diff --git "a/gcp_models/lora-logo_fix_full_deepseek33b_gpt35i_lr_0.0002_alpha_512_r_512/checkpoint-1362/trainer_state.json" "b/gcp_models/lora-logo_fix_full_deepseek33b_gpt35i_lr_0.0002_alpha_512_r_512/checkpoint-1362/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/gcp_models/lora-logo_fix_full_deepseek33b_gpt35i_lr_0.0002_alpha_512_r_512/checkpoint-1362/trainer_state.json"
@@ -0,0 +1,9651 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9779735682819384,
+  "eval_steps": 114,
+  "global_step": 1362,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5863164065184416,
+      "learning_rate": 1e-05,
+      "loss": 2.6411,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 2.7003138065338135,
+      "eval_runtime": 46.4646,
+      "eval_samples_per_second": 49.995,
+      "eval_steps_per_second": 0.796,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7346265404669823,
+      "learning_rate": 2e-05,
+      "loss": 2.7443,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6716196930916984,
+      "learning_rate": 3e-05,
+      "loss": 2.7234,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1526061519450894,
+      "learning_rate": 4e-05,
+      "loss": 2.4957,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4865554054320307,
+      "learning_rate": 5e-05,
+      "loss": 1.9933,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.155126076955286,
+      "learning_rate": 6e-05,
+      "loss": 1.5027,
+      "step": 6
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.278400042723415,
+      "learning_rate": 7e-05,
+      "loss": 1.1638,
+      "step": 7
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1176976992956562,
+      "learning_rate": 8e-05,
+      "loss": 0.8176,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7566457989485756,
+      "learning_rate": 9e-05,
+      "loss": 0.7411,
+      "step": 9
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5964188075673675,
+      "learning_rate": 0.0001,
+      "loss": 0.601,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.43137451530435944,
+      "learning_rate": 0.00011000000000000002,
+      "loss": 0.5498,
+      "step": 11
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.40031296511882886,
+      "learning_rate": 0.00012,
+      "loss": 0.5155,
+      "step": 12
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2719323147178774,
+      "learning_rate": 0.00013000000000000002,
+      "loss": 0.4721,
+      "step": 13
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28474205453285395,
+      "learning_rate": 0.00014,
+      "loss": 0.4884,
+      "step": 14
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2581057470221288,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.4643,
+      "step": 15
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.21564651844901164,
+      "learning_rate": 0.00016,
+      "loss": 0.4439,
+      "step": 16
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.23922983122711197,
+      "learning_rate": 0.00017,
+      "loss": 0.4433,
+      "step": 17
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.19539976078046137,
+      "learning_rate": 0.00018,
+      "loss": 0.4353,
+      "step": 18
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.23064420777740008,
+      "learning_rate": 0.00019,
+      "loss": 0.425,
+      "step": 19
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1861911029947257,
+      "learning_rate": 0.0002,
+      "loss": 0.4354,
+      "step": 20
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2176013403377687,
+      "learning_rate": 0.00019999972599119435,
+      "loss": 0.4248,
+      "step": 21
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.18439498526346876,
+      "learning_rate": 0.00019999890396627895,
+      "loss": 0.4348,
+      "step": 22
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.20550064883628566,
+      "learning_rate": 0.00019999753392975869,
+      "loss": 0.4219,
+      "step": 23
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1605970086299528,
+      "learning_rate": 0.00019999561588914156,
+      "loss": 0.4029,
+      "step": 24
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.16120032427832962,
+      "learning_rate": 0.00019999314985493885,
+      "loss": 0.3856,
+      "step": 25
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1843863079676701,
+      "learning_rate": 0.00019999013584066477,
+      "loss": 0.4103,
+      "step": 26
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12711897067641606,
+      "learning_rate": 0.00019998657386283668,
+      "loss": 0.3944,
+      "step": 27
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.15803894896156442,
+      "learning_rate": 0.00019998246394097485,
+      "loss": 0.3799,
+      "step": 28
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14337223811220748,
+      "learning_rate": 0.0001999778060976024,
+      "loss": 0.4021,
+      "step": 29
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14058905380496423,
+      "learning_rate": 0.0001999726003582451,
+      "loss": 0.3825,
+      "step": 30
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.13899671091332336,
+      "learning_rate": 0.0001999668467514313,
+      "loss": 0.3538,
+      "step": 31
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14151762743781257,
+      "learning_rate": 0.00019996054530869184,
+      "loss": 0.3706,
+      "step": 32
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1091466375973486,
+      "learning_rate": 0.0001999536960645597,
+      "loss": 0.3567,
+      "step": 33
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15473346268181623,
+      "learning_rate": 0.00019994629905656993,
+      "loss": 0.3882,
+      "step": 34
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15255868250724997,
+      "learning_rate": 0.00019993835432525944,
+      "loss": 0.4021,
+      "step": 35
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12591758352829893,
+      "learning_rate": 0.00019992986191416678,
+      "loss": 0.3674,
+      "step": 36
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12800053098222003,
+      "learning_rate": 0.00019992082186983184,
+      "loss": 0.3456,
+      "step": 37
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1561786488699848,
+      "learning_rate": 0.00019991123424179568,
+      "loss": 0.4018,
+      "step": 38
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12295103882282496,
+      "learning_rate": 0.00019990109908260014,
+      "loss": 0.3741,
+      "step": 39
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1349061495900084,
+      "learning_rate": 0.00019989041644778773,
+      "loss": 0.3971,
+      "step": 40
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1294512498451578,
+      "learning_rate": 0.00019987918639590116,
+      "loss": 0.3769,
+      "step": 41
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11380686986300748,
+      "learning_rate": 0.00019986740898848306,
+      "loss": 0.3716,
+      "step": 42
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1666574972332432,
+      "learning_rate": 0.0001998550842900757,
+      "loss": 0.4034,
+      "step": 43
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12108467052630499,
+      "learning_rate": 0.00019984221236822068,
+      "loss": 0.3742,
+      "step": 44
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10788347596246212,
+      "learning_rate": 0.0001998287932934583,
+      "loss": 0.3781,
+      "step": 45
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11650449875763275,
+      "learning_rate": 0.00019981482713932747,
+      "loss": 0.3767,
+      "step": 46
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10558453138729773,
+      "learning_rate": 0.0001998003139823652,
+      "loss": 0.3567,
+      "step": 47
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10144804043456603,
+      "learning_rate": 0.00019978525390210617,
+      "loss": 0.3632,
+      "step": 48
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10035769946828677,
+      "learning_rate": 0.00019976964698108218,
+      "loss": 0.3537,
+      "step": 49
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.126809031067748,
+      "learning_rate": 0.000199753493304822,
+      "loss": 0.3173,
+      "step": 50
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.19513381413671083,
+      "learning_rate": 0.0001997367929618506,
+      "loss": 0.3496,
+      "step": 51
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.17908545892800698,
+      "learning_rate": 0.00019971954604368876,
+      "loss": 0.3591,
+      "step": 52
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11264543669254343,
+      "learning_rate": 0.00019970175264485266,
+      "loss": 0.3867,
+      "step": 53
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11563180826092681,
+      "learning_rate": 0.00019968341286285329,
+      "loss": 0.3987,
+      "step": 54
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10681293159646547,
+      "learning_rate": 0.0001996645267981958,
+      "loss": 0.3667,
+      "step": 55
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.10701450787770315,
+      "learning_rate": 0.00019964509455437925,
+      "loss": 0.3602,
+      "step": 56
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10980014258294671,
+      "learning_rate": 0.00019962511623789568,
+      "loss": 0.3256,
+      "step": 57
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.12021640618530655,
+      "learning_rate": 0.0001996045919582298,
+      "loss": 0.367,
+      "step": 58
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.12597288821281044,
+      "learning_rate": 0.00019958352182785833,
+      "loss": 0.3455,
+      "step": 59
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11644708857930185,
+      "learning_rate": 0.0001995619059622492,
+      "loss": 0.3635,
+      "step": 60
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.10761054117133902,
+      "learning_rate": 0.0001995397444798613,
+      "loss": 0.3529,
+      "step": 61
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1109057285254121,
+      "learning_rate": 0.0001995170375021433,
+      "loss": 0.3609,
+      "step": 62
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.11567441522297107,
+      "learning_rate": 0.00019949378515353352,
+      "loss": 0.349,
+      "step": 63
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.11691885992187645,
+      "learning_rate": 0.0001994699875614589,
+      "loss": 0.3619,
+      "step": 64
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.10698464450938733,
+      "learning_rate": 0.00019944564485633448,
+      "loss": 0.3485,
+      "step": 65
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09435632457751685,
+      "learning_rate": 0.00019942075717156255,
+      "loss": 0.3324,
+      "step": 66
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.12167243583832113,
+      "learning_rate": 0.00019939532464353197,
+      "loss": 0.3662,
+      "step": 67
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09943615587328276,
+      "learning_rate": 0.00019936934741161748,
+      "loss": 0.3338,
+      "step": 68
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0973262121328283,
+      "learning_rate": 0.00019934282561817894,
+      "loss": 0.331,
+      "step": 69
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.11002091860459373,
+      "learning_rate": 0.0001993157594085604,
+      "loss": 0.3701,
+      "step": 70
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1142901065214701,
+      "learning_rate": 0.00019928814893108946,
+      "loss": 0.3262,
+      "step": 71
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.09546631592208919,
+      "learning_rate": 0.00019925999433707646,
+      "loss": 0.3405,
+      "step": 72
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1081166729380405,
+      "learning_rate": 0.00019923129578081343,
+      "loss": 0.3492,
+      "step": 73
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.11172505774397264,
+      "learning_rate": 0.00019920205341957356,
+      "loss": 0.3299,
+      "step": 74
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.12460023688050625,
+      "learning_rate": 0.00019917226741361015,
+      "loss": 0.3688,
+      "step": 75
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1069893756801982,
+      "learning_rate": 0.00019914193792615578,
+      "loss": 0.3512,
+      "step": 76
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.12430831151650398,
+      "learning_rate": 0.00019911106512342133,
+      "loss": 0.3549,
+      "step": 77
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.12311258529387459,
+      "learning_rate": 0.00019907964917459522,
+      "loss": 0.3349,
+      "step": 78
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.11088178488116049,
+      "learning_rate": 0.00019904769025184238,
+      "loss": 0.3378,
+      "step": 79
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.12291779635470604,
+      "learning_rate": 0.00019901518853030335,
+      "loss": 0.3394,
+      "step": 80
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09602386387271089,
+      "learning_rate": 0.0001989821441880933,
+      "loss": 0.3392,
+      "step": 81
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.12045916281672614,
+      "learning_rate": 0.00019894855740630097,
+      "loss": 0.3335,
+      "step": 82
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09350635737592786,
+      "learning_rate": 0.00019891442836898792,
+      "loss": 0.3229,
+      "step": 83
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1285676048115615,
+      "learning_rate": 0.00019887975726318727,
+      "loss": 0.3569,
+      "step": 84
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.10284105969506123,
+      "learning_rate": 0.00019884454427890273,
+      "loss": 0.3437,
+      "step": 85
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1112023948628312,
+      "learning_rate": 0.00019880878960910772,
+      "loss": 0.3287,
+      "step": 86
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.11126187930145588,
+      "learning_rate": 0.00019877249344974412,
+      "loss": 0.339,
+      "step": 87
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.12197619815567498,
+      "learning_rate": 0.00019873565599972125,
+      "loss": 0.3001,
+      "step": 88
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1196992861743786,
+      "learning_rate": 0.0001986982774609148,
+      "loss": 0.3502,
+      "step": 89
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.11154916636877048,
+      "learning_rate": 0.00019866035803816582,
+      "loss": 0.333,
+      "step": 90
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.10705061392796272,
+      "learning_rate": 0.00019862189793927936,
+      "loss": 0.3075,
+      "step": 91
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.10385638506846642,
+      "learning_rate": 0.0001985828973750236,
+      "loss": 0.3407,
+      "step": 92
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.10640488303002456,
+      "learning_rate": 0.0001985433565591284,
+      "loss": 0.3099,
+      "step": 93
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.10028365926799478,
+      "learning_rate": 0.00019850327570828448,
+      "loss": 0.3238,
+      "step": 94
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.11769222207986124,
+      "learning_rate": 0.00019846265504214192,
+      "loss": 0.3312,
+      "step": 95
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.08843620824207431,
+      "learning_rate": 0.00019842149478330916,
+      "loss": 0.3185,
+      "step": 96
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.10533045602926006,
+      "learning_rate": 0.00019837979515735166,
+      "loss": 0.3135,
+      "step": 97
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09390129358728014,
+      "learning_rate": 0.00019833755639279068,
+      "loss": 0.3219,
+      "step": 98
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.10197763696418795,
+      "learning_rate": 0.00019829477872110212,
+      "loss": 0.3376,
+      "step": 99
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09957845993500436,
+      "learning_rate": 0.00019825146237671513,
+      "loss": 0.3049,
+      "step": 100
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09991944428649677,
+      "learning_rate": 0.00019820760759701092,
+      "loss": 0.325,
+      "step": 101
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.09911919577900208,
+      "learning_rate": 0.00019816321462232141,
+      "loss": 0.3403,
+      "step": 102
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.09894182879026266,
+      "learning_rate": 0.00019811828369592792,
+      "loss": 0.3041,
+      "step": 103
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.10403013122601013,
+      "learning_rate": 0.00019807281506405985,
+      "loss": 0.3232,
+      "step": 104
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1140331888750124,
+      "learning_rate": 0.00019802680897589326,
+      "loss": 0.3003,
+      "step": 105
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0948238708694001,
+      "learning_rate": 0.0001979802656835497,
+      "loss": 0.3145,
+      "step": 106
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.10461514360193062,
+      "learning_rate": 0.00019793318544209453,
+      "loss": 0.3084,
+      "step": 107
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.09510968588008493,
+      "learning_rate": 0.0001978855685095358,
+      "loss": 0.3276,
+      "step": 108
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1308319646681228,
+      "learning_rate": 0.00019783741514682268,
+      "loss": 0.3142,
+      "step": 109
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.11389224347652244,
+      "learning_rate": 0.00019778872561784406,
+      "loss": 0.3101,
+      "step": 110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.12450943739053914,
+      "learning_rate": 0.0001977395001894272,
+      "loss": 0.3097,
+      "step": 111
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.10299716073441033,
+      "learning_rate": 0.00019768973913133602,
+      "loss": 0.3175,
+      "step": 112
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.10371674936249396,
+      "learning_rate": 0.00019763944271626994,
+      "loss": 0.3337,
+      "step": 113
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.08894218172938811,
+      "learning_rate": 0.00019758861121986222,
+      "loss": 0.3097,
+      "step": 114
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.32247820496559143,
+      "eval_runtime": 46.8718,
+      "eval_samples_per_second": 49.561,
+      "eval_steps_per_second": 0.789,
+      "step": 114
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.10358830815258127,
+      "learning_rate": 0.00019753724492067832,
+      "loss": 0.3284,
+      "step": 115
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0971743471362056,
+      "learning_rate": 0.00019748534410021461,
+      "loss": 0.3105,
+      "step": 116
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.10403048037023431,
+      "learning_rate": 0.0001974329090428968,
+      "loss": 0.3279,
+      "step": 117
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.10480969755785682,
+      "learning_rate": 0.0001973799400360782,
+      "loss": 0.3213,
+      "step": 118
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.09168995047664043,
+      "learning_rate": 0.00019732643737003827,
+      "loss": 0.3095,
+      "step": 119
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.10097243302421802,
+      "learning_rate": 0.00019727240133798107,
+      "loss": 0.321,
+      "step": 120
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.08975251893962247,
+      "learning_rate": 0.00019721783223603358,
+      "loss": 0.3004,
+      "step": 121
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.10454051067973331,
+      "learning_rate": 0.00019716273036324405,
+      "loss": 0.3071,
+      "step": 122
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1085035484630249,
+      "learning_rate": 0.0001971070960215805,
+      "loss": 0.3094,
+      "step": 123
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.10625999838403269,
+      "learning_rate": 0.0001970509295159289,
+      "loss": 0.3316,
+      "step": 124
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.11141757132451943,
+      "learning_rate": 0.00019699423115409156,
+      "loss": 0.3248,
+      "step": 125
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0961378403441233,
+      "learning_rate": 0.00019693700124678553,
+      "loss": 0.3209,
+      "step": 126
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.10232953052892273,
+      "learning_rate": 0.00019687924010764078,
+      "loss": 0.3314,
+      "step": 127
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.10140256025657933,
+      "learning_rate": 0.0001968209480531985,
+      "loss": 0.3188,
+      "step": 128
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.09730796218102972,
+      "learning_rate": 0.00019676212540290942,
+      "loss": 0.3052,
+      "step": 129
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.09433600443488169,
+      "learning_rate": 0.00019670277247913205,
+      "loss": 0.294,
+      "step": 130
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1201079489093659,
+      "learning_rate": 0.00019664288960713082,
+      "loss": 0.2879,
+      "step": 131
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.10457277310041056,
+      "learning_rate": 0.00019658247711507444,
+      "loss": 0.2958,
+      "step": 132
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.10567805044271565,
+      "learning_rate": 0.00019652153533403403,
+      "loss": 0.3289,
+      "step": 133
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0927741653012288,
+      "learning_rate": 0.00019646006459798123,
+      "loss": 0.3188,
+      "step": 134
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1051396515968855,
+      "learning_rate": 0.00019639806524378656,
+      "loss": 0.3053,
+      "step": 135
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.10548468310486735,
+      "learning_rate": 0.00019633553761121733,
+      "loss": 0.2807,
+      "step": 136
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.09192390985747663,
+      "learning_rate": 0.00019627248204293602,
+      "loss": 0.2934,
+      "step": 137
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.09797500563761673,
+      "learning_rate": 0.00019620889888449824,
+      "loss": 0.3032,
+      "step": 138
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0954822746021495,
+      "learning_rate": 0.00019614478848435093,
+      "loss": 0.2931,
+      "step": 139
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1107409914480272,
+      "learning_rate": 0.00019608015119383033,
+      "loss": 0.3146,
+      "step": 140
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.10994913773044612,
+      "learning_rate": 0.00019601498736716017,
+      "loss": 0.3008,
+      "step": 141
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1029786496548167,
+      "learning_rate": 0.00019594929736144976,
+      "loss": 0.2857,
+      "step": 142
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.09815352669101451,
+      "learning_rate": 0.0001958830815366918,
+      "loss": 0.317,
+      "step": 143
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1089273471045981,
+      "learning_rate": 0.00019581634025576077,
+      "loss": 0.2998,
+      "step": 144
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.10576455261294318,
+      "learning_rate": 0.0001957490738844106,
+      "loss": 0.3209,
+      "step": 145
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.08704284438783577,
+      "learning_rate": 0.00019568128279127283,
+      "loss": 0.2874,
+      "step": 146
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.09956041703676417,
+      "learning_rate": 0.0001956129673478546,
+      "loss": 0.3329,
+      "step": 147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.09777830126282391,
+      "learning_rate": 0.00019554412792853663,
+      "loss": 0.2956,
+      "step": 148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.10546178264916517,
+      "learning_rate": 0.00019547476491057098,
+      "loss": 0.2903,
+      "step": 149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.10669839590203294,
+      "learning_rate": 0.00019540487867407923,
+      "loss": 0.3049,
+      "step": 150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11471803200283248,
+      "learning_rate": 0.0001953344696020503,
+      "loss": 0.3083,
+      "step": 151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.10963369752875587,
+      "learning_rate": 0.00019526353808033825,
+      "loss": 0.2931,
+      "step": 152
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11554877178396908,
+      "learning_rate": 0.00019519208449766037,
+      "loss": 0.3026,
+      "step": 153
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.10607657303337156,
+      "learning_rate": 0.00019512010924559482,
+      "loss": 0.3211,
+      "step": 154
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0975269606190038,
+      "learning_rate": 0.0001950476127185787,
+      "loss": 0.2829,
+      "step": 155
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1035167810987296,
+      "learning_rate": 0.00019497459531390572,
+      "loss": 0.3138,
+      "step": 156
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.09514276759779315,
+      "learning_rate": 0.0001949010574317241,
+      "loss": 0.3187,
+      "step": 157
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.09026189607323905,
+      "learning_rate": 0.00019482699947503444,
+      "loss": 0.2968,
+      "step": 158
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08616104906595601,
+      "learning_rate": 0.00019475242184968737,
+      "loss": 0.3274,
+      "step": 159
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.10915950851398411,
+      "learning_rate": 0.00019467732496438137,
+      "loss": 0.304,
+      "step": 160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.08979641144073457,
+      "learning_rate": 0.00019460170923066059,
+      "loss": 0.3193,
+      "step": 161
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.11351817118341677,
+      "learning_rate": 0.00019452557506291265,
+      "loss": 0.2993,
+      "step": 162
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.09868559510897262,
+      "learning_rate": 0.00019444892287836613,
+      "loss": 0.2944,
+      "step": 163
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.09445445807583507,
+      "learning_rate": 0.00019437175309708853,
+      "loss": 0.2881,
+      "step": 164
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.10081082268724936,
+      "learning_rate": 0.0001942940661419838,
+      "loss": 0.2993,
+      "step": 165
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.09831846775753789,
+      "learning_rate": 0.00019421586243879022,
+      "loss": 0.2963,
+      "step": 166
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.10866947912568455,
+      "learning_rate": 0.00019413714241607778,
+      "loss": 0.3006,
+      "step": 167
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.09740833284710819,
+      "learning_rate": 0.0001940579065052461,
+      "loss": 0.2918,
+      "step": 168
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.10527034482500348,
+      "learning_rate": 0.00019397815514052188,
+      "loss": 0.2785,
+      "step": 169
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0927473633917097,
+      "learning_rate": 0.0001938978887589567,
+      "loss": 0.2965,
+      "step": 170
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10151934623914236,
+      "learning_rate": 0.00019381710780042445,
+      "loss": 0.2737,
+      "step": 171
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.09447061233249576,
+      "learning_rate": 0.00019373581270761903,
+      "loss": 0.2733,
+      "step": 172
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.08983521317268696,
+      "learning_rate": 0.0001936540039260518,
+      "loss": 0.2988,
+      "step": 173
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10398317855817302,
+      "learning_rate": 0.00019357168190404936,
+      "loss": 0.319,
+      "step": 174
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10209714484284646,
+      "learning_rate": 0.00019348884709275089,
+      "loss": 0.2931,
+      "step": 175
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0888032001647003,
+      "learning_rate": 0.00019340549994610568,
+      "loss": 0.2916,
+      "step": 176
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.09222111170041376,
+      "learning_rate": 0.0001933216409208708,
+      "loss": 0.2764,
+      "step": 177
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.08894839087559142,
+      "learning_rate": 0.00019323727047660852,
+      "loss": 0.2766,
+      "step": 178
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.08712085222748618,
+      "learning_rate": 0.00019315238907568368,
+      "loss": 0.288,
+      "step": 179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.09010570629087074,
+      "learning_rate": 0.00019306699718326132,
+      "loss": 0.3053,
+      "step": 180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.08898480460454242,
+      "learning_rate": 0.00019298109526730406,
+      "loss": 0.3106,
+      "step": 181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.09263026081266391,
+      "learning_rate": 0.00019289468379856948,
+      "loss": 0.282,
+      "step": 182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.09946322003302056,
+      "learning_rate": 0.00019280776325060776,
+      "loss": 0.3012,
+      "step": 183
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08921305454533981,
+      "learning_rate": 0.00019272033409975872,
+      "loss": 0.3228,
+      "step": 184
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08915832534364167,
+      "learning_rate": 0.00019263239682514952,
+      "loss": 0.2786,
+      "step": 185
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0990666300908459,
+      "learning_rate": 0.0001925439519086919,
+      "loss": 0.3141,
+      "step": 186
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08714626668622741,
+      "learning_rate": 0.0001924549998350796,
+      "loss": 0.2789,
+      "step": 187
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.08756339563123404,
+      "learning_rate": 0.00019236554109178564,
+      "loss": 0.2817,
+      "step": 188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10343734012257186,
+      "learning_rate": 0.0001922755761690597,
+      "loss": 0.2817,
+      "step": 189
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.09702887235790607,
+      "learning_rate": 0.0001921851055599254,
+      "loss": 0.3272,
+      "step": 190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.11193726600316871,
+      "learning_rate": 0.00019209412976017759,
+      "loss": 0.3235,
+      "step": 191
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.08351903176473893,
+      "learning_rate": 0.00019200264926837968,
+      "loss": 0.3079,
+      "step": 192
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1168886148139429,
+      "learning_rate": 0.0001919106645858609,
+      "loss": 0.3042,
+      "step": 193
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.09442179605189382,
+      "learning_rate": 0.00019181817621671346,
+      "loss": 0.3096,
+      "step": 194
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.09505975759784065,
+      "learning_rate": 0.00019172518466778997,
+      "loss": 0.3064,
+      "step": 195
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.09582644905193463,
+      "learning_rate": 0.0001916316904487005,
+      "loss": 0.3069,
+      "step": 196
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.09520837321670136,
+      "learning_rate": 0.00019153769407180974,
+      "loss": 0.2812,
+      "step": 197
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09346620998929686,
+      "learning_rate": 0.0001914431960522345,
+      "loss": 0.2831,
+      "step": 198
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11274848820567271,
+      "learning_rate": 0.00019134819690784052,
+      "loss": 0.321,
+      "step": 199
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10288972839016944,
+      "learning_rate": 0.00019125269715923983,
+      "loss": 0.2946,
+      "step": 200
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10177601781379259,
+      "learning_rate": 0.00019115669732978793,
+      "loss": 0.3179,
+      "step": 201
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09770325959747087,
+      "learning_rate": 0.0001910601979455807,
+      "loss": 0.3003,
+      "step": 202
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09339189681217434,
+      "learning_rate": 0.00019096319953545185,
+      "loss": 0.301,
+      "step": 203
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.08950118393887461,
+      "learning_rate": 0.0001908657026309697,
+      "loss": 0.273,
+      "step": 204
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.08455488028914505,
+      "learning_rate": 0.00019076770776643443,
+      "loss": 0.2753,
+      "step": 205
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.08274370419600903,
+      "learning_rate": 0.00019066921547887528,
+      "loss": 0.2799,
+      "step": 206
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.115370504339988,
+      "learning_rate": 0.00019057022630804716,
+      "loss": 0.3343,
+      "step": 207
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.09499601502436628,
+      "learning_rate": 0.00019047074079642825,
+      "loss": 0.2984,
+      "step": 208
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0871774881863037,
+      "learning_rate": 0.00019037075948921672,
+      "loss": 0.2993,
+      "step": 209
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.09447198263057913,
+      "learning_rate": 0.00019027028293432767,
+      "loss": 0.2861,
+      "step": 210
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.08965989145646927,
+      "learning_rate": 0.00019016931168239033,
+      "loss": 0.2843,
+      "step": 211
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.08854221830009207,
+      "learning_rate": 0.00019006784628674495,
+      "loss": 0.2838,
+      "step": 212
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10118873301091154,
+      "learning_rate": 0.0001899658873034398,
+      "loss": 0.3066,
+      "step": 213
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.09319369346721061,
+      "learning_rate": 0.000189863435291228,
+      "loss": 0.2744,
+      "step": 214
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.08534805030929074,
+      "learning_rate": 0.00018976049081156462,
+      "loss": 0.292,
+      "step": 215
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.09266172509477273,
+      "learning_rate": 0.0001896570544286036,
+      "loss": 0.3073,
+      "step": 216
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.09056726704958203,
+      "learning_rate": 0.0001895531267091945,
+      "loss": 0.2726,
+      "step": 217
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.08720624224913519,
+      "learning_rate": 0.00018944870822287956,
+      "loss": 0.2966,
+      "step": 218
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.08606448328590535,
+      "learning_rate": 0.00018934379954189042,
+      "loss": 0.2682,
+      "step": 219
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.09486741968607297,
+      "learning_rate": 0.00018923840124114517,
+      "loss": 0.2756,
+      "step": 220
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09813244114135808,
+      "learning_rate": 0.00018913251389824503,
+      "loss": 0.2803,
+      "step": 221
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09194220543702984,
+      "learning_rate": 0.00018902613809347133,
+      "loss": 0.2858,
+      "step": 222
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09849200253648169,
+      "learning_rate": 0.00018891927440978215,
+      "loss": 0.2466,
+      "step": 223
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09173729159995127,
+      "learning_rate": 0.00018881192343280935,
+      "loss": 0.2928,
+      "step": 224
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09794718320943081,
+      "learning_rate": 0.00018870408575085516,
+      "loss": 0.3038,
+      "step": 225
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.08250539985960245,
+      "learning_rate": 0.00018859576195488908,
+      "loss": 0.2748,
+      "step": 226
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09642356604986511,
+      "learning_rate": 0.0001884869526385446,
+      "loss": 0.2942,
+      "step": 227
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.08980724380857069,
+      "learning_rate": 0.00018837765839811594,
+      "loss": 0.2859,
+      "step": 228
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.28947827219963074,
+      "eval_runtime": 46.9014,
+      "eval_samples_per_second": 49.529,
+      "eval_steps_per_second": 0.789,
+      "step": 228
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09018009398559709,
+      "learning_rate": 0.00018826787983255473,
+      "loss": 0.2964,
+      "step": 229
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09028670230436779,
+      "learning_rate": 0.00018815761754346696,
+      "loss": 0.2817,
+      "step": 230
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0900338276894956,
+      "learning_rate": 0.00018804687213510923,
+      "loss": 0.288,
+      "step": 231
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0992327235361792,
+      "learning_rate": 0.00018793564421438604,
+      "loss": 0.2833,
+      "step": 232
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09929733862269687,
+      "learning_rate": 0.00018782393439084586,
+      "loss": 0.3016,
+      "step": 233
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.11392645185127047,
+      "learning_rate": 0.00018771174327667826,
+      "loss": 0.2974,
+      "step": 234
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09348795946069262,
+      "learning_rate": 0.0001875990714867103,
+      "loss": 0.2799,
+      "step": 235
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08897432743614427,
+      "learning_rate": 0.00018748591963840323,
+      "loss": 0.2804,
+      "step": 236
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0907797774229916,
+      "learning_rate": 0.0001873722883518491,
+      "loss": 0.2757,
+      "step": 237
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10148105657025418,
+      "learning_rate": 0.00018725817824976737,
+      "loss": 0.3061,
+      "step": 238
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09380012404079309,
+      "learning_rate": 0.00018714358995750152,
+      "loss": 0.2818,
+      "step": 239
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09137309952682544,
+      "learning_rate": 0.00018702852410301554,
+      "loss": 0.2972,
+      "step": 240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09299401112373723,
+      "learning_rate": 0.00018691298131689056,
+      "loss": 0.2794,
+      "step": 241
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09551399382576438,
+      "learning_rate": 0.00018679696223232147,
+      "loss": 0.3172,
+      "step": 242
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07818498394845357,
+      "learning_rate": 0.00018668046748511322,
+      "loss": 0.2608,
+      "step": 243
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08463554443447543,
+      "learning_rate": 0.00018656349771367756,
+      "loss": 0.256,
+      "step": 244
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09327367136790388,
+      "learning_rate": 0.00018644605355902947,
+      "loss": 0.2913,
+      "step": 245
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09411906573249411,
+      "learning_rate": 0.00018632813566478356,
+      "loss": 0.2606,
+      "step": 246
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08711074156608702,
+      "learning_rate": 0.00018620974467715065,
+      "loss": 0.2708,
+      "step": 247
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.08663229058203331,
+      "learning_rate": 0.00018609088124493423,
+      "loss": 0.2591,
+      "step": 248
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10619707209248297,
+      "learning_rate": 0.00018597154601952682,
+      "loss": 0.297,
+      "step": 249
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10839120308044262,
+      "learning_rate": 0.00018585173965490653,
+      "loss": 0.329,
+      "step": 250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.095543991834507,
+      "learning_rate": 0.00018573146280763324,
+      "loss": 0.2956,
+      "step": 251
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09721453906914458,
+      "learning_rate": 0.00018561071613684533,
+      "loss": 0.2647,
+      "step": 252
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10857672249105992,
+      "learning_rate": 0.0001854895003042558,
+      "loss": 0.2859,
+      "step": 253
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09082162507452043,
+      "learning_rate": 0.00018536781597414877,
+      "loss": 0.2742,
+      "step": 254
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09579065271278289,
+      "learning_rate": 0.00018524566381337575,
+      "loss": 0.2813,
+      "step": 255
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09903891214899607,
+      "learning_rate": 0.00018512304449135218,
+      "loss": 0.2865,
+      "step": 256
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09389582359490105,
+      "learning_rate": 0.00018499995868005346,
+      "loss": 0.3096,
+      "step": 257
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09324607418319461,
+      "learning_rate": 0.00018487640705401156,
+      "loss": 0.2905,
+      "step": 258
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10148213489782407,
+      "learning_rate": 0.00018475239029031113,
+      "loss": 0.3009,
+      "step": 259
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0846471582728864,
+      "learning_rate": 0.0001846279090685859,
+      "loss": 0.2633,
+      "step": 260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08540013674768626,
+      "learning_rate": 0.00018450296407101485,
+      "loss": 0.2687,
+      "step": 261
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08880630965071441,
+      "learning_rate": 0.00018437755598231856,
+      "loss": 0.2844,
+      "step": 262
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08784531644054196,
+      "learning_rate": 0.00018425168548975553,
+      "loss": 0.2769,
+      "step": 263
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08271339318400822,
+      "learning_rate": 0.00018412535328311814,
+      "loss": 0.284,
+      "step": 264
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09649656272854211,
+      "learning_rate": 0.00018399856005472915,
+      "loss": 0.2704,
+      "step": 265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07902885557034076,
+      "learning_rate": 0.0001838713064994378,
+      "loss": 0.2696,
+      "step": 266
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08579666887653513,
+      "learning_rate": 0.00018374359331461596,
+      "loss": 0.2691,
+      "step": 267
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0866125991810605,
+      "learning_rate": 0.00018361542120015444,
+      "loss": 0.2731,
+      "step": 268
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07925193234153907,
+      "learning_rate": 0.0001834867908584589,
+      "loss": 0.2561,
+      "step": 269
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.10042955686396764,
+      "learning_rate": 0.00018335770299444633,
+      "loss": 0.2744,
+      "step": 270
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09303964268068415,
+      "learning_rate": 0.00018322815831554095,
+      "loss": 0.2847,
+      "step": 271
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08995962962741563,
+      "learning_rate": 0.0001830981575316704,
+      "loss": 0.2703,
+      "step": 272
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09465338643258427,
+      "learning_rate": 0.0001829677013552619,
+      "loss": 0.2771,
+      "step": 273
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0891484535952062,
+      "learning_rate": 0.00018283679050123825,
+      "loss": 0.2643,
+      "step": 274
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1004402160615555,
+      "learning_rate": 0.00018270542568701398,
+      "loss": 0.3046,
+      "step": 275
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09623505236310947,
+      "learning_rate": 0.00018257360763249142,
+      "loss": 0.2751,
+      "step": 276
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09312261323142398,
+      "learning_rate": 0.00018244133706005672,
+      "loss": 0.268,
+      "step": 277
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08944408004542047,
+      "learning_rate": 0.00018230861469457587,
+      "loss": 0.3095,
+      "step": 278
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.10320455668551522,
+      "learning_rate": 0.00018217544126339087,
+      "loss": 0.2899,
+      "step": 279
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08699760279696164,
+      "learning_rate": 0.00018204181749631557,
+      "loss": 0.2814,
+      "step": 280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08898658382129276,
+      "learning_rate": 0.00018190774412563175,
+      "loss": 0.2609,
+      "step": 281
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09664290624376146,
+      "learning_rate": 0.00018177322188608504,
+      "loss": 0.2682,
+      "step": 282
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.10047512987861203,
+      "learning_rate": 0.00018163825151488102,
+      "loss": 0.2589,
+      "step": 283
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1045016435996989,
+      "learning_rate": 0.00018150283375168114,
+      "loss": 0.2697,
+      "step": 284
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09514303189191896,
+      "learning_rate": 0.00018136696933859856,
+      "loss": 0.2639,
+      "step": 285
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0880861253799837,
+      "learning_rate": 0.0001812306590201942,
+      "loss": 0.276,
+      "step": 286
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08905109303188885,
+      "learning_rate": 0.00018109390354347258,
+      "loss": 0.2708,
+      "step": 287
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09341254196818168,
+      "learning_rate": 0.00018095670365787784,
+      "loss": 0.2643,
+      "step": 288
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09389008546351392,
+      "learning_rate": 0.0001808190601152895,
+      "loss": 0.3009,
+      "step": 289
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09573632267856752,
+      "learning_rate": 0.00018068097367001843,
+      "loss": 0.2836,
+      "step": 290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08005491943936306,
+      "learning_rate": 0.0001805424450788026,
+      "loss": 0.2766,
+      "step": 291
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08581845751612466,
+      "learning_rate": 0.00018040347510080317,
+      "loss": 0.2649,
+      "step": 292
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0895258836652537,
+      "learning_rate": 0.00018026406449760004,
+      "loss": 0.277,
+      "step": 293
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09617168204925791,
+      "learning_rate": 0.00018012421403318794,
+      "loss": 0.3049,
+      "step": 294
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08981851555091117,
+      "learning_rate": 0.00017998392447397197,
+      "loss": 0.279,
+      "step": 295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0895854044783868,
+      "learning_rate": 0.0001798431965887636,
+      "loss": 0.26,
+      "step": 296
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0910303803790978,
+      "learning_rate": 0.00017970203114877652,
+      "loss": 0.2871,
+      "step": 297
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.09277535486749651,
+      "learning_rate": 0.00017956042892762212,
+      "loss": 0.2843,
+      "step": 298
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08655364873971913,
+      "learning_rate": 0.00017941839070130557,
+      "loss": 0.2685,
+      "step": 299
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.09487799701744236,
+      "learning_rate": 0.0001792759172482213,
+      "loss": 0.2871,
+      "step": 300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.09956376115924534,
+      "learning_rate": 0.000179133009349149,
+      "loss": 0.2689,
+      "step": 301
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0970485758861932,
+      "learning_rate": 0.00017898966778724909,
+      "loss": 0.2531,
+      "step": 302
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08338075054949988,
+      "learning_rate": 0.00017884589334805857,
+      "loss": 0.2778,
+      "step": 303
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08854705663244716,
+      "learning_rate": 0.00017870168681948667,
+      "loss": 0.277,
+      "step": 304
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09591918514989665,
+      "learning_rate": 0.00017855704899181056,
+      "loss": 0.2673,
+      "step": 305
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09225407839045617,
+      "learning_rate": 0.00017841198065767107,
+      "loss": 0.2679,
+      "step": 306
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09041482893974015,
+      "learning_rate": 0.00017826648261206812,
+      "loss": 0.2849,
+      "step": 307
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1043417417981917,
+      "learning_rate": 0.00017812055565235674,
+      "loss": 0.3,
+      "step": 308
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08556408926676959,
+      "learning_rate": 0.0001779742005782423,
+      "loss": 0.2715,
+      "step": 309
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08299224651467078,
+      "learning_rate": 0.00017782741819177638,
+      "loss": 0.2813,
+      "step": 310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0892826262859673,
+      "learning_rate": 0.0001776802092973523,
+      "loss": 0.2756,
+      "step": 311
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08451795190473695,
+      "learning_rate": 0.0001775325747017008,
+      "loss": 0.2797,
+      "step": 312
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0893836524166378,
+      "learning_rate": 0.00017738451521388538,
+      "loss": 0.2673,
+      "step": 313
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08220379144869369,
+      "learning_rate": 0.00017723603164529817,
+      "loss": 0.269,
+      "step": 314
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08683409387310503,
+      "learning_rate": 0.0001770871248096552,
+      "loss": 0.2567,
+      "step": 315
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08179525647934902,
+      "learning_rate": 0.00017693779552299224,
+      "loss": 0.2721,
+      "step": 316
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08901531847813472,
+      "learning_rate": 0.00017678804460366,
+      "loss": 0.2722,
+      "step": 317
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08961734544216175,
+      "learning_rate": 0.00017663787287231997,
+      "loss": 0.2765,
+      "step": 318
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.10100944489074588,
+      "learning_rate": 0.0001764872811519396,
+      "loss": 0.2792,
+      "step": 319
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08364204845271891,
+      "learning_rate": 0.00017633627026778816,
+      "loss": 0.2651,
+      "step": 320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1101871971190039,
+      "learning_rate": 0.00017618484104743175,
+      "loss": 0.2899,
+      "step": 321
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07954472743319763,
+      "learning_rate": 0.00017603299432072925,
+      "loss": 0.2824,
+      "step": 322
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.09784510947172133,
+      "learning_rate": 0.00017588073091982746,
+      "loss": 0.2938,
+      "step": 323
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.09192983149976516,
+      "learning_rate": 0.00017572805167915664,
+      "loss": 0.2694,
+      "step": 324
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.09199281335929607,
+      "learning_rate": 0.00017557495743542585,
+      "loss": 0.2968,
+      "step": 325
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07890871464307705,
+      "learning_rate": 0.00017542144902761857,
+      "loss": 0.2502,
+      "step": 326
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08672824025848662,
+      "learning_rate": 0.00017526752729698787,
+      "loss": 0.267,
+      "step": 327
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08641232908483212,
+      "learning_rate": 0.00017511319308705198,
+      "loss": 0.2764,
+      "step": 328
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.09404642706797793,
+      "learning_rate": 0.00017495844724358956,
+      "loss": 0.2511,
+      "step": 329
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.09066701071505331,
+      "learning_rate": 0.00017480329061463504,
+      "loss": 0.2738,
+      "step": 330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0897688916610469,
+      "learning_rate": 0.00017464772405047411,
+      "loss": 0.2698,
+      "step": 331
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.088294834910458,
+      "learning_rate": 0.0001744917484036389,
+      "loss": 0.2615,
+      "step": 332
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08716962776853286,
+      "learning_rate": 0.00017433536452890347,
+      "loss": 0.2698,
+      "step": 333
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08454613439101259,
+      "learning_rate": 0.00017417857328327893,
+      "loss": 0.2714,
+      "step": 334
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0883862060356333,
+      "learning_rate": 0.00017402137552600898,
+      "loss": 0.2673,
+      "step": 335
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08783879871351531,
+      "learning_rate": 0.000173863772118565,
+      "loss": 0.2745,
+      "step": 336
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09929554250319823,
+      "learning_rate": 0.00017370576392464145,
+      "loss": 0.2925,
+      "step": 337
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08717726117092518,
+      "learning_rate": 0.000173547351810151,
+      "loss": 0.2856,
+      "step": 338
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0920284525748897,
+      "learning_rate": 0.00017338853664321992,
+      "loss": 0.2709,
+      "step": 339
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08483096298611031,
+      "learning_rate": 0.00017322931929418336,
+      "loss": 0.2528,
+      "step": 340
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08835820224263634,
+      "learning_rate": 0.0001730697006355804,
+      "loss": 0.2567,
+      "step": 341
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.09716565402547053,
+      "learning_rate": 0.0001729096815421494,
+      "loss": 0.274,
+      "step": 342
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 0.2732306718826294,
+      "eval_runtime": 46.8792,
+      "eval_samples_per_second": 49.553,
+      "eval_steps_per_second": 0.789,
+      "step": 342
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08632430083232402,
+      "learning_rate": 0.00017274926289082321,
+      "loss": 0.272,
+      "step": 343
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.099556394586788,
+      "learning_rate": 0.00017258844556072424,
+      "loss": 0.2888,
+      "step": 344
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09253918033848381,
+      "learning_rate": 0.0001724272304331598,
+      "loss": 0.2702,
+      "step": 345
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09031968934571989,
+      "learning_rate": 0.0001722656183916172,
+      "loss": 0.3057,
+      "step": 346
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0802570211372807,
+      "learning_rate": 0.00017210361032175885,
+      "loss": 0.2631,
+      "step": 347
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09860942109973829,
+      "learning_rate": 0.00017194120711141753,
+      "loss": 0.2756,
+      "step": 348
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08394950042811544,
+      "learning_rate": 0.00017177840965059145,
+      "loss": 0.2779,
+      "step": 349
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0903039630894169,
+      "learning_rate": 0.00017161521883143934,
+      "loss": 0.2545,
+      "step": 350
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09705481818131326,
+      "learning_rate": 0.00017145163554827564,
+      "loss": 0.2689,
+      "step": 351
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09986080432209749,
+      "learning_rate": 0.00017128766069756553,
+      "loss": 0.2691,
+      "step": 352
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09963488252203793,
+      "learning_rate": 0.0001711232951779201,
+      "loss": 0.2565,
+      "step": 353
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0910255512669536,
+      "learning_rate": 0.00017095853989009135,
+      "loss": 0.283,
+      "step": 354
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.10321718506401961,
+      "learning_rate": 0.00017079339573696723,
+      "loss": 0.2937,
+      "step": 355
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08911300030537235,
+      "learning_rate": 0.00017062786362356683,
+      "loss": 0.2708,
+      "step": 356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.09046623982519672,
+      "learning_rate": 0.00017046194445703524,
+      "loss": 0.2721,
+      "step": 357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08268251913394827,
+      "learning_rate": 0.00017029563914663874,
+      "loss": 0.2606,
+      "step": 358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08631575040459773,
+      "learning_rate": 0.00017012894860375975,
+      "loss": 0.2854,
+      "step": 359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.08961358857828595,
+      "learning_rate": 0.00016996187374189169,
+      "loss": 0.2505,
+      "step": 360
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0824199617534427,
+      "learning_rate": 0.00016979441547663435,
+      "loss": 0.2763,
+      "step": 361
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.09009411384569915,
+      "learning_rate": 0.0001696265747256885,
+      "loss": 0.2822,
+      "step": 362
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08428400948200425,
+      "learning_rate": 0.00016945835240885092,
+      "loss": 0.2695,
+      "step": 363
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08783878546022265,
+      "learning_rate": 0.00016928974944800963,
+      "loss": 0.2913,
+      "step": 364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08713956040465161,
+      "learning_rate": 0.00016912076676713853,
+      "loss": 0.2687,
+      "step": 365
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.09293094114634146,
+      "learning_rate": 0.00016895140529229243,
+      "loss": 0.2694,
+      "step": 366
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08260563866629433,
+      "learning_rate": 0.0001687816659516021,
+      "loss": 0.2708,
+      "step": 367
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08298706749135985,
+      "learning_rate": 0.00016861154967526894,
+      "loss": 0.2664,
+      "step": 368
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08091193577887784,
+      "learning_rate": 0.00016844105739556014,
+      "loss": 0.2629,
+      "step": 369
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07734372977054216,
+      "learning_rate": 0.00016827019004680348,
+      "loss": 0.2459,
+      "step": 370
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09272504849603619,
+      "learning_rate": 0.00016809894856538205,
+      "loss": 0.2843,
+      "step": 371
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09378162982950193,
+      "learning_rate": 0.00016792733388972932,
+      "loss": 0.2623,
+      "step": 372
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.09265377961286181,
+      "learning_rate": 0.000167755346960324,
+      "loss": 0.2681,
+      "step": 373
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08552399496999884,
+      "learning_rate": 0.0001675829887196847,
+      "loss": 0.2412,
+      "step": 374
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.09950622191516496,
+      "learning_rate": 0.00016741026011236494,
+      "loss": 0.2583,
+      "step": 375
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08495164140728746,
+      "learning_rate": 0.0001672371620849479,
+      "loss": 0.2615,
+      "step": 376
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08726491192703084,
+      "learning_rate": 0.0001670636955860413,
+      "loss": 0.2692,
+      "step": 377
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08887302463350889,
+      "learning_rate": 0.00016688986156627208,
+      "loss": 0.2721,
+      "step": 378
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07924548348860172,
+      "learning_rate": 0.00016671566097828126,
+      "loss": 0.2622,
+      "step": 379
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08714408386158717,
+      "learning_rate": 0.00016654109477671878,
+      "loss": 0.2662,
+      "step": 380
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.09102824721178376,
+      "learning_rate": 0.00016636616391823815,
+      "loss": 0.2872,
+      "step": 381
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08287301899896311,
+      "learning_rate": 0.00016619086936149128,
+      "loss": 0.2718,
+      "step": 382
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0884161917927334,
+      "learning_rate": 0.00016601521206712318,
+      "loss": 0.2658,
+      "step": 383
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09346259787153757,
+      "learning_rate": 0.0001658391929977668,
+      "loss": 0.2999,
+      "step": 384
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07983406895157653,
+      "learning_rate": 0.00016566281311803765,
+      "loss": 0.247,
+      "step": 385
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09043262654053759,
+      "learning_rate": 0.00016548607339452853,
+      "loss": 0.2595,
+      "step": 386
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09037518372688506,
+      "learning_rate": 0.00016530897479580419,
+      "loss": 0.2761,
+      "step": 387
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.09206377957083482,
+      "learning_rate": 0.00016513151829239623,
+      "loss": 0.2706,
+      "step": 388
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08353449958791122,
+      "learning_rate": 0.00016495370485679746,
+      "loss": 0.2494,
+      "step": 389
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.10227810031721918,
+      "learning_rate": 0.0001647755354634569,
+      "loss": 0.2866,
+      "step": 390
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08302674918992708,
+      "learning_rate": 0.0001645970110887741,
+      "loss": 0.2797,
+      "step": 391
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.09203890588097992,
+      "learning_rate": 0.00016441813271109417,
+      "loss": 0.243,
+      "step": 392
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08014399738341266,
+      "learning_rate": 0.0001642389013107021,
+      "loss": 0.2716,
+      "step": 393
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08328860139070615,
+      "learning_rate": 0.00016405931786981755,
+      "loss": 0.2481,
+      "step": 394
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.12109499976836058,
+      "learning_rate": 0.0001638793833725893,
+      "loss": 0.2922,
+      "step": 395
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.07664261149438296,
+      "learning_rate": 0.0001636990988050902,
+      "loss": 0.2598,
+      "step": 396
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.08388618340282568,
+      "learning_rate": 0.00016351846515531138,
+      "loss": 0.2745,
+      "step": 397
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.09079858048189099,
+      "learning_rate": 0.00016333748341315705,
+      "loss": 0.2749,
+      "step": 398
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.07643161407291658,
+      "learning_rate": 0.00016315615457043902,
+      "loss": 0.2357,
+      "step": 399
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0822688165183256,
+      "learning_rate": 0.00016297447962087134,
+      "loss": 0.2754,
+      "step": 400
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.08058792462674925,
+      "learning_rate": 0.00016279245956006468,
+      "loss": 0.2671,
+      "step": 401
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08013161175544058,
+      "learning_rate": 0.00016261009538552106,
+      "loss": 0.2617,
+      "step": 402
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08762024510593239,
+      "learning_rate": 0.0001624273880966282,
+      "loss": 0.2686,
+      "step": 403
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.10337641640393926,
+      "learning_rate": 0.0001622443386946543,
+      "loss": 0.2906,
+      "step": 404
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08263875316855607,
+      "learning_rate": 0.00016206094818274229,
+      "loss": 0.2829,
+      "step": 405
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.08735257601885658,
+      "learning_rate": 0.00016187721756590447,
+      "loss": 0.2486,
+      "step": 406
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08262708246211714,
+      "learning_rate": 0.000161693147851017,
+      "loss": 0.2532,
+      "step": 407
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.07937904244364251,
+      "learning_rate": 0.00016150874004681432,
+      "loss": 0.2456,
+      "step": 408
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.07334396919604567,
+      "learning_rate": 0.00016132399516388365,
+      "loss": 0.2326,
+      "step": 409
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.08877275500932807,
+      "learning_rate": 0.00016113891421465953,
+      "loss": 0.266,
+      "step": 410
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.07733103112922254,
+      "learning_rate": 0.00016095349821341811,
+      "loss": 0.2563,
+      "step": 411
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08256677539192653,
+      "learning_rate": 0.0001607677481762718,
+      "loss": 0.2476,
+      "step": 412
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09039191761098214,
+      "learning_rate": 0.00016058166512116344,
+      "loss": 0.265,
+      "step": 413
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.08264019005611048,
+      "learning_rate": 0.00016039525006786098,
+      "loss": 0.2425,
+      "step": 414
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.09295766006970703,
+      "learning_rate": 0.00016020850403795174,
+      "loss": 0.2892,
+      "step": 415
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.08315117741107862,
+      "learning_rate": 0.00016002142805483685,
+      "loss": 0.261,
+      "step": 416
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09020273388728756,
+      "learning_rate": 0.00015983402314372565,
+      "loss": 0.2807,
+      "step": 417
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.09765923495499138,
+      "learning_rate": 0.00015964629033163002,
+      "loss": 0.2933,
+      "step": 418
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0824294557972185,
+      "learning_rate": 0.00015945823064735887,
+      "loss": 0.2838,
+      "step": 419
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08371904511024147,
+      "learning_rate": 0.00015926984512151243,
+      "loss": 0.2797,
+      "step": 420
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08090565089802919,
+      "learning_rate": 0.00015908113478647646,
+      "loss": 0.2522,
+      "step": 421
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08221999660925741,
+      "learning_rate": 0.00015889210067641688,
+      "loss": 0.2552,
+      "step": 422
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08568247158080035,
+      "learning_rate": 0.00015870274382727396,
+      "loss": 0.2461,
+      "step": 423
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.08981252210506627,
+      "learning_rate": 0.00015851306527675649,
+      "loss": 0.2669,
+      "step": 424
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08077133299073487,
+      "learning_rate": 0.0001583230660643364,
+      "loss": 0.2591,
+      "step": 425
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08451584169624858,
+      "learning_rate": 0.00015813274723124277,
+      "loss": 0.2725,
+      "step": 426
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.09262062336675674,
+      "learning_rate": 0.00015794210982045636,
+      "loss": 0.2733,
+      "step": 427
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08449323216152702,
+      "learning_rate": 0.0001577511548767038,
+      "loss": 0.2618,
+      "step": 428
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.08776628122581846,
+      "learning_rate": 0.00015755988344645175,
+      "loss": 0.2777,
+      "step": 429
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08194483167307404,
+      "learning_rate": 0.0001573682965779013,
+      "loss": 0.2549,
+      "step": 430
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08026094037717894,
+      "learning_rate": 0.00015717639532098233,
+      "loss": 0.2345,
+      "step": 431
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.07605761431277488,
+      "learning_rate": 0.00015698418072734746,
+      "loss": 0.2571,
+      "step": 432
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.08076667459082296,
+      "learning_rate": 0.00015679165385036655,
+      "loss": 0.2394,
+      "step": 433
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09381651505577016,
+      "learning_rate": 0.0001565988157451207,
+      "loss": 0.2824,
+      "step": 434
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08298723863434691,
+      "learning_rate": 0.00015640566746839683,
+      "loss": 0.2763,
+      "step": 435
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09794013721966201,
+      "learning_rate": 0.00015621221007868142,
+      "loss": 0.2631,
+      "step": 436
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.08464739207373433,
+      "learning_rate": 0.00015601844463615505,
+      "loss": 0.2592,
+      "step": 437
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.09035904454654518,
+      "learning_rate": 0.00015582437220268647,
+      "loss": 0.2916,
+      "step": 438
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08096249634411586,
+      "learning_rate": 0.00015562999384182683,
+      "loss": 0.254,
+      "step": 439
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08181582274220905,
+      "learning_rate": 0.00015543531061880375,
+      "loss": 0.2484,
+      "step": 440
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.09617638833791838,
+      "learning_rate": 0.00015524032360051554,
+      "loss": 0.279,
+      "step": 441
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.08138784054888658,
+      "learning_rate": 0.00015504503385552546,
+      "loss": 0.2597,
+      "step": 442
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.07936100738823337,
+      "learning_rate": 0.00015484944245405571,
+      "loss": 0.2779,
+      "step": 443
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.08424353811116546,
+      "learning_rate": 0.00015465355046798157,
+      "loss": 0.2623,
+      "step": 444
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09267748840174621,
+      "learning_rate": 0.00015445735897082566,
+      "loss": 0.2676,
+      "step": 445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.09111076997524134,
+      "learning_rate": 0.00015426086903775192,
+      "loss": 0.2664,
+      "step": 446
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.07739856468175603,
+      "learning_rate": 0.00015406408174555976,
+      "loss": 0.2653,
+      "step": 447
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08163998594444614,
+      "learning_rate": 0.00015386699817267825,
+      "loss": 0.239,
+      "step": 448
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.10310242615344037,
+      "learning_rate": 0.00015366961939916008,
+      "loss": 0.2801,
+      "step": 449
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.09494229948608109,
+      "learning_rate": 0.00015347194650667563,
+      "loss": 0.2751,
+      "step": 450
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.08238892687648729,
+      "learning_rate": 0.0001532739805785072,
+      "loss": 0.2766,
+      "step": 451
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.09085904597766505,
+      "learning_rate": 0.00015307572269954294,
+      "loss": 0.2917,
+      "step": 452
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1060448383795029,
+      "learning_rate": 0.00015287717395627096,
+      "loss": 0.2853,
+      "step": 453
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0823271246337277,
+      "learning_rate": 0.0001526783354367733,
+      "loss": 0.2522,
+      "step": 454
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.08598239224647009,
+      "learning_rate": 0.00015247920823072012,
+      "loss": 0.2648,
+      "step": 455
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.07817010173696376,
+      "learning_rate": 0.0001522797934293635,
+      "loss": 0.2552,
+      "step": 456
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.2607235014438629,
+      "eval_runtime": 46.8368,
+      "eval_samples_per_second": 49.598,
+      "eval_steps_per_second": 0.79,
+      "step": 456
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.08742186259115368,
+      "learning_rate": 0.00015208009212553173,
+      "loss": 0.2539,
+      "step": 457
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.08280334253331892,
+      "learning_rate": 0.00015188010541362313,
+      "loss": 0.2454,
+      "step": 458
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.0783069399407246,
+      "learning_rate": 0.0001516798343896001,
+      "loss": 0.2495,
+      "step": 459
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.07221522055473084,
+      "learning_rate": 0.0001514792801509831,
+      "loss": 0.2231,
+      "step": 460
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0897428941623312,
+      "learning_rate": 0.00015127844379684465,
+      "loss": 0.2395,
+      "step": 461
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.07835083447615553,
+      "learning_rate": 0.0001510773264278034,
+      "loss": 0.2416,
+      "step": 462
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.09141253099298859,
+      "learning_rate": 0.0001508759291460179,
+      "loss": 0.2405,
+      "step": 463
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.08155354128614145,
+      "learning_rate": 0.0001506742530551808,
+      "loss": 0.2306,
+      "step": 464
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.09076483888395669,
+      "learning_rate": 0.0001504722992605125,
+      "loss": 0.2524,
+      "step": 465
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.08321808100767036,
+      "learning_rate": 0.0001502700688687554,
+      "loss": 0.2366,
+      "step": 466
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.08822200346311028,
+      "learning_rate": 0.00015006756298816768,
+      "loss": 0.252,
+      "step": 467
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.07831568395108493,
+      "learning_rate": 0.00014986478272851721,
+      "loss": 0.222,
+      "step": 468
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.08740469860122592,
+      "learning_rate": 0.00014966172920107557,
+      "loss": 0.2355,
+      "step": 469
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.09001639612506895,
+      "learning_rate": 0.0001494584035186118,
+      "loss": 0.249,
+      "step": 470
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.09851916535963984,
+      "learning_rate": 0.00014925480679538647,
+      "loss": 0.2527,
+      "step": 471
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.08974182525545715,
+      "learning_rate": 0.00014905094014714543,
+      "loss": 0.2408,
+      "step": 472
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.08165668222409328,
+      "learning_rate": 0.0001488468046911139,
+      "loss": 0.2608,
+      "step": 473
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.0802969762302658,
+      "learning_rate": 0.00014864240154599005,
+      "loss": 0.2363,
+      "step": 474
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.09002098969569035,
+      "learning_rate": 0.0001484377318319392,
+      "loss": 0.2433,
+      "step": 475
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.07921003039749554,
+      "learning_rate": 0.00014823279667058731,
+      "loss": 0.246,
+      "step": 476
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.09210461915658659,
+      "learning_rate": 0.0001480275971850153,
+      "loss": 0.2455,
+      "step": 477
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.08038224354992239,
+      "learning_rate": 0.00014782213449975235,
+      "loss": 0.2329,
+      "step": 478
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.08623976543939627,
+      "learning_rate": 0.0001476164097407703,
+      "loss": 0.2569,
+      "step": 479
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.08799506446018629,
+      "learning_rate": 0.00014741042403547692,
+      "loss": 0.2543,
+      "step": 480
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.08696364210212125,
+      "learning_rate": 0.0001472041785127102,
+      "loss": 0.2414,
+      "step": 481
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.08795970215519909,
+      "learning_rate": 0.000146997674302732,
+      "loss": 0.2367,
+      "step": 482
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.08328524676120717,
+      "learning_rate": 0.0001467909125372217,
+      "loss": 0.253,
+      "step": 483
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.08031124688721997,
+      "learning_rate": 0.00014658389434927017,
+      "loss": 0.2109,
+      "step": 484
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.09046760900960552,
+      "learning_rate": 0.00014637662087337356,
+      "loss": 0.2349,
+      "step": 485
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.09377292511703166,
+      "learning_rate": 0.00014616909324542703,
+      "loss": 0.2515,
+      "step": 486
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.07788800712000839,
+      "learning_rate": 0.0001459613126027185,
+      "loss": 0.2085,
+      "step": 487
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.07896486291311948,
+      "learning_rate": 0.00014575328008392254,
+      "loss": 0.2245,
+      "step": 488
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.08633577110123299,
+      "learning_rate": 0.00014554499682909396,
+      "loss": 0.2439,
+      "step": 489
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.08287761238239943,
+      "learning_rate": 0.00014533646397966164,
+      "loss": 0.2124,
+      "step": 490
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.07966913182203844,
+      "learning_rate": 0.00014512768267842239,
+      "loss": 0.2357,
+      "step": 491
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.09124779844730965,
+      "learning_rate": 0.0001449186540695344,
+      "loss": 0.2447,
+      "step": 492
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.0897516369275213,
+      "learning_rate": 0.0001447093792985114,
+      "loss": 0.253,
+      "step": 493
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.0986251922534234,
+      "learning_rate": 0.00014449985951221593,
+      "loss": 0.2688,
+      "step": 494
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.08372662047113374,
+      "learning_rate": 0.0001442900958588533,
+      "loss": 0.2409,
+      "step": 495
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.09425294593993103,
+      "learning_rate": 0.00014408008948796526,
+      "loss": 0.2348,
+      "step": 496
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.07746515270754586,
+      "learning_rate": 0.00014386984155042372,
+      "loss": 0.2375,
+      "step": 497
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.07932235667742167,
+      "learning_rate": 0.00014365935319842447,
+      "loss": 0.2134,
+      "step": 498
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.08830748617598361,
+      "learning_rate": 0.0001434486255854807,
+      "loss": 0.2482,
+      "step": 499
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.07851401571740697,
+      "learning_rate": 0.0001432376598664168,
+      "loss": 0.2212,
+      "step": 500
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.08216834249790027,
+      "learning_rate": 0.00014302645719736212,
+      "loss": 0.2382,
+      "step": 501
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.07841885852427691,
+      "learning_rate": 0.00014281501873574448,
+      "loss": 0.2356,
+      "step": 502
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.0841592918877469,
+      "learning_rate": 0.00014260334564028383,
+      "loss": 0.2441,
+      "step": 503
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.09045113109493719,
+      "learning_rate": 0.0001423914390709861,
+      "loss": 0.2403,
+      "step": 504
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.0936179224109749,
+      "learning_rate": 0.00014217930018913652,
+      "loss": 0.2473,
+      "step": 505
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.08428504815806069,
+      "learning_rate": 0.0001419669301572936,
+      "loss": 0.2416,
+      "step": 506
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.08904900134838277,
+      "learning_rate": 0.00014175433013928243,
+      "loss": 0.2463,
+      "step": 507
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.08751662205017641,
+      "learning_rate": 0.00014154150130018866,
+      "loss": 0.2494,
+      "step": 508
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.08731360432620024,
+      "learning_rate": 0.0001413284448063517,
+      "loss": 0.251,
+      "step": 509
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.07667993788403807,
+      "learning_rate": 0.00014111516182535878,
+      "loss": 0.223,
+      "step": 510
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.08451659652883611,
+      "learning_rate": 0.00014090165352603805,
+      "loss": 0.2306,
+      "step": 511
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.08616458356604456,
+      "learning_rate": 0.00014068792107845273,
+      "loss": 0.239,
+      "step": 512
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.09518185557601673,
+      "learning_rate": 0.0001404739656538942,
+      "loss": 0.2609,
+      "step": 513
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.08090276728656375,
+      "learning_rate": 0.00014025978842487584,
+      "loss": 0.2217,
+      "step": 514
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.08084800232507564,
+      "learning_rate": 0.00014004539056512667,
+      "loss": 0.2333,
+      "step": 515
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.08756040155151476,
+      "learning_rate": 0.00013983077324958464,
+      "loss": 0.2302,
+      "step": 516
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.09543371719833908,
+      "learning_rate": 0.0001396159376543905,
+      "loss": 0.2528,
+      "step": 517
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.08959960417786461,
+      "learning_rate": 0.0001394008849568811,
+      "loss": 0.2521,
+      "step": 518
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.10764276162567371,
+      "learning_rate": 0.0001391856163355831,
+      "loss": 0.2489,
+      "step": 519
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.10606558025475264,
+      "learning_rate": 0.00013897013297020652,
+      "loss": 0.2439,
+      "step": 520
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.08695085154095022,
+      "learning_rate": 0.00013875443604163807,
+      "loss": 0.2551,
+      "step": 521
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.08218995421066876,
+      "learning_rate": 0.00013853852673193496,
+      "loss": 0.2412,
+      "step": 522
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.08108714040012284,
+      "learning_rate": 0.00013832240622431815,
+      "loss": 0.2401,
+      "step": 523
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.0862280447847096,
+      "learning_rate": 0.0001381060757031662,
+      "loss": 0.2278,
+      "step": 524
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.08545920177454297,
+      "learning_rate": 0.00013788953635400841,
+      "loss": 0.2462,
+      "step": 525
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.08480896731629045,
+      "learning_rate": 0.00013767278936351854,
+      "loss": 0.2437,
+      "step": 526
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.08946464726730828,
+      "learning_rate": 0.00013745583591950823,
+      "loss": 0.2506,
+      "step": 527
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.08535637852853982,
+      "learning_rate": 0.00013723867721092064,
+      "loss": 0.2245,
+      "step": 528
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.08098106602696958,
+      "learning_rate": 0.0001370213144278237,
+      "loss": 0.2318,
+      "step": 529
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.08830854607884242,
+      "learning_rate": 0.00013680374876140378,
+      "loss": 0.2372,
+      "step": 530
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.09604268267784377,
+      "learning_rate": 0.000136585981403959,
+      "loss": 0.2502,
+      "step": 531
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.08494935559660675,
+      "learning_rate": 0.00013636801354889285,
+      "loss": 0.25,
+      "step": 532
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.08845186646743208,
+      "learning_rate": 0.00013614984639070753,
+      "loss": 0.2404,
+      "step": 533
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.08297034988998019,
+      "learning_rate": 0.0001359314811249975,
+      "loss": 0.2297,
+      "step": 534
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.07659983843585298,
+      "learning_rate": 0.00013571291894844297,
+      "loss": 0.2277,
+      "step": 535
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.08687764915879914,
+      "learning_rate": 0.00013549416105880301,
+      "loss": 0.2324,
+      "step": 536
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.08756819498594,
+      "learning_rate": 0.0001352752086549095,
+      "loss": 0.2438,
+      "step": 537
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.08454872300675646,
+      "learning_rate": 0.00013505606293666008,
+      "loss": 0.2346,
+      "step": 538
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.09043402712895539,
+      "learning_rate": 0.00013483672510501194,
+      "loss": 0.2396,
+      "step": 539
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.08780366781470661,
+      "learning_rate": 0.00013461719636197504,
+      "loss": 0.2645,
+      "step": 540
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.09389126585108279,
+      "learning_rate": 0.0001343974779106055,
+      "loss": 0.225,
+      "step": 541
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.08393753486573638,
+      "learning_rate": 0.00013417757095499915,
+      "loss": 0.228,
+      "step": 542
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.09581975739229241,
+      "learning_rate": 0.00013395747670028486,
+      "loss": 0.2568,
+      "step": 543
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.09238917931680911,
+      "learning_rate": 0.0001337371963526179,
+      "loss": 0.2592,
+      "step": 544
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.08934986117751913,
+      "learning_rate": 0.00013351673111917331,
+      "loss": 0.2362,
+      "step": 545
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.09409634866676543,
+      "learning_rate": 0.00013329608220813947,
+      "loss": 0.2304,
+      "step": 546
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.09439594046742317,
+      "learning_rate": 0.00013307525082871123,
+      "loss": 0.253,
+      "step": 547
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.09528839009632512,
+      "learning_rate": 0.0001328542381910835,
+      "loss": 0.2422,
+      "step": 548
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.09226237121742821,
+      "learning_rate": 0.00013263304550644435,
+      "loss": 0.2606,
+      "step": 549
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.08801443598238876,
+      "learning_rate": 0.00013241167398696875,
+      "loss": 0.2271,
+      "step": 550
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.09453520668040735,
+      "learning_rate": 0.00013219012484581155,
+      "loss": 0.2454,
+      "step": 551
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.09267134821653143,
+      "learning_rate": 0.00013196839929710113,
+      "loss": 0.2605,
+      "step": 552
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.07958994028752563,
+      "learning_rate": 0.00013174649855593243,
+      "loss": 0.2404,
+      "step": 553
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.08379968306893623,
+      "learning_rate": 0.0001315244238383607,
+      "loss": 0.2561,
+      "step": 554
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.08291630188593745,
+      "learning_rate": 0.00013130217636139445,
+      "loss": 0.2266,
+      "step": 555
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.08556890552373712,
+      "learning_rate": 0.00013107975734298903,
+      "loss": 0.2298,
+      "step": 556
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.07965624853282312,
+      "learning_rate": 0.0001308571680020398,
+      "loss": 0.227,
+      "step": 557
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.08517939609536171,
+      "learning_rate": 0.00013063440955837555,
+      "loss": 0.2569,
+      "step": 558
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.09051963295961941,
+      "learning_rate": 0.0001304114832327518,
+      "loss": 0.2435,
+      "step": 559
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.09282119926951815,
+      "learning_rate": 0.00013018839024684406,
+      "loss": 0.2501,
+      "step": 560
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.0854068841968234,
+      "learning_rate": 0.0001299651318232412,
+      "loss": 0.2493,
+      "step": 561
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.0829882458457593,
+      "learning_rate": 0.00012974170918543868,
+      "loss": 0.2355,
+      "step": 562
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.08451556620822356,
+      "learning_rate": 0.0001295181235578319,
+      "loss": 0.2385,
+      "step": 563
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.08348298945964819,
+      "learning_rate": 0.0001292943761657095,
+      "loss": 0.2391,
+      "step": 564
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.07476546302466949,
+      "learning_rate": 0.00012907046823524656,
+      "loss": 0.2271,
+      "step": 565
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.09116314712203458,
+      "learning_rate": 0.00012884640099349802,
+      "loss": 0.2161,
+      "step": 566
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.08636886263361133,
+      "learning_rate": 0.0001286221756683918,
+      "loss": 0.2254,
+      "step": 567
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.08993659715915059,
+      "learning_rate": 0.0001283977934887222,
+      "loss": 0.2227,
+      "step": 568
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.08648265432870464,
+      "learning_rate": 0.00012817325568414297,
+      "loss": 0.2378,
+      "step": 569
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.0799851109947209,
+      "learning_rate": 0.00012794856348516095,
+      "loss": 0.226,
+      "step": 570
+    },
+    {
+      "epoch": 1.24,
+      "eval_loss": 0.25866472721099854,
+      "eval_runtime": 46.967,
+      "eval_samples_per_second": 49.46,
+      "eval_steps_per_second": 0.788,
+      "step": 570
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.09116565000730968,
+      "learning_rate": 0.0001277237181231289,
+      "loss": 0.2421,
+      "step": 571
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.08765537514394504,
+      "learning_rate": 0.00012749872083023906,
+      "loss": 0.2319,
+      "step": 572
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.09024073652656935,
+      "learning_rate": 0.00012727357283951615,
+      "loss": 0.2297,
+      "step": 573
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.08776531293079054,
+      "learning_rate": 0.00012704827538481084,
+      "loss": 0.2473,
+      "step": 574
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.0826875116798064,
+      "learning_rate": 0.00012682282970079284,
+      "loss": 0.2239,
+      "step": 575
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.08900534715923833,
+      "learning_rate": 0.00012659723702294423,
+      "loss": 0.235,
+      "step": 576
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.07969397770975745,
+      "learning_rate": 0.0001263714985875526,
+      "loss": 0.2272,
+      "step": 577
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.08487532950454624,
+      "learning_rate": 0.00012614561563170437,
+      "loss": 0.2327,
+      "step": 578
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.07768200207480558,
+      "learning_rate": 0.00012591958939327786,
+      "loss": 0.2353,
+      "step": 579
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.0737846156739731,
+      "learning_rate": 0.0001256934211109367,
+      "loss": 0.2332,
+      "step": 580
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.07688468804216825,
+      "learning_rate": 0.00012546711202412287,
+      "loss": 0.2413,
+      "step": 581
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.07931162469142657,
+      "learning_rate": 0.00012524066337305008,
+      "loss": 0.2335,
+      "step": 582
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.07624335175497067,
+      "learning_rate": 0.0001250140763986968,
+      "loss": 0.2348,
+      "step": 583
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.091328537134851,
+      "learning_rate": 0.0001247873523427995,
+      "loss": 0.2297,
+      "step": 584
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.0811859125697209,
+      "learning_rate": 0.000124560492447846,
+      "loss": 0.2294,
+      "step": 585
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.0841047732515047,
+      "learning_rate": 0.00012433349795706845,
+      "loss": 0.2324,
+      "step": 586
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.08386654683451615,
+      "learning_rate": 0.00012410637011443665,
+      "loss": 0.2409,
+      "step": 587
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.08200748985054801,
+      "learning_rate": 0.00012387911016465116,
+      "loss": 0.2222,
+      "step": 588
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.08614235541835254,
+      "learning_rate": 0.00012365171935313653,
+      "loss": 0.2372,
+      "step": 589
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.07991734215438401,
+      "learning_rate": 0.00012342419892603446,
+      "loss": 0.2184,
+      "step": 590
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.08685812596724986,
+      "learning_rate": 0.000123196550130197,
+      "loss": 0.2398,
+      "step": 591
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.08660466867381622,
+      "learning_rate": 0.0001229687742131796,
+      "loss": 0.246,
+      "step": 592
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.08552802130689283,
+      "learning_rate": 0.00012274087242323433,
+      "loss": 0.2289,
+      "step": 593
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.08590776026426133,
+      "learning_rate": 0.00012251284600930327,
+      "loss": 0.2438,
+      "step": 594
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.08181855750061642,
+      "learning_rate": 0.00012228469622101127,
+      "loss": 0.2377,
+      "step": 595
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.08271221654383495,
+      "learning_rate": 0.0001220564243086593,
+      "loss": 0.2303,
+      "step": 596
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.08511455941486844,
+      "learning_rate": 0.00012182803152321774,
+      "loss": 0.2369,
+      "step": 597
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.08505574778931962,
+      "learning_rate": 0.0001215995191163192,
+      "loss": 0.2586,
+      "step": 598
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.08330562125810212,
+      "learning_rate": 0.00012137088834025197,
+      "loss": 0.2249,
+      "step": 599
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.08647535001759606,
+      "learning_rate": 0.00012114214044795287,
+      "loss": 0.2366,
+      "step": 600
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.08567378372092092,
+      "learning_rate": 0.00012091327669300075,
+      "loss": 0.2373,
+      "step": 601
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.07581176364703038,
+      "learning_rate": 0.00012068429832960924,
+      "loss": 0.2221,
+      "step": 602
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.0864257408400619,
+      "learning_rate": 0.0001204552066126201,
+      "loss": 0.239,
+      "step": 603
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.09417645496720935,
+      "learning_rate": 0.00012022600279749627,
+      "loss": 0.2467,
+      "step": 604
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.07650578500185509,
+      "learning_rate": 0.00011999668814031506,
+      "loss": 0.2076,
+      "step": 605
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.08883258649010815,
+      "learning_rate": 0.00011976726389776114,
+      "loss": 0.2435,
+      "step": 606
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.08188907473195263,
+      "learning_rate": 0.00011953773132711981,
+      "loss": 0.2328,
+      "step": 607
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.08134121978135175,
+      "learning_rate": 0.00011930809168626994,
+      "loss": 0.2339,
+      "step": 608
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.07365163162927094,
+      "learning_rate": 0.0001190783462336772,
+      "loss": 0.2187,
+      "step": 609
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.08252374367430065,
+      "learning_rate": 0.00011884849622838717,
+      "loss": 0.2318,
+      "step": 610
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.08454334190175966,
+      "learning_rate": 0.00011861854293001833,
+      "loss": 0.2555,
+      "step": 611
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.08675671113872546,
+      "learning_rate": 0.00011838848759875526,
+      "loss": 0.2494,
+      "step": 612
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.08276522474604452,
+      "learning_rate": 0.00011815833149534165,
+      "loss": 0.2424,
+      "step": 613
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.08775138952406279,
+      "learning_rate": 0.00011792807588107357,
+      "loss": 0.2459,
+      "step": 614
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.08362653927034998,
+      "learning_rate": 0.00011769772201779228,
+      "loss": 0.2405,
+      "step": 615
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.08406348993204213,
+      "learning_rate": 0.00011746727116787752,
+      "loss": 0.2253,
+      "step": 616
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.08860230531541739,
+      "learning_rate": 0.00011723672459424055,
+      "loss": 0.2335,
+      "step": 617
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.08709792716237423,
+      "learning_rate": 0.00011700608356031716,
+      "loss": 0.2417,
+      "step": 618
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.07943643810600176,
+      "learning_rate": 0.00011677534933006091,
+      "loss": 0.2317,
+      "step": 619
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.08692247639337791,
+      "learning_rate": 0.00011654452316793592,
+      "loss": 0.2382,
+      "step": 620
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.0819181275348235,
+      "learning_rate": 0.00011631360633891028,
+      "loss": 0.2361,
+      "step": 621
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.08918709864923731,
+      "learning_rate": 0.00011608260010844885,
+      "loss": 0.2557,
+      "step": 622
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.0912318289477136,
+      "learning_rate": 0.00011585150574250647,
+      "loss": 0.2536,
+      "step": 623
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.07763619992544649,
+      "learning_rate": 0.00011562032450752093,
+      "loss": 0.2171,
+      "step": 624
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.08355913793741429,
+      "learning_rate": 0.0001153890576704062,
+      "loss": 0.2755,
+      "step": 625
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.08169631473159601,
+      "learning_rate": 0.00011515770649854518,
+      "loss": 0.246,
+      "step": 626
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.08589495613655412,
+      "learning_rate": 0.00011492627225978307,
+      "loss": 0.2238,
+      "step": 627
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.080984656977802,
+      "learning_rate": 0.00011469475622242028,
+      "loss": 0.2213,
+      "step": 628
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.08790946069852021,
+      "learning_rate": 0.00011446315965520544,
+      "loss": 0.2487,
+      "step": 629
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.08447595821387797,
+      "learning_rate": 0.00011423148382732853,
+      "loss": 0.2297,
+      "step": 630
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.08062468357264355,
+      "learning_rate": 0.00011399973000841386,
+      "loss": 0.2439,
+      "step": 631
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.0815294677156492,
+      "learning_rate": 0.00011376789946851327,
+      "loss": 0.2318,
+      "step": 632
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.08538149738651506,
+      "learning_rate": 0.00011353599347809885,
+      "loss": 0.2207,
+      "step": 633
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.0914017207498894,
+      "learning_rate": 0.00011330401330805632,
+      "loss": 0.2497,
+      "step": 634
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.08416335861370626,
+      "learning_rate": 0.00011307196022967778,
+      "loss": 0.2416,
+      "step": 635
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.08270490016866311,
+      "learning_rate": 0.00011283983551465511,
+      "loss": 0.2178,
+      "step": 636
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.08030280135368922,
+      "learning_rate": 0.0001126076404350725,
+      "loss": 0.2092,
+      "step": 637
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.08479790007191501,
+      "learning_rate": 0.00011237537626339997,
+      "loss": 0.235,
+      "step": 638
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.0852885506885306,
+      "learning_rate": 0.00011214304427248605,
+      "loss": 0.2598,
+      "step": 639
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.08583843687898746,
+      "learning_rate": 0.00011191064573555095,
+      "loss": 0.2405,
+      "step": 640
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.0806904531401263,
+      "learning_rate": 0.0001116781819261796,
+      "loss": 0.2254,
+      "step": 641
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.07653654665821262,
+      "learning_rate": 0.0001114456541183146,
+      "loss": 0.2212,
+      "step": 642
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.07557026468967774,
+      "learning_rate": 0.0001112130635862493,
+      "loss": 0.2136,
+      "step": 643
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.08889051801843051,
+      "learning_rate": 0.00011098041160462076,
+      "loss": 0.2581,
+      "step": 644
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.07483166539341271,
+      "learning_rate": 0.00011074769944840286,
+      "loss": 0.2251,
+      "step": 645
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.08480451148958008,
+      "learning_rate": 0.00011051492839289916,
+      "loss": 0.2228,
+      "step": 646
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.08154257471594718,
+      "learning_rate": 0.00011028209971373605,
+      "loss": 0.2588,
+      "step": 647
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.07395490692433179,
+      "learning_rate": 0.00011004921468685567,
+      "loss": 0.2334,
+      "step": 648
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.08905339382700107,
+      "learning_rate": 0.00010981627458850906,
+      "loss": 0.2454,
+      "step": 649
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.07490913007249862,
+      "learning_rate": 0.00010958328069524888,
+      "loss": 0.2054,
+      "step": 650
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.07774909601326149,
+      "learning_rate": 0.00010935023428392275,
+      "loss": 0.2202,
+      "step": 651
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.08656286155198262,
+      "learning_rate": 0.00010911713663166607,
+      "loss": 0.2315,
+      "step": 652
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.08095967227507062,
+      "learning_rate": 0.00010888398901589499,
+      "loss": 0.2303,
+      "step": 653
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.0901555734919747,
+      "learning_rate": 0.0001086507927142995,
+      "loss": 0.2537,
+      "step": 654
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.08565947957876019,
+      "learning_rate": 0.0001084175490048364,
+      "loss": 0.2299,
+      "step": 655
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.0817439947786202,
+      "learning_rate": 0.00010818425916572235,
+      "loss": 0.2529,
+      "step": 656
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.08227803653345145,
+      "learning_rate": 0.0001079509244754267,
+      "loss": 0.2294,
+      "step": 657
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.08086315758929115,
+      "learning_rate": 0.00010771754621266466,
+      "loss": 0.2262,
+      "step": 658
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.08031174843061695,
+      "learning_rate": 0.00010748412565639019,
+      "loss": 0.2132,
+      "step": 659
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.08239523990604684,
+      "learning_rate": 0.0001072506640857891,
+      "loss": 0.2304,
+      "step": 660
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.0794827671233348,
+      "learning_rate": 0.00010701716278027187,
+      "loss": 0.2334,
+      "step": 661
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.07901196055621171,
+      "learning_rate": 0.00010678362301946676,
+      "loss": 0.2367,
+      "step": 662
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.0731368547238744,
+      "learning_rate": 0.00010655004608321286,
+      "loss": 0.2098,
+      "step": 663
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.08211250198046752,
+      "learning_rate": 0.00010631643325155286,
+      "loss": 0.2271,
+      "step": 664
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.08787661016835348,
+      "learning_rate": 0.00010608278580472624,
+      "loss": 0.2379,
+      "step": 665
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.08141439832280062,
+      "learning_rate": 0.00010584910502316209,
+      "loss": 0.2232,
+      "step": 666
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.08869600051845548,
+      "learning_rate": 0.00010561539218747238,
+      "loss": 0.2523,
+      "step": 667
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.08140563535824365,
+      "learning_rate": 0.0001053816485784445,
+      "loss": 0.223,
+      "step": 668
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.0839156518362482,
+      "learning_rate": 0.00010514787547703466,
+      "loss": 0.2524,
+      "step": 669
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.08232115709306517,
+      "learning_rate": 0.00010491407416436058,
+      "loss": 0.2464,
+      "step": 670
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.07773730821203234,
+      "learning_rate": 0.00010468024592169465,
+      "loss": 0.2242,
+      "step": 671
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.0740616352598192,
+      "learning_rate": 0.00010444639203045683,
+      "loss": 0.2211,
+      "step": 672
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.07851257745322647,
+      "learning_rate": 0.00010421251377220758,
+      "loss": 0.2307,
+      "step": 673
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.08711361746442828,
+      "learning_rate": 0.00010397861242864098,
+      "loss": 0.2403,
+      "step": 674
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.08379272283993074,
+      "learning_rate": 0.00010374468928157759,
+      "loss": 0.2342,
+      "step": 675
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.08280896323227015,
+      "learning_rate": 0.00010351074561295749,
+      "loss": 0.2463,
+      "step": 676
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.08646477880878306,
+      "learning_rate": 0.00010327678270483308,
+      "loss": 0.2666,
+      "step": 677
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.07982998000901836,
+      "learning_rate": 0.00010304280183936238,
+      "loss": 0.2244,
+      "step": 678
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.07958900123571516,
+      "learning_rate": 0.00010280880429880174,
+      "loss": 0.2294,
+      "step": 679
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.08128634634800733,
+      "learning_rate": 0.00010257479136549889,
+      "loss": 0.2166,
+      "step": 680
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.08045012723831614,
+      "learning_rate": 0.00010234076432188584,
+      "loss": 0.2299,
+      "step": 681
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.08288664950743586,
+      "learning_rate": 0.00010210672445047212,
+      "loss": 0.2315,
+      "step": 682
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.08754211842600262,
+      "learning_rate": 0.00010187267303383738,
+      "loss": 0.2359,
+      "step": 683
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.08196614440673552,
+      "learning_rate": 0.0001016386113546246,
+      "loss": 0.2206,
+      "step": 684
+    },
+    {
+      "epoch": 1.5,
+      "eval_loss": 0.24948738515377045,
+      "eval_runtime": 46.9159,
+      "eval_samples_per_second": 49.514,
+      "eval_steps_per_second": 0.789,
+      "step": 684
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.08394731213907013,
+      "learning_rate": 0.00010140454069553302,
+      "loss": 0.2351,
+      "step": 685
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.07826908652856117,
+      "learning_rate": 0.00010117046233931105,
+      "loss": 0.2317,
+      "step": 686
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.08353319236243155,
+      "learning_rate": 0.00010093637756874936,
+      "loss": 0.2343,
+      "step": 687
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.08501699943812632,
+      "learning_rate": 0.00010070228766667367,
+      "loss": 0.2236,
+      "step": 688
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.0812418081894846,
+      "learning_rate": 0.00010046819391593789,
+      "loss": 0.2336,
+      "step": 689
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.0752079476231479,
+      "learning_rate": 0.000100234097599417,
+      "loss": 0.22,
+      "step": 690
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.08795078126255108,
+      "learning_rate": 0.0001,
+      "loss": 0.2515,
+      "step": 691
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.08322978353788983,
+      "learning_rate": 9.976590240058303e-05,
+      "loss": 0.2194,
+      "step": 692
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.08900182083261693,
+      "learning_rate": 9.953180608406215e-05,
+      "loss": 0.2548,
+      "step": 693
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.08305235677320687,
+      "learning_rate": 9.929771233332635e-05,
+      "loss": 0.2314,
+      "step": 694
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.07577584090051011,
+      "learning_rate": 9.906362243125063e-05,
+      "loss": 0.2325,
+      "step": 695
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.07700019186130685,
+      "learning_rate": 9.882953766068897e-05,
+      "loss": 0.2393,
+      "step": 696
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.0789659155330154,
+      "learning_rate": 9.859545930446699e-05,
+      "loss": 0.2145,
+      "step": 697
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.08049985275269897,
+      "learning_rate": 9.836138864537545e-05,
+      "loss": 0.2494,
+      "step": 698
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.08280562291290458,
+      "learning_rate": 9.812732696616265e-05,
+      "loss": 0.2418,
+      "step": 699
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.07971495758368538,
+      "learning_rate": 9.789327554952789e-05,
+      "loss": 0.2241,
+      "step": 700
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.08677316614381443,
+      "learning_rate": 9.765923567811417e-05,
+      "loss": 0.2374,
+      "step": 701
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.07865796685040652,
+      "learning_rate": 9.742520863450115e-05,
+      "loss": 0.2193,
+      "step": 702
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.08435355317876911,
+      "learning_rate": 9.719119570119827e-05,
+      "loss": 0.2444,
+      "step": 703
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.08020919157111929,
+      "learning_rate": 9.695719816063764e-05,
+      "loss": 0.2379,
+      "step": 704
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.08084679565620428,
+      "learning_rate": 9.672321729516693e-05,
+      "loss": 0.2174,
+      "step": 705
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.08193105584579982,
+      "learning_rate": 9.648925438704258e-05,
+      "loss": 0.2478,
+      "step": 706
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.0787510020943492,
+      "learning_rate": 9.625531071842243e-05,
+      "loss": 0.2359,
+      "step": 707
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.07496104346263367,
+      "learning_rate": 9.6021387571359e-05,
+      "loss": 0.2215,
+      "step": 708
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.08303409609082166,
+      "learning_rate": 9.578748622779247e-05,
+      "loss": 0.2422,
+      "step": 709
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.08159988381506675,
+      "learning_rate": 9.55536079695432e-05,
+      "loss": 0.2361,
+      "step": 710
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.08541600262957548,
+      "learning_rate": 9.531975407830536e-05,
+      "loss": 0.2515,
+      "step": 711
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.07744197592700136,
+      "learning_rate": 9.508592583563944e-05,
+      "loss": 0.2071,
+      "step": 712
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.07427943409042963,
+      "learning_rate": 9.485212452296535e-05,
+      "loss": 0.222,
+      "step": 713
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.08901403158160176,
+      "learning_rate": 9.46183514215555e-05,
+      "loss": 0.241,
+      "step": 714
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.08202698389949295,
+      "learning_rate": 9.438460781252764e-05,
+      "loss": 0.2255,
+      "step": 715
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.07968389769434038,
+      "learning_rate": 9.41508949768379e-05,
+      "loss": 0.2251,
+      "step": 716
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.07779985957139146,
+      "learning_rate": 9.391721419527381e-05,
+      "loss": 0.2126,
+      "step": 717
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.08426830347137788,
+      "learning_rate": 9.368356674844717e-05,
+      "loss": 0.2226,
+      "step": 718
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.0826010808958624,
+      "learning_rate": 9.344995391678713e-05,
+      "loss": 0.2101,
+      "step": 719
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.08896387520523545,
+      "learning_rate": 9.321637698053327e-05,
+      "loss": 0.2555,
+      "step": 720
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.07641322462938643,
+      "learning_rate": 9.298283721972814e-05,
+      "loss": 0.2242,
+      "step": 721
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.09066531069946981,
+      "learning_rate": 9.274933591421096e-05,
+      "loss": 0.2282,
+      "step": 722
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.08143523803443777,
+      "learning_rate": 9.251587434360982e-05,
+      "loss": 0.2348,
+      "step": 723
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.09407900162519509,
+      "learning_rate": 9.228245378733537e-05,
+      "loss": 0.2409,
+      "step": 724
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.08217621097523965,
+      "learning_rate": 9.204907552457333e-05,
+      "loss": 0.2393,
+      "step": 725
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.07869827225405131,
+      "learning_rate": 9.181574083427767e-05,
+      "loss": 0.233,
+      "step": 726
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.08072559062923186,
+      "learning_rate": 9.158245099516361e-05,
+      "loss": 0.2315,
+      "step": 727
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.08016098782738346,
+      "learning_rate": 9.134920728570054e-05,
+      "loss": 0.2199,
+      "step": 728
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.0758869883628278,
+      "learning_rate": 9.111601098410505e-05,
+      "loss": 0.2287,
+      "step": 729
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.08299918971434583,
+      "learning_rate": 9.088286336833395e-05,
+      "loss": 0.1931,
+      "step": 730
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.08531328014918346,
+      "learning_rate": 9.064976571607727e-05,
+      "loss": 0.2437,
+      "step": 731
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.08304079186022913,
+      "learning_rate": 9.041671930475113e-05,
+      "loss": 0.2254,
+      "step": 732
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.08140290388892739,
+      "learning_rate": 9.0183725411491e-05,
+      "loss": 0.2258,
+      "step": 733
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.0916146336158295,
+      "learning_rate": 8.995078531314434e-05,
+      "loss": 0.2363,
+      "step": 734
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.0827281532478695,
+      "learning_rate": 8.971790028626395e-05,
+      "loss": 0.2473,
+      "step": 735
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.08509302559196476,
+      "learning_rate": 8.948507160710085e-05,
+      "loss": 0.218,
+      "step": 736
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.08447867576011206,
+      "learning_rate": 8.925230055159715e-05,
+      "loss": 0.2259,
+      "step": 737
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.08383971330619845,
+      "learning_rate": 8.901958839537924e-05,
+      "loss": 0.2332,
+      "step": 738
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.08333135828518855,
+      "learning_rate": 8.878693641375073e-05,
+      "loss": 0.236,
+      "step": 739
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.07782818917587933,
+      "learning_rate": 8.855434588168543e-05,
+      "loss": 0.2172,
+      "step": 740
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.07890578756102465,
+      "learning_rate": 8.832181807382045e-05,
+      "loss": 0.2325,
+      "step": 741
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.08477364372525044,
+      "learning_rate": 8.808935426444909e-05,
+      "loss": 0.2451,
+      "step": 742
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.08118728913565396,
+      "learning_rate": 8.785695572751398e-05,
+      "loss": 0.2192,
+      "step": 743
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.09296381139861269,
+      "learning_rate": 8.762462373660005e-05,
+      "loss": 0.2715,
+      "step": 744
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.08769595256393421,
+      "learning_rate": 8.739235956492751e-05,
+      "loss": 0.2421,
+      "step": 745
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.08650400641050761,
+      "learning_rate": 8.71601644853449e-05,
+      "loss": 0.2456,
+      "step": 746
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.08006497600807996,
+      "learning_rate": 8.692803977032224e-05,
+      "loss": 0.2406,
+      "step": 747
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.08014780013960171,
+      "learning_rate": 8.669598669194372e-05,
+      "loss": 0.224,
+      "step": 748
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.07870358262355977,
+      "learning_rate": 8.646400652190121e-05,
+      "loss": 0.2281,
+      "step": 749
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.08152931635889546,
+      "learning_rate": 8.623210053148676e-05,
+      "loss": 0.2324,
+      "step": 750
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.08092137369389883,
+      "learning_rate": 8.600026999158613e-05,
+      "loss": 0.2232,
+      "step": 751
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.08052613089751662,
+      "learning_rate": 8.57685161726715e-05,
+      "loss": 0.2184,
+      "step": 752
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.08127484239836073,
+      "learning_rate": 8.55368403447946e-05,
+      "loss": 0.2319,
+      "step": 753
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.0811016308762424,
+      "learning_rate": 8.530524377757973e-05,
+      "loss": 0.2142,
+      "step": 754
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.08219419485644304,
+      "learning_rate": 8.507372774021696e-05,
+      "loss": 0.2233,
+      "step": 755
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.08371413077059317,
+      "learning_rate": 8.484229350145486e-05,
+      "loss": 0.2385,
+      "step": 756
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.08467643336890418,
+      "learning_rate": 8.461094232959381e-05,
+      "loss": 0.2453,
+      "step": 757
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.08177231327416434,
+      "learning_rate": 8.437967549247908e-05,
+      "loss": 0.2367,
+      "step": 758
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.07503070401489957,
+      "learning_rate": 8.414849425749354e-05,
+      "loss": 0.2173,
+      "step": 759
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.08131567748910913,
+      "learning_rate": 8.391739989155119e-05,
+      "loss": 0.2311,
+      "step": 760
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.08550768078479756,
+      "learning_rate": 8.368639366108974e-05,
+      "loss": 0.2588,
+      "step": 761
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.07310738874661277,
+      "learning_rate": 8.345547683206409e-05,
+      "loss": 0.2243,
+      "step": 762
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.08046253907205837,
+      "learning_rate": 8.322465066993914e-05,
+      "loss": 0.2227,
+      "step": 763
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.08440283552390136,
+      "learning_rate": 8.299391643968284e-05,
+      "loss": 0.2458,
+      "step": 764
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.08238206507640838,
+      "learning_rate": 8.276327540575947e-05,
+      "loss": 0.2486,
+      "step": 765
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.07702335043438999,
+      "learning_rate": 8.25327288321225e-05,
+      "loss": 0.2202,
+      "step": 766
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.07733785939181663,
+      "learning_rate": 8.230227798220774e-05,
+      "loss": 0.2458,
+      "step": 767
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.07437992163810951,
+      "learning_rate": 8.207192411892646e-05,
+      "loss": 0.2064,
+      "step": 768
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.08021366464971075,
+      "learning_rate": 8.184166850465836e-05,
+      "loss": 0.2219,
+      "step": 769
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.08438503814960518,
+      "learning_rate": 8.161151240124475e-05,
+      "loss": 0.2296,
+      "step": 770
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.0880588352352524,
+      "learning_rate": 8.138145706998173e-05,
+      "loss": 0.2398,
+      "step": 771
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.07867719587290391,
+      "learning_rate": 8.115150377161284e-05,
+      "loss": 0.2272,
+      "step": 772
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.08467665609034607,
+      "learning_rate": 8.09216537663228e-05,
+      "loss": 0.2333,
+      "step": 773
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.07316546236135894,
+      "learning_rate": 8.069190831373009e-05,
+      "loss": 0.1965,
+      "step": 774
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.07681455068240883,
+      "learning_rate": 8.04622686728802e-05,
+      "loss": 0.201,
+      "step": 775
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.10293167879614176,
+      "learning_rate": 8.023273610223887e-05,
+      "loss": 0.2308,
+      "step": 776
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.09321308865333812,
+      "learning_rate": 8.000331185968495e-05,
+      "loss": 0.25,
+      "step": 777
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.08517927035817639,
+      "learning_rate": 7.977399720250374e-05,
+      "loss": 0.2235,
+      "step": 778
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.08390235499801926,
+      "learning_rate": 7.954479338737995e-05,
+      "loss": 0.2327,
+      "step": 779
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.08749398163976607,
+      "learning_rate": 7.93157016703908e-05,
+      "loss": 0.2449,
+      "step": 780
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.08598649189038254,
+      "learning_rate": 7.908672330699925e-05,
+      "loss": 0.2352,
+      "step": 781
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.07825838563607705,
+      "learning_rate": 7.885785955204715e-05,
+      "loss": 0.2298,
+      "step": 782
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.07464840090504722,
+      "learning_rate": 7.862911165974807e-05,
+      "loss": 0.2183,
+      "step": 783
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.08214255853368775,
+      "learning_rate": 7.840048088368084e-05,
+      "loss": 0.238,
+      "step": 784
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.08040849063603861,
+      "learning_rate": 7.817196847678227e-05,
+      "loss": 0.2251,
+      "step": 785
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.07187602337855015,
+      "learning_rate": 7.79435756913407e-05,
+      "loss": 0.1903,
+      "step": 786
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.07317720391420812,
+      "learning_rate": 7.771530377898876e-05,
+      "loss": 0.207,
+      "step": 787
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.08666941682910177,
+      "learning_rate": 7.748715399069675e-05,
+      "loss": 0.2337,
+      "step": 788
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.08026877828288766,
+      "learning_rate": 7.725912757676567e-05,
+      "loss": 0.2382,
+      "step": 789
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.07535426876553636,
+      "learning_rate": 7.703122578682046e-05,
+      "loss": 0.2096,
+      "step": 790
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.08058052517258636,
+      "learning_rate": 7.680344986980303e-05,
+      "loss": 0.2205,
+      "step": 791
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.08506969637844296,
+      "learning_rate": 7.657580107396551e-05,
+      "loss": 0.226,
+      "step": 792
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.09473181739467829,
+      "learning_rate": 7.634828064686349e-05,
+      "loss": 0.2502,
+      "step": 793
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.0809871754359907,
+      "learning_rate": 7.612088983534885e-05,
+      "loss": 0.2344,
+      "step": 794
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.08617304632111902,
+      "learning_rate": 7.589362988556339e-05,
+      "loss": 0.2266,
+      "step": 795
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.09018857235214806,
+      "learning_rate": 7.566650204293156e-05,
+      "loss": 0.2458,
+      "step": 796
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.07932249169863058,
+      "learning_rate": 7.543950755215401e-05,
+      "loss": 0.2193,
+      "step": 797
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.07668196616962725,
+      "learning_rate": 7.521264765720052e-05,
+      "loss": 0.2181,
+      "step": 798
+    },
+    {
+      "epoch": 1.75,
+      "eval_loss": 0.24336878955364227,
+      "eval_runtime": 46.9117,
+      "eval_samples_per_second": 49.519,
+      "eval_steps_per_second": 0.789,
+      "step": 798
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.08876682210214744,
+      "learning_rate": 7.498592360130323e-05,
+      "loss": 0.2324,
+      "step": 799
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.0799245990874372,
+      "learning_rate": 7.475933662694993e-05,
+      "loss": 0.2337,
+      "step": 800
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.07568756831407811,
+      "learning_rate": 7.453288797587714e-05,
+      "loss": 0.2149,
+      "step": 801
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.06935218835505103,
+      "learning_rate": 7.430657888906333e-05,
+      "loss": 0.2084,
+      "step": 802
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.07951558047732071,
+      "learning_rate": 7.408041060672218e-05,
+      "loss": 0.2253,
+      "step": 803
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.07913220659710016,
+      "learning_rate": 7.385438436829566e-05,
+      "loss": 0.219,
+      "step": 804
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.07780795137676555,
+      "learning_rate": 7.362850141244738e-05,
+      "loss": 0.2161,
+      "step": 805
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.07817554031203039,
+      "learning_rate": 7.340276297705579e-05,
+      "loss": 0.2133,
+      "step": 806
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.08916967389217759,
+      "learning_rate": 7.317717029920717e-05,
+      "loss": 0.2407,
+      "step": 807
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.07990946977813441,
+      "learning_rate": 7.295172461518918e-05,
+      "loss": 0.2221,
+      "step": 808
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.07948516913134088,
+      "learning_rate": 7.272642716048387e-05,
+      "loss": 0.2384,
+      "step": 809
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.0857186891337925,
+      "learning_rate": 7.250127916976097e-05,
+      "loss": 0.2234,
+      "step": 810
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.08124087103580349,
+      "learning_rate": 7.227628187687111e-05,
+      "loss": 0.2211,
+      "step": 811
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.08464885262676666,
+      "learning_rate": 7.205143651483906e-05,
+      "loss": 0.2225,
+      "step": 812
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.07509030094016476,
+      "learning_rate": 7.182674431585704e-05,
+      "loss": 0.2155,
+      "step": 813
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.09177394844426678,
+      "learning_rate": 7.160220651127787e-05,
+      "loss": 0.2768,
+      "step": 814
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.07673817509570781,
+      "learning_rate": 7.137782433160823e-05,
+      "loss": 0.2114,
+      "step": 815
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.07844304376270883,
+      "learning_rate": 7.115359900650198e-05,
+      "loss": 0.2295,
+      "step": 816
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.08236902382008632,
+      "learning_rate": 7.092953176475346e-05,
+      "loss": 0.2392,
+      "step": 817
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.07555587821419002,
+      "learning_rate": 7.070562383429052e-05,
+      "loss": 0.2222,
+      "step": 818
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.0781063385371518,
+      "learning_rate": 7.04818764421681e-05,
+      "loss": 0.214,
+      "step": 819
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.07678976461648601,
+      "learning_rate": 7.025829081456137e-05,
+      "loss": 0.2208,
+      "step": 820
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.0728342521494589,
+      "learning_rate": 7.003486817675881e-05,
+      "loss": 0.2266,
+      "step": 821
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.08365682000780762,
+      "learning_rate": 6.981160975315597e-05,
+      "loss": 0.2215,
+      "step": 822
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.0819797581324554,
+      "learning_rate": 6.958851676724823e-05,
+      "loss": 0.2194,
+      "step": 823
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.08468082994678829,
+      "learning_rate": 6.936559044162445e-05,
+      "loss": 0.2412,
+      "step": 824
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.0852273157825867,
+      "learning_rate": 6.914283199796023e-05,
+      "loss": 0.2179,
+      "step": 825
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.08194038206130427,
+      "learning_rate": 6.892024265701099e-05,
+      "loss": 0.2394,
+      "step": 826
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.08238822602690216,
+      "learning_rate": 6.869782363860556e-05,
+      "loss": 0.2254,
+      "step": 827
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.08401318266736248,
+      "learning_rate": 6.847557616163934e-05,
+      "loss": 0.2251,
+      "step": 828
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.08438938369272424,
+      "learning_rate": 6.825350144406758e-05,
+      "loss": 0.2321,
+      "step": 829
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.07933966393596179,
+      "learning_rate": 6.803160070289894e-05,
+      "loss": 0.2377,
+      "step": 830
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.07938136839756817,
+      "learning_rate": 6.780987515418846e-05,
+      "loss": 0.218,
+      "step": 831
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.07361088142825277,
+      "learning_rate": 6.758832601303126e-05,
+      "loss": 0.2145,
+      "step": 832
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.08445570891500642,
+      "learning_rate": 6.736695449355568e-05,
+      "loss": 0.2386,
+      "step": 833
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.07914077116716449,
+      "learning_rate": 6.714576180891654e-05,
+      "loss": 0.2094,
+      "step": 834
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.07467713412271462,
+      "learning_rate": 6.692474917128876e-05,
+      "loss": 0.2055,
+      "step": 835
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.0771698891717675,
+      "learning_rate": 6.670391779186055e-05,
+      "loss": 0.21,
+      "step": 836
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.0852844011141155,
+      "learning_rate": 6.64832688808267e-05,
+      "loss": 0.2209,
+      "step": 837
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.08679725266193321,
+      "learning_rate": 6.626280364738215e-05,
+      "loss": 0.2022,
+      "step": 838
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.08574903749723038,
+      "learning_rate": 6.604252329971517e-05,
+      "loss": 0.241,
+      "step": 839
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.08732116381600609,
+      "learning_rate": 6.582242904500084e-05,
+      "loss": 0.2402,
+      "step": 840
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.08072061230600541,
+      "learning_rate": 6.560252208939453e-05,
+      "loss": 0.204,
+      "step": 841
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.08347708223526511,
+      "learning_rate": 6.538280363802499e-05,
+      "loss": 0.2175,
+      "step": 842
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.07752744914413041,
+      "learning_rate": 6.516327489498804e-05,
+      "loss": 0.2328,
+      "step": 843
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.07884514716075042,
+      "learning_rate": 6.494393706333995e-05,
+      "loss": 0.2276,
+      "step": 844
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.07538113917330841,
+      "learning_rate": 6.472479134509052e-05,
+      "loss": 0.2257,
+      "step": 845
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.0755268134292597,
+      "learning_rate": 6.450583894119702e-05,
+      "loss": 0.2075,
+      "step": 846
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.08583763942123201,
+      "learning_rate": 6.428708105155706e-05,
+      "loss": 0.2394,
+      "step": 847
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.07746270008457591,
+      "learning_rate": 6.406851887500248e-05,
+      "loss": 0.2352,
+      "step": 848
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.07868625381617642,
+      "learning_rate": 6.38501536092925e-05,
+      "loss": 0.2086,
+      "step": 849
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.07096151981914894,
+      "learning_rate": 6.36319864511072e-05,
+      "loss": 0.2158,
+      "step": 850
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.08205381718798445,
+      "learning_rate": 6.341401859604102e-05,
+      "loss": 0.2258,
+      "step": 851
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.07661307158470883,
+      "learning_rate": 6.319625123859625e-05,
+      "loss": 0.2168,
+      "step": 852
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.07015306888241452,
+      "learning_rate": 6.29786855721763e-05,
+      "loss": 0.1934,
+      "step": 853
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.08513296732083729,
+      "learning_rate": 6.276132278907934e-05,
+      "loss": 0.2144,
+      "step": 854
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.08487284686541557,
+      "learning_rate": 6.254416408049179e-05,
+      "loss": 0.2123,
+      "step": 855
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.08304921127632507,
+      "learning_rate": 6.232721063648148e-05,
+      "loss": 0.2218,
+      "step": 856
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.08400616252310818,
+      "learning_rate": 6.211046364599165e-05,
+      "loss": 0.231,
+      "step": 857
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.07635964883244331,
+      "learning_rate": 6.18939242968338e-05,
+      "loss": 0.2317,
+      "step": 858
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.07584145909450023,
+      "learning_rate": 6.167759377568183e-05,
+      "loss": 0.2243,
+      "step": 859
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.07584652436800784,
+      "learning_rate": 6.146147326806508e-05,
+      "loss": 0.2229,
+      "step": 860
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.0751093614729807,
+      "learning_rate": 6.124556395836194e-05,
+      "loss": 0.2074,
+      "step": 861
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.0742844961736531,
+      "learning_rate": 6.1029867029793496e-05,
+      "loss": 0.199,
+      "step": 862
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.07550110812846336,
+      "learning_rate": 6.08143836644169e-05,
+      "loss": 0.2025,
+      "step": 863
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.07376067067190423,
+      "learning_rate": 6.059911504311891e-05,
+      "loss": 0.2227,
+      "step": 864
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.07538792785701962,
+      "learning_rate": 6.038406234560954e-05,
+      "loss": 0.2081,
+      "step": 865
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.08635457656350462,
+      "learning_rate": 6.0169226750415386e-05,
+      "loss": 0.2476,
+      "step": 866
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.08330839654474603,
+      "learning_rate": 5.9954609434873344e-05,
+      "loss": 0.248,
+      "step": 867
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.07593676767859536,
+      "learning_rate": 5.9740211575124195e-05,
+      "loss": 0.2262,
+      "step": 868
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.07814013811313314,
+      "learning_rate": 5.952603434610584e-05,
+      "loss": 0.2165,
+      "step": 869
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.07583035114678865,
+      "learning_rate": 5.931207892154729e-05,
+      "loss": 0.216,
+      "step": 870
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.07599364115714057,
+      "learning_rate": 5.909834647396195e-05,
+      "loss": 0.2113,
+      "step": 871
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.08201145887049803,
+      "learning_rate": 5.888483817464125e-05,
+      "loss": 0.2356,
+      "step": 872
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.07660908729188763,
+      "learning_rate": 5.8671555193648306e-05,
+      "loss": 0.2321,
+      "step": 873
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.08060695899157026,
+      "learning_rate": 5.845849869981137e-05,
+      "loss": 0.2293,
+      "step": 874
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.0805750177578399,
+      "learning_rate": 5.8245669860717556e-05,
+      "loss": 0.2154,
+      "step": 875
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.0768656042945751,
+      "learning_rate": 5.8033069842706446e-05,
+      "loss": 0.209,
+      "step": 876
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.07982817754129624,
+      "learning_rate": 5.782069981086351e-05,
+      "loss": 0.2283,
+      "step": 877
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.07654810874814488,
+      "learning_rate": 5.7608560929013946e-05,
+      "loss": 0.2246,
+      "step": 878
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.08013386356605007,
+      "learning_rate": 5.7396654359716176e-05,
+      "loss": 0.2271,
+      "step": 879
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.07772867287276429,
+      "learning_rate": 5.7184981264255554e-05,
+      "loss": 0.2281,
+      "step": 880
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.0741030330542206,
+      "learning_rate": 5.6973542802637925e-05,
+      "loss": 0.2083,
+      "step": 881
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.07449413019476621,
+      "learning_rate": 5.6762340133583235e-05,
+      "loss": 0.218,
+      "step": 882
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.0811237908133261,
+      "learning_rate": 5.655137441451934e-05,
+      "loss": 0.2327,
+      "step": 883
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.07623140136593945,
+      "learning_rate": 5.634064680157555e-05,
+      "loss": 0.2116,
+      "step": 884
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.07544703275756244,
+      "learning_rate": 5.6130158449576274e-05,
+      "loss": 0.1944,
+      "step": 885
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.07327661341007288,
+      "learning_rate": 5.591991051203476e-05,
+      "loss": 0.2045,
+      "step": 886
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.07669199142811815,
+      "learning_rate": 5.5709904141146776e-05,
+      "loss": 0.2414,
+      "step": 887
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.07598512326726549,
+      "learning_rate": 5.550014048778409e-05,
+      "loss": 0.239,
+      "step": 888
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.08279375075474679,
+      "learning_rate": 5.5290620701488594e-05,
+      "loss": 0.2365,
+      "step": 889
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.08085062007223938,
+      "learning_rate": 5.508134593046561e-05,
+      "loss": 0.2352,
+      "step": 890
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.07479237375427672,
+      "learning_rate": 5.487231732157767e-05,
+      "loss": 0.2188,
+      "step": 891
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.07337494580300977,
+      "learning_rate": 5.4663536020338394e-05,
+      "loss": 0.212,
+      "step": 892
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.07393526535474362,
+      "learning_rate": 5.445500317090608e-05,
+      "loss": 0.2178,
+      "step": 893
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.08072792681644342,
+      "learning_rate": 5.424671991607746e-05,
+      "loss": 0.2289,
+      "step": 894
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.0801226939639071,
+      "learning_rate": 5.403868739728153e-05,
+      "loss": 0.2119,
+      "step": 895
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.08708409489959527,
+      "learning_rate": 5.383090675457301e-05,
+      "loss": 0.2299,
+      "step": 896
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.08101317480165578,
+      "learning_rate": 5.3623379126626426e-05,
+      "loss": 0.2274,
+      "step": 897
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.07908162407312125,
+      "learning_rate": 5.3416105650729855e-05,
+      "loss": 0.2286,
+      "step": 898
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.07968340973080358,
+      "learning_rate": 5.320908746277832e-05,
+      "loss": 0.2427,
+      "step": 899
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.07642266363408927,
+      "learning_rate": 5.300232569726804e-05,
+      "loss": 0.2221,
+      "step": 900
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.07579467681527446,
+      "learning_rate": 5.2795821487289765e-05,
+      "loss": 0.228,
+      "step": 901
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.07701034802506825,
+      "learning_rate": 5.258957596452307e-05,
+      "loss": 0.2247,
+      "step": 902
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.0795754466462776,
+      "learning_rate": 5.2383590259229755e-05,
+      "loss": 0.248,
+      "step": 903
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.08323784258336307,
+      "learning_rate": 5.2177865500247655e-05,
+      "loss": 0.2416,
+      "step": 904
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.07542690617191034,
+      "learning_rate": 5.1972402814984733e-05,
+      "loss": 0.2325,
+      "step": 905
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.07770761515689899,
+      "learning_rate": 5.176720332941268e-05,
+      "loss": 0.2339,
+      "step": 906
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.07603886496540664,
+      "learning_rate": 5.156226816806082e-05,
+      "loss": 0.1998,
+      "step": 907
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.07537679322997529,
+      "learning_rate": 5.1357598454009984e-05,
+      "loss": 0.2235,
+      "step": 908
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.07900614929787292,
+      "learning_rate": 5.1153195308886146e-05,
+      "loss": 0.2262,
+      "step": 909
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.07588147167645987,
+      "learning_rate": 5.094905985285456e-05,
+      "loss": 0.2095,
+      "step": 910
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.07850439219047502,
+      "learning_rate": 5.074519320461357e-05,
+      "loss": 0.2336,
+      "step": 911
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.07999541209427953,
+      "learning_rate": 5.054159648138823e-05,
+      "loss": 0.2145,
+      "step": 912
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.2374275028705597,
+      "eval_runtime": 46.8588,
+      "eval_samples_per_second": 49.574,
+      "eval_steps_per_second": 0.79,
+      "step": 912
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.08494971190207824,
+      "learning_rate": 5.0338270798924434e-05,
+      "loss": 0.2205,
+      "step": 913
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.08044193073202309,
+      "learning_rate": 5.013521727148277e-05,
+      "loss": 0.2262,
+      "step": 914
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.07872818611127354,
+      "learning_rate": 4.993243701183231e-05,
+      "loss": 0.2081,
+      "step": 915
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.07558012010923276,
+      "learning_rate": 4.972993113124459e-05,
+      "loss": 0.2116,
+      "step": 916
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.08315873730597646,
+      "learning_rate": 4.9527700739487536e-05,
+      "loss": 0.2273,
+      "step": 917
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.07691311558614734,
+      "learning_rate": 4.932574694481925e-05,
+      "loss": 0.2234,
+      "step": 918
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.06919656432980084,
+      "learning_rate": 4.9124070853982105e-05,
+      "loss": 0.194,
+      "step": 919
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.07065224357842954,
+      "learning_rate": 4.892267357219662e-05,
+      "loss": 0.1899,
+      "step": 920
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.0713428660073573,
+      "learning_rate": 4.872155620315536e-05,
+      "loss": 0.1942,
+      "step": 921
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.07659918796490793,
+      "learning_rate": 4.852071984901696e-05,
+      "loss": 0.1907,
+      "step": 922
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.07164694820566198,
+      "learning_rate": 4.8320165610399904e-05,
+      "loss": 0.1904,
+      "step": 923
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.07910286643283973,
+      "learning_rate": 4.8119894586376844e-05,
+      "loss": 0.2001,
+      "step": 924
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.08146712493629228,
+      "learning_rate": 4.791990787446827e-05,
+      "loss": 0.2054,
+      "step": 925
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.07506387543589367,
+      "learning_rate": 4.77202065706365e-05,
+      "loss": 0.1952,
+      "step": 926
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.08437415813538211,
+      "learning_rate": 4.752079176927991e-05,
+      "loss": 0.2055,
+      "step": 927
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.08519431650138368,
+      "learning_rate": 4.732166456322671e-05,
+      "loss": 0.2002,
+      "step": 928
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.08297988758477137,
+      "learning_rate": 4.712282604372905e-05,
+      "loss": 0.1816,
+      "step": 929
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.08923670688241434,
+      "learning_rate": 4.69242773004571e-05,
+      "loss": 0.1874,
+      "step": 930
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.08419500596584514,
+      "learning_rate": 4.672601942149284e-05,
+      "loss": 0.1873,
+      "step": 931
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.08582510968197693,
+      "learning_rate": 4.652805349332437e-05,
+      "loss": 0.1849,
+      "step": 932
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.08438655842537737,
+      "learning_rate": 4.633038060083996e-05,
+      "loss": 0.189,
+      "step": 933
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.09081370319395073,
+      "learning_rate": 4.613300182732176e-05,
+      "loss": 0.1984,
+      "step": 934
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.07978925706090226,
+      "learning_rate": 4.593591825444028e-05,
+      "loss": 0.1838,
+      "step": 935
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.08557857259316759,
+      "learning_rate": 4.57391309622481e-05,
+      "loss": 0.1965,
+      "step": 936
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.09155272928873684,
+      "learning_rate": 4.554264102917434e-05,
+      "loss": 0.1946,
+      "step": 937
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.08994319843573481,
+      "learning_rate": 4.534644953201845e-05,
+      "loss": 0.2034,
+      "step": 938
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.08747994584789397,
+      "learning_rate": 4.5150557545944315e-05,
+      "loss": 0.1976,
+      "step": 939
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.09345072423048112,
+      "learning_rate": 4.495496614447454e-05,
+      "loss": 0.1894,
+      "step": 940
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.08696578126501477,
+      "learning_rate": 4.475967639948446e-05,
+      "loss": 0.1809,
+      "step": 941
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.09658718500378473,
+      "learning_rate": 4.456468938119627e-05,
+      "loss": 0.1987,
+      "step": 942
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.08404978710958308,
+      "learning_rate": 4.4370006158173205e-05,
+      "loss": 0.1866,
+      "step": 943
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.08146487282437172,
+      "learning_rate": 4.417562779731355e-05,
+      "loss": 0.1784,
+      "step": 944
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.08881869407787875,
+      "learning_rate": 4.398155536384497e-05,
+      "loss": 0.1951,
+      "step": 945
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.09057852132788513,
+      "learning_rate": 4.378778992131861e-05,
+      "loss": 0.1779,
+      "step": 946
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.09553533559254306,
+      "learning_rate": 4.359433253160319e-05,
+      "loss": 0.1783,
+      "step": 947
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.08742612909358861,
+      "learning_rate": 4.340118425487929e-05,
+      "loss": 0.1909,
+      "step": 948
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.08935384801009703,
+      "learning_rate": 4.320834614963353e-05,
+      "loss": 0.1882,
+      "step": 949
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.09402532679423165,
+      "learning_rate": 4.301581927265255e-05,
+      "loss": 0.1961,
+      "step": 950
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.09846159738441958,
+      "learning_rate": 4.282360467901767e-05,
+      "loss": 0.1965,
+      "step": 951
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.09148613605547108,
+      "learning_rate": 4.2631703422098725e-05,
+      "loss": 0.1882,
+      "step": 952
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.09850194010118866,
+      "learning_rate": 4.2440116553548294e-05,
+      "loss": 0.1998,
+      "step": 953
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.08474300374931883,
+      "learning_rate": 4.2248845123296234e-05,
+      "loss": 0.167,
+      "step": 954
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.09267220244014548,
+      "learning_rate": 4.205789017954364e-05,
+      "loss": 0.1953,
+      "step": 955
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.08876364827989355,
+      "learning_rate": 4.186725276875724e-05,
+      "loss": 0.1878,
+      "step": 956
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.09309338561854612,
+      "learning_rate": 4.1676933935663654e-05,
+      "loss": 0.1935,
+      "step": 957
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.09251347308369356,
+      "learning_rate": 4.1486934723243545e-05,
+      "loss": 0.1928,
+      "step": 958
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.08543876038991123,
+      "learning_rate": 4.129725617272604e-05,
+      "loss": 0.174,
+      "step": 959
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.09575551895434961,
+      "learning_rate": 4.1107899323583124e-05,
+      "loss": 0.1911,
+      "step": 960
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.0985416897894017,
+      "learning_rate": 4.091886521352356e-05,
+      "loss": 0.1928,
+      "step": 961
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.08998883495184473,
+      "learning_rate": 4.073015487848765e-05,
+      "loss": 0.184,
+      "step": 962
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.0976158249154309,
+      "learning_rate": 4.0541769352641135e-05,
+      "loss": 0.1859,
+      "step": 963
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.10514630814129479,
+      "learning_rate": 4.0353709668369986e-05,
+      "loss": 0.1904,
+      "step": 964
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.0857491379616673,
+      "learning_rate": 4.016597685627441e-05,
+      "loss": 0.1677,
+      "step": 965
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.08918832561649435,
+      "learning_rate": 3.997857194516319e-05,
+      "loss": 0.174,
+      "step": 966
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.0915443518297898,
+      "learning_rate": 3.97914959620483e-05,
+      "loss": 0.1763,
+      "step": 967
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.09941982016537511,
+      "learning_rate": 3.9604749932139055e-05,
+      "loss": 0.2021,
+      "step": 968
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.0881745652765701,
+      "learning_rate": 3.941833487883658e-05,
+      "loss": 0.1811,
+      "step": 969
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.089917437769552,
+      "learning_rate": 3.923225182372825e-05,
+      "loss": 0.1813,
+      "step": 970
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.09334122429461554,
+      "learning_rate": 3.9046501786581916e-05,
+      "loss": 0.1861,
+      "step": 971
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.0945747968767825,
+      "learning_rate": 3.886108578534047e-05,
+      "loss": 0.1933,
+      "step": 972
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.08922519463480218,
+      "learning_rate": 3.8676004836116364e-05,
+      "loss": 0.1878,
+      "step": 973
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.08977645447647488,
+      "learning_rate": 3.849125995318571e-05,
+      "loss": 0.1903,
+      "step": 974
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.08928842061488207,
+      "learning_rate": 3.830685214898302e-05,
+      "loss": 0.1794,
+      "step": 975
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.09269511781841656,
+      "learning_rate": 3.8122782434095527e-05,
+      "loss": 0.1897,
+      "step": 976
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.09192856007579465,
+      "learning_rate": 3.793905181725772e-05,
+      "loss": 0.1847,
+      "step": 977
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.0935488270972769,
+      "learning_rate": 3.7755661305345737e-05,
+      "loss": 0.1767,
+      "step": 978
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.10012697860343904,
+      "learning_rate": 3.7572611903371836e-05,
+      "loss": 0.1751,
+      "step": 979
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.0953075007003056,
+      "learning_rate": 3.7389904614479e-05,
+      "loss": 0.1792,
+      "step": 980
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.0971246459842794,
+      "learning_rate": 3.720754043993534e-05,
+      "loss": 0.1953,
+      "step": 981
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.1005716730391889,
+      "learning_rate": 3.7025520379128675e-05,
+      "loss": 0.2013,
+      "step": 982
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.09646192745134229,
+      "learning_rate": 3.684384542956097e-05,
+      "loss": 0.1872,
+      "step": 983
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.09864151793483182,
+      "learning_rate": 3.6662516586843e-05,
+      "loss": 0.181,
+      "step": 984
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.10047779009951295,
+      "learning_rate": 3.6481534844688624e-05,
+      "loss": 0.2025,
+      "step": 985
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.0951495166566708,
+      "learning_rate": 3.6300901194909795e-05,
+      "loss": 0.169,
+      "step": 986
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.08980599418608746,
+      "learning_rate": 3.612061662741072e-05,
+      "loss": 0.1851,
+      "step": 987
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.09336200036064458,
+      "learning_rate": 3.594068213018249e-05,
+      "loss": 0.18,
+      "step": 988
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.08876162326953495,
+      "learning_rate": 3.57610986892979e-05,
+      "loss": 0.1856,
+      "step": 989
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.08574659373142962,
+      "learning_rate": 3.5581867288905824e-05,
+      "loss": 0.1824,
+      "step": 990
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.09927274349976051,
+      "learning_rate": 3.540298891122589e-05,
+      "loss": 0.1958,
+      "step": 991
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.09181296937957556,
+      "learning_rate": 3.522446453654316e-05,
+      "loss": 0.1853,
+      "step": 992
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.09240765993528147,
+      "learning_rate": 3.5046295143202565e-05,
+      "loss": 0.1988,
+      "step": 993
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.10121490528157377,
+      "learning_rate": 3.486848170760377e-05,
+      "loss": 0.1886,
+      "step": 994
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.09288359677200793,
+      "learning_rate": 3.4691025204195824e-05,
+      "loss": 0.1897,
+      "step": 995
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.0886840190217752,
+      "learning_rate": 3.45139266054715e-05,
+      "loss": 0.1932,
+      "step": 996
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.09534163713661604,
+      "learning_rate": 3.433718688196238e-05,
+      "loss": 0.2044,
+      "step": 997
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.08808254903632433,
+      "learning_rate": 3.4160807002233185e-05,
+      "loss": 0.1899,
+      "step": 998
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.08996554080914541,
+      "learning_rate": 3.3984787932876814e-05,
+      "loss": 0.1811,
+      "step": 999
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.09254890173812214,
+      "learning_rate": 3.380913063850877e-05,
+      "loss": 0.1926,
+      "step": 1000
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.08614123372454993,
+      "learning_rate": 3.363383608176188e-05,
+      "loss": 0.167,
+      "step": 1001
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.09839543018841357,
+      "learning_rate": 3.345890522328123e-05,
+      "loss": 0.1957,
+      "step": 1002
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.09764356530777654,
+      "learning_rate": 3.3284339021718734e-05,
+      "loss": 0.1912,
+      "step": 1003
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.09203800024461105,
+      "learning_rate": 3.311013843372792e-05,
+      "loss": 0.1822,
+      "step": 1004
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.1062649633936401,
+      "learning_rate": 3.293630441395872e-05,
+      "loss": 0.2239,
+      "step": 1005
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.09661288795457461,
+      "learning_rate": 3.276283791505211e-05,
+      "loss": 0.1859,
+      "step": 1006
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.0995187534170564,
+      "learning_rate": 3.258973988763506e-05,
+      "loss": 0.1872,
+      "step": 1007
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.0883888481407275,
+      "learning_rate": 3.241701128031534e-05,
+      "loss": 0.1822,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.09719494105031129,
+      "learning_rate": 3.224465303967603e-05,
+      "loss": 0.1897,
+      "step": 1009
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.09287605633398062,
+      "learning_rate": 3.207266611027069e-05,
+      "loss": 0.1892,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.09060705948699806,
+      "learning_rate": 3.190105143461798e-05,
+      "loss": 0.1873,
+      "step": 1011
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.0982344296314008,
+      "learning_rate": 3.172980995319654e-05,
+      "loss": 0.1966,
+      "step": 1012
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.09921501201436793,
+      "learning_rate": 3.155894260443985e-05,
+      "loss": 0.1876,
+      "step": 1013
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.10218535680594067,
+      "learning_rate": 3.13884503247311e-05,
+      "loss": 0.1844,
+      "step": 1014
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.09567246627188801,
+      "learning_rate": 3.121833404839794e-05,
+      "loss": 0.1822,
+      "step": 1015
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.08747479297087953,
+      "learning_rate": 3.104859470770759e-05,
+      "loss": 0.1825,
+      "step": 1016
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.09192595132215417,
+      "learning_rate": 3.087923323286149e-05,
+      "loss": 0.1736,
+      "step": 1017
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.09312671356576784,
+      "learning_rate": 3.071025055199037e-05,
+      "loss": 0.1694,
+      "step": 1018
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.0943007336358788,
+      "learning_rate": 3.0541647591149106e-05,
+      "loss": 0.1948,
+      "step": 1019
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.08884398188847548,
+      "learning_rate": 3.0373425274311516e-05,
+      "loss": 0.1825,
+      "step": 1020
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.09469161421269594,
+      "learning_rate": 3.0205584523365626e-05,
+      "loss": 0.187,
+      "step": 1021
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.10115293763747425,
+      "learning_rate": 3.003812625810831e-05,
+      "loss": 0.1933,
+      "step": 1022
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.09469339438935565,
+      "learning_rate": 2.98710513962403e-05,
+      "loss": 0.1918,
+      "step": 1023
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.09753262801481763,
+      "learning_rate": 2.970436085336127e-05,
+      "loss": 0.1894,
+      "step": 1024
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.10483582312468735,
+      "learning_rate": 2.953805554296477e-05,
+      "loss": 0.1918,
+      "step": 1025
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.09733351306910255,
+      "learning_rate": 2.937213637643319e-05,
+      "loss": 0.1956,
+      "step": 1026
+    },
+    {
+      "epoch": 2.24,
+      "eval_loss": 0.24242068827152252,
+      "eval_runtime": 46.9235,
+      "eval_samples_per_second": 49.506,
+      "eval_steps_per_second": 0.789,
+      "step": 1026
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.09834625617204684,
+      "learning_rate": 2.92066042630328e-05,
+      "loss": 0.1841,
+      "step": 1027
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.0961460258210564,
+      "learning_rate": 2.904146010990868e-05,
+      "loss": 0.1922,
+      "step": 1028
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.09769602442424419,
+      "learning_rate": 2.8876704822079914e-05,
+      "loss": 0.1885,
+      "step": 1029
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.09208940503854306,
+      "learning_rate": 2.8712339302434475e-05,
+      "loss": 0.1801,
+      "step": 1030
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.0977909256219045,
+      "learning_rate": 2.8548364451724376e-05,
+      "loss": 0.1849,
+      "step": 1031
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.09453169452152878,
+      "learning_rate": 2.8384781168560693e-05,
+      "loss": 0.1842,
+      "step": 1032
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.10183728623603636,
+      "learning_rate": 2.822159034940858e-05,
+      "loss": 0.202,
+      "step": 1033
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.0953546866483737,
+      "learning_rate": 2.8058792888582452e-05,
+      "loss": 0.1851,
+      "step": 1034
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.1033836549165198,
+      "learning_rate": 2.7896389678241174e-05,
+      "loss": 0.1907,
+      "step": 1035
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.10163218601739536,
+      "learning_rate": 2.7734381608382832e-05,
+      "loss": 0.1866,
+      "step": 1036
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.09146817689402649,
+      "learning_rate": 2.7572769566840208e-05,
+      "loss": 0.175,
+      "step": 1037
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.1039311463423461,
+      "learning_rate": 2.7411554439275767e-05,
+      "loss": 0.1868,
+      "step": 1038
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.10539902042122816,
+      "learning_rate": 2.7250737109176795e-05,
+      "loss": 0.2047,
+      "step": 1039
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.09861721727566744,
+      "learning_rate": 2.7090318457850615e-05,
+      "loss": 0.1843,
+      "step": 1040
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.09198681477531094,
+      "learning_rate": 2.693029936441962e-05,
+      "loss": 0.1712,
+      "step": 1041
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.09488052951206745,
+      "learning_rate": 2.6770680705816664e-05,
+      "loss": 0.1912,
+      "step": 1042
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.09425261986314033,
+      "learning_rate": 2.6611463356780096e-05,
+      "loss": 0.1976,
+      "step": 1043
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.08930434134867352,
+      "learning_rate": 2.6452648189849038e-05,
+      "loss": 0.1651,
+      "step": 1044
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.09752264420109448,
+      "learning_rate": 2.6294236075358568e-05,
+      "loss": 0.1815,
+      "step": 1045
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.09245642155044898,
+      "learning_rate": 2.6136227881435017e-05,
+      "loss": 0.1856,
+      "step": 1046
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.09390950541637487,
+      "learning_rate": 2.5978624473991e-05,
+      "loss": 0.1827,
+      "step": 1047
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.10161043833059343,
+      "learning_rate": 2.5821426716721053e-05,
+      "loss": 0.1754,
+      "step": 1048
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.0954031427279133,
+      "learning_rate": 2.5664635471096565e-05,
+      "loss": 0.1758,
+      "step": 1049
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.09397349936011976,
+      "learning_rate": 2.550825159636111e-05,
+      "loss": 0.1917,
+      "step": 1050
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.09652214025492545,
+      "learning_rate": 2.5352275949525905e-05,
+      "loss": 0.1825,
+      "step": 1051
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.09389997447049563,
+      "learning_rate": 2.5196709385364958e-05,
+      "loss": 0.1951,
+      "step": 1052
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.10158267187654747,
+      "learning_rate": 2.5041552756410447e-05,
+      "loss": 0.1917,
+      "step": 1053
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.09345318578065326,
+      "learning_rate": 2.4886806912948035e-05,
+      "loss": 0.1921,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.08914153749699319,
+      "learning_rate": 2.4732472703012143e-05,
+      "loss": 0.1932,
+      "step": 1055
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.09460803609793658,
+      "learning_rate": 2.4578550972381442e-05,
+      "loss": 0.1742,
+      "step": 1056
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.09751006151669249,
+      "learning_rate": 2.4425042564574184e-05,
+      "loss": 0.1894,
+      "step": 1057
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.09064582758023604,
+      "learning_rate": 2.4271948320843407e-05,
+      "loss": 0.1642,
+      "step": 1058
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.10696922649173776,
+      "learning_rate": 2.411926908017258e-05,
+      "loss": 0.1971,
+      "step": 1059
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.09715364820297807,
+      "learning_rate": 2.3967005679270736e-05,
+      "loss": 0.1952,
+      "step": 1060
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.09166459434327026,
+      "learning_rate": 2.381515895256824e-05,
+      "loss": 0.1753,
+      "step": 1061
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.09931255509403206,
+      "learning_rate": 2.3663729732211883e-05,
+      "loss": 0.1826,
+      "step": 1062
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.09963652296926058,
+      "learning_rate": 2.351271884806039e-05,
+      "loss": 0.1906,
+      "step": 1063
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.1005470747082438,
+      "learning_rate": 2.3362127127680044e-05,
+      "loss": 0.194,
+      "step": 1064
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.10241286368340984,
+      "learning_rate": 2.3211955396340002e-05,
+      "loss": 0.2016,
+      "step": 1065
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.10112850832040313,
+      "learning_rate": 2.3062204477007775e-05,
+      "loss": 0.194,
+      "step": 1066
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.09972095317582468,
+      "learning_rate": 2.2912875190344828e-05,
+      "loss": 0.1899,
+      "step": 1067
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.08950430559478709,
+      "learning_rate": 2.276396835470188e-05,
+      "loss": 0.1505,
+      "step": 1068
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.10271077084341011,
+      "learning_rate": 2.2615484786114616e-05,
+      "loss": 0.1937,
+      "step": 1069
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.09591271843791505,
+      "learning_rate": 2.2467425298299227e-05,
+      "loss": 0.1719,
+      "step": 1070
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.1025513232939018,
+      "learning_rate": 2.2319790702647693e-05,
+      "loss": 0.1976,
+      "step": 1071
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.10228578075999872,
+      "learning_rate": 2.2172581808223636e-05,
+      "loss": 0.2139,
+      "step": 1072
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.09793032809163253,
+      "learning_rate": 2.202579942175772e-05,
+      "loss": 0.1841,
+      "step": 1073
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.09833165734167365,
+      "learning_rate": 2.1879444347643263e-05,
+      "loss": 0.1989,
+      "step": 1074
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.09161584840330853,
+      "learning_rate": 2.1733517387931858e-05,
+      "loss": 0.1713,
+      "step": 1075
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.09606547269118214,
+      "learning_rate": 2.1588019342328968e-05,
+      "loss": 0.1919,
+      "step": 1076
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.0994420797143376,
+      "learning_rate": 2.144295100818946e-05,
+      "loss": 0.2069,
+      "step": 1077
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.09836914682819063,
+      "learning_rate": 2.1298313180513364e-05,
+      "loss": 0.1937,
+      "step": 1078
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.09568820564600349,
+      "learning_rate": 2.1154106651941464e-05,
+      "loss": 0.1743,
+      "step": 1079
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.09804045439590922,
+      "learning_rate": 2.1010332212750928e-05,
+      "loss": 0.1894,
+      "step": 1080
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.08951884587393363,
+      "learning_rate": 2.0866990650851014e-05,
+      "loss": 0.177,
+      "step": 1081
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.09344140562934025,
+      "learning_rate": 2.0724082751778672e-05,
+      "loss": 0.1919,
+      "step": 1082
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.1008113908422888,
+      "learning_rate": 2.0581609298694426e-05,
+      "loss": 0.1988,
+      "step": 1083
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.09210237436003878,
+      "learning_rate": 2.0439571072377882e-05,
+      "loss": 0.1774,
+      "step": 1084
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.0999360198372951,
+      "learning_rate": 2.0297968851223504e-05,
+      "loss": 0.194,
+      "step": 1085
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.09508138252953177,
+      "learning_rate": 2.01568034112364e-05,
+      "loss": 0.1827,
+      "step": 1086
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.0895347828442651,
+      "learning_rate": 2.0016075526028065e-05,
+      "loss": 0.1766,
+      "step": 1087
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.09480757542704112,
+      "learning_rate": 1.9875785966812076e-05,
+      "loss": 0.1868,
+      "step": 1088
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.09194019978706514,
+      "learning_rate": 1.9735935502399962e-05,
+      "loss": 0.1844,
+      "step": 1089
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.08501617784325628,
+      "learning_rate": 1.9596524899196845e-05,
+      "loss": 0.1707,
+      "step": 1090
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.09611210240751791,
+      "learning_rate": 1.9457554921197386e-05,
+      "loss": 0.183,
+      "step": 1091
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.09080898562934359,
+      "learning_rate": 1.9319026329981604e-05,
+      "loss": 0.1669,
+      "step": 1092
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.09689035486233792,
+      "learning_rate": 1.9180939884710512e-05,
+      "loss": 0.1841,
+      "step": 1093
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.09438021748059282,
+      "learning_rate": 1.9043296342122184e-05,
+      "loss": 0.1739,
+      "step": 1094
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.10207303274583229,
+      "learning_rate": 1.8906096456527412e-05,
+      "loss": 0.1946,
+      "step": 1095
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.09341741218377173,
+      "learning_rate": 1.8769340979805805e-05,
+      "loss": 0.1915,
+      "step": 1096
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.10289016286880881,
+      "learning_rate": 1.8633030661401463e-05,
+      "loss": 0.2009,
+      "step": 1097
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.09795652953654237,
+      "learning_rate": 1.8497166248318876e-05,
+      "loss": 0.1848,
+      "step": 1098
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.09369085850494888,
+      "learning_rate": 1.836174848511899e-05,
+      "loss": 0.1677,
+      "step": 1099
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.101108706390889,
+      "learning_rate": 1.8226778113914987e-05,
+      "loss": 0.1916,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.0923731802905025,
+      "learning_rate": 1.809225587436828e-05,
+      "loss": 0.1738,
+      "step": 1101
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.09311296673290433,
+      "learning_rate": 1.795818250368445e-05,
+      "loss": 0.1726,
+      "step": 1102
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.09779033039450798,
+      "learning_rate": 1.7824558736609142e-05,
+      "loss": 0.1883,
+      "step": 1103
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.08752917299407634,
+      "learning_rate": 1.7691385305424123e-05,
+      "loss": 0.1652,
+      "step": 1104
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.0897662512688499,
+      "learning_rate": 1.755866293994333e-05,
+      "loss": 0.1892,
+      "step": 1105
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.10139822512292197,
+      "learning_rate": 1.7426392367508613e-05,
+      "loss": 0.1874,
+      "step": 1106
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.0974112297917095,
+      "learning_rate": 1.7294574312986044e-05,
+      "loss": 0.1733,
+      "step": 1107
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.10616718932416265,
+      "learning_rate": 1.7163209498761768e-05,
+      "loss": 0.1858,
+      "step": 1108
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.09496435325595999,
+      "learning_rate": 1.703229864473811e-05,
+      "loss": 0.1839,
+      "step": 1109
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.09592873877335717,
+      "learning_rate": 1.6901842468329588e-05,
+      "loss": 0.1799,
+      "step": 1110
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.09034665479083971,
+      "learning_rate": 1.6771841684459078e-05,
+      "loss": 0.1765,
+      "step": 1111
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.09601454766012843,
+      "learning_rate": 1.6642297005553687e-05,
+      "loss": 0.1719,
+      "step": 1112
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.08843056344067668,
+      "learning_rate": 1.651320914154112e-05,
+      "loss": 0.1688,
+      "step": 1113
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.10172484334224395,
+      "learning_rate": 1.6384578799845594e-05,
+      "loss": 0.189,
+      "step": 1114
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.1011809600889664,
+      "learning_rate": 1.625640668538403e-05,
+      "loss": 0.1846,
+      "step": 1115
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.09578789148905303,
+      "learning_rate": 1.6128693500562232e-05,
+      "loss": 0.1907,
+      "step": 1116
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.10122720685528408,
+      "learning_rate": 1.600143994527088e-05,
+      "loss": 0.1854,
+      "step": 1117
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.10102517785722426,
+      "learning_rate": 1.587464671688187e-05,
+      "loss": 0.1926,
+      "step": 1118
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.09811253761616896,
+      "learning_rate": 1.57483145102445e-05,
+      "loss": 0.1869,
+      "step": 1119
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.09903748749295115,
+      "learning_rate": 1.562244401768144e-05,
+      "loss": 0.1938,
+      "step": 1120
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.10198521218760591,
+      "learning_rate": 1.5497035928985204e-05,
+      "loss": 0.1921,
+      "step": 1121
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.09982554036853282,
+      "learning_rate": 1.537209093141413e-05,
+      "loss": 0.1957,
+      "step": 1122
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.10335569352811393,
+      "learning_rate": 1.524760970968887e-05,
+      "loss": 0.1823,
+      "step": 1123
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.09363866692240999,
+      "learning_rate": 1.5123592945988463e-05,
+      "loss": 0.1678,
+      "step": 1124
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.10201323881848162,
+      "learning_rate": 1.5000041319946556e-05,
+      "loss": 0.1875,
+      "step": 1125
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.1021844909674077,
+      "learning_rate": 1.4876955508647839e-05,
+      "loss": 0.1847,
+      "step": 1126
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.09731111573402795,
+      "learning_rate": 1.475433618662425e-05,
+      "loss": 0.1756,
+      "step": 1127
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.09770489434830722,
+      "learning_rate": 1.4632184025851247e-05,
+      "loss": 0.178,
+      "step": 1128
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.09581155376159926,
+      "learning_rate": 1.4510499695744228e-05,
+      "loss": 0.1842,
+      "step": 1129
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.09215314192525487,
+      "learning_rate": 1.4389283863154701e-05,
+      "loss": 0.1739,
+      "step": 1130
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.0974223243883501,
+      "learning_rate": 1.426853719236676e-05,
+      "loss": 0.17,
+      "step": 1131
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.09410339050449333,
+      "learning_rate": 1.4148260345093512e-05,
+      "loss": 0.1892,
+      "step": 1132
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.08992806396762007,
+      "learning_rate": 1.4028453980473189e-05,
+      "loss": 0.1896,
+      "step": 1133
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.10196430704027959,
+      "learning_rate": 1.3909118755065787e-05,
+      "loss": 0.1922,
+      "step": 1134
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.1025070958036504,
+      "learning_rate": 1.3790255322849365e-05,
+      "loss": 0.1901,
+      "step": 1135
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.10041099909347231,
+      "learning_rate": 1.3671864335216467e-05,
+      "loss": 0.1762,
+      "step": 1136
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.10097962251228719,
+      "learning_rate": 1.3553946440970567e-05,
+      "loss": 0.1828,
+      "step": 1137
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.09735921030515804,
+      "learning_rate": 1.343650228632245e-05,
+      "loss": 0.1792,
+      "step": 1138
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.09430552835589713,
+      "learning_rate": 1.331953251488679e-05,
+      "loss": 0.1836,
+      "step": 1139
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.09472351462583166,
+      "learning_rate": 1.320303776767855e-05,
+      "loss": 0.1832,
+      "step": 1140
+    },
+    {
+      "epoch": 2.49,
+      "eval_loss": 0.2407429963350296,
+      "eval_runtime": 46.8725,
+      "eval_samples_per_second": 49.56,
+      "eval_steps_per_second": 0.789,
+      "step": 1140
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.09985759139195344,
+      "learning_rate": 1.308701868310943e-05,
+      "loss": 0.1751,
+      "step": 1141
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.09970077360507341,
+      "learning_rate": 1.2971475896984475e-05,
+      "loss": 0.1872,
+      "step": 1142
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.10379509716989263,
+      "learning_rate": 1.2856410042498513e-05,
+      "loss": 0.1933,
+      "step": 1143
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.08976121407736387,
+      "learning_rate": 1.274182175023263e-05,
+      "loss": 0.1624,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.09971878643776405,
+      "learning_rate": 1.2627711648150897e-05,
+      "loss": 0.1869,
+      "step": 1145
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.10622387785821616,
+      "learning_rate": 1.2514080361596791e-05,
+      "loss": 0.1978,
+      "step": 1146
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.11021942058864327,
+      "learning_rate": 1.2400928513289711e-05,
+      "loss": 0.1899,
+      "step": 1147
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.09688228995058676,
+      "learning_rate": 1.2288256723321756e-05,
+      "loss": 0.1757,
+      "step": 1148
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.09705062470833231,
+      "learning_rate": 1.2176065609154164e-05,
+      "loss": 0.1912,
+      "step": 1149
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.0953133831217767,
+      "learning_rate": 1.2064355785614002e-05,
+      "loss": 0.186,
+      "step": 1150
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.09358754088864682,
+      "learning_rate": 1.195312786489079e-05,
+      "loss": 0.1854,
+      "step": 1151
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.09815277235816793,
+      "learning_rate": 1.1842382456533097e-05,
+      "loss": 0.17,
+      "step": 1152
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.10562225932171529,
+      "learning_rate": 1.1732120167445248e-05,
+      "loss": 0.1925,
+      "step": 1153
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.09954570624108924,
+      "learning_rate": 1.162234160188408e-05,
+      "loss": 0.1886,
+      "step": 1154
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.09192094636033367,
+      "learning_rate": 1.1513047361455398e-05,
+      "loss": 0.1649,
+      "step": 1155
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.09494771899919272,
+      "learning_rate": 1.1404238045110937e-05,
+      "loss": 0.1772,
+      "step": 1156
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.09517280922329258,
+      "learning_rate": 1.1295914249144846e-05,
+      "loss": 0.1731,
+      "step": 1157
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.09873450579440651,
+      "learning_rate": 1.1188076567190653e-05,
+      "loss": 0.1956,
+      "step": 1158
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.10328154819231948,
+      "learning_rate": 1.1080725590217866e-05,
+      "loss": 0.2037,
+      "step": 1159
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.10052537020474239,
+      "learning_rate": 1.0973861906528692e-05,
+      "loss": 0.1797,
+      "step": 1160
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.09610347795480867,
+      "learning_rate": 1.0867486101754965e-05,
+      "loss": 0.1804,
+      "step": 1161
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.09534447965996033,
+      "learning_rate": 1.076159875885483e-05,
+      "loss": 0.1874,
+      "step": 1162
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.0984559881922783,
+      "learning_rate": 1.0656200458109577e-05,
+      "loss": 0.1821,
+      "step": 1163
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.0977576539943681,
+      "learning_rate": 1.0551291777120464e-05,
+      "loss": 0.1767,
+      "step": 1164
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.09121908298528363,
+      "learning_rate": 1.0446873290805504e-05,
+      "loss": 0.1719,
+      "step": 1165
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.09480482977694164,
+      "learning_rate": 1.034294557139639e-05,
+      "loss": 0.1773,
+      "step": 1166
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.10273046341038103,
+      "learning_rate": 1.0239509188435392e-05,
+      "loss": 0.1816,
+      "step": 1167
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.10349805700826828,
+      "learning_rate": 1.0136564708772044e-05,
+      "loss": 0.1787,
+      "step": 1168
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.09668771633371812,
+      "learning_rate": 1.0034112696560238e-05,
+      "loss": 0.1857,
+      "step": 1169
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.11179013339228272,
+      "learning_rate": 9.932153713255055e-06,
+      "loss": 0.2134,
+      "step": 1170
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.09944942666014846,
+      "learning_rate": 9.830688317609682e-06,
+      "loss": 0.1875,
+      "step": 1171
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.1013398756597619,
+      "learning_rate": 9.729717065672339e-06,
+      "loss": 0.1807,
+      "step": 1172
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.1033644111571255,
+      "learning_rate": 9.629240510783299e-06,
+      "loss": 0.1809,
+      "step": 1173
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.10178466486509374,
+      "learning_rate": 9.52925920357175e-06,
+      "loss": 0.1781,
+      "step": 1174
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.09382366226979735,
+      "learning_rate": 9.429773691952858e-06,
+      "loss": 0.1799,
+      "step": 1175
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.08891659191265348,
+      "learning_rate": 9.33078452112477e-06,
+      "loss": 0.1668,
+      "step": 1176
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.09714657341886997,
+      "learning_rate": 9.232292233565554e-06,
+      "loss": 0.1796,
+      "step": 1177
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.10167377593228938,
+      "learning_rate": 9.13429736903033e-06,
+      "loss": 0.1887,
+      "step": 1178
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.08712282451156825,
+      "learning_rate": 9.036800464548157e-06,
+      "loss": 0.166,
+      "step": 1179
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.10460116594937302,
+      "learning_rate": 8.939802054419288e-06,
+      "loss": 0.1851,
+      "step": 1180
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.09775219495649233,
+      "learning_rate": 8.843302670212095e-06,
+      "loss": 0.1817,
+      "step": 1181
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.10017385606260155,
+      "learning_rate": 8.747302840760162e-06,
+      "loss": 0.1907,
+      "step": 1182
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.0915647524257185,
+      "learning_rate": 8.651803092159482e-06,
+      "loss": 0.1731,
+      "step": 1183
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.0980085788269613,
+      "learning_rate": 8.556803947765501e-06,
+      "loss": 0.1814,
+      "step": 1184
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.10348067503346223,
+      "learning_rate": 8.462305928190251e-06,
+      "loss": 0.1977,
+      "step": 1185
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.10788501688326456,
+      "learning_rate": 8.368309551299536e-06,
+      "loss": 0.1823,
+      "step": 1186
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.09725835324876417,
+      "learning_rate": 8.274815332210039e-06,
+      "loss": 0.1838,
+      "step": 1187
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.11240842423286292,
+      "learning_rate": 8.181823783286547e-06,
+      "loss": 0.2165,
+      "step": 1188
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.09898010336518936,
+      "learning_rate": 8.08933541413912e-06,
+      "loss": 0.1795,
+      "step": 1189
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.10089452868136724,
+      "learning_rate": 7.997350731620335e-06,
+      "loss": 0.1871,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.0981812479601362,
+      "learning_rate": 7.905870239822444e-06,
+      "loss": 0.1637,
+      "step": 1191
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.10328200356587999,
+      "learning_rate": 7.814894440074616e-06,
+      "loss": 0.1785,
+      "step": 1192
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.09864235822356453,
+      "learning_rate": 7.724423830940309e-06,
+      "loss": 0.1734,
+      "step": 1193
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.0966888866479244,
+      "learning_rate": 7.63445890821438e-06,
+      "loss": 0.1804,
+      "step": 1194
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.09789224719189894,
+      "learning_rate": 7.545000164920424e-06,
+      "loss": 0.1887,
+      "step": 1195
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.09063636445446958,
+      "learning_rate": 7.456048091308121e-06,
+      "loss": 0.1716,
+      "step": 1196
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.10735496257680281,
+      "learning_rate": 7.367603174850502e-06,
+      "loss": 0.1796,
+      "step": 1197
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.101555902010002,
+      "learning_rate": 7.279665900241295e-06,
+      "loss": 0.1759,
+      "step": 1198
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.10154017331398542,
+      "learning_rate": 7.192236749392256e-06,
+      "loss": 0.1951,
+      "step": 1199
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.09787380304585115,
+      "learning_rate": 7.105316201430512e-06,
+      "loss": 0.1838,
+      "step": 1200
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.09725676365974052,
+      "learning_rate": 7.018904732695963e-06,
+      "loss": 0.1815,
+      "step": 1201
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.08880672389105437,
+      "learning_rate": 6.933002816738699e-06,
+      "loss": 0.1715,
+      "step": 1202
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.1004245212862719,
+      "learning_rate": 6.8476109243163496e-06,
+      "loss": 0.1842,
+      "step": 1203
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.10629017487588402,
+      "learning_rate": 6.762729523391498e-06,
+      "loss": 0.1934,
+      "step": 1204
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.10334935870694845,
+      "learning_rate": 6.678359079129215e-06,
+      "loss": 0.1997,
+      "step": 1205
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.09822940477819025,
+      "learning_rate": 6.5945000538943366e-06,
+      "loss": 0.1871,
+      "step": 1206
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.09812328241448662,
+      "learning_rate": 6.511152907249129e-06,
+      "loss": 0.1853,
+      "step": 1207
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.1013383087373085,
+      "learning_rate": 6.428318095950647e-06,
+      "loss": 0.1907,
+      "step": 1208
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.10048676203254954,
+      "learning_rate": 6.345996073948201e-06,
+      "loss": 0.177,
+      "step": 1209
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.08914723130268362,
+      "learning_rate": 6.264187292380985e-06,
+      "loss": 0.1632,
+      "step": 1210
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.10409218439942107,
+      "learning_rate": 6.182892199575552e-06,
+      "loss": 0.1879,
+      "step": 1211
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.09260272331028377,
+      "learning_rate": 6.102111241043296e-06,
+      "loss": 0.1791,
+      "step": 1212
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.09760712341598825,
+      "learning_rate": 6.021844859478143e-06,
+      "loss": 0.1857,
+      "step": 1213
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.10510158649150257,
+      "learning_rate": 5.942093494753942e-06,
+      "loss": 0.1882,
+      "step": 1214
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.09977641778398352,
+      "learning_rate": 5.862857583922232e-06,
+      "loss": 0.1862,
+      "step": 1215
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.09598414850278122,
+      "learning_rate": 5.784137561209801e-06,
+      "loss": 0.1726,
+      "step": 1216
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.10020032595013476,
+      "learning_rate": 5.705933858016188e-06,
+      "loss": 0.1743,
+      "step": 1217
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.10057124706507646,
+      "learning_rate": 5.628246902911494e-06,
+      "loss": 0.1728,
+      "step": 1218
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.09035560197289758,
+      "learning_rate": 5.551077121633874e-06,
+      "loss": 0.1805,
+      "step": 1219
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.09514863083939815,
+      "learning_rate": 5.474424937087353e-06,
+      "loss": 0.1704,
+      "step": 1220
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.1073273546352554,
+      "learning_rate": 5.39829076933942e-06,
+      "loss": 0.2072,
+      "step": 1221
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.09897277803989789,
+      "learning_rate": 5.322675035618674e-06,
+      "loss": 0.1868,
+      "step": 1222
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.1029758550699064,
+      "learning_rate": 5.2475781503126705e-06,
+      "loss": 0.1811,
+      "step": 1223
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.0974427798754134,
+      "learning_rate": 5.173000524965577e-06,
+      "loss": 0.1739,
+      "step": 1224
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.0949660939958928,
+      "learning_rate": 5.0989425682759e-06,
+      "loss": 0.1797,
+      "step": 1225
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.09328370059930155,
+      "learning_rate": 5.025404686094315e-06,
+      "loss": 0.1941,
+      "step": 1226
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.09352026547046174,
+      "learning_rate": 4.952387281421334e-06,
+      "loss": 0.1772,
+      "step": 1227
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.09838990714762562,
+      "learning_rate": 4.8798907544051805e-06,
+      "loss": 0.1693,
+      "step": 1228
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.10471531850661873,
+      "learning_rate": 4.807915502339655e-06,
+      "loss": 0.1837,
+      "step": 1229
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.10089946696088178,
+      "learning_rate": 4.7364619196617495e-06,
+      "loss": 0.1871,
+      "step": 1230
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.10524177411269857,
+      "learning_rate": 4.665530397949713e-06,
+      "loss": 0.1884,
+      "step": 1231
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.10529112870680633,
+      "learning_rate": 4.59512132592077e-06,
+      "loss": 0.1779,
+      "step": 1232
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.10317971049851754,
+      "learning_rate": 4.525235089429026e-06,
+      "loss": 0.1928,
+      "step": 1233
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.10080974179808999,
+      "learning_rate": 4.4558720714633965e-06,
+      "loss": 0.1783,
+      "step": 1234
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.10516149305165114,
+      "learning_rate": 4.387032652145395e-06,
+      "loss": 0.187,
+      "step": 1235
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.0922924926909573,
+      "learning_rate": 4.318717208727186e-06,
+      "loss": 0.1751,
+      "step": 1236
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.1007998885433209,
+      "learning_rate": 4.250926115589427e-06,
+      "loss": 0.1911,
+      "step": 1237
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.09075505465526879,
+      "learning_rate": 4.1836597442392435e-06,
+      "loss": 0.1759,
+      "step": 1238
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.09758904051678226,
+      "learning_rate": 4.116918463308195e-06,
+      "loss": 0.1794,
+      "step": 1239
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.09487326067842981,
+      "learning_rate": 4.050702638550275e-06,
+      "loss": 0.1743,
+      "step": 1240
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.09631561354283748,
+      "learning_rate": 3.985012632839824e-06,
+      "loss": 0.1817,
+      "step": 1241
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.09932584708324509,
+      "learning_rate": 3.919848806169679e-06,
+      "loss": 0.1832,
+      "step": 1242
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.09770047616868205,
+      "learning_rate": 3.855211515649082e-06,
+      "loss": 0.1748,
+      "step": 1243
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.0975873083904438,
+      "learning_rate": 3.7911011155017606e-06,
+      "loss": 0.1782,
+      "step": 1244
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.09528196365505809,
+      "learning_rate": 3.7275179570639863e-06,
+      "loss": 0.1796,
+      "step": 1245
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.09783033113988192,
+      "learning_rate": 3.6644623887826857e-06,
+      "loss": 0.1841,
+      "step": 1246
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.09453507860958489,
+      "learning_rate": 3.6019347562134655e-06,
+      "loss": 0.1819,
+      "step": 1247
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.10171037692226044,
+      "learning_rate": 3.5399354020187792e-06,
+      "loss": 0.1839,
+      "step": 1248
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.09684179182925626,
+      "learning_rate": 3.478464665965997e-06,
+      "loss": 0.1736,
+      "step": 1249
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.09343517111583868,
+      "learning_rate": 3.4175228849255503e-06,
+      "loss": 0.1748,
+      "step": 1250
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.10213658499288596,
+      "learning_rate": 3.3571103928691892e-06,
+      "loss": 0.1951,
+      "step": 1251
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.09190531169804102,
+      "learning_rate": 3.2972275208679625e-06,
+      "loss": 0.1771,
+      "step": 1252
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.0969965779435791,
+      "learning_rate": 3.2378745970905953e-06,
+      "loss": 0.1717,
+      "step": 1253
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.10092979285102317,
+      "learning_rate": 3.1790519468015034e-06,
+      "loss": 0.1922,
+      "step": 1254
+    },
+    {
+      "epoch": 2.74,
+      "eval_loss": 0.23931144177913666,
+      "eval_runtime": 46.8773,
+      "eval_samples_per_second": 49.555,
+      "eval_steps_per_second": 0.789,
+      "step": 1254
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.09713962275660072,
+      "learning_rate": 3.1207598923592263e-06,
+      "loss": 0.1694,
+      "step": 1255
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.0982284566638539,
+      "learning_rate": 3.0629987532144744e-06,
+      "loss": 0.1811,
+      "step": 1256
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.1094266863019417,
+      "learning_rate": 3.005768845908441e-06,
+      "loss": 0.1935,
+      "step": 1257
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.09247566923680253,
+      "learning_rate": 2.9490704840711147e-06,
+      "loss": 0.1651,
+      "step": 1258
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.0987773393483217,
+      "learning_rate": 2.8929039784195033e-06,
+      "loss": 0.183,
+      "step": 1259
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.11322025185967428,
+      "learning_rate": 2.837269636755946e-06,
+      "loss": 0.1957,
+      "step": 1260
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.10189892453562262,
+      "learning_rate": 2.782167763966448e-06,
+      "loss": 0.192,
+      "step": 1261
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.09577367089857132,
+      "learning_rate": 2.7275986620189485e-06,
+      "loss": 0.1745,
+      "step": 1262
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.11083666759091057,
+      "learning_rate": 2.6735626299617457e-06,
+      "loss": 0.2013,
+      "step": 1263
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.10017339120841098,
+      "learning_rate": 2.6200599639218393e-06,
+      "loss": 0.1791,
+      "step": 1264
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.09877046572948672,
+      "learning_rate": 2.567090957103213e-06,
+      "loss": 0.1747,
+      "step": 1265
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.09726494593555605,
+      "learning_rate": 2.5146558997853897e-06,
+      "loss": 0.1772,
+      "step": 1266
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.10268570087764146,
+      "learning_rate": 2.46275507932171e-06,
+      "loss": 0.1895,
+      "step": 1267
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.10395441982334992,
+      "learning_rate": 2.411388780137802e-06,
+      "loss": 0.1785,
+      "step": 1268
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.101817283714402,
+      "learning_rate": 2.360557283730047e-06,
+      "loss": 0.1835,
+      "step": 1269
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.09316783693453097,
+      "learning_rate": 2.3102608686639824e-06,
+      "loss": 0.178,
+      "step": 1270
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.09749612555233375,
+      "learning_rate": 2.260499810572825e-06,
+      "loss": 0.1757,
+      "step": 1271
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.10035463237113246,
+      "learning_rate": 2.2112743821559366e-06,
+      "loss": 0.1719,
+      "step": 1272
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.10367820226712023,
+      "learning_rate": 2.1625848531773406e-06,
+      "loss": 0.1753,
+      "step": 1273
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.09099410461622891,
+      "learning_rate": 2.1144314904642195e-06,
+      "loss": 0.1706,
+      "step": 1274
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.10441930403791613,
+      "learning_rate": 2.0668145579054943e-06,
+      "loss": 0.1854,
+      "step": 1275
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.09062751692532434,
+      "learning_rate": 2.019734316450317e-06,
+      "loss": 0.1755,
+      "step": 1276
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.10685755146655547,
+      "learning_rate": 1.973191024106724e-06,
+      "loss": 0.1948,
+      "step": 1277
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.100033981725985,
+      "learning_rate": 1.927184935940174e-06,
+      "loss": 0.1793,
+      "step": 1278
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.10179900179593263,
+      "learning_rate": 1.8817163040720897e-06,
+      "loss": 0.1786,
+      "step": 1279
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.10144746207455829,
+      "learning_rate": 1.836785377678596e-06,
+      "loss": 0.1816,
+      "step": 1280
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.10167883207655251,
+      "learning_rate": 1.7923924029890848e-06,
+      "loss": 0.181,
+      "step": 1281
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.103710005553058,
+      "learning_rate": 1.748537623284885e-06,
+      "loss": 0.191,
+      "step": 1282
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.10485706173063664,
+      "learning_rate": 1.705221278897917e-06,
+      "loss": 0.1904,
+      "step": 1283
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.09457891795363077,
+      "learning_rate": 1.6624436072093409e-06,
+      "loss": 0.1756,
+      "step": 1284
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.0973546090276532,
+      "learning_rate": 1.6202048426483651e-06,
+      "loss": 0.1849,
+      "step": 1285
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.08889593455905097,
+      "learning_rate": 1.578505216690851e-06,
+      "loss": 0.1633,
+      "step": 1286
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.0931570954186878,
+      "learning_rate": 1.537344957858089e-06,
+      "loss": 0.1928,
+      "step": 1287
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.09128169821309522,
+      "learning_rate": 1.4967242917155456e-06,
+      "loss": 0.1764,
+      "step": 1288
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.09247330252123818,
+      "learning_rate": 1.4566434408716301e-06,
+      "loss": 0.1742,
+      "step": 1289
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.10376444188968435,
+      "learning_rate": 1.4171026249764407e-06,
+      "loss": 0.1774,
+      "step": 1290
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.09827948065288569,
+      "learning_rate": 1.378102060720654e-06,
+      "loss": 0.1992,
+      "step": 1291
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.09794748272408028,
+      "learning_rate": 1.3396419618342037e-06,
+      "loss": 0.1793,
+      "step": 1292
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.09591725294979761,
+      "learning_rate": 1.301722539085204e-06,
+      "loss": 0.1767,
+      "step": 1293
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.10657897921081319,
+      "learning_rate": 1.2643440002787832e-06,
+      "loss": 0.1961,
+      "step": 1294
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.09715161593291378,
+      "learning_rate": 1.2275065502558968e-06,
+      "loss": 0.1729,
+      "step": 1295
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.10372572396541298,
+      "learning_rate": 1.1912103908922945e-06,
+      "loss": 0.195,
+      "step": 1296
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.1041686629064334,
+      "learning_rate": 1.155455721097276e-06,
+      "loss": 0.193,
+      "step": 1297
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.09720136557346325,
+      "learning_rate": 1.12024273681276e-06,
+      "loss": 0.1748,
+      "step": 1298
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.10693945063049883,
+      "learning_rate": 1.0855716310120945e-06,
+      "loss": 0.1732,
+      "step": 1299
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.0921298895664722,
+      "learning_rate": 1.0514425936990368e-06,
+      "loss": 0.1729,
+      "step": 1300
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.09757485812592509,
+      "learning_rate": 1.0178558119067315e-06,
+      "loss": 0.1769,
+      "step": 1301
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.10076394117299803,
+      "learning_rate": 9.848114696966559e-07,
+      "loss": 0.1886,
+      "step": 1302
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.09691233530777216,
+      "learning_rate": 9.523097481576204e-07,
+      "loss": 0.1823,
+      "step": 1303
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.10144946336975605,
+      "learning_rate": 9.20350825404781e-07,
+      "loss": 0.1762,
+      "step": 1304
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.09598304721970188,
+      "learning_rate": 8.889348765786842e-07,
+      "loss": 0.1737,
+      "step": 1305
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.10914192740961,
+      "learning_rate": 8.580620738442457e-07,
+      "loss": 0.1866,
+      "step": 1306
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.09988742709549234,
+      "learning_rate": 8.277325863898511e-07,
+      "loss": 0.1851,
+      "step": 1307
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.10374395811056726,
+      "learning_rate": 7.979465804264452e-07,
+      "loss": 0.1914,
+      "step": 1308
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.10011704885061652,
+      "learning_rate": 7.68704219186589e-07,
+      "loss": 0.2002,
+      "step": 1309
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.09340629057214901,
+      "learning_rate": 7.400056629235819e-07,
+      "loss": 0.1654,
+      "step": 1310
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.09717721751300551,
+      "learning_rate": 7.118510689105407e-07,
+      "loss": 0.1763,
+      "step": 1311
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.09786734653562601,
+      "learning_rate": 6.842405914396e-07,
+      "loss": 0.185,
+      "step": 1312
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.11476421618911542,
+      "learning_rate": 6.571743818210685e-07,
+      "loss": 0.2176,
+      "step": 1313
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.09893753174024014,
+      "learning_rate": 6.306525883825187e-07,
+      "loss": 0.1855,
+      "step": 1314
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.10103787584256144,
+      "learning_rate": 6.046753564680541e-07,
+      "loss": 0.185,
+      "step": 1315
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.09735287751348289,
+      "learning_rate": 5.792428284374762e-07,
+      "loss": 0.1757,
+      "step": 1316
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.09662521638821335,
+      "learning_rate": 5.543551436655304e-07,
+      "loss": 0.1719,
+      "step": 1317
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.0937402491664388,
+      "learning_rate": 5.300124385410943e-07,
+      "loss": 0.1634,
+      "step": 1318
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.0943923195284422,
+      "learning_rate": 5.062148464665018e-07,
+      "loss": 0.1678,
+      "step": 1319
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.09605547697154704,
+      "learning_rate": 4.829624978567204e-07,
+      "loss": 0.1845,
+      "step": 1320
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.10604605472759493,
+      "learning_rate": 4.6025552013874153e-07,
+      "loss": 0.1863,
+      "step": 1321
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.11627461187616332,
+      "learning_rate": 4.380940377507914e-07,
+      "loss": 0.2072,
+      "step": 1322
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.09344587568392909,
+      "learning_rate": 4.164781721416877e-07,
+      "loss": 0.1564,
+      "step": 1323
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.1035185534584825,
+      "learning_rate": 3.954080417701955e-07,
+      "loss": 0.1922,
+      "step": 1324
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.09388611270557366,
+      "learning_rate": 3.7488376210432775e-07,
+      "loss": 0.1744,
+      "step": 1325
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.0906573237947595,
+      "learning_rate": 3.549054456207679e-07,
+      "loss": 0.1631,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.09576627535699536,
+      "learning_rate": 3.354732018041928e-07,
+      "loss": 0.1709,
+      "step": 1327
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.0975270465634383,
+      "learning_rate": 3.165871371467288e-07,
+      "loss": 0.1921,
+      "step": 1328
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.10267767393048689,
+      "learning_rate": 2.9824735514732974e-07,
+      "loss": 0.1748,
+      "step": 1329
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.09867210360973995,
+      "learning_rate": 2.8045395631124406e-07,
+      "loss": 0.1821,
+      "step": 1330
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.10674530156285135,
+      "learning_rate": 2.632070381494156e-07,
+      "loss": 0.1862,
+      "step": 1331
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.09912023872795964,
+      "learning_rate": 2.4650669517800596e-07,
+      "loss": 0.1759,
+      "step": 1332
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.09849285828440721,
+      "learning_rate": 2.3035301891781712e-07,
+      "loss": 0.1671,
+      "step": 1333
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.09742602867932287,
+      "learning_rate": 2.147460978938587e-07,
+      "loss": 0.1731,
+      "step": 1334
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.09673242382952495,
+      "learning_rate": 1.9968601763480367e-07,
+      "loss": 0.1806,
+      "step": 1335
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.09405193781245572,
+      "learning_rate": 1.8517286067254446e-07,
+      "loss": 0.1696,
+      "step": 1336
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.09749504134822441,
+      "learning_rate": 1.7120670654172667e-07,
+      "loss": 0.1797,
+      "step": 1337
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.09687208048205159,
+      "learning_rate": 1.577876317793381e-07,
+      "loss": 0.1787,
+      "step": 1338
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.10047272913099506,
+      "learning_rate": 1.4491570992428704e-07,
+      "loss": 0.1831,
+      "step": 1339
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.10154651552739213,
+      "learning_rate": 1.3259101151694708e-07,
+      "loss": 0.1809,
+      "step": 1340
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.11423762868641689,
+      "learning_rate": 1.2081360409885722e-07,
+      "loss": 0.1937,
+      "step": 1341
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.10684344441402814,
+      "learning_rate": 1.0958355221227789e-07,
+      "loss": 0.1816,
+      "step": 1342
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.10344046437202124,
+      "learning_rate": 9.890091739985785e-08,
+      "loss": 0.188,
+      "step": 1343
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.09755292671705987,
+      "learning_rate": 8.876575820433441e-08,
+      "loss": 0.183,
+      "step": 1344
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.09436505936380679,
+      "learning_rate": 7.917813016816711e-08,
+      "loss": 0.1851,
+      "step": 1345
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.09533844285977386,
+      "learning_rate": 7.013808583323788e-08,
+      "loss": 0.1746,
+      "step": 1346
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.1041941492265022,
+      "learning_rate": 6.164567474057359e-08,
+      "loss": 0.1717,
+      "step": 1347
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.09995337531824866,
+      "learning_rate": 5.370094343009058e-08,
+      "loss": 0.1927,
+      "step": 1348
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.10540790922304928,
+      "learning_rate": 4.630393544031719e-08,
+      "loss": 0.1907,
+      "step": 1349
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.09289428970801497,
+      "learning_rate": 3.94546913081717e-08,
+      "loss": 0.178,
+      "step": 1350
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.0988901671068859,
+      "learning_rate": 3.3153248568695835e-08,
+      "loss": 0.1763,
+      "step": 1351
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.10742679497982714,
+      "learning_rate": 2.73996417549105e-08,
+      "loss": 0.1857,
+      "step": 1352
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.09425408719981482,
+      "learning_rate": 2.219390239760477e-08,
+      "loss": 0.1648,
+      "step": 1353
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.10373918665841911,
+      "learning_rate": 1.7536059025136108e-08,
+      "loss": 0.178,
+      "step": 1354
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.09569261940885007,
+      "learning_rate": 1.3426137163319307e-08,
+      "loss": 0.1801,
+      "step": 1355
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.08776856529861832,
+      "learning_rate": 9.864159335237766e-09,
+      "loss": 0.1677,
+      "step": 1356
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.11240296003237354,
+      "learning_rate": 6.8501450611657695e-09,
+      "loss": 0.1961,
+      "step": 1357
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.09932744575293442,
+      "learning_rate": 4.3841108584241616e-09,
+      "loss": 0.1828,
+      "step": 1358
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.09979353504615274,
+      "learning_rate": 2.466070241324836e-09,
+      "loss": 0.1745,
+      "step": 1359
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.10144576828699613,
+      "learning_rate": 1.0960337210597083e-09,
+      "loss": 0.1783,
+      "step": 1360
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.0950426108940041,
+      "learning_rate": 2.7400880566741395e-10,
+      "loss": 0.1934,
+      "step": 1361
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.09541847266944416,
+      "learning_rate": 0.0,
+      "loss": 0.1752,
+      "step": 1362
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1362,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 454,
+  "total_flos": 7.464942074308067e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}