4509052572054086,
-      "learning_rate": 6.117619462015534e-06,
-      "loss": 0.3797,
-      "step": 5531
-    },
-    {
-      "epoch": 2.7163133640552997,
-      "grad_norm": 0.4207411945120334,
-      "learning_rate": 6.096430721253343e-06,
-      "loss": 0.4015,
-      "step": 5532
-    },
-    {
-      "epoch": 2.716804915514593,
-      "grad_norm": 0.4090107518202127,
-      "learning_rate": 6.075277584665862e-06,
-      "loss": 0.4146,
-      "step": 5533
-    },
-    {
-      "epoch": 2.7172964669738864,
-      "grad_norm": 0.42453049869647747,
-      "learning_rate": 6.054160060273406e-06,
-      "loss": 0.4052,
-      "step": 5534
-    },
-    {
-      "epoch": 2.71778801843318,
-      "grad_norm": 0.392115007468133,
-      "learning_rate": 6.033078156082872e-06,
-      "loss": 0.3834,
-      "step": 5535
-    },
-    {
-      "epoch": 2.718279569892473,
-      "grad_norm": 0.4491975234066452,
-      "learning_rate": 6.012031880087576e-06,
-      "loss": 0.4261,
-      "step": 5536
-    },
-    {
-      "epoch": 2.7187711213517662,
-      "grad_norm": 0.46133803637202614,
-      "learning_rate": 5.991021240267369e-06,
-      "loss": 0.4404,
-      "step": 5537
-    },
-    {
-      "epoch": 2.71926267281106,
-      "grad_norm": 0.41266991788978646,
-      "learning_rate": 5.970046244588557e-06,
-      "loss": 0.4547,
-      "step": 5538
-    },
-    {
-      "epoch": 2.7197542242703534,
-      "grad_norm": 0.43379401767716325,
-      "learning_rate": 5.949106901004009e-06,
-      "loss": 0.4175,
-      "step": 5539
-    },
-    {
-      "epoch": 2.7202457757296465,
-      "grad_norm": 0.41715520539584056,
-      "learning_rate": 5.928203217452944e-06,
-      "loss": 0.4077,
-      "step": 5540
-    },
-    {
-      "epoch": 2.72073732718894,
-      "grad_norm": 0.39979302017139584,
-      "learning_rate": 5.907335201861175e-06,
-      "loss": 0.4012,
-      "step": 5541
-    },
-    {
-      "epoch": 2.7212288786482333,
-      "grad_norm": 0.4492443218638585,
-      "learning_rate": 5.886502862140952e-06,
-      "loss": 0.3901,
-      "step": 5542
-    },
-    {
-      "epoch": 2.721720430107527,
-      "grad_norm": 0.41330868165963774,
-      "learning_rate": 5.865706206191002e-06,
-      "loss": 0.4062,
-      "step": 5543
-    },
-    {
-      "epoch": 2.7222119815668204,
-      "grad_norm": 0.4061204852095859,
-      "learning_rate": 5.844945241896505e-06,
-      "loss": 0.408,
-      "step": 5544
-    },
-    {
-      "epoch": 2.7227035330261136,
-      "grad_norm": 0.4303086768341994,
-      "learning_rate": 5.824219977129119e-06,
-      "loss": 0.4494,
-      "step": 5545
-    },
-    {
-      "epoch": 2.723195084485407,
-      "grad_norm": 0.43234124947647756,
-      "learning_rate": 5.803530419746972e-06,
-      "loss": 0.4412,
-      "step": 5546
-    },
-    {
-      "epoch": 2.7236866359447003,
-      "grad_norm": 0.42856716154026187,
-      "learning_rate": 5.782876577594643e-06,
-      "loss": 0.4816,
-      "step": 5547
-    },
-    {
-      "epoch": 2.724178187403994,
-      "grad_norm": 0.43517964835085765,
-      "learning_rate": 5.762258458503223e-06,
-      "loss": 0.41,
-      "step": 5548
-    },
-    {
-      "epoch": 2.7246697388632874,
-      "grad_norm": 0.4087681873162632,
-      "learning_rate": 5.741676070290136e-06,
-      "loss": 0.4221,
-      "step": 5549
-    },
-    {
-      "epoch": 2.7251612903225806,
-      "grad_norm": 0.413728022603252,
-      "learning_rate": 5.721129420759386e-06,
-      "loss": 0.3996,
-      "step": 5550
-    },
-    {
-      "epoch": 2.725652841781874,
-      "grad_norm": 0.42567245625415584,
-      "learning_rate": 5.700618517701361e-06,
-      "loss": 0.4377,
-      "step": 5551
-    },
-    {
-      "epoch": 2.7261443932411673,
-      "grad_norm": 0.4027120128347094,
-      "learning_rate": 5.68014336889291e-06,
-      "loss": 0.4071,
-      "step": 5552
-    },
-    {
-      "epoch": 2.726635944700461,
-      "grad_norm": 0.4173600085987435,
-      "learning_rate": 5.6597039820973195e-06,
-      "loss": 0.4278,
-      "step": 5553
-    },
-    {
-      "epoch": 2.7271274961597545,
-      "grad_norm": 0.42602378606067504,
-      "learning_rate": 5.639300365064337e-06,
-      "loss": 0.4129,
-      "step": 5554
-    },
-    {
-      "epoch": 2.7276190476190476,
-      "grad_norm": 0.4063955012542778,
-      "learning_rate": 5.618932525530107e-06,
-      "loss": 0.3785,
-      "step": 5555
-    },
-    {
-      "epoch": 2.7281105990783407,
-      "grad_norm": 0.3895672747801335,
-      "learning_rate": 5.598600471217253e-06,
-      "loss": 0.4021,
-      "step": 5556
-    },
-    {
-      "epoch": 2.7286021505376343,
-      "grad_norm": 0.3788407920337871,
-      "learning_rate": 5.578304209834806e-06,
-      "loss": 0.3791,
-      "step": 5557
-    },
-    {
-      "epoch": 2.729093701996928,
-      "grad_norm": 0.3885984546093099,
-      "learning_rate": 5.558043749078213e-06,
-      "loss": 0.4157,
-      "step": 5558
-    },
-    {
-      "epoch": 2.729585253456221,
-      "grad_norm": 0.461999180680633,
-      "learning_rate": 5.537819096629415e-06,
-      "loss": 0.4222,
-      "step": 5559
-    },
-    {
-      "epoch": 2.7300768049155146,
-      "grad_norm": 0.4321143441787244,
-      "learning_rate": 5.517630260156659e-06,
-      "loss": 0.4631,
-      "step": 5560
-    },
-    {
-      "epoch": 2.7305683563748078,
-      "grad_norm": 0.42942568715414287,
-      "learning_rate": 5.497477247314731e-06,
-      "loss": 0.3754,
-      "step": 5561
-    },
-    {
-      "epoch": 2.7310599078341014,
-      "grad_norm": 0.41924713137026487,
-      "learning_rate": 5.477360065744764e-06,
-      "loss": 0.4149,
-      "step": 5562
-    },
-    {
-      "epoch": 2.731551459293395,
-      "grad_norm": 0.4211760013667313,
-      "learning_rate": 5.4572787230743325e-06,
-      "loss": 0.4064,
-      "step": 5563
-    },
-    {
-      "epoch": 2.732043010752688,
-      "grad_norm": 0.40925325423567027,
-      "learning_rate": 5.437233226917393e-06,
-      "loss": 0.4127,
-      "step": 5564
-    },
-    {
-      "epoch": 2.7325345622119817,
-      "grad_norm": 0.39017418155738515,
-      "learning_rate": 5.4172235848743536e-06,
-      "loss": 0.3871,
-      "step": 5565
-    },
-    {
-      "epoch": 2.733026113671275,
-      "grad_norm": 0.42387860572388514,
-      "learning_rate": 5.39724980453199e-06,
-      "loss": 0.4306,
-      "step": 5566
-    },
-    {
-      "epoch": 2.7335176651305684,
-      "grad_norm": 0.41859759911190514,
-      "learning_rate": 5.377311893463499e-06,
-      "loss": 0.4497,
-      "step": 5567
-    },
-    {
-      "epoch": 2.734009216589862,
-      "grad_norm": 0.44551197469589415,
-      "learning_rate": 5.3574098592284906e-06,
-      "loss": 0.4359,
-      "step": 5568
-    },
-    {
-      "epoch": 2.734500768049155,
-      "grad_norm": 0.4146158022437149,
-      "learning_rate": 5.337543709372928e-06,
-      "loss": 0.4125,
-      "step": 5569
-    },
-    {
-      "epoch": 2.7349923195084487,
-      "grad_norm": 0.4395996463882175,
-      "learning_rate": 5.317713451429218e-06,
-      "loss": 0.4374,
-      "step": 5570
-    },
-    {
-      "epoch": 2.735483870967742,
-      "grad_norm": 0.4063059656204468,
-      "learning_rate": 5.297919092916137e-06,
-      "loss": 0.3952,
-      "step": 5571
-    },
-    {
-      "epoch": 2.7359754224270354,
-      "grad_norm": 0.42422509347871734,
-      "learning_rate": 5.2781606413388475e-06,
-      "loss": 0.4637,
-      "step": 5572
-    },
-    {
-      "epoch": 2.736466973886329,
-      "grad_norm": 0.41608683387413403,
-      "learning_rate": 5.258438104188879e-06,
-      "loss": 0.4597,
-      "step": 5573
-    },
-    {
-      "epoch": 2.736958525345622,
-      "grad_norm": 0.3923512448829839,
-      "learning_rate": 5.2387514889442045e-06,
-      "loss": 0.3537,
-      "step": 5574
-    },
-    {
-      "epoch": 2.7374500768049153,
-      "grad_norm": 0.4105197544010745,
-      "learning_rate": 5.2191008030691105e-06,
-      "loss": 0.4044,
-      "step": 5575
-    },
-    {
-      "epoch": 2.737941628264209,
-      "grad_norm": 0.5221672938411795,
-      "learning_rate": 5.199486054014291e-06,
-      "loss": 0.4793,
-      "step": 5576
-    },
-    {
-      "epoch": 2.7384331797235024,
-      "grad_norm": 0.43335904506723294,
-      "learning_rate": 5.17990724921682e-06,
-      "loss": 0.3987,
-      "step": 5577
-    },
-    {
-      "epoch": 2.7389247311827956,
-      "grad_norm": 0.4197191888563136,
-      "learning_rate": 5.160364396100115e-06,
-      "loss": 0.4309,
-      "step": 5578
-    },
-    {
-      "epoch": 2.739416282642089,
-      "grad_norm": 0.4221371761334608,
-      "learning_rate": 5.140857502074015e-06,
-      "loss": 0.4478,
-      "step": 5579
-    },
-    {
-      "epoch": 2.7399078341013823,
-      "grad_norm": 0.43891218920831676,
-      "learning_rate": 5.121386574534648e-06,
-      "loss": 0.4146,
-      "step": 5580
-    },
-    {
-      "epoch": 2.740399385560676,
-      "grad_norm": 0.39607461571077496,
-      "learning_rate": 5.101951620864576e-06,
-      "loss": 0.3921,
-      "step": 5581
-    },
-    {
-      "epoch": 2.7408909370199694,
-      "grad_norm": 0.4088936138727943,
-      "learning_rate": 5.082552648432693e-06,
-      "loss": 0.3994,
-      "step": 5582
-    },
-    {
-      "epoch": 2.7413824884792626,
-      "grad_norm": 0.40441391191003495,
-      "learning_rate": 5.063189664594248e-06,
-      "loss": 0.4282,
-      "step": 5583
-    },
-    {
-      "epoch": 2.741874039938556,
-      "grad_norm": 0.4221894321675064,
-      "learning_rate": 5.043862676690825e-06,
-      "loss": 0.4678,
-      "step": 5584
-    },
-    {
-      "epoch": 2.7423655913978493,
-      "grad_norm": 0.40481345629281384,
-      "learning_rate": 5.0245716920504395e-06,
-      "loss": 0.3859,
-      "step": 5585
-    },
-    {
-      "epoch": 2.742857142857143,
-      "grad_norm": 0.4470732877744559,
-      "learning_rate": 5.005316717987329e-06,
-      "loss": 0.4092,
-      "step": 5586
-    },
-    {
-      "epoch": 2.7433486943164365,
-      "grad_norm": 0.3928363936142154,
-      "learning_rate": 4.986097761802189e-06,
-      "loss": 0.416,
-      "step": 5587
-    },
-    {
-      "epoch": 2.7438402457757296,
-      "grad_norm": 0.39688771386393373,
-      "learning_rate": 4.96691483078201e-06,
-      "loss": 0.3787,
-      "step": 5588
-    },
-    {
-      "epoch": 2.744331797235023,
-      "grad_norm": 0.4113568451176203,
-      "learning_rate": 4.9477679322001425e-06,
-      "loss": 0.4412,
-      "step": 5589
-    },
-    {
-      "epoch": 2.7448233486943163,
-      "grad_norm": 0.4056724191898062,
-      "learning_rate": 4.928657073316234e-06,
-      "loss": 0.4094,
-      "step": 5590
-    },
-    {
-      "epoch": 2.74531490015361,
-      "grad_norm": 0.44799921697791584,
-      "learning_rate": 4.909582261376322e-06,
-      "loss": 0.4247,
-      "step": 5591
-    },
-    {
-      "epoch": 2.7458064516129035,
-      "grad_norm": 0.40343199165739774,
-      "learning_rate": 4.890543503612733e-06,
-      "loss": 0.3976,
-      "step": 5592
-    },
-    {
-      "epoch": 2.7462980030721966,
-      "grad_norm": 0.40623013345428693,
-      "learning_rate": 4.8715408072441346e-06,
-      "loss": 0.4035,
-      "step": 5593
-    },
-    {
-      "epoch": 2.7467895545314898,
-      "grad_norm": 0.3702036442100213,
-      "learning_rate": 4.852574179475566e-06,
-      "loss": 0.3864,
-      "step": 5594
-    },
-    {
-      "epoch": 2.7472811059907833,
-      "grad_norm": 0.42621981328885633,
-      "learning_rate": 4.8336436274983075e-06,
-      "loss": 0.4485,
-      "step": 5595
-    },
-    {
-      "epoch": 2.747772657450077,
-      "grad_norm": 0.41055722258421873,
-      "learning_rate": 4.8147491584900395e-06,
-      "loss": 0.4082,
-      "step": 5596
-    },
-    {
-      "epoch": 2.74826420890937,
-      "grad_norm": 0.40601680743235696,
-      "learning_rate": 4.79589077961472e-06,
-      "loss": 0.4172,
-      "step": 5597
-    },
-    {
-      "epoch": 2.7487557603686636,
-      "grad_norm": 0.3956705174303359,
-      "learning_rate": 4.777068498022619e-06,
-      "loss": 0.4425,
-      "step": 5598
-    },
-    {
-      "epoch": 2.749247311827957,
-      "grad_norm": 0.4321025631895987,
-      "learning_rate": 4.758282320850338e-06,
-      "loss": 0.4387,
-      "step": 5599
-    },
-    {
-      "epoch": 2.7497388632872504,
-      "grad_norm": 0.4093227044262583,
-      "learning_rate": 4.739532255220791e-06,
-      "loss": 0.4014,
-      "step": 5600
-    },
-    {
-      "epoch": 2.750230414746544,
-      "grad_norm": 0.40095999040258756,
-      "learning_rate": 4.720818308243191e-06,
-      "loss": 0.3794,
-      "step": 5601
-    },
-    {
-      "epoch": 2.750721966205837,
-      "grad_norm": 0.4192009076581332,
-      "learning_rate": 4.70214048701304e-06,
-      "loss": 0.5041,
-      "step": 5602
-    },
-    {
-      "epoch": 2.7512135176651307,
-      "grad_norm": 0.4140192895860933,
-      "learning_rate": 4.683498798612185e-06,
-      "loss": 0.4127,
-      "step": 5603
-    },
-    {
-      "epoch": 2.751705069124424,
-      "grad_norm": 0.5398164419166952,
-      "learning_rate": 4.664893250108715e-06,
-      "loss": 0.4327,
-      "step": 5604
-    },
-    {
-      "epoch": 2.7521966205837174,
-      "grad_norm": 0.4107205382854118,
-      "learning_rate": 4.6463238485570995e-06,
-      "loss": 0.4145,
-      "step": 5605
-    },
-    {
-      "epoch": 2.752688172043011,
-      "grad_norm": 0.42792597729641585,
-      "learning_rate": 4.627790600998005e-06,
-      "loss": 0.437,
-      "step": 5606
-    },
-    {
-      "epoch": 2.753179723502304,
-      "grad_norm": 0.42765023402346036,
-      "learning_rate": 4.6092935144584685e-06,
-      "loss": 0.3702,
-      "step": 5607
-    },
-    {
-      "epoch": 2.7536712749615977,
-      "grad_norm": 0.4026404175611252,
-      "learning_rate": 4.590832595951777e-06,
-      "loss": 0.4171,
-      "step": 5608
-    },
-    {
-      "epoch": 2.754162826420891,
-      "grad_norm": 0.40278418056646736,
-      "learning_rate": 4.572407852477511e-06,
-      "loss": 0.419,
-      "step": 5609
-    },
-    {
-      "epoch": 2.7546543778801844,
-      "grad_norm": 0.4238053397903939,
-      "learning_rate": 4.554019291021538e-06,
-      "loss": 0.4679,
-      "step": 5610
-    },
-    {
-      "epoch": 2.755145929339478,
-      "grad_norm": 0.40736301230045824,
-      "learning_rate": 4.535666918556003e-06,
-      "loss": 0.3774,
-      "step": 5611
-    },
-    {
-      "epoch": 2.755637480798771,
-      "grad_norm": 0.4264247142085447,
-      "learning_rate": 4.517350742039339e-06,
-      "loss": 0.4313,
-      "step": 5612
-    },
-    {
-      "epoch": 2.7561290322580643,
-      "grad_norm": 0.39867246605189943,
-      "learning_rate": 4.499070768416225e-06,
-      "loss": 0.3814,
-      "step": 5613
-    },
-    {
-      "epoch": 2.756620583717358,
-      "grad_norm": 0.42287412396206325,
-      "learning_rate": 4.480827004617682e-06,
-      "loss": 0.4587,
-      "step": 5614
-    },
-    {
-      "epoch": 2.7571121351766514,
-      "grad_norm": 0.38744592745914636,
-      "learning_rate": 4.462619457560913e-06,
-      "loss": 0.4094,
-      "step": 5615
-    },
-    {
-      "epoch": 2.7576036866359446,
-      "grad_norm": 0.4438337583512388,
-      "learning_rate": 4.4444481341494595e-06,
-      "loss": 0.4245,
-      "step": 5616
-    },
-    {
-      "epoch": 2.758095238095238,
-      "grad_norm": 0.41462000850833064,
-      "learning_rate": 4.426313041273089e-06,
-      "loss": 0.4046,
-      "step": 5617
-    },
-    {
-      "epoch": 2.7585867895545313,
-      "grad_norm": 0.4128230230223947,
-      "learning_rate": 4.408214185807846e-06,
-      "loss": 0.4101,
-      "step": 5618
-    },
-    {
-      "epoch": 2.759078341013825,
-      "grad_norm": 0.43699350736254866,
-      "learning_rate": 4.390151574616031e-06,
-      "loss": 0.4778,
-      "step": 5619
-    },
-    {
-      "epoch": 2.7595698924731185,
-      "grad_norm": 0.4279502071987037,
-      "learning_rate": 4.37212521454623e-06,
-      "loss": 0.468,
-      "step": 5620
-    },
-    {
-      "epoch": 2.7600614439324116,
-      "grad_norm": 0.43204123008984363,
-      "learning_rate": 4.354135112433233e-06,
-      "loss": 0.395,
-      "step": 5621
-    },
-    {
-      "epoch": 2.760552995391705,
-      "grad_norm": 0.3914407994339302,
-      "learning_rate": 4.336181275098106e-06,
-      "loss": 0.3932,
-      "step": 5622
-    },
-    {
-      "epoch": 2.7610445468509983,
-      "grad_norm": 0.4130162880384426,
-      "learning_rate": 4.318263709348203e-06,
-      "loss": 0.4206,
-      "step": 5623
-    },
-    {
-      "epoch": 2.761536098310292,
-      "grad_norm": 0.39264128575213614,
-      "learning_rate": 4.300382421977034e-06,
-      "loss": 0.3882,
-      "step": 5624
-    },
-    {
-      "epoch": 2.7620276497695855,
-      "grad_norm": 0.428753777234044,
-      "learning_rate": 4.2825374197644764e-06,
-      "loss": 0.423,
-      "step": 5625
-    },
-    {
-      "epoch": 2.7625192012288786,
-      "grad_norm": 0.39258652168306096,
-      "learning_rate": 4.264728709476529e-06,
-      "loss": 0.3975,
-      "step": 5626
-    },
-    {
-      "epoch": 2.763010752688172,
-      "grad_norm": 0.4221328929868399,
-      "learning_rate": 4.246956297865512e-06,
-      "loss": 0.4395,
-      "step": 5627
-    },
-    {
-      "epoch": 2.7635023041474653,
-      "grad_norm": 0.38856552140373496,
-      "learning_rate": 4.229220191669947e-06,
-      "loss": 0.3835,
-      "step": 5628
-    },
-    {
-      "epoch": 2.763993855606759,
-      "grad_norm": 0.43274506749309355,
-      "learning_rate": 4.21152039761461e-06,
-      "loss": 0.3924,
-      "step": 5629
-    },
-    {
-      "epoch": 2.7644854070660525,
-      "grad_norm": 0.43860814003791937,
-      "learning_rate": 4.193856922410466e-06,
-      "loss": 0.4258,
-      "step": 5630
-    },
-    {
-      "epoch": 2.7649769585253456,
-      "grad_norm": 0.40890393389705854,
-      "learning_rate": 4.176229772754803e-06,
-      "loss": 0.4141,
-      "step": 5631
-    },
-    {
-      "epoch": 2.765468509984639,
-      "grad_norm": 0.38812692467616483,
-      "learning_rate": 4.158638955331007e-06,
-      "loss": 0.4021,
-      "step": 5632
-    },
-    {
-      "epoch": 2.7659600614439324,
-      "grad_norm": 0.4090721550067412,
-      "learning_rate": 4.141084476808799e-06,
-      "loss": 0.4157,
-      "step": 5633
-    },
-    {
-      "epoch": 2.766451612903226,
-      "grad_norm": 0.4541326062938184,
-      "learning_rate": 4.123566343844054e-06,
-      "loss": 0.4578,
-      "step": 5634
-    },
-    {
-      "epoch": 2.766943164362519,
-      "grad_norm": 0.39586760511797603,
-      "learning_rate": 4.106084563078916e-06,
-      "loss": 0.4015,
-      "step": 5635
-    },
-    {
-      "epoch": 2.7674347158218127,
-      "grad_norm": 0.43934476361263725,
-      "learning_rate": 4.088639141141692e-06,
-      "loss": 0.4286,
-      "step": 5636
-    },
-    {
-      "epoch": 2.767926267281106,
-      "grad_norm": 0.4277219137317346,
-      "learning_rate": 4.071230084646949e-06,
-      "loss": 0.4711,
-      "step": 5637
-    },
-    {
-      "epoch": 2.7684178187403994,
-      "grad_norm": 0.3924805623423362,
-      "learning_rate": 4.0538574001954485e-06,
-      "loss": 0.4207,
-      "step": 5638
-    },
-    {
-      "epoch": 2.768909370199693,
-      "grad_norm": 0.41177963644920046,
-      "learning_rate": 4.036521094374146e-06,
-      "loss": 0.4144,
-      "step": 5639
-    },
-    {
-      "epoch": 2.769400921658986,
-      "grad_norm": 0.4583722326715423,
-      "learning_rate": 4.01922117375626e-06,
-      "loss": 0.4573,
-      "step": 5640
-    },
-    {
-      "epoch": 2.7698924731182797,
-      "grad_norm": 0.4195845603490557,
-      "learning_rate": 4.001957644901122e-06,
-      "loss": 0.3999,
-      "step": 5641
-    },
-    {
-      "epoch": 2.770384024577573,
-      "grad_norm": 0.410935593144402,
-      "learning_rate": 3.98473051435434e-06,
-      "loss": 0.4129,
-      "step": 5642
-    },
-    {
-      "epoch": 2.7708755760368664,
-      "grad_norm": 0.37260710139029496,
-      "learning_rate": 3.9675397886477e-06,
-      "loss": 0.3788,
-      "step": 5643
-    },
-    {
-      "epoch": 2.77136712749616,
-      "grad_norm": 0.42630720479982703,
-      "learning_rate": 3.95038547429919e-06,
-      "loss": 0.4037,
-      "step": 5644
-    },
-    {
-      "epoch": 2.771858678955453,
-      "grad_norm": 0.401648370479641,
-      "learning_rate": 3.93326757781296e-06,
-      "loss": 0.3904,
-      "step": 5645
-    },
-    {
-      "epoch": 2.7723502304147467,
-      "grad_norm": 0.4067435753332641,
-      "learning_rate": 3.916186105679387e-06,
-      "loss": 0.4243,
-      "step": 5646
-    },
-    {
-      "epoch": 2.77284178187404,
-      "grad_norm": 0.4179595838433307,
-      "learning_rate": 3.8991410643750335e-06,
-      "loss": 0.4237,
-      "step": 5647
-    },
-    {
-      "epoch": 2.7733333333333334,
-      "grad_norm": 0.4186244279095402,
-      "learning_rate": 3.882132460362631e-06,
-      "loss": 0.3977,
-      "step": 5648
-    },
-    {
-      "epoch": 2.7738248847926266,
-      "grad_norm": 0.387260809699995,
-      "learning_rate": 3.865160300091131e-06,
-      "loss": 0.4247,
-      "step": 5649
-    },
-    {
-      "epoch": 2.77431643625192,
-      "grad_norm": 0.4401140091413137,
-      "learning_rate": 3.8482245899956085e-06,
-      "loss": 0.4292,
-      "step": 5650
-    },
-    {
-      "epoch": 2.7748079877112133,
-      "grad_norm": 0.3990434151324543,
-      "learning_rate": 3.8313253364973935e-06,
-      "loss": 0.4188,
-      "step": 5651
-    },
-    {
-      "epoch": 2.775299539170507,
-      "grad_norm": 0.39905339579254295,
-      "learning_rate": 3.814462546003894e-06,
-      "loss": 0.4141,
-      "step": 5652
-    },
-    {
-      "epoch": 2.7757910906298005,
-      "grad_norm": 0.41792089931492427,
-      "learning_rate": 3.797636224908807e-06,
-      "loss": 0.4011,
-      "step": 5653
-    },
-    {
-      "epoch": 2.7762826420890936,
-      "grad_norm": 0.4114884127456886,
-      "learning_rate": 3.780846379591929e-06,
-      "loss": 0.4159,
-      "step": 5654
-    },
-    {
-      "epoch": 2.776774193548387,
-      "grad_norm": 0.4385959643072983,
-      "learning_rate": 3.7640930164192345e-06,
-      "loss": 0.4119,
-      "step": 5655
-    },
-    {
-      "epoch": 2.7772657450076803,
-      "grad_norm": 0.45062375467929294,
-      "learning_rate": 3.7473761417428865e-06,
-      "loss": 0.4371,
-      "step": 5656
-    },
-    {
-      "epoch": 2.777757296466974,
-      "grad_norm": 0.4034404366841015,
-      "learning_rate": 3.730695761901193e-06,
-      "loss": 0.4112,
-      "step": 5657
-    },
-    {
-      "epoch": 2.7782488479262675,
-      "grad_norm": 0.3915928507522561,
-      "learning_rate": 3.714051883218628e-06,
-      "loss": 0.4269,
-      "step": 5658
-    },
-    {
-      "epoch": 2.7787403993855606,
-      "grad_norm": 0.41321988653670255,
-      "learning_rate": 3.697444512005832e-06,
-      "loss": 0.4226,
-      "step": 5659
-    },
-    {
-      "epoch": 2.779231950844854,
-      "grad_norm": 0.4199415240561289,
-      "learning_rate": 3.6808736545596356e-06,
-      "loss": 0.4381,
-      "step": 5660
-    },
-    {
-      "epoch": 2.7797235023041473,
-      "grad_norm": 0.4141725330701319,
-      "learning_rate": 3.664339317162935e-06,
-      "loss": 0.3977,
-      "step": 5661
-    },
-    {
-      "epoch": 2.780215053763441,
-      "grad_norm": 0.4235933545288212,
-      "learning_rate": 3.6478415060848812e-06,
-      "loss": 0.4248,
-      "step": 5662
-    },
-    {
-      "epoch": 2.7807066052227345,
-      "grad_norm": 0.42384951392101605,
-      "learning_rate": 3.631380227580716e-06,
-      "loss": 0.4235,
-      "step": 5663
-    },
-    {
-      "epoch": 2.7811981566820276,
-      "grad_norm": 0.40767940713545203,
-      "learning_rate": 3.6149554878918466e-06,
-      "loss": 0.4271,
-      "step": 5664
-    },
-    {
-      "epoch": 2.7816897081413208,
-      "grad_norm": 0.4423360687014389,
-      "learning_rate": 3.5985672932458136e-06,
-      "loss": 0.4234,
-      "step": 5665
-    },
-    {
-      "epoch": 2.7821812596006144,
-      "grad_norm": 0.40300178693165634,
-      "learning_rate": 3.582215649856335e-06,
-      "loss": 0.3895,
-      "step": 5666
-    },
-    {
-      "epoch": 2.782672811059908,
-      "grad_norm": 0.48277576185163745,
-      "learning_rate": 3.56590056392323e-06,
-      "loss": 0.405,
-      "step": 5667
-    },
-    {
-      "epoch": 2.783164362519201,
-      "grad_norm": 0.434678155230467,
-      "learning_rate": 3.549622041632461e-06,
-      "loss": 0.5053,
-      "step": 5668
-    },
-    {
-      "epoch": 2.7836559139784947,
-      "grad_norm": 0.4191663329057678,
-      "learning_rate": 3.5333800891561798e-06,
-      "loss": 0.4165,
-      "step": 5669
-    },
-    {
-      "epoch": 2.784147465437788,
-      "grad_norm": 0.4007109985950894,
-      "learning_rate": 3.5171747126525823e-06,
-      "loss": 0.3701,
-      "step": 5670
-    },
-    {
-      "epoch": 2.7846390168970814,
-      "grad_norm": 0.42808063830432996,
-      "learning_rate": 3.501005918266087e-06,
-      "loss": 0.4057,
-      "step": 5671
-    },
-    {
-      "epoch": 2.785130568356375,
-      "grad_norm": 0.41227216033941905,
-      "learning_rate": 3.4848737121271903e-06,
-      "loss": 0.4517,
-      "step": 5672
-    },
-    {
-      "epoch": 2.785622119815668,
-      "grad_norm": 0.4105400060865333,
-      "learning_rate": 3.468778100352532e-06,
-      "loss": 0.3813,
-      "step": 5673
-    },
-    {
-      "epoch": 2.7861136712749617,
-      "grad_norm": 0.39665087383048686,
-      "learning_rate": 3.4527190890448535e-06,
-      "loss": 0.3737,
-      "step": 5674
-    },
-    {
-      "epoch": 2.786605222734255,
-      "grad_norm": 0.4369044762828287,
-      "learning_rate": 3.4366966842930614e-06,
-      "loss": 0.4399,
-      "step": 5675
-    },
-    {
-      "epoch": 2.7870967741935484,
-      "grad_norm": 0.412136990828404,
-      "learning_rate": 3.4207108921721296e-06,
-      "loss": 0.3865,
-      "step": 5676
-    },
-    {
-      "epoch": 2.787588325652842,
-      "grad_norm": 0.4064981349120306,
-      "learning_rate": 3.4047617187432213e-06,
-      "loss": 0.4053,
-      "step": 5677
-    },
-    {
-      "epoch": 2.788079877112135,
-      "grad_norm": 0.426782122079462,
-      "learning_rate": 3.388849170053532e-06,
-      "loss": 0.4463,
-      "step": 5678
-    },
-    {
-      "epoch": 2.7885714285714287,
-      "grad_norm": 0.42049598883352435,
-      "learning_rate": 3.3729732521364355e-06,
-      "loss": 0.4622,
-      "step": 5679
-    },
-    {
-      "epoch": 2.789062980030722,
-      "grad_norm": 0.41414703388627117,
-      "learning_rate": 3.357133971011395e-06,
-      "loss": 0.3732,
-      "step": 5680
-    },
-    {
-      "epoch": 2.7895545314900154,
-      "grad_norm": 0.41599541935514955,
-      "learning_rate": 3.341331332683972e-06,
-      "loss": 0.4099,
-      "step": 5681
-    },
-    {
-      "epoch": 2.790046082949309,
-      "grad_norm": 0.44442572494419924,
-      "learning_rate": 3.3255653431458533e-06,
-      "loss": 0.3876,
-      "step": 5682
-    },
-    {
-      "epoch": 2.790537634408602,
-      "grad_norm": 0.4288570826018885,
-      "learning_rate": 3.3098360083748005e-06,
-      "loss": 0.4342,
-      "step": 5683
-    },
-    {
-      "epoch": 2.7910291858678953,
-      "grad_norm": 5.013046468247719,
-      "learning_rate": 3.2941433343347205e-06,
-      "loss": 0.4303,
-      "step": 5684
-    },
-    {
-      "epoch": 2.791520737327189,
-      "grad_norm": 0.38606143503997387,
-      "learning_rate": 3.2784873269755766e-06,
-      "loss": 0.3961,
-      "step": 5685
-    },
-    {
-      "epoch": 2.7920122887864824,
-      "grad_norm": 0.4076876178372838,
-      "learning_rate": 3.2628679922334872e-06,
-      "loss": 0.3995,
-      "step": 5686
-    },
-    {
-      "epoch": 2.7925038402457756,
-      "grad_norm": 0.3847657745305065,
-      "learning_rate": 3.2472853360305813e-06,
-      "loss": 0.3966,
-      "step": 5687
-    },
-    {
-      "epoch": 2.792995391705069,
-      "grad_norm": 0.40583672033690993,
-      "learning_rate": 3.231739364275155e-06,
-      "loss": 0.4108,
-      "step": 5688
-    },
-    {
-      "epoch": 2.7934869431643623,
-      "grad_norm": 0.4022656754911858,
-      "learning_rate": 3.21623008286156e-06,
-      "loss": 0.4315,
-      "step": 5689
-    },
-    {
-      "epoch": 2.793978494623656,
-      "grad_norm": 0.4084874921285853,
-      "learning_rate": 3.200757497670259e-06,
-      "loss": 0.4291,
-      "step": 5690
-    },
-    {
-      "epoch": 2.7944700460829495,
-      "grad_norm": 0.42821729610813675,
-      "learning_rate": 3.1853216145677824e-06,
-      "loss": 0.4211,
-      "step": 5691
-    },
-    {
-      "epoch": 2.7949615975422426,
-      "grad_norm": 0.4590091225565753,
-      "learning_rate": 3.169922439406736e-06,
-      "loss": 0.3995,
-      "step": 5692
-    },
-    {
-      "epoch": 2.795453149001536,
-      "grad_norm": 0.4174708604215899,
-      "learning_rate": 3.1545599780258393e-06,
-      "loss": 0.3952,
-      "step": 5693
-    },
-    {
-      "epoch": 2.7959447004608293,
-      "grad_norm": 0.42596384169116536,
-      "learning_rate": 3.139234236249844e-06,
-      "loss": 0.4209,
-      "step": 5694
-    },
-    {
-      "epoch": 2.796436251920123,
-      "grad_norm": 0.3928036464119011,
-      "learning_rate": 3.123945219889657e-06,
-      "loss": 0.3819,
-      "step": 5695
-    },
-    {
-      "epoch": 2.7969278033794165,
-      "grad_norm": 0.3969523013562946,
-      "learning_rate": 3.1086929347421635e-06,
-      "loss": 0.4205,
-      "step": 5696
-    },
-    {
-      "epoch": 2.7974193548387096,
-      "grad_norm": 0.40642537323319916,
-      "learning_rate": 3.0934773865904155e-06,
-      "loss": 0.4446,
-      "step": 5697
-    },
-    {
-      "epoch": 2.797910906298003,
-      "grad_norm": 0.4162400605584403,
-      "learning_rate": 3.0782985812034536e-06,
-      "loss": 0.3912,
-      "step": 5698
-    },
-    {
-      "epoch": 2.7984024577572963,
-      "grad_norm": 0.38699502084153975,
-      "learning_rate": 3.0631565243364525e-06,
-      "loss": 0.4146,
-      "step": 5699
-    },
-    {
-      "epoch": 2.79889400921659,
-      "grad_norm": 0.3850118916386613,
-      "learning_rate": 3.048051221730597e-06,
-      "loss": 0.4017,
-      "step": 5700
-    },
-    {
-      "epoch": 2.7993855606758835,
-      "grad_norm": 0.4031761455330946,
-      "learning_rate": 3.0329826791131945e-06,
-      "loss": 0.399,
-      "step": 5701
-    },
-    {
-      "epoch": 2.7998771121351766,
-      "grad_norm": 0.4332592857247782,
-      "learning_rate": 3.017950902197575e-06,
-      "loss": 0.4331,
-      "step": 5702
-    },
-    {
-      "epoch": 2.80036866359447,
-      "grad_norm": 0.4063561963438648,
-      "learning_rate": 3.002955896683124e-06,
-      "loss": 0.3997,
-      "step": 5703
-    },
-    {
-      "epoch": 2.8008602150537634,
-      "grad_norm": 0.3917812144849345,
-      "learning_rate": 2.9879976682553157e-06,
-      "loss": 0.4258,
-      "step": 5704
-    },
-    {
-      "epoch": 2.801351766513057,
-      "grad_norm": 0.4248388162898716,
-      "learning_rate": 2.973076222585647e-06,
-      "loss": 0.3986,
-      "step": 5705
-    },
-    {
-      "epoch": 2.80184331797235,
-      "grad_norm": 0.4275315357467191,
-      "learning_rate": 2.958191565331725e-06,
-      "loss": 0.4156,
-      "step": 5706
-    },
-    {
-      "epoch": 2.8023348694316437,
-      "grad_norm": 0.42512752944847276,
-      "learning_rate": 2.9433437021371136e-06,
-      "loss": 0.3989,
-      "step": 5707
-    },
-    {
-      "epoch": 2.802826420890937,
-      "grad_norm": 0.3935426785444233,
-      "learning_rate": 2.9285326386315203e-06,
-      "loss": 0.3935,
-      "step": 5708
-    },
-    {
-      "epoch": 2.8033179723502304,
-      "grad_norm": 0.37791572685851643,
-      "learning_rate": 2.9137583804306423e-06,
-      "loss": 0.4022,
-      "step": 5709
-    },
-    {
-      "epoch": 2.803809523809524,
-      "grad_norm": 1.7350169441101304,
-      "learning_rate": 2.899020933136254e-06,
-      "loss": 0.4593,
-      "step": 5710
-    },
-    {
-      "epoch": 2.804301075268817,
-      "grad_norm": 0.38052940572165195,
-      "learning_rate": 2.88432030233613e-06,
-      "loss": 0.3923,
-      "step": 5711
-    },
-    {
-      "epoch": 2.8047926267281107,
-      "grad_norm": 0.4248541707131468,
-      "learning_rate": 2.869656493604156e-06,
-      "loss": 0.427,
-      "step": 5712
-    },
-    {
-      "epoch": 2.805284178187404,
-      "grad_norm": 0.43210340363530475,
-      "learning_rate": 2.8550295125001845e-06,
-      "loss": 0.443,
-      "step": 5713
-    },
-    {
-      "epoch": 2.8057757296466974,
-      "grad_norm": 0.43254811666951304,
-      "learning_rate": 2.8404393645701243e-06,
-      "loss": 0.4192,
-      "step": 5714
-    },
-    {
-      "epoch": 2.806267281105991,
-      "grad_norm": 0.43105457684232945,
-      "learning_rate": 2.825886055345983e-06,
-      "loss": 0.4302,
-      "step": 5715
-    },
-    {
-      "epoch": 2.806758832565284,
-      "grad_norm": 0.4305166931578773,
-      "learning_rate": 2.8113695903456804e-06,
-      "loss": 0.4204,
-      "step": 5716
-    },
-    {
-      "epoch": 2.8072503840245777,
-      "grad_norm": 0.4666338078365619,
-      "learning_rate": 2.7968899750732692e-06,
-      "loss": 0.4434,
-      "step": 5717
-    },
-    {
-      "epoch": 2.807741935483871,
-      "grad_norm": 0.4144834502994846,
-      "learning_rate": 2.7824472150187907e-06,
-      "loss": 0.4289,
-      "step": 5718
-    },
-    {
-      "epoch": 2.8082334869431644,
-      "grad_norm": 0.4190753678398001,
-      "learning_rate": 2.7680413156583094e-06,
-      "loss": 0.4299,
-      "step": 5719
-    },
-    {
-      "epoch": 2.808725038402458,
-      "grad_norm": 0.411122997987756,
-      "learning_rate": 2.753672282453912e-06,
-      "loss": 0.3866,
-      "step": 5720
-    },
-    {
-      "epoch": 2.809216589861751,
-      "grad_norm": 0.40404884890729637,
-      "learning_rate": 2.7393401208537394e-06,
-      "loss": 0.3839,
-      "step": 5721
-    },
-    {
-      "epoch": 2.8097081413210443,
-      "grad_norm": 0.408315575972837,
-      "learning_rate": 2.7250448362919013e-06,
-      "loss": 0.3927,
-      "step": 5722
-    },
-    {
-      "epoch": 2.810199692780338,
-      "grad_norm": 0.3986494417059805,
-      "learning_rate": 2.7107864341885725e-06,
-      "loss": 0.3804,
-      "step": 5723
-    },
-    {
-      "epoch": 2.8106912442396315,
-      "grad_norm": 0.413264261548953,
-      "learning_rate": 2.6965649199499064e-06,
-      "loss": 0.4156,
-      "step": 5724
-    },
-    {
-      "epoch": 2.8111827956989246,
-      "grad_norm": 0.39473403982437766,
-      "learning_rate": 2.6823802989680903e-06,
-      "loss": 0.3787,
-      "step": 5725
-    },
-    {
-      "epoch": 2.811674347158218,
-      "grad_norm": 0.36852636823131574,
-      "learning_rate": 2.6682325766213323e-06,
-      "loss": 0.3661,
-      "step": 5726
-    },
-    {
-      "epoch": 2.8121658986175113,
-      "grad_norm": 0.4069134695206052,
-      "learning_rate": 2.65412175827382e-06,
-      "loss": 0.4346,
-      "step": 5727
-    },
-    {
-      "epoch": 2.812657450076805,
-      "grad_norm": 0.39604692783506845,
-      "learning_rate": 2.640047849275784e-06,
-      "loss": 0.4139,
-      "step": 5728
-    },
-    {
-      "epoch": 2.8131490015360985,
-      "grad_norm": 0.41716189376104984,
-      "learning_rate": 2.6260108549634234e-06,
-      "loss": 0.4283,
-      "step": 5729
-    },
-    {
-      "epoch": 2.8136405529953916,
-      "grad_norm": 0.39628912016162815,
-      "learning_rate": 2.612010780658969e-06,
-      "loss": 0.4006,
-      "step": 5730
-    },
-    {
-      "epoch": 2.814132104454685,
-      "grad_norm": 0.4063347417350446,
-      "learning_rate": 2.5980476316706303e-06,
-      "loss": 0.439,
-      "step": 5731
-    },
-    {
-      "epoch": 2.8146236559139783,
-      "grad_norm": 0.3868052037666175,
-      "learning_rate": 2.5841214132926728e-06,
-      "loss": 0.4166,
-      "step": 5732
-    },
-    {
-      "epoch": 2.815115207373272,
-      "grad_norm": 0.41930073936898876,
-      "learning_rate": 2.5702321308052504e-06,
-      "loss": 0.4005,
-      "step": 5733
-    },
-    {
-      "epoch": 2.8156067588325655,
-      "grad_norm": 0.3729774585601725,
-      "learning_rate": 2.5563797894746388e-06,
-      "loss": 0.3771,
-      "step": 5734
-    },
-    {
-      "epoch": 2.8160983102918586,
-      "grad_norm": 0.40110216338081484,
-      "learning_rate": 2.5425643945530153e-06,
-      "loss": 0.4124,
-      "step": 5735
-    },
-    {
-      "epoch": 2.8165898617511522,
-      "grad_norm": 0.44083808008431297,
-      "learning_rate": 2.5287859512785894e-06,
-      "loss": 0.4326,
-      "step": 5736
-    },
-    {
-      "epoch": 2.8170814132104454,
-      "grad_norm": 0.41176873783279533,
-      "learning_rate": 2.5150444648755487e-06,
-      "loss": 0.4083,
-      "step": 5737
-    },
-    {
-      "epoch": 2.817572964669739,
-      "grad_norm": 0.4323226659220806,
-      "learning_rate": 2.5013399405540706e-06,
-      "loss": 0.4449,
-      "step": 5738
-    },
-    {
-      "epoch": 2.8180645161290325,
-      "grad_norm": 0.4203322896243011,
-      "learning_rate": 2.4876723835103198e-06,
-      "loss": 0.4174,
-      "step": 5739
-    },
-    {
-      "epoch": 2.8185560675883257,
-      "grad_norm": 0.40702076832462786,
-      "learning_rate": 2.4740417989264408e-06,
-      "loss": 0.3962,
-      "step": 5740
-    },
-    {
-      "epoch": 2.819047619047619,
-      "grad_norm": 0.6976067687268606,
-      "learning_rate": 2.4604481919705767e-06,
-      "loss": 0.434,
-      "step": 5741
-    },
-    {
-      "epoch": 2.8195391705069124,
-      "grad_norm": 0.3947677501029888,
-      "learning_rate": 2.446891567796805e-06,
-      "loss": 0.4266,
-      "step": 5742
-    },
-    {
-      "epoch": 2.820030721966206,
-      "grad_norm": 0.4386189304416501,
-      "learning_rate": 2.433371931545236e-06,
-      "loss": 0.4462,
-      "step": 5743
-    },
-    {
-      "epoch": 2.820522273425499,
-      "grad_norm": 0.3683482292160676,
-      "learning_rate": 2.4198892883419256e-06,
-      "loss": 0.3626,
-      "step": 5744
-    },
-    {
-      "epoch": 2.8210138248847927,
-      "grad_norm": 0.4366764917625803,
-      "learning_rate": 2.4064436432989166e-06,
-      "loss": 0.4299,
-      "step": 5745
-    },
-    {
-      "epoch": 2.821505376344086,
-      "grad_norm": 0.4602580056734708,
-      "learning_rate": 2.39303500151421e-06,
-      "loss": 0.4079,
-      "step": 5746
-    },
-    {
-      "epoch": 2.8219969278033794,
-      "grad_norm": 0.41933360010206094,
-      "learning_rate": 2.3796633680717716e-06,
-      "loss": 0.4164,
-      "step": 5747
-    },
-    {
-      "epoch": 2.822488479262673,
-      "grad_norm": 0.40077504255731666,
-      "learning_rate": 2.3663287480415688e-06,
-      "loss": 0.4145,
-      "step": 5748
-    },
-    {
-      "epoch": 2.822980030721966,
-      "grad_norm": 0.43007781093624364,
-      "learning_rate": 2.3530311464794895e-06,
-      "loss": 0.4136,
-      "step": 5749
-    },
-    {
-      "epoch": 2.8234715821812597,
-      "grad_norm": 0.41898990373912737,
-      "learning_rate": 2.3397705684274353e-06,
-      "loss": 0.4609,
-      "step": 5750
-    },
-    {
-      "epoch": 2.823963133640553,
-      "grad_norm": 0.3971932682184444,
-      "learning_rate": 2.3265470189132165e-06,
-      "loss": 0.3969,
-      "step": 5751
-    },
-    {
-      "epoch": 2.8244546850998464,
-      "grad_norm": 0.41822934855331323,
-      "learning_rate": 2.3133605029506567e-06,
-      "loss": 0.3937,
-      "step": 5752
-    },
-    {
-      "epoch": 2.82494623655914,
-      "grad_norm": 0.41080572721083625,
-      "learning_rate": 2.3002110255394894e-06,
-      "loss": 0.4258,
-      "step": 5753
-    },
-    {
-      "epoch": 2.825437788018433,
-      "grad_norm": 0.37415712009570373,
-      "learning_rate": 2.2870985916654487e-06,
-      "loss": 0.3777,
-      "step": 5754
-    },
-    {
-      "epoch": 2.8259293394777267,
-      "grad_norm": 0.41756880594908885,
-      "learning_rate": 2.274023206300202e-06,
-      "loss": 0.4352,
-      "step": 5755
-    },
-    {
-      "epoch": 2.82642089093702,
-      "grad_norm": 0.45221324718717576,
-      "learning_rate": 2.2609848744013505e-06,
-      "loss": 0.4087,
-      "step": 5756
-    },
-    {
-      "epoch": 2.8269124423963135,
-      "grad_norm": 0.42645704169630394,
-      "learning_rate": 2.247983600912473e-06,
-      "loss": 0.4395,
-      "step": 5757
-    },
-    {
-      "epoch": 2.827403993855607,
-      "grad_norm": 0.3915655944202693,
-      "learning_rate": 2.2350193907631157e-06,
-      "loss": 0.3729,
-      "step": 5758
-    },
-    {
-      "epoch": 2.8278955453149,
-      "grad_norm": 0.39172795396693244,
-      "learning_rate": 2.2220922488687235e-06,
-      "loss": 0.4122,
-      "step": 5759
-    },
-    {
-      "epoch": 2.8283870967741933,
-      "grad_norm": 0.42898796785421117,
-      "learning_rate": 2.2092021801306983e-06,
-      "loss": 0.4358,
-      "step": 5760
-    },
-    {
-      "epoch": 2.828878648233487,
-      "grad_norm": 0.4170541030741761,
-      "learning_rate": 2.196349189436431e-06,
-      "loss": 0.4144,
-      "step": 5761
-    },
-    {
-      "epoch": 2.8293701996927805,
-      "grad_norm": 0.3871174685755417,
-      "learning_rate": 2.183533281659178e-06,
-      "loss": 0.403,
-      "step": 5762
-    },
-    {
-      "epoch": 2.8298617511520736,
-      "grad_norm": 0.43973428451434293,
-      "learning_rate": 2.17075446165822e-06,
-      "loss": 0.4252,
-      "step": 5763
-    },
-    {
-      "epoch": 2.830353302611367,
-      "grad_norm": 0.4360226716825237,
-      "learning_rate": 2.158012734278703e-06,
-      "loss": 0.419,
-      "step": 5764
-    },
-    {
-      "epoch": 2.8308448540706603,
-      "grad_norm": 0.40371625764307756,
-      "learning_rate": 2.145308104351762e-06,
-      "loss": 0.4155,
-      "step": 5765
-    },
-    {
-      "epoch": 2.831336405529954,
-      "grad_norm": 0.39080285430860423,
-      "learning_rate": 2.132640576694411e-06,
-      "loss": 0.3813,
-      "step": 5766
-    },
-    {
-      "epoch": 2.8318279569892475,
-      "grad_norm": 0.4175151123713847,
-      "learning_rate": 2.120010156109675e-06,
-      "loss": 0.4102,
-      "step": 5767
-    },
-    {
-      "epoch": 2.8323195084485406,
-      "grad_norm": 0.41894567973153435,
-      "learning_rate": 2.107416847386423e-06,
-      "loss": 0.4373,
-      "step": 5768
-    },
-    {
-      "epoch": 2.832811059907834,
-      "grad_norm": 0.4240410046404566,
-      "learning_rate": 2.094860655299513e-06,
-      "loss": 0.402,
-      "step": 5769
-    },
-    {
-      "epoch": 2.8333026113671274,
-      "grad_norm": 0.3986226573320233,
-      "learning_rate": 2.0823415846097037e-06,
-      "loss": 0.3772,
-      "step": 5770
-    },
-    {
-      "epoch": 2.833794162826421,
-      "grad_norm": 0.3849087000154648,
-      "learning_rate": 2.069859640063687e-06,
-      "loss": 0.3979,
-      "step": 5771
-    },
-    {
-      "epoch": 2.8342857142857145,
-      "grad_norm": 0.45059079762835375,
-      "learning_rate": 2.057414826394077e-06,
-      "loss": 0.4152,
-      "step": 5772
-    },
-    {
-      "epoch": 2.8347772657450077,
-      "grad_norm": 0.42056566150784974,
-      "learning_rate": 2.045007148319422e-06,
-      "loss": 0.4135,
-      "step": 5773
-    },
-    {
-      "epoch": 2.8352688172043012,
-      "grad_norm": 0.412838646332942,
-      "learning_rate": 2.0326366105441584e-06,
-      "loss": 0.4212,
-      "step": 5774
-    },
-    {
-      "epoch": 2.8357603686635944,
-      "grad_norm": 0.43651126081288794,
-      "learning_rate": 2.0203032177586566e-06,
-      "loss": 0.41,
-      "step": 5775
-    },
-    {
-      "epoch": 2.836251920122888,
-      "grad_norm": 0.41279042173374253,
-      "learning_rate": 2.0080069746392314e-06,
-      "loss": 0.4085,
-      "step": 5776
-    },
-    {
-      "epoch": 2.836743471582181,
-      "grad_norm": 0.4188024180669136,
-      "learning_rate": 1.995747885848054e-06,
-      "loss": 0.3796,
-      "step": 5777
-    },
-    {
-      "epoch": 2.8372350230414747,
-      "grad_norm": 0.40816271053360703,
-      "learning_rate": 1.983525956033272e-06,
-      "loss": 0.415,
-      "step": 5778
-    },
-    {
-      "epoch": 2.837726574500768,
-      "grad_norm": 0.4586830317603499,
-      "learning_rate": 1.9713411898288792e-06,
-      "loss": 0.4474,
-      "step": 5779
-    },
-    {
-      "epoch": 2.8382181259600614,
-      "grad_norm": 0.4147635867066052,
-      "learning_rate": 1.9591935918548464e-06,
-      "loss": 0.413,
-      "step": 5780
-    },
-    {
-      "epoch": 2.838709677419355,
-      "grad_norm": 0.4170261163736035,
-      "learning_rate": 1.9470831667170008e-06,
-      "loss": 0.4405,
-      "step": 5781
-    },
-    {
-      "epoch": 2.839201228878648,
-      "grad_norm": 0.40901257467283164,
-      "learning_rate": 1.9350099190071026e-06,
-      "loss": 0.433,
-      "step": 5782
-    },
-    {
-      "epoch": 2.8396927803379417,
-      "grad_norm": 0.4242527956777,
-      "learning_rate": 1.9229738533027897e-06,
-      "loss": 0.413,
-      "step": 5783
-    },
-    {
-      "epoch": 2.840184331797235,
-      "grad_norm": 0.4373874865628205,
-      "learning_rate": 1.9109749741676232e-06,
-      "loss": 0.4246,
-      "step": 5784
-    },
-    {
-      "epoch": 2.8406758832565284,
-      "grad_norm": 0.4231041511488596,
-      "learning_rate": 1.8990132861510635e-06,
-      "loss": 0.3885,
-      "step": 5785
-    },
-    {
-      "epoch": 2.841167434715822,
-      "grad_norm": 0.42181889900433195,
-      "learning_rate": 1.8870887937884606e-06,
-      "loss": 0.3774,
-      "step": 5786
-    },
-    {
-      "epoch": 2.841658986175115,
-      "grad_norm": 0.38920154502716,
-      "learning_rate": 1.875201501601087e-06,
-      "loss": 0.3991,
-      "step": 5787
-    },
-    {
-      "epoch": 2.8421505376344087,
-      "grad_norm": 0.46782773120654203,
-      "learning_rate": 1.8633514140960484e-06,
-      "loss": 0.4834,
-      "step": 5788
-    },
-    {
-      "epoch": 2.842642089093702,
-      "grad_norm": 0.386818692823787,
-      "learning_rate": 1.8515385357664283e-06,
-      "loss": 0.4026,
-      "step": 5789
-    },
-    {
-      "epoch": 2.8431336405529954,
-      "grad_norm": 0.40553074709774,
-      "learning_rate": 1.839762871091133e-06,
-      "loss": 0.3832,
-      "step": 5790
-    },
-    {
-      "epoch": 2.843625192012289,
-      "grad_norm": 0.41270964206202504,
-      "learning_rate": 1.828024424535002e-06,
-      "loss": 0.4305,
-      "step": 5791
-    },
-    {
-      "epoch": 2.844116743471582,
-      "grad_norm": 0.4465637385251171,
-      "learning_rate": 1.8163232005487418e-06,
-      "loss": 0.4342,
-      "step": 5792
-    },
-    {
-      "epoch": 2.8446082949308753,
-      "grad_norm": 0.4256923210296834,
-      "learning_rate": 1.804659203568937e-06,
-      "loss": 0.4174,
-      "step": 5793
-    },
-    {
-      "epoch": 2.845099846390169,
-      "grad_norm": 0.4432405410292091,
-      "learning_rate": 1.793032438018083e-06,
-      "loss": 0.4314,
-      "step": 5794
-    },
-    {
-      "epoch": 2.8455913978494625,
-      "grad_norm": 0.4304274915714804,
-      "learning_rate": 1.7814429083045424e-06,
-      "loss": 0.3916,
-      "step": 5795
-    },
-    {
-      "epoch": 2.8460829493087556,
-      "grad_norm": 0.3946844609662595,
-      "learning_rate": 1.7698906188225895e-06,
-      "loss": 0.4111,
-      "step": 5796
-    },
-    {
-      "epoch": 2.846574500768049,
-      "grad_norm": 0.42696319911981817,
-      "learning_rate": 1.758375573952309e-06,
-      "loss": 0.4272,
-      "step": 5797
-    },
-    {
-      "epoch": 2.8470660522273423,
-      "grad_norm": 0.4201958560260909,
-      "learning_rate": 1.7468977780597528e-06,
-      "loss": 0.3863,
-      "step": 5798
-    },
-    {
-      "epoch": 2.847557603686636,
-      "grad_norm": 0.3986373925787837,
-      "learning_rate": 1.7354572354967513e-06,
-      "loss": 0.3932,
-      "step": 5799
-    },
-    {
-      "epoch": 2.8480491551459295,
-      "grad_norm": 0.4004726172192835,
-      "learning_rate": 1.7240539506011234e-06,
-      "loss": 0.4057,
-      "step": 5800
-    },
-    {
-      "epoch": 2.8485407066052226,
-      "grad_norm": 0.3957133068653019,
-      "learning_rate": 1.7126879276964547e-06,
-      "loss": 0.4002,
-      "step": 5801
-    },
-    {
-      "epoch": 2.849032258064516,
-      "grad_norm": 0.44720652225597013,
-      "learning_rate": 1.7013591710922761e-06,
-      "loss": 0.4693,
-      "step": 5802
-    },
-    {
-      "epoch": 2.8495238095238093,
-      "grad_norm": 0.4045320606754897,
-      "learning_rate": 1.6900676850839515e-06,
-      "loss": 0.4166,
-      "step": 5803
-    },
-    {
-      "epoch": 2.850015360983103,
-      "grad_norm": 0.41359523976204987,
-      "learning_rate": 1.678813473952745e-06,
-      "loss": 0.3971,
-      "step": 5804
-    },
-    {
-      "epoch": 2.8505069124423965,
-      "grad_norm": 0.42316833962175493,
-      "learning_rate": 1.6675965419657325e-06,
-      "loss": 0.4342,
-      "step": 5805
-    },
-    {
-      "epoch": 2.8509984639016897,
-      "grad_norm": 0.4274162811885969,
-      "learning_rate": 1.6564168933759006e-06,
-      "loss": 0.4066,
-      "step": 5806
-    },
-    {
-      "epoch": 2.8514900153609832,
-      "grad_norm": 0.42895370674760874,
-      "learning_rate": 1.645274532422114e-06,
-      "loss": 0.4639,
-      "step": 5807
-    },
-    {
-      "epoch": 2.8519815668202764,
-      "grad_norm": 0.4082614372312397,
-      "learning_rate": 1.6341694633290495e-06,
-      "loss": 0.3878,
-      "step": 5808
-    },
-    {
-      "epoch": 2.85247311827957,
-      "grad_norm": 0.3925800970725663,
-      "learning_rate": 1.6231016903072715e-06,
-      "loss": 0.387,
-      "step": 5809
-    },
-    {
-      "epoch": 2.8529646697388635,
-      "grad_norm": 0.4224411583072324,
-      "learning_rate": 1.6120712175532128e-06,
-      "loss": 0.4214,
-      "step": 5810
-    },
-    {
-      "epoch": 2.8534562211981567,
-      "grad_norm": 0.4162720212427797,
-      "learning_rate": 1.6010780492491383e-06,
-      "loss": 0.4497,
-      "step": 5811
-    },
-    {
-      "epoch": 2.85394777265745,
-      "grad_norm": 0.4134220209944446,
-      "learning_rate": 1.5901221895631814e-06,
-      "loss": 0.4415,
-      "step": 5812
-    },
-    {
-      "epoch": 2.8544393241167434,
-      "grad_norm": 0.4178793963449324,
-      "learning_rate": 1.5792036426493517e-06,
-      "loss": 0.4488,
-      "step": 5813
-    },
-    {
-      "epoch": 2.854930875576037,
-      "grad_norm": 0.3958854792473097,
-      "learning_rate": 1.5683224126474604e-06,
-      "loss": 0.3668,
-      "step": 5814
-    },
-    {
-      "epoch": 2.85542242703533,
-      "grad_norm": 0.42444608032520237,
-      "learning_rate": 1.5574785036832297e-06,
-      "loss": 0.4643,
-      "step": 5815
-    },
-    {
-      "epoch": 2.8559139784946237,
-      "grad_norm": 0.40419884209556983,
-      "learning_rate": 1.546671919868181e-06,
-      "loss": 0.4401,
-      "step": 5816
-    },
-    {
-      "epoch": 2.856405529953917,
-      "grad_norm": 0.42815907611256276,
-      "learning_rate": 1.5359026652997044e-06,
-      "loss": 0.432,
-      "step": 5817
-    },
-    {
-      "epoch": 2.8568970814132104,
-      "grad_norm": 0.4094154752308921,
-      "learning_rate": 1.5251707440610552e-06,
-      "loss": 0.3749,
-      "step": 5818
-    },
-    {
-      "epoch": 2.857388632872504,
-      "grad_norm": 0.4217212400015568,
-      "learning_rate": 1.5144761602212899e-06,
-      "loss": 0.4236,
-      "step": 5819
-    },
-    {
-      "epoch": 2.857880184331797,
-      "grad_norm": 0.402599232840116,
-      "learning_rate": 1.5038189178353424e-06,
-      "loss": 0.3897,
-      "step": 5820
-    },
-    {
-      "epoch": 2.8583717357910907,
-      "grad_norm": 0.4471144273015755,
-      "learning_rate": 1.4931990209439805e-06,
-      "loss": 0.401,
-      "step": 5821
-    },
-    {
-      "epoch": 2.858863287250384,
-      "grad_norm": 0.42827422892425915,
-      "learning_rate": 1.4826164735738057e-06,
-      "loss": 0.4186,
-      "step": 5822
-    },
-    {
-      "epoch": 2.8593548387096774,
-      "grad_norm": 0.41381727659146794,
-      "learning_rate": 1.4720712797372638e-06,
-      "loss": 0.3949,
-      "step": 5823
-    },
-    {
-      "epoch": 2.859846390168971,
-      "grad_norm": 0.4076505586040814,
-      "learning_rate": 1.4615634434326453e-06,
-      "loss": 0.4389,
-      "step": 5824
-    },
-    {
-      "epoch": 2.860337941628264,
-      "grad_norm": 0.42843366544843803,
-      "learning_rate": 1.4510929686440412e-06,
-      "loss": 0.4608,
-      "step": 5825
-    },
-    {
-      "epoch": 2.8608294930875577,
-      "grad_norm": 0.42937302912340913,
-      "learning_rate": 1.4406598593414311e-06,
-      "loss": 0.3884,
-      "step": 5826
-    },
-    {
-      "epoch": 2.861321044546851,
-      "grad_norm": 0.39636500572588623,
-      "learning_rate": 1.4302641194805955e-06,
-      "loss": 0.3653,
-      "step": 5827
-    },
-    {
-      "epoch": 2.8618125960061445,
-      "grad_norm": 0.4058676567250964,
-      "learning_rate": 1.4199057530031367e-06,
-      "loss": 0.4351,
-      "step": 5828
-    },
-    {
-      "epoch": 2.862304147465438,
-      "grad_norm": 0.45160943605787573,
-      "learning_rate": 1.4095847638365133e-06,
-      "loss": 0.456,
-      "step": 5829
-    },
-    {
-      "epoch": 2.862795698924731,
-      "grad_norm": 0.40248423112533765,
-      "learning_rate": 1.399301155893995e-06,
-      "loss": 0.3999,
-      "step": 5830
-    },
-    {
-      "epoch": 2.8632872503840243,
-      "grad_norm": 0.45058274488231065,
-      "learning_rate": 1.3890549330746626e-06,
-      "loss": 0.4303,
-      "step": 5831
-    },
-    {
-      "epoch": 2.863778801843318,
-      "grad_norm": 0.417595846867743,
-      "learning_rate": 1.3788460992634644e-06,
-      "loss": 0.4307,
-      "step": 5832
-    },
-    {
-      "epoch": 2.8642703533026115,
-      "grad_norm": 0.42524426477909055,
-      "learning_rate": 1.3686746583311483e-06,
-      "loss": 0.3951,
-      "step": 5833
-    },
-    {
-      "epoch": 2.8647619047619046,
-      "grad_norm": 0.41566176406668565,
-      "learning_rate": 1.3585406141342517e-06,
-      "loss": 0.4337,
-      "step": 5834
-    },
-    {
-      "epoch": 2.865253456221198,
-      "grad_norm": 0.38436414496786353,
-      "learning_rate": 1.348443970515212e-06,
-      "loss": 0.3908,
-      "step": 5835
-    },
-    {
-      "epoch": 2.8657450076804913,
-      "grad_norm": 0.41940734233633625,
-      "learning_rate": 1.3383847313022224e-06,
-      "loss": 0.4346,
-      "step": 5836
-    },
-    {
-      "epoch": 2.866236559139785,
-      "grad_norm": 0.40964351098155755,
-      "learning_rate": 1.328362900309299e-06,
-      "loss": 0.4394,
-      "step": 5837
-    },
-    {
-      "epoch": 2.8667281105990785,
-      "grad_norm": 0.38558104252725756,
-      "learning_rate": 1.318378481336302e-06,
-      "loss": 0.4239,
-      "step": 5838
-    },
-    {
-      "epoch": 2.8672196620583716,
-      "grad_norm": 0.4397126380793387,
-      "learning_rate": 1.3084314781688923e-06,
-      "loss": 0.441,
-      "step": 5839
-    },
-    {
-      "epoch": 2.8677112135176652,
-      "grad_norm": 0.39452989709684694,
-      "learning_rate": 1.2985218945785304e-06,
-      "loss": 0.4113,
-      "step": 5840
-    },
-    {
-      "epoch": 2.8682027649769584,
-      "grad_norm": 0.4077174202666877,
-      "learning_rate": 1.2886497343225002e-06,
-      "loss": 0.3975,
-      "step": 5841
-    },
-    {
-      "epoch": 2.868694316436252,
-      "grad_norm": 0.40520758288034336,
-      "learning_rate": 1.2788150011439294e-06,
-      "loss": 0.3973,
-      "step": 5842
-    },
-    {
-      "epoch": 2.8691858678955455,
-      "grad_norm": 0.4158537857966892,
-      "learning_rate": 1.2690176987716907e-06,
-      "loss": 0.411,
-      "step": 5843
-    },
-    {
-      "epoch": 2.8696774193548387,
-      "grad_norm": 0.43070013333943247,
-      "learning_rate": 1.2592578309205017e-06,
-      "loss": 0.4,
-      "step": 5844
-    },
-    {
-      "epoch": 2.8701689708141322,
-      "grad_norm": 0.3975217116767363,
-      "learning_rate": 1.249535401290902e-06,
-      "loss": 0.4092,
-      "step": 5845
-    },
-    {
-      "epoch": 2.8706605222734254,
-      "grad_norm": 0.42259862112336966,
-      "learning_rate": 1.2398504135691991e-06,
-      "loss": 0.4187,
-      "step": 5846
-    },
-    {
-      "epoch": 2.871152073732719,
-      "grad_norm": 0.38713490890227364,
-      "learning_rate": 1.2302028714275215e-06,
-      "loss": 0.3954,
-      "step": 5847
-    },
-    {
-      "epoch": 2.8716436251920125,
-      "grad_norm": 0.38207755238597946,
-      "learning_rate": 1.2205927785238213e-06,
-      "loss": 0.3769,
-      "step": 5848
-    },
-    {
-      "epoch": 2.8721351766513057,
-      "grad_norm": 0.40466802391199014,
-      "learning_rate": 1.211020138501795e-06,
-      "loss": 0.3993,
-      "step": 5849
-    },
-    {
-      "epoch": 2.872626728110599,
-      "grad_norm": 0.3888509310388858,
-      "learning_rate": 1.2014849549910056e-06,
-      "loss": 0.3767,
-      "step": 5850
-    },
-    {
-      "epoch": 2.8731182795698924,
-      "grad_norm": 0.39918166555643847,
-      "learning_rate": 1.1919872316067726e-06,
-      "loss": 0.4239,
-      "step": 5851
-    },
-    {
-      "epoch": 2.873609831029186,
-      "grad_norm": 0.42365474077158444,
-      "learning_rate": 1.1825269719502041e-06,
-      "loss": 0.3996,
-      "step": 5852
-    },
-    {
-      "epoch": 2.874101382488479,
-      "grad_norm": 0.40383653043538387,
-      "learning_rate": 1.1731041796082531e-06,
-      "loss": 0.3815,
-      "step": 5853
-    },
-    {
-      "epoch": 2.8745929339477727,
-      "grad_norm": 0.3664272594363523,
-      "learning_rate": 1.1637188581536175e-06,
-      "loss": 0.3477,
-      "step": 5854
-    },
-    {
-      "epoch": 2.875084485407066,
-      "grad_norm": 0.4014211416926223,
-      "learning_rate": 1.1543710111448059e-06,
-      "loss": 0.3835,
-      "step": 5855
-    },
-    {
-      "epoch": 2.8755760368663594,
-      "grad_norm": 0.42697665438228166,
-      "learning_rate": 1.1450606421261167e-06,
-      "loss": 0.4205,
-      "step": 5856
-    },
-    {
-      "epoch": 2.876067588325653,
-      "grad_norm": 0.43225707718851697,
-      "learning_rate": 1.1357877546276485e-06,
-      "loss": 0.421,
-      "step": 5857
-    },
-    {
-      "epoch": 2.876559139784946,
-      "grad_norm": 0.38102777069228094,
-      "learning_rate": 1.1265523521652666e-06,
-      "loss": 0.3746,
-      "step": 5858
-    },
-    {
-      "epoch": 2.8770506912442397,
-      "grad_norm": 0.4062530572452384,
-      "learning_rate": 1.1173544382406476e-06,
-      "loss": 0.4073,
-      "step": 5859
-    },
-    {
-      "epoch": 2.877542242703533,
-      "grad_norm": 0.4367819112894127,
-      "learning_rate": 1.1081940163412352e-06,
-      "loss": 0.433,
-      "step": 5860
-    },
-    {
-      "epoch": 2.8780337941628265,
-      "grad_norm": 0.4337088517858264,
-      "learning_rate": 1.0990710899402733e-06,
-      "loss": 0.4007,
-      "step": 5861
-    },
-    {
-      "epoch": 2.87852534562212,
-      "grad_norm": 0.41737324630561373,
-      "learning_rate": 1.089985662496773e-06,
-      "loss": 0.4469,
-      "step": 5862
-    },
-    {
-      "epoch": 2.879016897081413,
-      "grad_norm": 0.3890062337328297,
-      "learning_rate": 1.080937737455534e-06,
-      "loss": 0.3611,
-      "step": 5863
-    },
-    {
-      "epoch": 2.8795084485407068,
-      "grad_norm": 0.42159932402698724,
-      "learning_rate": 1.0719273182471569e-06,
-      "loss": 0.4296,
-      "step": 5864
-    },
-    {
-      "epoch": 2.88,
-      "grad_norm": 0.40591650799539847,
-      "learning_rate": 1.0629544082879861e-06,
-      "loss": 0.391,
-      "step": 5865
-    },
-    {
-      "epoch": 2.8804915514592935,
-      "grad_norm": 0.4053640976693638,
-      "learning_rate": 1.0540190109801562e-06,
-      "loss": 0.4571,
-      "step": 5866
-    },
-    {
-      "epoch": 2.880983102918587,
-      "grad_norm": 0.38604230174811205,
-      "learning_rate": 1.0451211297116015e-06,
-      "loss": 0.3873,
-      "step": 5867
-    },
-    {
-      "epoch": 2.88147465437788,
-      "grad_norm": 0.4278692361086267,
-      "learning_rate": 1.036260767856001e-06,
-      "loss": 0.3741,
-      "step": 5868
-    },
-    {
-      "epoch": 2.8819662058371733,
-      "grad_norm": 0.4256563762435781,
-      "learning_rate": 1.0274379287728232e-06,
-      "loss": 0.3957,
-      "step": 5869
-    },
-    {
-      "epoch": 2.882457757296467,
-      "grad_norm": 0.46252690525039364,
-      "learning_rate": 1.0186526158073251e-06,
-      "loss": 0.4108,
-      "step": 5870
-    },
-    {
-      "epoch": 2.8829493087557605,
-      "grad_norm": 0.4047418843201301,
-      "learning_rate": 1.009904832290487e-06,
-      "loss": 0.3955,
-      "step": 5871
-    },
-    {
-      "epoch": 2.8834408602150536,
-      "grad_norm": 0.4077592414319878,
-      "learning_rate": 1.0011945815391e-06,
-      "loss": 0.3968,
-      "step": 5872
-    },
-    {
-      "epoch": 2.883932411674347,
-      "grad_norm": 0.3862666192750378,
-      "learning_rate": 9.925218668557334e-07,
-      "loss": 0.3736,
-      "step": 5873
-    },
-    {
-      "epoch": 2.8844239631336404,
-      "grad_norm": 0.4254112125731013,
-      "learning_rate": 9.838866915286903e-07,
-      "loss": 0.4322,
-      "step": 5874
-    },
-    {
-      "epoch": 2.884915514592934,
-      "grad_norm": 0.38933308848071546,
-      "learning_rate": 9.752890588320518e-07,
-      "loss": 0.3825,
-      "step": 5875
-    },
-    {
-      "epoch": 2.8854070660522275,
-      "grad_norm": 0.39411985329179455,
-      "learning_rate": 9.667289720256766e-07,
-      "loss": 0.3818,
-      "step": 5876
-    },
-    {
-      "epoch": 2.8858986175115207,
-      "grad_norm": 0.41074578261619943,
-      "learning_rate": 9.5820643435518e-07,
-      "loss": 0.4238,
-      "step": 5877
-    },
-    {
-      "epoch": 2.8863901689708142,
-      "grad_norm": 0.443401602818785,
-      "learning_rate": 9.497214490519213e-07,
-      "loss": 0.4315,
-      "step": 5878
-    },
-    {
-      "epoch": 2.8868817204301074,
-      "grad_norm": 0.40812302458216504,
-      "learning_rate": 9.412740193330827e-07,
-      "loss": 0.4324,
-      "step": 5879
-    },
-    {
-      "epoch": 2.887373271889401,
-      "grad_norm": 0.42011468714114697,
-      "learning_rate": 9.328641484015244e-07,
-      "loss": 0.4032,
-      "step": 5880
-    },
-    {
-      "epoch": 2.8878648233486945,
-      "grad_norm": 0.3953711918512807,
-      "learning_rate": 9.244918394459179e-07,
-      "loss": 0.4026,
-      "step": 5881
-    },
-    {
-      "epoch": 2.8883563748079877,
-      "grad_norm": 0.45792210525002497,
-      "learning_rate": 9.161570956406907e-07,
-      "loss": 0.4647,
-      "step": 5882
-    },
-    {
-      "epoch": 2.8888479262672813,
-      "grad_norm": 0.41823347848545295,
-      "learning_rate": 9.078599201460036e-07,
-      "loss": 0.4495,
-      "step": 5883
-    },
-    {
-      "epoch": 2.8893394777265744,
-      "grad_norm": 0.4296176804479501,
-      "learning_rate": 8.99600316107796e-07,
-      "loss": 0.4081,
-      "step": 5884
-    },
-    {
-      "epoch": 2.889831029185868,
-      "grad_norm": 0.4209430754008335,
-      "learning_rate": 8.913782866577403e-07,
-      "loss": 0.4202,
-      "step": 5885
-    },
-    {
-      "epoch": 2.8903225806451616,
-      "grad_norm": 0.4162519675064981,
-      "learning_rate": 8.831938349132984e-07,
-      "loss": 0.4464,
-      "step": 5886
-    },
-    {
-      "epoch": 2.8908141321044547,
-      "grad_norm": 0.4113365207521592,
-      "learning_rate": 8.750469639776326e-07,
-      "loss": 0.444,
-      "step": 5887
-    },
-    {
-      "epoch": 2.891305683563748,
-      "grad_norm": 0.41122673214944083,
-      "learning_rate": 8.669376769397053e-07,
-      "loss": 0.4466,
-      "step": 5888
-    },
-    {
-      "epoch": 2.8917972350230414,
-      "grad_norm": 0.3850543479460839,
-      "learning_rate": 8.588659768741903e-07,
-      "loss": 0.401,
-      "step": 5889
-    },
-    {
-      "epoch": 2.892288786482335,
-      "grad_norm": 0.38540758790413565,
-      "learning_rate": 8.508318668415505e-07,
-      "loss": 0.3831,
-      "step": 5890
-    },
-    {
-      "epoch": 2.892780337941628,
-      "grad_norm": 0.4245614822640265,
-      "learning_rate": 8.428353498879493e-07,
-      "loss": 0.4397,
-      "step": 5891
-    },
-    {
-      "epoch": 2.8932718894009217,
-      "grad_norm": 0.4498402247464881,
-      "learning_rate": 8.348764290453392e-07,
-      "loss": 0.4868,
-      "step": 5892
-    },
-    {
-      "epoch": 2.893763440860215,
-      "grad_norm": 0.4442629847525694,
-      "learning_rate": 8.269551073313841e-07,
-      "loss": 0.4432,
-      "step": 5893
-    },
-    {
-      "epoch": 2.8942549923195084,
-      "grad_norm": 0.38416652535860824,
-      "learning_rate": 8.19071387749526e-07,
-      "loss": 0.424,
-      "step": 5894
-    },
-    {
-      "epoch": 2.894746543778802,
-      "grad_norm": 0.417534663952261,
-      "learning_rate": 8.112252732888959e-07,
-      "loss": 0.3897,
-      "step": 5895
-    },
-    {
-      "epoch": 2.895238095238095,
-      "grad_norm": 0.41071717447054606,
-      "learning_rate": 8.034167669244475e-07,
-      "loss": 0.4453,
-      "step": 5896
-    },
-    {
-      "epoch": 2.8957296466973887,
-      "grad_norm": 3.342217382367793,
-      "learning_rate": 7.956458716167902e-07,
-      "loss": 0.5777,
-      "step": 5897
-    },
-    {
-      "epoch": 2.896221198156682,
-      "grad_norm": 0.4178805006042299,
-      "learning_rate": 7.87912590312323e-07,
-      "loss": 0.4048,
-      "step": 5898
-    },
-    {
-      "epoch": 2.8967127496159755,
-      "grad_norm": 0.40527456086358044,
-      "learning_rate": 7.80216925943189e-07,
-      "loss": 0.3881,
-      "step": 5899
-    },
-    {
-      "epoch": 2.897204301075269,
-      "grad_norm": 0.3943595674242165,
-      "learning_rate": 7.725588814272211e-07,
-      "loss": 0.4436,
-      "step": 5900
-    },
-    {
-      "epoch": 2.897695852534562,
-      "grad_norm": 0.4246094853771633,
-      "learning_rate": 7.64938459668052e-07,
-      "loss": 0.3692,
-      "step": 5901
-    },
-    {
-      "epoch": 2.8981874039938558,
-      "grad_norm": 0.37153255095226706,
-      "learning_rate": 7.573556635549928e-07,
-      "loss": 0.394,
-      "step": 5902
-    },
-    {
-      "epoch": 2.898678955453149,
-      "grad_norm": 0.4302436644540405,
-      "learning_rate": 7.498104959631103e-07,
-      "loss": 0.4505,
-      "step": 5903
-    },
-    {
-      "epoch": 2.8991705069124425,
-      "grad_norm": 0.420636643242112,
-      "learning_rate": 7.423029597532161e-07,
-      "loss": 0.4129,
-      "step": 5904
-    },
-    {
-      "epoch": 2.8996620583717356,
-      "grad_norm": 0.4212966683961716,
-      "learning_rate": 7.348330577718554e-07,
-      "loss": 0.3766,
-      "step": 5905
-    },
-    {
-      "epoch": 2.900153609831029,
-      "grad_norm": 0.4150062744422125,
-      "learning_rate": 7.274007928512627e-07,
-      "loss": 0.4442,
-      "step": 5906
-    },
-    {
-      "epoch": 2.9006451612903223,
-      "grad_norm": 0.426745939723344,
-      "learning_rate": 7.200061678094505e-07,
-      "loss": 0.4156,
-      "step": 5907
-    },
-    {
-      "epoch": 2.901136712749616,
-      "grad_norm": 0.404007805671517,
-      "learning_rate": 7.126491854501427e-07,
-      "loss": 0.3804,
-      "step": 5908
-    },
-    {
-      "epoch": 2.9016282642089095,
-      "grad_norm": 0.42348993273163427,
-      "learning_rate": 7.053298485627857e-07,
-      "loss": 0.4138,
-      "step": 5909
-    },
-    {
-      "epoch": 2.9021198156682027,
-      "grad_norm": 0.4360258952759854,
-      "learning_rate": 6.980481599225486e-07,
-      "loss": 0.4028,
-      "step": 5910
-    },
-    {
-      "epoch": 2.9026113671274962,
-      "grad_norm": 0.41588938401658465,
-      "learning_rate": 6.908041222903449e-07,
-      "loss": 0.3815,
-      "step": 5911
-    },
-    {
-      "epoch": 2.9031029185867894,
-      "grad_norm": 0.39590760874778863,
-      "learning_rate": 6.83597738412789e-07,
-      "loss": 0.3871,
-      "step": 5912
-    },
-    {
-      "epoch": 2.903594470046083,
-      "grad_norm": 0.4070223484307068,
-      "learning_rate": 6.764290110222394e-07,
-      "loss": 0.4295,
-      "step": 5913
-    },
-    {
-      "epoch": 2.9040860215053765,
-      "grad_norm": 0.4183721327239439,
-      "learning_rate": 6.692979428367663e-07,
-      "loss": 0.4251,
-      "step": 5914
-    },
-    {
-      "epoch": 2.9045775729646697,
-      "grad_norm": 0.38848263692664525,
-      "learning_rate": 6.622045365601515e-07,
-      "loss": 0.4233,
-      "step": 5915
-    },
-    {
-      "epoch": 2.9050691244239633,
-      "grad_norm": 0.4248028009965904,
-      "learning_rate": 6.551487948819212e-07,
-      "loss": 0.454,
-      "step": 5916
-    },
-    {
-      "epoch": 2.9055606758832564,
-      "grad_norm": 0.4265998555776826,
-      "learning_rate": 6.481307204773024e-07,
-      "loss": 0.4264,
-      "step": 5917
-    },
-    {
-      "epoch": 2.90605222734255,
-      "grad_norm": 0.4414698848708529,
-      "learning_rate": 6.411503160072441e-07,
-      "loss": 0.4647,
-      "step": 5918
-    },
-    {
-      "epoch": 2.9065437788018436,
-      "grad_norm": 0.42171120321986805,
-      "learning_rate": 6.34207584118418e-07,
-      "loss": 0.3938,
-      "step": 5919
-    },
-    {
-      "epoch": 2.9070353302611367,
-      "grad_norm": 0.42297836590279686,
-      "learning_rate": 6.273025274431965e-07,
-      "loss": 0.4059,
-      "step": 5920
-    },
-    {
-      "epoch": 2.90752688172043,
-      "grad_norm": 0.4198860005329027,
-      "learning_rate": 6.204351485997073e-07,
-      "loss": 0.4155,
-      "step": 5921
-    },
-    {
-      "epoch": 2.9080184331797234,
-      "grad_norm": 0.38911050250036117,
-      "learning_rate": 6.136054501917232e-07,
-      "loss": 0.3943,
-      "step": 5922
-    },
-    {
-      "epoch": 2.908509984639017,
-      "grad_norm": 0.3881482600331499,
-      "learning_rate": 6.068134348088061e-07,
-      "loss": 0.4115,
-      "step": 5923
-    },
-    {
-      "epoch": 2.90900153609831,
-      "grad_norm": 0.4156476478433157,
-      "learning_rate": 6.000591050261739e-07,
-      "loss": 0.3963,
-      "step": 5924
-    },
-    {
-      "epoch": 2.9094930875576037,
-      "grad_norm": 0.43041357985433026,
-      "learning_rate": 5.933424634047891e-07,
-      "loss": 0.4928,
-      "step": 5925
-    },
-    {
-      "epoch": 2.909984639016897,
-      "grad_norm": 0.4383580898047529,
-      "learning_rate": 5.866635124913034e-07,
-      "loss": 0.4156,
-      "step": 5926
-    },
-    {
-      "epoch": 2.9104761904761904,
-      "grad_norm": 0.430766224847513,
-      "learning_rate": 5.8002225481808e-07,
-      "loss": 0.4973,
-      "step": 5927
-    },
-    {
-      "epoch": 2.910967741935484,
-      "grad_norm": 0.5198957679999848,
-      "learning_rate": 5.734186929032159e-07,
-      "loss": 0.4376,
-      "step": 5928
-    },
-    {
-      "epoch": 2.911459293394777,
-      "grad_norm": 0.39222659680119215,
-      "learning_rate": 5.66852829250486e-07,
-      "loss": 0.378,
-      "step": 5929
-    },
-    {
-      "epoch": 2.9119508448540707,
-      "grad_norm": 0.3962717075758154,
-      "learning_rate": 5.603246663493766e-07,
-      "loss": 0.3825,
-      "step": 5930
-    },
-    {
-      "epoch": 2.912442396313364,
-      "grad_norm": 0.4420441291907064,
-      "learning_rate": 5.538342066750968e-07,
-      "loss": 0.443,
-      "step": 5931
-    },
-    {
-      "epoch": 2.9129339477726575,
-      "grad_norm": 0.4085987606406521,
-      "learning_rate": 5.473814526885335e-07,
-      "loss": 0.4153,
-      "step": 5932
-    },
-    {
-      "epoch": 2.913425499231951,
-      "grad_norm": 0.40317330597251816,
-      "learning_rate": 5.409664068362963e-07,
-      "loss": 0.4333,
-      "step": 5933
-    },
-    {
-      "epoch": 2.913917050691244,
-      "grad_norm": 0.4327706832914622,
-      "learning_rate": 5.345890715507173e-07,
-      "loss": 0.4125,
-      "step": 5934
-    },
-    {
-      "epoch": 2.9144086021505378,
-      "grad_norm": 0.41311355124205346,
-      "learning_rate": 5.282494492497736e-07,
-      "loss": 0.3668,
-      "step": 5935
-    },
-    {
-      "epoch": 2.914900153609831,
-      "grad_norm": 0.44293607894209375,
-      "learning_rate": 5.219475423371867e-07,
-      "loss": 0.4479,
-      "step": 5936
-    },
-    {
-      "epoch": 2.9153917050691245,
-      "grad_norm": 0.4232264828938185,
-      "learning_rate": 5.156833532023675e-07,
-      "loss": 0.3848,
-      "step": 5937
-    },
-    {
-      "epoch": 2.915883256528418,
-      "grad_norm": 0.431431143628101,
-      "learning_rate": 5.094568842204383e-07,
-      "loss": 0.428,
-      "step": 5938
-    },
-    {
-      "epoch": 2.916374807987711,
-      "grad_norm": 0.4435159455107381,
-      "learning_rate": 5.032681377521886e-07,
-      "loss": 0.4346,
-      "step": 5939
-    },
-    {
-      "epoch": 2.9168663594470043,
-      "grad_norm": 0.43443250079825363,
-      "learning_rate": 4.971171161441302e-07,
-      "loss": 0.4418,
-      "step": 5940
-    },
-    {
-      "epoch": 2.917357910906298,
-      "grad_norm": 0.40421011489480574,
-      "learning_rate": 4.910038217284752e-07,
-      "loss": 0.4368,
-      "step": 5941
-    },
-    {
-      "epoch": 2.9178494623655915,
-      "grad_norm": 0.42642056514858656,
-      "learning_rate": 4.849282568231028e-07,
-      "loss": 0.4321,
-      "step": 5942
-    },
-    {
-      "epoch": 2.9183410138248846,
-      "grad_norm": 0.39507647525376055,
-      "learning_rate": 4.788904237316149e-07,
-      "loss": 0.41,
-      "step": 5943
-    },
-    {
-      "epoch": 2.9188325652841782,
-      "grad_norm": 0.3858639562932033,
-      "learning_rate": 4.7289032474329143e-07,
-      "loss": 0.4189,
-      "step": 5944
-    },
-    {
-      "epoch": 2.9193241167434714,
-      "grad_norm": 0.47691969890438757,
-      "learning_rate": 4.669279621331235e-07,
-      "loss": 0.4101,
-      "step": 5945
-    },
-    {
-      "epoch": 2.919815668202765,
-      "grad_norm": 0.4105611443112569,
-      "learning_rate": 4.610033381617695e-07,
-      "loss": 0.4647,
-      "step": 5946
-    },
-    {
-      "epoch": 2.9203072196620585,
-      "grad_norm": 0.4063027536070045,
-      "learning_rate": 4.551164550755882e-07,
-      "loss": 0.3664,
-      "step": 5947
-    },
-    {
-      "epoch": 2.9207987711213517,
-      "grad_norm": 0.39798519458046383,
-      "learning_rate": 4.4926731510663843e-07,
-      "loss": 0.397,
-      "step": 5948
-    },
-    {
-      "epoch": 2.9212903225806452,
-      "grad_norm": 0.42032651916849023,
-      "learning_rate": 4.434559204726574e-07,
-      "loss": 0.4208,
-      "step": 5949
-    },
-    {
-      "epoch": 2.9217818740399384,
-      "grad_norm": 0.41567196570469567,
-      "learning_rate": 4.3768227337707135e-07,
-      "loss": 0.4274,
-      "step": 5950
-    },
-    {
-      "epoch": 2.922273425499232,
-      "grad_norm": 0.4027071416383512,
-      "learning_rate": 4.3194637600901813e-07,
-      "loss": 0.4226,
-      "step": 5951
-    },
-    {
-      "epoch": 2.9227649769585256,
-      "grad_norm": 0.4083293654665912,
-      "learning_rate": 4.2624823054328024e-07,
-      "loss": 0.4383,
-      "step": 5952
-    },
-    {
-      "epoch": 2.9232565284178187,
-      "grad_norm": 0.4202209247395239,
-      "learning_rate": 4.205878391403517e-07,
-      "loss": 0.4068,
-      "step": 5953
-    },
-    {
-      "epoch": 2.9237480798771123,
-      "grad_norm": 0.4144287902263344,
-      "learning_rate": 4.149652039464047e-07,
-      "loss": 0.3911,
-      "step": 5954
-    },
-    {
-      "epoch": 2.9242396313364054,
-      "grad_norm": 0.41504942741631884,
-      "learning_rate": 4.0938032709332275e-07,
-      "loss": 0.4082,
-      "step": 5955
-    },
-    {
-      "epoch": 2.924731182795699,
-      "grad_norm": 0.3846908548282004,
-      "learning_rate": 4.038332106986231e-07,
-      "loss": 0.3927,
-      "step": 5956
-    },
-    {
-      "epoch": 2.9252227342549926,
-      "grad_norm": 0.41705628166588243,
-      "learning_rate": 3.983238568655345e-07,
-      "loss": 0.3615,
-      "step": 5957
-    },
-    {
-      "epoch": 2.9257142857142857,
-      "grad_norm": 0.4507590009565246,
-      "learning_rate": 3.92852267682986e-07,
-      "loss": 0.4831,
-      "step": 5958
-    },
-    {
-      "epoch": 2.926205837173579,
-      "grad_norm": 0.39985549841289475,
-      "learning_rate": 3.8741844522555136e-07,
-      "loss": 0.4306,
-      "step": 5959
-    },
-    {
-      "epoch": 2.9266973886328724,
-      "grad_norm": 0.4201925462347482,
-      "learning_rate": 3.820223915535048e-07,
-      "loss": 0.4048,
-      "step": 5960
-    },
-    {
-      "epoch": 2.927188940092166,
-      "grad_norm": 0.40623733028686093,
-      "learning_rate": 3.7666410871279867e-07,
-      "loss": 0.4137,
-      "step": 5961
-    },
-    {
-      "epoch": 2.927680491551459,
-      "grad_norm": 0.4298677565113412,
-      "learning_rate": 3.713435987350522e-07,
-      "loss": 0.4336,
-      "step": 5962
-    },
-    {
-      "epoch": 2.9281720430107527,
-      "grad_norm": 0.451860189302632,
-      "learning_rate": 3.6606086363759615e-07,
-      "loss": 0.4045,
-      "step": 5963
-    },
-    {
-      "epoch": 2.928663594470046,
-      "grad_norm": 0.42494963642857425,
-      "learning_rate": 3.608159054233951e-07,
-      "loss": 0.4167,
-      "step": 5964
-    },
-    {
-      "epoch": 2.9291551459293395,
-      "grad_norm": 0.40904406489952366,
-      "learning_rate": 3.556087260811136e-07,
-      "loss": 0.4094,
-      "step": 5965
-    },
-    {
-      "epoch": 2.929646697388633,
-      "grad_norm": 0.42327856786153,
-      "learning_rate": 3.504393275850948e-07,
-      "loss": 0.4143,
-      "step": 5966
-    },
-    {
-      "epoch": 2.930138248847926,
-      "grad_norm": 0.3941127188442293,
-      "learning_rate": 3.453077118953374e-07,
-      "loss": 0.3891,
-      "step": 5967
-    },
-    {
-      "epoch": 2.9306298003072198,
-      "grad_norm": 0.41281696074150137,
-      "learning_rate": 3.402138809575517e-07,
-      "loss": 0.3806,
-      "step": 5968
-    },
-    {
-      "epoch": 2.931121351766513,
-      "grad_norm": 0.4092918330993273,
-      "learning_rate": 3.3515783670307057e-07,
-      "loss": 0.3871,
-      "step": 5969
-    },
-    {
-      "epoch": 2.9316129032258065,
-      "grad_norm": 0.40678432857497393,
-      "learning_rate": 3.301395810489494e-07,
-      "loss": 0.3949,
-      "step": 5970
-    },
-    {
-      "epoch": 2.9321044546851,
-      "grad_norm": 0.433874021223381,
-      "learning_rate": 3.251591158978884e-07,
-      "loss": 0.4499,
-      "step": 5971
-    },
-    {
-      "epoch": 2.932596006144393,
-      "grad_norm": 0.3880236694309334,
-      "learning_rate": 3.202164431382659e-07,
-      "loss": 0.4031,
-      "step": 5972
-    },
-    {
-      "epoch": 2.933087557603687,
-      "grad_norm": 0.4140154485434843,
-      "learning_rate": 3.1531156464411624e-07,
-      "loss": 0.4029,
-      "step": 5973
-    },
-    {
-      "epoch": 2.93357910906298,
-      "grad_norm": 0.4203633666443706,
-      "learning_rate": 3.10444482275174e-07,
-      "loss": 0.3991,
-      "step": 5974
-    },
-    {
-      "epoch": 2.9340706605222735,
-      "grad_norm": 0.423022842267573,
-      "learning_rate": 3.056151978768185e-07,
-      "loss": 0.4091,
-      "step": 5975
-    },
-    {
-      "epoch": 2.934562211981567,
-      "grad_norm": 0.4319111693033177,
-      "learning_rate": 3.0082371328010727e-07,
-      "loss": 0.4636,
-      "step": 5976
-    },
-    {
-      "epoch": 2.93505376344086,
-      "grad_norm": 0.3996701173390591,
-      "learning_rate": 2.9607003030176494e-07,
-      "loss": 0.3551,
-      "step": 5977
-    },
-    {
-      "epoch": 2.9355453149001534,
-      "grad_norm": 0.4021917342472955,
-      "learning_rate": 2.91354150744183e-07,
-      "loss": 0.4354,
-      "step": 5978
-    },
-    {
-      "epoch": 2.936036866359447,
-      "grad_norm": 0.4070519322847705,
-      "learning_rate": 2.8667607639542016e-07,
-      "loss": 0.3898,
-      "step": 5979
-    },
-    {
-      "epoch": 2.9365284178187405,
-      "grad_norm": 0.4052114101319615,
-      "learning_rate": 2.820358090291908e-07,
-      "loss": 0.424,
-      "step": 5980
-    },
-    {
-      "epoch": 2.9370199692780337,
-      "grad_norm": 0.4231356716503333,
-      "learning_rate": 2.774333504048876e-07,
-      "loss": 0.4334,
-      "step": 5981
-    },
-    {
-      "epoch": 2.9375115207373272,
-      "grad_norm": 0.4210842344560463,
-      "learning_rate": 2.7286870226758135e-07,
-      "loss": 0.4146,
-      "step": 5982
-    },
-    {
-      "epoch": 2.9380030721966204,
-      "grad_norm": 0.38704619471286666,
-      "learning_rate": 2.6834186634796534e-07,
-      "loss": 0.3902,
-      "step": 5983
-    },
-    {
-      "epoch": 2.938494623655914,
-      "grad_norm": 0.3988897628911073,
-      "learning_rate": 2.638528443624333e-07,
-      "loss": 0.3953,
-      "step": 5984
-    },
-    {
-      "epoch": 2.9389861751152075,
-      "grad_norm": 0.4200837658240136,
-      "learning_rate": 2.5940163801301267e-07,
-      "loss": 0.3957,
-      "step": 5985
-    },
-    {
-      "epoch": 2.9394777265745007,
-      "grad_norm": 0.4228404926309639,
-      "learning_rate": 2.5498824898744224e-07,
-      "loss": 0.4198,
-      "step": 5986
-    },
-    {
-      "epoch": 2.9399692780337943,
-      "grad_norm": 0.38672086814750295,
-      "learning_rate": 2.5061267895905016e-07,
-      "loss": 0.4091,
-      "step": 5987
-    },
-    {
-      "epoch": 2.9404608294930874,
-      "grad_norm": 0.41196772689289507,
-      "learning_rate": 2.4627492958688714e-07,
-      "loss": 0.4327,
-      "step": 5988
-    },
-    {
-      "epoch": 2.940952380952381,
-      "grad_norm": 0.38198369126642734,
-      "learning_rate": 2.4197500251563753e-07,
-      "loss": 0.3585,
-      "step": 5989
-    },
-    {
-      "epoch": 2.9414439324116746,
-      "grad_norm": 0.416107428976138,
-      "learning_rate": 2.3771289937563056e-07,
-      "loss": 0.4041,
-      "step": 5990
-    },
-    {
-      "epoch": 2.9419354838709677,
-      "grad_norm": 0.4014966002475941,
-      "learning_rate": 2.3348862178289575e-07,
-      "loss": 0.4166,
-      "step": 5991
-    },
-    {
-      "epoch": 2.9424270353302613,
-      "grad_norm": 0.42158686121084765,
-      "learning_rate": 2.2930217133907418e-07,
-      "loss": 0.4571,
-      "step": 5992
-    },
-    {
-      "epoch": 2.9429185867895544,
-      "grad_norm": 0.5447030154486988,
-      "learning_rate": 2.2515354963150715e-07,
-      "loss": 0.4161,
-      "step": 5993
-    },
-    {
-      "epoch": 2.943410138248848,
-      "grad_norm": 0.4003026766203279,
-      "learning_rate": 2.2104275823315868e-07,
-      "loss": 0.4411,
-      "step": 5994
-    },
-    {
-      "epoch": 2.9439016897081416,
-      "grad_norm": 0.3894061583770945,
-      "learning_rate": 2.1696979870267087e-07,
-      "loss": 0.3544,
-      "step": 5995
-    },
-    {
-      "epoch": 2.9443932411674347,
-      "grad_norm": 0.41164779323823,
-      "learning_rate": 2.1293467258433065e-07,
-      "loss": 0.4686,
-      "step": 5996
-    },
-    {
-      "epoch": 2.944884792626728,
-      "grad_norm": 0.4119417860239911,
-      "learning_rate": 2.0893738140808083e-07,
-      "loss": 0.4151,
-      "step": 5997
-    },
-    {
-      "epoch": 2.9453763440860214,
-      "grad_norm": 0.42082802046114,
-      "learning_rate": 2.0497792668953132e-07,
-      "loss": 0.4193,
-      "step": 5998
-    },
-    {
-      "epoch": 2.945867895545315,
-      "grad_norm": 0.4200033745830691,
-      "learning_rate": 2.0105630992992564e-07,
-      "loss": 0.4361,
-      "step": 5999
-    },
-    {
-      "epoch": 2.946359447004608,
-      "grad_norm": 0.43196118605709805,
-      "learning_rate": 1.9717253261617442e-07,
-      "loss": 0.4441,
-      "step": 6000
-    },
-    {
-      "epoch": 2.9468509984639017,
-      "grad_norm": 0.4144578066395236,
-      "learning_rate": 1.9332659622083304e-07,
-      "loss": 0.4449,
-      "step": 6001
-    },
-    {
-      "epoch": 2.947342549923195,
-      "grad_norm": 0.3873889749151009,
-      "learning_rate": 1.8951850220213508e-07,
-      "loss": 0.4288,
-      "step": 6002
-    },
-    {
-      "epoch": 2.9478341013824885,
-      "grad_norm": 0.4010679148855431,
-      "learning_rate": 1.8574825200391445e-07,
-      "loss": 0.3936,
-      "step": 6003
-    },
-    {
-      "epoch": 2.948325652841782,
-      "grad_norm": 0.40243074836209924,
-      "learning_rate": 1.820158470557165e-07,
-      "loss": 0.3806,
-      "step": 6004
-    },
-    {
-      "epoch": 2.948817204301075,
-      "grad_norm": 0.4274781332589534,
-      "learning_rate": 1.7832128877268705e-07,
-      "loss": 0.3942,
-      "step": 6005
-    },
-    {
-      "epoch": 2.9493087557603688,
-      "grad_norm": 0.4069318726121919,
-      "learning_rate": 1.7466457855565e-07,
-      "loss": 0.4424,
-      "step": 6006
-    },
-    {
-      "epoch": 2.949800307219662,
-      "grad_norm": 0.40870341173156804,
-      "learning_rate": 1.7104571779107402e-07,
-      "loss": 0.3824,
-      "step": 6007
-    },
-    {
-      "epoch": 2.9502918586789555,
-      "grad_norm": 0.40053018901382154,
-      "learning_rate": 1.674647078510727e-07,
-      "loss": 0.4182,
-      "step": 6008
-    },
-    {
-      "epoch": 2.950783410138249,
-      "grad_norm": 0.44085495296066624,
-      "learning_rate": 1.6392155009340437e-07,
-      "loss": 0.4126,
-      "step": 6009
-    },
-    {
-      "epoch": 2.951274961597542,
-      "grad_norm": 0.42778884084759744,
-      "learning_rate": 1.604162458614944e-07,
-      "loss": 0.4384,
-      "step": 6010
-    },
-    {
-      "epoch": 2.951766513056836,
-      "grad_norm": 0.3860862006471472,
-      "learning_rate": 1.5694879648439076e-07,
-      "loss": 0.435,
-      "step": 6011
-    },
-    {
-      "epoch": 2.952258064516129,
-      "grad_norm": 0.44385517994740425,
-      "learning_rate": 1.5351920327680847e-07,
-      "loss": 0.4027,
-      "step": 6012
-    },
-    {
-      "epoch": 2.9527496159754225,
-      "grad_norm": 0.39830272358746444,
-      "learning_rate": 1.5012746753909624e-07,
-      "loss": 0.4052,
-      "step": 6013
-    },
-    {
-      "epoch": 2.953241167434716,
-      "grad_norm": 0.4267178954516768,
-      "learning_rate": 1.467735905572476e-07,
-      "loss": 0.4187,
-      "step": 6014
-    },
-    {
-      "epoch": 2.9537327188940092,
-      "grad_norm": 0.4107082385394562,
-      "learning_rate": 1.434575736029231e-07,
-      "loss": 0.4084,
-      "step": 6015
-    },
-    {
-      "epoch": 2.9542242703533024,
-      "grad_norm": 0.4402235094289675,
-      "learning_rate": 1.4017941793340593e-07,
-      "loss": 0.4449,
-      "step": 6016
-    },
-    {
-      "epoch": 2.954715821812596,
-      "grad_norm": 0.3957201828010452,
-      "learning_rate": 1.3693912479162408e-07,
-      "loss": 0.3873,
-      "step": 6017
-    },
-    {
-      "epoch": 2.9552073732718895,
-      "grad_norm": 0.418258384332858,
-      "learning_rate": 1.337366954061725e-07,
-      "loss": 0.4369,
-      "step": 6018
-    },
-    {
-      "epoch": 2.9556989247311827,
-      "grad_norm": 0.4099216051892321,
-      "learning_rate": 1.3057213099125776e-07,
-      "loss": 0.4164,
-      "step": 6019
-    },
-    {
-      "epoch": 2.9561904761904763,
-      "grad_norm": 0.3698579870804078,
-      "learning_rate": 1.2744543274675334e-07,
-      "loss": 0.363,
-      "step": 6020
-    },
-    {
-      "epoch": 2.9566820276497694,
-      "grad_norm": 0.44962052139173175,
-      "learning_rate": 1.2435660185816655e-07,
-      "loss": 0.4557,
-      "step": 6021
-    },
-    {
-      "epoch": 2.957173579109063,
-      "grad_norm": 0.39983886649349315,
-      "learning_rate": 1.213056394966494e-07,
-      "loss": 0.4032,
-      "step": 6022
-    },
-    {
-      "epoch": 2.9576651305683566,
-      "grad_norm": 0.39411613942516277,
-      "learning_rate": 1.1829254681898772e-07,
-      "loss": 0.407,
-      "step": 6023
-    },
-    {
-      "epoch": 2.9581566820276497,
-      "grad_norm": 0.38850085462320016,
-      "learning_rate": 1.1531732496763425e-07,
-      "loss": 0.359,
-      "step": 6024
-    },
-    {
-      "epoch": 2.9586482334869433,
-      "grad_norm": 0.40159884802082935,
-      "learning_rate": 1.1237997507064224e-07,
-      "loss": 0.426,
-      "step": 6025
-    },
-    {
-      "epoch": 2.9591397849462364,
-      "grad_norm": 0.40125688005210397,
-      "learning_rate": 1.09480498241743e-07,
-      "loss": 0.4098,
-      "step": 6026
-    },
-    {
-      "epoch": 2.95963133640553,
-      "grad_norm": 0.3808112577083854,
-      "learning_rate": 1.0661889558029048e-07,
-      "loss": 0.3759,
-      "step": 6027
-    },
-    {
-      "epoch": 2.9601228878648236,
-      "grad_norm": 0.4266575266481944,
-      "learning_rate": 1.0379516817128343e-07,
-      "loss": 0.3861,
-      "step": 6028
-    },
-    {
-      "epoch": 2.9606144393241167,
-      "grad_norm": 0.45668055392825835,
-      "learning_rate": 1.0100931708534323e-07,
-      "loss": 0.4616,
-      "step": 6029
-    },
-    {
-      "epoch": 2.9611059907834103,
-      "grad_norm": 0.4639244945697901,
-      "learning_rate": 9.826134337875826e-08,
-      "loss": 0.4125,
-      "step": 6030
-    },
-    {
-      "epoch": 2.9615975422427034,
-      "grad_norm": 0.4031595099385033,
-      "learning_rate": 9.555124809343952e-08,
-      "loss": 0.4026,
-      "step": 6031
-    },
-    {
-      "epoch": 2.962089093701997,
-      "grad_norm": 0.4209953740906405,
-      "learning_rate": 9.287903225693173e-08,
-      "loss": 0.4086,
-      "step": 6032
-    },
-    {
-      "epoch": 2.96258064516129,
-      "grad_norm": 0.4203806817396384,
-      "learning_rate": 9.024469688242443e-08,
-      "loss": 0.4447,
-      "step": 6033
-    },
-    {
-      "epoch": 2.9630721966205837,
-      "grad_norm": 0.4130465246202411,
-      "learning_rate": 8.764824296875196e-08,
-      "loss": 0.445,
-      "step": 6034
-    },
-    {
-      "epoch": 2.963563748079877,
-      "grad_norm": 0.38443477373932644,
-      "learning_rate": 8.508967150037128e-08,
-      "loss": 0.4041,
-      "step": 6035
-    },
-    {
-      "epoch": 2.9640552995391705,
-      "grad_norm": 0.42372931505908396,
-      "learning_rate": 8.256898344737307e-08,
-      "loss": 0.4387,
-      "step": 6036
-    },
-    {
-      "epoch": 2.964546850998464,
-      "grad_norm": 0.4080582091433443,
-      "learning_rate": 8.008617976551502e-08,
-      "loss": 0.3915,
-      "step": 6037
-    },
-    {
-      "epoch": 2.965038402457757,
-      "grad_norm": 0.39739507320924133,
-      "learning_rate": 7.764126139615524e-08,
-      "loss": 0.3999,
-      "step": 6038
-    },
-    {
-      "epoch": 2.9655299539170508,
-      "grad_norm": 0.41723383285383575,
-      "learning_rate": 7.523422926629664e-08,
-      "loss": 0.4235,
-      "step": 6039
-    },
-    {
-      "epoch": 2.966021505376344,
-      "grad_norm": 0.44034358447187427,
-      "learning_rate": 7.286508428858696e-08,
-      "loss": 0.4395,
-      "step": 6040
-    },
-    {
-      "epoch": 2.9665130568356375,
-      "grad_norm": 0.42902757844756756,
-      "learning_rate": 7.053382736130764e-08,
-      "loss": 0.4085,
-      "step": 6041
-    },
-    {
-      "epoch": 2.967004608294931,
-      "grad_norm": 0.43774027421218564,
-      "learning_rate": 6.824045936836276e-08,
-      "loss": 0.3751,
-      "step": 6042
-    },
-    {
-      "epoch": 2.967496159754224,
-      "grad_norm": 0.40683340468552215,
-      "learning_rate": 6.598498117931228e-08,
-      "loss": 0.4142,
-      "step": 6043
-    },
-    {
-      "epoch": 2.967987711213518,
-      "grad_norm": 0.40725696543062384,
-      "learning_rate": 6.376739364932772e-08,
-      "loss": 0.4354,
-      "step": 6044
-    },
-    {
-      "epoch": 2.968479262672811,
-      "grad_norm": 0.43894079285299115,
-      "learning_rate": 6.158769761921423e-08,
-      "loss": 0.4263,
-      "step": 6045
-    },
-    {
-      "epoch": 2.9689708141321045,
-      "grad_norm": 0.44513675567143146,
-      "learning_rate": 5.944589391542188e-08,
-      "loss": 0.4409,
-      "step": 6046
-    },
-    {
-      "epoch": 2.969462365591398,
-      "grad_norm": 0.42649597478472673,
-      "learning_rate": 5.734198335004548e-08,
-      "loss": 0.4355,
-      "step": 6047
-    },
-    {
-      "epoch": 2.9699539170506912,
-      "grad_norm": 0.4088928304993863,
-      "learning_rate": 5.527596672078028e-08,
-      "loss": 0.4255,
-      "step": 6048
-    },
-    {
-      "epoch": 2.9704454685099844,
-      "grad_norm": 0.4444738183882466,
-      "learning_rate": 5.324784481096634e-08,
-      "loss": 0.3721,
-      "step": 6049
-    },
-    {
-      "epoch": 2.970937019969278,
-      "grad_norm": 0.4126968027163256,
-      "learning_rate": 5.125761838959964e-08,
-      "loss": 0.3688,
-      "step": 6050
-    },
-    {
-      "epoch": 2.9714285714285715,
-      "grad_norm": 0.4003310634673174,
-      "learning_rate": 4.930528821126545e-08,
-      "loss": 0.3939,
-      "step": 6051
-    },
-    {
-      "epoch": 2.9719201228878647,
-      "grad_norm": 0.41670499712303666,
-      "learning_rate": 4.73908550162272e-08,
-      "loss": 0.4265,
-      "step": 6052
-    },
-    {
-      "epoch": 2.9724116743471583,
-      "grad_norm": 0.41468367323871685,
-      "learning_rate": 4.551431953033758e-08,
-      "loss": 0.3913,
-      "step": 6053
-    },
-    {
-      "epoch": 2.9729032258064514,
-      "grad_norm": 0.3954875606446369,
-      "learning_rate": 4.367568246510523e-08,
-      "loss": 0.3993,
-      "step": 6054
-    },
-    {
-      "epoch": 2.973394777265745,
-      "grad_norm": 0.42037941517522137,
-      "learning_rate": 4.1874944517661385e-08,
-      "loss": 0.411,
-      "step": 6055
-    },
-    {
-      "epoch": 2.9738863287250386,
-      "grad_norm": 0.40689557517667146,
-      "learning_rate": 4.01121063707599e-08,
-      "loss": 0.4276,
-      "step": 6056
-    },
-    {
-      "epoch": 2.9743778801843317,
-      "grad_norm": 0.43832708700955153,
-      "learning_rate": 3.8387168692799456e-08,
-      "loss": 0.3919,
-      "step": 6057
-    },
-    {
-      "epoch": 2.9748694316436253,
-      "grad_norm": 0.4170563697685453,
-      "learning_rate": 3.6700132137812426e-08,
-      "loss": 0.4068,
-      "step": 6058
-    },
-    {
-      "epoch": 2.9753609831029184,
-      "grad_norm": 0.440455237868665,
-      "learning_rate": 3.50509973454316e-08,
-      "loss": 0.4298,
-      "step": 6059
-    },
-    {
-      "epoch": 2.975852534562212,
-      "grad_norm": 0.3762155148789076,
-      "learning_rate": 3.3439764940934594e-08,
-      "loss": 0.3898,
-      "step": 6060
-    },
-    {
-      "epoch": 2.9763440860215056,
-      "grad_norm": 0.39882875673227924,
-      "learning_rate": 3.186643553525492e-08,
-      "loss": 0.4033,
-      "step": 6061
-    },
-    {
-      "epoch": 2.9768356374807987,
-      "grad_norm": 0.40712218639627423,
-      "learning_rate": 3.033100972491543e-08,
-      "loss": 0.4052,
-      "step": 6062
-    },
-    {
-      "epoch": 2.9773271889400923,
-      "grad_norm": 0.46334739447452034,
-      "learning_rate": 2.883348809208375e-08,
-      "loss": 0.4313,
-      "step": 6063
-    },
-    {
-      "epoch": 2.9778187403993854,
-      "grad_norm": 0.41473301999742473,
-      "learning_rate": 2.7373871204561252e-08,
-      "loss": 0.4343,
-      "step": 6064
-    },
-    {
-      "epoch": 2.978310291858679,
-      "grad_norm": 0.45174724302909136,
-      "learning_rate": 2.5952159615760806e-08,
-      "loss": 0.4046,
-      "step": 6065
-    },
-    {
-      "epoch": 2.9788018433179726,
-      "grad_norm": 0.4497066917671094,
-      "learning_rate": 2.4568353864751204e-08,
-      "loss": 0.4291,
-      "step": 6066
-    },
-    {
-      "epoch": 2.9792933947772657,
-      "grad_norm": 0.4499684444059429,
-      "learning_rate": 2.3222454476190535e-08,
-      "loss": 0.3975,
-      "step": 6067
-    },
-    {
-      "epoch": 2.979784946236559,
-      "grad_norm": 0.4279091247595014,
-      "learning_rate": 2.191446196040392e-08,
-      "loss": 0.3934,
-      "step": 6068
-    },
-    {
-      "epoch": 2.9802764976958525,
-      "grad_norm": 0.4543201244976313,
-      "learning_rate": 2.064437681331688e-08,
-      "loss": 0.3991,
-      "step": 6069
-    },
-    {
-      "epoch": 2.980768049155146,
-      "grad_norm": 0.40319902320709167,
-      "learning_rate": 1.941219951648865e-08,
-      "loss": 0.4245,
-      "step": 6070
-    },
-    {
-      "epoch": 2.981259600614439,
-      "grad_norm": 0.4040314613900974,
-      "learning_rate": 1.8217930537112184e-08,
-      "loss": 0.4243,
-      "step": 6071
-    },
-    {
-      "epoch": 2.9817511520737328,
-      "grad_norm": 0.41014280331005337,
-      "learning_rate": 1.706157032800304e-08,
-      "loss": 0.4117,
-      "step": 6072
-    },
-    {
-      "epoch": 2.982242703533026,
-      "grad_norm": 0.4397016377924984,
-      "learning_rate": 1.59431193275994e-08,
-      "loss": 0.4313,
-      "step": 6073
-    },
-    {
-      "epoch": 2.9827342549923195,
-      "grad_norm": 0.4048797114510302,
-      "learning_rate": 1.4862577959973145e-08,
-      "loss": 0.4263,
-      "step": 6074
-    },
-    {
-      "epoch": 2.983225806451613,
-      "grad_norm": 0.38905967221178384,
-      "learning_rate": 1.3819946634818782e-08,
-      "loss": 0.3735,
-      "step": 6075
-    },
-    {
-      "epoch": 2.983717357910906,
-      "grad_norm": 0.37764869902734094,
-      "learning_rate": 1.281522574745342e-08,
-      "loss": 0.4038,
-      "step": 6076
-    },
-    {
-      "epoch": 2.9842089093702,
-      "grad_norm": 0.4026355877938672,
-      "learning_rate": 1.1848415678827885e-08,
-      "loss": 0.4531,
-      "step": 6077
-    },
-    {
-      "epoch": 2.984700460829493,
-      "grad_norm": 0.40417802514297885,
-      "learning_rate": 1.0919516795515617e-08,
-      "loss": 0.444,
-      "step": 6078
-    },
-    {
-      "epoch": 2.9851920122887865,
-      "grad_norm": 0.41842658128623017,
-      "learning_rate": 1.0028529449701563e-08,
-      "loss": 0.4232,
-      "step": 6079
-    },
-    {
-      "epoch": 2.98568356374808,
-      "grad_norm": 0.38651846873424983,
-      "learning_rate": 9.17545397922659e-09,
-      "loss": 0.3678,
-      "step": 6080
-    },
-    {
-      "epoch": 2.986175115207373,
-      "grad_norm": 0.37837707655414704,
-      "learning_rate": 8.360290707543073e-09,
-      "loss": 0.3956,
-      "step": 6081
-    },
-    {
-      "epoch": 2.986666666666667,
-      "grad_norm": 0.4215845213690753,
-      "learning_rate": 7.583039943703795e-09,
-      "loss": 0.4249,
-      "step": 6082
-    },
-    {
-      "epoch": 2.98715821812596,
-      "grad_norm": 0.39909141904110923,
-      "learning_rate": 6.843701982428563e-09,
-      "loss": 0.4096,
-      "step": 6083
-    },
-    {
-      "epoch": 2.9876497695852535,
-      "grad_norm": 0.43292174211698903,
-      "learning_rate": 6.142277104026484e-09,
-      "loss": 0.4054,
-      "step": 6084
-    },
-    {
-      "epoch": 2.988141321044547,
-      "grad_norm": 0.40029677964018207,
-      "learning_rate": 5.478765574462586e-09,
-      "loss": 0.4008,
-      "step": 6085
-    },
-    {
-      "epoch": 2.9886328725038402,
-      "grad_norm": 0.4294975110718298,
-      "learning_rate": 4.853167645302303e-09,
-      "loss": 0.3856,
-      "step": 6086
-    },
-    {
-      "epoch": 2.9891244239631334,
-      "grad_norm": 0.4286046977545209,
-      "learning_rate": 4.265483553755889e-09,
-      "loss": 0.4107,
-      "step": 6087
-    },
-    {
-      "epoch": 2.989615975422427,
-      "grad_norm": 0.40104635914997994,
-      "learning_rate": 3.715713522622899e-09,
-      "loss": 0.3473,
-      "step": 6088
-    },
-    {
-      "epoch": 2.9901075268817205,
-      "grad_norm": 0.38453583888031445,
-      "learning_rate": 3.2038577603810127e-09,
-      "loss": 0.3976,
-      "step": 6089
-    },
-    {
-      "epoch": 2.9905990783410137,
-      "grad_norm": 0.45510364577481827,
-      "learning_rate": 2.729916461097215e-09,
-      "loss": 0.4103,
-      "step": 6090
-    },
-    {
-      "epoch": 2.9910906298003073,
-      "grad_norm": 0.4027303305051323,
-      "learning_rate": 2.2938898044611023e-09,
-      "loss": 0.4263,
-      "step": 6091
-    },
-    {
-      "epoch": 2.9915821812596004,
-      "grad_norm": 0.40878866001483,
-      "learning_rate": 1.8957779557959855e-09,
-      "loss": 0.3921,
-      "step": 6092
-    },
-    {
-      "epoch": 2.992073732718894,
-      "grad_norm": 0.42846769694886017,
-      "learning_rate": 1.5355810660477864e-09,
-      "loss": 0.4352,
-      "step": 6093
-    },
-    {
-      "epoch": 2.9925652841781876,
-      "grad_norm": 0.39993957432953003,
-      "learning_rate": 1.2132992717961422e-09,
-      "loss": 0.4217,
-      "step": 6094
-    },
-    {
-      "epoch": 2.9930568356374807,
-      "grad_norm": 0.3936801466298494,
-      "learning_rate": 9.289326952321986e-10,
-      "loss": 0.3839,
-      "step": 6095
-    },
-    {
-      "epoch": 2.9935483870967743,
-      "grad_norm": 0.41596733354476106,
-      "learning_rate": 6.824814441808159e-10,
-      "loss": 0.4171,
-      "step": 6096
-    },
-    {
-      "epoch": 2.9940399385560674,
-      "grad_norm": 0.410800646796449,
-      "learning_rate": 4.739456120672614e-10,
-      "loss": 0.4112,
-      "step": 6097
-    },
-    {
-      "epoch": 2.994531490015361,
-      "grad_norm": 0.4170747581370785,
-      "learning_rate": 3.033252779838236e-10,
-      "loss": 0.4316,
-      "step": 6098
-    },
-    {
-      "epoch": 2.9950230414746546,
-      "grad_norm": 0.4304517925812922,
-      "learning_rate": 1.706205066009936e-10,
-      "loss": 0.4562,
-      "step": 6099
-    },
-    {
-      "epoch": 2.9955145929339477,
-      "grad_norm": 0.40260588675218906,
-      "learning_rate": 7.583134824518112e-11,
-      "loss": 0.4133,
-      "step": 6100
-    },
-    {
-      "epoch": 2.9960061443932413,
-      "grad_norm": 0.38341260419568185,
-      "learning_rate": 1.8957838854305465e-11,
-      "loss": 0.4098,
-      "step": 6101
-    },
-    {
-      "epoch": 2.9964976958525344,
-      "grad_norm": 0.39987369746027823,
-      "learning_rate": 0.0,
-      "loss": 0.4278,
-      "step": 6102
-    },
-    {
-      "epoch": 2.9964976958525344,
-      "eval_loss": 0.8067362904548645,
-      "eval_runtime": 6662.6718,
-      "eval_samples_per_second": 4.277,
-      "eval_steps_per_second": 2.139,
-      "step": 6102
-    }
-  ],
-  "logging_steps": 1,
-  "max_steps": 6102,
-  "num_input_tokens_seen": 0,
-  "num_train_epochs": 3,
-  "save_steps": 2034,
-  "stateful_callbacks": {
-    "TrainerControl": {
-      "args": {
-        "should_epoch_stop": false,
-        "should_evaluate": false,
-        "should_log": false,
-        "should_save": true,
-        "should_training_stop": true
-      },
-      "attributes": {}
-    }
-  },
-  "total_flos": 840273273815040.0,
-  "train_batch_size": 1,
-  "trial_name": null,
-  "trial_params": null
-}