diff --git "a/checkpoint-1020/trainer_state.json" "b/checkpoint-1020/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1020/trainer_state.json"
@@ -0,0 +1,8193 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 16.73104880581516,
+  "eval_steps": 500,
+  "global_step": 1020,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.016614745586708203,
+      "grad_norm": 0.050998032093048096,
+      "learning_rate": 4.999991432639962e-05,
+      "loss": 0.5487,
+      "num_input_tokens_seen": 70408,
+      "step": 1
+    },
+    {
+      "epoch": 0.033229491173416406,
+      "grad_norm": 0.049370743334293365,
+      "learning_rate": 4.999965730618567e-05,
+      "loss": 0.4981,
+      "num_input_tokens_seen": 139640,
+      "step": 2
+    },
+    {
+      "epoch": 0.04984423676012461,
+      "grad_norm": 0.05077400803565979,
+      "learning_rate": 4.9999228941119745e-05,
+      "loss": 0.5505,
+      "num_input_tokens_seen": 223656,
+      "step": 3
+    },
+    {
+      "epoch": 0.06645898234683281,
+      "grad_norm": 0.04397282376885414,
+      "learning_rate": 4.999862923413781e-05,
+      "loss": 0.504,
+      "num_input_tokens_seen": 300688,
+      "step": 4
+    },
+    {
+      "epoch": 0.08307372793354102,
+      "grad_norm": 0.05225864797830582,
+      "learning_rate": 4.999785818935018e-05,
+      "loss": 0.4925,
+      "num_input_tokens_seen": 366368,
+      "step": 5
+    },
+    {
+      "epoch": 0.09968847352024922,
+      "grad_norm": 0.049482282251119614,
+      "learning_rate": 4.999691581204152e-05,
+      "loss": 0.4771,
+      "num_input_tokens_seen": 445808,
+      "step": 6
+    },
+    {
+      "epoch": 0.11630321910695743,
+      "grad_norm": 0.05594080314040184,
+      "learning_rate": 4.9995802108670775e-05,
+      "loss": 0.4986,
+      "num_input_tokens_seen": 522800,
+      "step": 7
+    },
+    {
+      "epoch": 0.13291796469366562,
+      "grad_norm": 0.051852282136678696,
+      "learning_rate": 4.999451708687114e-05,
+      "loss": 0.5171,
+      "num_input_tokens_seen": 599608,
+      "step": 8
+    },
+    {
+      "epoch": 0.14953271028037382,
+      "grad_norm": 0.045517683029174805,
+      "learning_rate": 4.9993060755450015e-05,
+      "loss": 0.5669,
+      "num_input_tokens_seen": 681424,
+      "step": 9
+    },
+    {
+      "epoch": 0.16614745586708204,
+      "grad_norm": 0.044325754046440125,
+      "learning_rate": 4.999143312438893e-05,
+      "loss": 0.4218,
+      "num_input_tokens_seen": 756744,
+      "step": 10
+    },
+    {
+      "epoch": 0.18276220145379024,
+      "grad_norm": 0.04328459873795509,
+      "learning_rate": 4.998963420484349e-05,
+      "loss": 0.434,
+      "num_input_tokens_seen": 842576,
+      "step": 11
+    },
+    {
+      "epoch": 0.19937694704049844,
+      "grad_norm": 0.04725787043571472,
+      "learning_rate": 4.998766400914329e-05,
+      "loss": 0.4287,
+      "num_input_tokens_seen": 917232,
+      "step": 12
+    },
+    {
+      "epoch": 0.21599169262720663,
+      "grad_norm": 0.03806879371404648,
+      "learning_rate": 4.9985522550791825e-05,
+      "loss": 0.3454,
+      "num_input_tokens_seen": 1006800,
+      "step": 13
+    },
+    {
+      "epoch": 0.23260643821391486,
+      "grad_norm": 0.05201176926493645,
+      "learning_rate": 4.998320984446641e-05,
+      "loss": 0.436,
+      "num_input_tokens_seen": 1085824,
+      "step": 14
+    },
+    {
+      "epoch": 0.24922118380062305,
+      "grad_norm": 0.047955628484487534,
+      "learning_rate": 4.9980725906018074e-05,
+      "loss": 0.4625,
+      "num_input_tokens_seen": 1164160,
+      "step": 15
+    },
+    {
+      "epoch": 0.26583592938733125,
+      "grad_norm": 0.05529098957777023,
+      "learning_rate": 4.997807075247146e-05,
+      "loss": 0.5035,
+      "num_input_tokens_seen": 1242264,
+      "step": 16
+    },
+    {
+      "epoch": 0.2824506749740395,
+      "grad_norm": 0.04751162976026535,
+      "learning_rate": 4.997524440202469e-05,
+      "loss": 0.4354,
+      "num_input_tokens_seen": 1325904,
+      "step": 17
+    },
+    {
+      "epoch": 0.29906542056074764,
+      "grad_norm": 0.06726882606744766,
+      "learning_rate": 4.9972246874049254e-05,
+      "loss": 0.5439,
+      "num_input_tokens_seen": 1385632,
+      "step": 18
+    },
+    {
+      "epoch": 0.31568016614745587,
+      "grad_norm": 0.05245920270681381,
+      "learning_rate": 4.996907818908987e-05,
+      "loss": 0.3727,
+      "num_input_tokens_seen": 1470632,
+      "step": 19
+    },
+    {
+      "epoch": 0.3322949117341641,
+      "grad_norm": 0.05745376646518707,
+      "learning_rate": 4.996573836886435e-05,
+      "loss": 0.4894,
+      "num_input_tokens_seen": 1547536,
+      "step": 20
+    },
+    {
+      "epoch": 0.34890965732087226,
+      "grad_norm": 0.056607529520988464,
+      "learning_rate": 4.9962227436263453e-05,
+      "loss": 0.3846,
+      "num_input_tokens_seen": 1615528,
+      "step": 21
+    },
+    {
+      "epoch": 0.3655244029075805,
+      "grad_norm": 0.06150667741894722,
+      "learning_rate": 4.995854541535071e-05,
+      "loss": 0.4362,
+      "num_input_tokens_seen": 1694352,
+      "step": 22
+    },
+    {
+      "epoch": 0.3821391484942887,
+      "grad_norm": 0.056484442204236984,
+      "learning_rate": 4.9954692331362294e-05,
+      "loss": 0.4438,
+      "num_input_tokens_seen": 1753776,
+      "step": 23
+    },
+    {
+      "epoch": 0.3987538940809969,
+      "grad_norm": 0.0704159140586853,
+      "learning_rate": 4.995066821070679e-05,
+      "loss": 0.4496,
+      "num_input_tokens_seen": 1809048,
+      "step": 24
+    },
+    {
+      "epoch": 0.4153686396677051,
+      "grad_norm": 0.06202029809355736,
+      "learning_rate": 4.994647308096509e-05,
+      "loss": 0.5096,
+      "num_input_tokens_seen": 1884264,
+      "step": 25
+    },
+    {
+      "epoch": 0.43198338525441327,
+      "grad_norm": 0.04237145930528641,
+      "learning_rate": 4.994210697089014e-05,
+      "loss": 0.3722,
+      "num_input_tokens_seen": 1981704,
+      "step": 26
+    },
+    {
+      "epoch": 0.4485981308411215,
+      "grad_norm": 0.06920398026704788,
+      "learning_rate": 4.9937569910406756e-05,
+      "loss": 0.4103,
+      "num_input_tokens_seen": 2044144,
+      "step": 27
+    },
+    {
+      "epoch": 0.4652128764278297,
+      "grad_norm": 0.062432270497083664,
+      "learning_rate": 4.9932861930611454e-05,
+      "loss": 0.357,
+      "num_input_tokens_seen": 2107584,
+      "step": 28
+    },
+    {
+      "epoch": 0.4818276220145379,
+      "grad_norm": 0.06791180372238159,
+      "learning_rate": 4.9927983063772196e-05,
+      "loss": 0.3889,
+      "num_input_tokens_seen": 2169248,
+      "step": 29
+    },
+    {
+      "epoch": 0.4984423676012461,
+      "grad_norm": 0.07219590991735458,
+      "learning_rate": 4.99229333433282e-05,
+      "loss": 0.3543,
+      "num_input_tokens_seen": 2230344,
+      "step": 30
+    },
+    {
+      "epoch": 0.5150571131879543,
+      "grad_norm": 0.0647474005818367,
+      "learning_rate": 4.9917712803889674e-05,
+      "loss": 0.3453,
+      "num_input_tokens_seen": 2302368,
+      "step": 31
+    },
+    {
+      "epoch": 0.5316718587746625,
+      "grad_norm": 0.07434642314910889,
+      "learning_rate": 4.991232148123761e-05,
+      "loss": 0.435,
+      "num_input_tokens_seen": 2369984,
+      "step": 32
+    },
+    {
+      "epoch": 0.5482866043613707,
+      "grad_norm": 0.05302443355321884,
+      "learning_rate": 4.990675941232353e-05,
+      "loss": 0.3981,
+      "num_input_tokens_seen": 2453032,
+      "step": 33
+    },
+    {
+      "epoch": 0.564901349948079,
+      "grad_norm": 0.053745292127132416,
+      "learning_rate": 4.990102663526924e-05,
+      "loss": 0.3755,
+      "num_input_tokens_seen": 2527464,
+      "step": 34
+    },
+    {
+      "epoch": 0.5815160955347871,
+      "grad_norm": 0.06717613339424133,
+      "learning_rate": 4.989512318936655e-05,
+      "loss": 0.3699,
+      "num_input_tokens_seen": 2597032,
+      "step": 35
+    },
+    {
+      "epoch": 0.5981308411214953,
+      "grad_norm": 0.071847103536129,
+      "learning_rate": 4.9889049115077005e-05,
+      "loss": 0.3705,
+      "num_input_tokens_seen": 2671704,
+      "step": 36
+    },
+    {
+      "epoch": 0.6147455867082036,
+      "grad_norm": 0.0460306741297245,
+      "learning_rate": 4.988280445403164e-05,
+      "loss": 0.3797,
+      "num_input_tokens_seen": 2767640,
+      "step": 37
+    },
+    {
+      "epoch": 0.6313603322949117,
+      "grad_norm": 0.053273387253284454,
+      "learning_rate": 4.987638924903067e-05,
+      "loss": 0.3799,
+      "num_input_tokens_seen": 2843720,
+      "step": 38
+    },
+    {
+      "epoch": 0.6479750778816199,
+      "grad_norm": 0.05600422993302345,
+      "learning_rate": 4.9869803544043166e-05,
+      "loss": 0.2866,
+      "num_input_tokens_seen": 2921472,
+      "step": 39
+    },
+    {
+      "epoch": 0.6645898234683282,
+      "grad_norm": 0.06414052098989487,
+      "learning_rate": 4.9863047384206835e-05,
+      "loss": 0.4115,
+      "num_input_tokens_seen": 2998400,
+      "step": 40
+    },
+    {
+      "epoch": 0.6812045690550363,
+      "grad_norm": 0.09214208275079727,
+      "learning_rate": 4.985612081582764e-05,
+      "loss": 0.3804,
+      "num_input_tokens_seen": 3059648,
+      "step": 41
+    },
+    {
+      "epoch": 0.6978193146417445,
+      "grad_norm": 0.0555964931845665,
+      "learning_rate": 4.98490238863795e-05,
+      "loss": 0.3121,
+      "num_input_tokens_seen": 3140184,
+      "step": 42
+    },
+    {
+      "epoch": 0.7144340602284528,
+      "grad_norm": 0.06256969273090363,
+      "learning_rate": 4.984175664450397e-05,
+      "loss": 0.3271,
+      "num_input_tokens_seen": 3207184,
+      "step": 43
+    },
+    {
+      "epoch": 0.731048805815161,
+      "grad_norm": 0.0543232187628746,
+      "learning_rate": 4.983431914000991e-05,
+      "loss": 0.364,
+      "num_input_tokens_seen": 3292344,
+      "step": 44
+    },
+    {
+      "epoch": 0.7476635514018691,
+      "grad_norm": 0.06077824532985687,
+      "learning_rate": 4.982671142387316e-05,
+      "loss": 0.3894,
+      "num_input_tokens_seen": 3365384,
+      "step": 45
+    },
+    {
+      "epoch": 0.7642782969885774,
+      "grad_norm": 0.06091070920228958,
+      "learning_rate": 4.981893354823614e-05,
+      "loss": 0.3354,
+      "num_input_tokens_seen": 3440720,
+      "step": 46
+    },
+    {
+      "epoch": 0.7808930425752856,
+      "grad_norm": 0.054153311997652054,
+      "learning_rate": 4.9810985566407544e-05,
+      "loss": 0.3058,
+      "num_input_tokens_seen": 3533576,
+      "step": 47
+    },
+    {
+      "epoch": 0.7975077881619937,
+      "grad_norm": 0.06662417948246002,
+      "learning_rate": 4.980286753286195e-05,
+      "loss": 0.4658,
+      "num_input_tokens_seen": 3599744,
+      "step": 48
+    },
+    {
+      "epoch": 0.814122533748702,
+      "grad_norm": 0.05790851265192032,
+      "learning_rate": 4.979457950323945e-05,
+      "loss": 0.3647,
+      "num_input_tokens_seen": 3689520,
+      "step": 49
+    },
+    {
+      "epoch": 0.8307372793354102,
+      "grad_norm": 0.10742159187793732,
+      "learning_rate": 4.9786121534345265e-05,
+      "loss": 0.343,
+      "num_input_tokens_seen": 3751808,
+      "step": 50
+    },
+    {
+      "epoch": 0.8473520249221184,
+      "grad_norm": 0.05565556138753891,
+      "learning_rate": 4.9777493684149375e-05,
+      "loss": 0.3317,
+      "num_input_tokens_seen": 3839096,
+      "step": 51
+    },
+    {
+      "epoch": 0.8639667705088265,
+      "grad_norm": 0.05752381682395935,
+      "learning_rate": 4.976869601178609e-05,
+      "loss": 0.38,
+      "num_input_tokens_seen": 3919824,
+      "step": 52
+    },
+    {
+      "epoch": 0.8805815160955348,
+      "grad_norm": 0.06406434625387192,
+      "learning_rate": 4.975972857755369e-05,
+      "loss": 0.2676,
+      "num_input_tokens_seen": 3989312,
+      "step": 53
+    },
+    {
+      "epoch": 0.897196261682243,
+      "grad_norm": 0.0653691440820694,
+      "learning_rate": 4.975059144291394e-05,
+      "loss": 0.3516,
+      "num_input_tokens_seen": 4060528,
+      "step": 54
+    },
+    {
+      "epoch": 0.9138110072689511,
+      "grad_norm": 0.06272953748703003,
+      "learning_rate": 4.974128467049176e-05,
+      "loss": 0.3004,
+      "num_input_tokens_seen": 4129368,
+      "step": 55
+    },
+    {
+      "epoch": 0.9304257528556594,
+      "grad_norm": 0.08054930716753006,
+      "learning_rate": 4.9731808324074717e-05,
+      "loss": 0.3009,
+      "num_input_tokens_seen": 4175208,
+      "step": 56
+    },
+    {
+      "epoch": 0.9470404984423676,
+      "grad_norm": 0.07523038238286972,
+      "learning_rate": 4.972216246861262e-05,
+      "loss": 0.2814,
+      "num_input_tokens_seen": 4218096,
+      "step": 57
+    },
+    {
+      "epoch": 0.9636552440290758,
+      "grad_norm": 0.07347433269023895,
+      "learning_rate": 4.971234717021709e-05,
+      "loss": 0.3321,
+      "num_input_tokens_seen": 4275968,
+      "step": 58
+    },
+    {
+      "epoch": 0.980269989615784,
+      "grad_norm": 0.05830248445272446,
+      "learning_rate": 4.9702362496161085e-05,
+      "loss": 0.2881,
+      "num_input_tokens_seen": 4346616,
+      "step": 59
+    },
+    {
+      "epoch": 0.9968847352024922,
+      "grad_norm": 0.061629410833120346,
+      "learning_rate": 4.9692208514878444e-05,
+      "loss": 0.2993,
+      "num_input_tokens_seen": 4425064,
+      "step": 60
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.13380740582942963,
+      "learning_rate": 4.968188529596342e-05,
+      "loss": 0.2511,
+      "num_input_tokens_seen": 4435328,
+      "step": 61
+    },
+    {
+      "epoch": 1.0166147455867083,
+      "grad_norm": 0.0726238414645195,
+      "learning_rate": 4.9671392910170185e-05,
+      "loss": 0.3127,
+      "num_input_tokens_seen": 4500104,
+      "step": 62
+    },
+    {
+      "epoch": 1.0332294911734163,
+      "grad_norm": 0.05980083718895912,
+      "learning_rate": 4.966073142941239e-05,
+      "loss": 0.3601,
+      "num_input_tokens_seen": 4581976,
+      "step": 63
+    },
+    {
+      "epoch": 1.0498442367601246,
+      "grad_norm": 0.06445376574993134,
+      "learning_rate": 4.964990092676263e-05,
+      "loss": 0.3049,
+      "num_input_tokens_seen": 4652160,
+      "step": 64
+    },
+    {
+      "epoch": 1.066458982346833,
+      "grad_norm": 0.07824505120515823,
+      "learning_rate": 4.9638901476451946e-05,
+      "loss": 0.3099,
+      "num_input_tokens_seen": 4709368,
+      "step": 65
+    },
+    {
+      "epoch": 1.083073727933541,
+      "grad_norm": 0.058268457651138306,
+      "learning_rate": 4.962773315386935e-05,
+      "loss": 0.3273,
+      "num_input_tokens_seen": 4798256,
+      "step": 66
+    },
+    {
+      "epoch": 1.0996884735202492,
+      "grad_norm": 0.07069691270589828,
+      "learning_rate": 4.961639603556127e-05,
+      "loss": 0.282,
+      "num_input_tokens_seen": 4859200,
+      "step": 67
+    },
+    {
+      "epoch": 1.1163032191069575,
+      "grad_norm": 0.0775996670126915,
+      "learning_rate": 4.960489019923105e-05,
+      "loss": 0.3642,
+      "num_input_tokens_seen": 4925992,
+      "step": 68
+    },
+    {
+      "epoch": 1.1329179646936656,
+      "grad_norm": 0.07044171541929245,
+      "learning_rate": 4.9593215723738404e-05,
+      "loss": 0.2896,
+      "num_input_tokens_seen": 4998808,
+      "step": 69
+    },
+    {
+      "epoch": 1.1495327102803738,
+      "grad_norm": 0.05971802771091461,
+      "learning_rate": 4.958137268909887e-05,
+      "loss": 0.2578,
+      "num_input_tokens_seen": 5089672,
+      "step": 70
+    },
+    {
+      "epoch": 1.1661474558670821,
+      "grad_norm": 0.07145556062459946,
+      "learning_rate": 4.9569361176483286e-05,
+      "loss": 0.3243,
+      "num_input_tokens_seen": 5166744,
+      "step": 71
+    },
+    {
+      "epoch": 1.1827622014537902,
+      "grad_norm": 0.07455787807703018,
+      "learning_rate": 4.9557181268217227e-05,
+      "loss": 0.3949,
+      "num_input_tokens_seen": 5228264,
+      "step": 72
+    },
+    {
+      "epoch": 1.1993769470404985,
+      "grad_norm": 0.055582575500011444,
+      "learning_rate": 4.9544833047780394e-05,
+      "loss": 0.2877,
+      "num_input_tokens_seen": 5338224,
+      "step": 73
+    },
+    {
+      "epoch": 1.2159916926272065,
+      "grad_norm": 0.07675391435623169,
+      "learning_rate": 4.9532316599806124e-05,
+      "loss": 0.3152,
+      "num_input_tokens_seen": 5399848,
+      "step": 74
+    },
+    {
+      "epoch": 1.2326064382139148,
+      "grad_norm": 0.08048644661903381,
+      "learning_rate": 4.951963201008076e-05,
+      "loss": 0.2976,
+      "num_input_tokens_seen": 5468624,
+      "step": 75
+    },
+    {
+      "epoch": 1.249221183800623,
+      "grad_norm": 0.07579060643911362,
+      "learning_rate": 4.9506779365543046e-05,
+      "loss": 0.2982,
+      "num_input_tokens_seen": 5536776,
+      "step": 76
+    },
+    {
+      "epoch": 1.2658359293873311,
+      "grad_norm": 0.07828006893396378,
+      "learning_rate": 4.949375875428357e-05,
+      "loss": 0.3272,
+      "num_input_tokens_seen": 5609296,
+      "step": 77
+    },
+    {
+      "epoch": 1.2824506749740394,
+      "grad_norm": 0.08079098165035248,
+      "learning_rate": 4.9480570265544144e-05,
+      "loss": 0.2768,
+      "num_input_tokens_seen": 5663824,
+      "step": 78
+    },
+    {
+      "epoch": 1.2990654205607477,
+      "grad_norm": 0.07579358667135239,
+      "learning_rate": 4.94672139897172e-05,
+      "loss": 0.318,
+      "num_input_tokens_seen": 5742032,
+      "step": 79
+    },
+    {
+      "epoch": 1.3156801661474558,
+      "grad_norm": 0.07588379085063934,
+      "learning_rate": 4.9453690018345144e-05,
+      "loss": 0.3007,
+      "num_input_tokens_seen": 5816864,
+      "step": 80
+    },
+    {
+      "epoch": 1.332294911734164,
+      "grad_norm": 0.08709035068750381,
+      "learning_rate": 4.943999844411977e-05,
+      "loss": 0.2797,
+      "num_input_tokens_seen": 5881624,
+      "step": 81
+    },
+    {
+      "epoch": 1.3489096573208723,
+      "grad_norm": 0.05975884944200516,
+      "learning_rate": 4.94261393608816e-05,
+      "loss": 0.2591,
+      "num_input_tokens_seen": 5970272,
+      "step": 82
+    },
+    {
+      "epoch": 1.3655244029075804,
+      "grad_norm": 0.07372818142175674,
+      "learning_rate": 4.941211286361922e-05,
+      "loss": 0.2687,
+      "num_input_tokens_seen": 6058752,
+      "step": 83
+    },
+    {
+      "epoch": 1.3821391484942886,
+      "grad_norm": 0.09071576595306396,
+      "learning_rate": 4.939791904846869e-05,
+      "loss": 0.2979,
+      "num_input_tokens_seen": 6120064,
+      "step": 84
+    },
+    {
+      "epoch": 1.398753894080997,
+      "grad_norm": 0.0849960595369339,
+      "learning_rate": 4.938355801271282e-05,
+      "loss": 0.2927,
+      "num_input_tokens_seen": 6182072,
+      "step": 85
+    },
+    {
+      "epoch": 1.415368639667705,
+      "grad_norm": 0.08258760720491409,
+      "learning_rate": 4.936902985478055e-05,
+      "loss": 0.295,
+      "num_input_tokens_seen": 6269680,
+      "step": 86
+    },
+    {
+      "epoch": 1.4319833852544133,
+      "grad_norm": 0.0851503536105156,
+      "learning_rate": 4.935433467424624e-05,
+      "loss": 0.2925,
+      "num_input_tokens_seen": 6347424,
+      "step": 87
+    },
+    {
+      "epoch": 1.4485981308411215,
+      "grad_norm": 0.08852345496416092,
+      "learning_rate": 4.933947257182901e-05,
+      "loss": 0.3153,
+      "num_input_tokens_seen": 6412584,
+      "step": 88
+    },
+    {
+      "epoch": 1.4652128764278296,
+      "grad_norm": 0.08184897154569626,
+      "learning_rate": 4.932444364939205e-05,
+      "loss": 0.292,
+      "num_input_tokens_seen": 6482728,
+      "step": 89
+    },
+    {
+      "epoch": 1.4818276220145379,
+      "grad_norm": 0.08270515501499176,
+      "learning_rate": 4.9309248009941914e-05,
+      "loss": 0.3472,
+      "num_input_tokens_seen": 6562104,
+      "step": 90
+    },
+    {
+      "epoch": 1.4984423676012462,
+      "grad_norm": 0.07407747954130173,
+      "learning_rate": 4.929388575762782e-05,
+      "loss": 0.2995,
+      "num_input_tokens_seen": 6656552,
+      "step": 91
+    },
+    {
+      "epoch": 1.5150571131879542,
+      "grad_norm": 0.08710360527038574,
+      "learning_rate": 4.9278356997740904e-05,
+      "loss": 0.2549,
+      "num_input_tokens_seen": 6714184,
+      "step": 92
+    },
+    {
+      "epoch": 1.5316718587746625,
+      "grad_norm": 0.0773790255188942,
+      "learning_rate": 4.9262661836713564e-05,
+      "loss": 0.2814,
+      "num_input_tokens_seen": 6793552,
+      "step": 93
+    },
+    {
+      "epoch": 1.5482866043613708,
+      "grad_norm": 0.1002134457230568,
+      "learning_rate": 4.924680038211867e-05,
+      "loss": 0.2876,
+      "num_input_tokens_seen": 6865256,
+      "step": 94
+    },
+    {
+      "epoch": 1.5649013499480788,
+      "grad_norm": 0.09670394659042358,
+      "learning_rate": 4.9230772742668866e-05,
+      "loss": 0.2846,
+      "num_input_tokens_seen": 6931152,
+      "step": 95
+    },
+    {
+      "epoch": 1.5815160955347871,
+      "grad_norm": 0.08910100907087326,
+      "learning_rate": 4.9214579028215776e-05,
+      "loss": 0.2944,
+      "num_input_tokens_seen": 6998408,
+      "step": 96
+    },
+    {
+      "epoch": 1.5981308411214954,
+      "grad_norm": 0.09202459454536438,
+      "learning_rate": 4.919821934974933e-05,
+      "loss": 0.251,
+      "num_input_tokens_seen": 7053008,
+      "step": 97
+    },
+    {
+      "epoch": 1.6147455867082035,
+      "grad_norm": 0.10218881815671921,
+      "learning_rate": 4.918169381939692e-05,
+      "loss": 0.2851,
+      "num_input_tokens_seen": 7106440,
+      "step": 98
+    },
+    {
+      "epoch": 1.6313603322949117,
+      "grad_norm": 0.09290914982557297,
+      "learning_rate": 4.916500255042268e-05,
+      "loss": 0.2959,
+      "num_input_tokens_seen": 7167032,
+      "step": 99
+    },
+    {
+      "epoch": 1.64797507788162,
+      "grad_norm": 0.07791033387184143,
+      "learning_rate": 4.914814565722671e-05,
+      "loss": 0.2481,
+      "num_input_tokens_seen": 7245720,
+      "step": 100
+    },
+    {
+      "epoch": 1.664589823468328,
+      "grad_norm": 0.08885534107685089,
+      "learning_rate": 4.913112325534426e-05,
+      "loss": 0.3168,
+      "num_input_tokens_seen": 7326320,
+      "step": 101
+    },
+    {
+      "epoch": 1.6812045690550363,
+      "grad_norm": 0.08569750934839249,
+      "learning_rate": 4.9113935461444955e-05,
+      "loss": 0.2805,
+      "num_input_tokens_seen": 7442232,
+      "step": 102
+    },
+    {
+      "epoch": 1.6978193146417446,
+      "grad_norm": 0.1112508773803711,
+      "learning_rate": 4.9096582393332025e-05,
+      "loss": 0.2675,
+      "num_input_tokens_seen": 7502496,
+      "step": 103
+    },
+    {
+      "epoch": 1.7144340602284527,
+      "grad_norm": 0.09654372185468674,
+      "learning_rate": 4.907906416994146e-05,
+      "loss": 0.3038,
+      "num_input_tokens_seen": 7566496,
+      "step": 104
+    },
+    {
+      "epoch": 1.731048805815161,
+      "grad_norm": 0.10022995620965958,
+      "learning_rate": 4.906138091134118e-05,
+      "loss": 0.3639,
+      "num_input_tokens_seen": 7629056,
+      "step": 105
+    },
+    {
+      "epoch": 1.7476635514018692,
+      "grad_norm": 0.08336564153432846,
+      "learning_rate": 4.9043532738730284e-05,
+      "loss": 0.2944,
+      "num_input_tokens_seen": 7706096,
+      "step": 106
+    },
+    {
+      "epoch": 1.7642782969885773,
+      "grad_norm": 0.08539658784866333,
+      "learning_rate": 4.9025519774438136e-05,
+      "loss": 0.2392,
+      "num_input_tokens_seen": 7780072,
+      "step": 107
+    },
+    {
+      "epoch": 1.7808930425752856,
+      "grad_norm": 0.09139693528413773,
+      "learning_rate": 4.900734214192358e-05,
+      "loss": 0.2685,
+      "num_input_tokens_seen": 7857712,
+      "step": 108
+    },
+    {
+      "epoch": 1.7975077881619939,
+      "grad_norm": 0.1043916717171669,
+      "learning_rate": 4.898899996577407e-05,
+      "loss": 0.2513,
+      "num_input_tokens_seen": 7916832,
+      "step": 109
+    },
+    {
+      "epoch": 1.814122533748702,
+      "grad_norm": 0.09203662723302841,
+      "learning_rate": 4.8970493371704826e-05,
+      "loss": 0.2974,
+      "num_input_tokens_seen": 7993056,
+      "step": 110
+    },
+    {
+      "epoch": 1.8307372793354102,
+      "grad_norm": 0.09319474548101425,
+      "learning_rate": 4.8951822486557986e-05,
+      "loss": 0.3096,
+      "num_input_tokens_seen": 8090056,
+      "step": 111
+    },
+    {
+      "epoch": 1.8473520249221185,
+      "grad_norm": 0.10193445533514023,
+      "learning_rate": 4.893298743830168e-05,
+      "loss": 0.2633,
+      "num_input_tokens_seen": 8164808,
+      "step": 112
+    },
+    {
+      "epoch": 1.8639667705088265,
+      "grad_norm": 0.11407948285341263,
+      "learning_rate": 4.891398835602925e-05,
+      "loss": 0.2584,
+      "num_input_tokens_seen": 8223568,
+      "step": 113
+    },
+    {
+      "epoch": 1.8805815160955348,
+      "grad_norm": 0.11977085471153259,
+      "learning_rate": 4.8894825369958255e-05,
+      "loss": 0.2619,
+      "num_input_tokens_seen": 8276160,
+      "step": 114
+    },
+    {
+      "epoch": 1.897196261682243,
+      "grad_norm": 0.10925433784723282,
+      "learning_rate": 4.8875498611429674e-05,
+      "loss": 0.2762,
+      "num_input_tokens_seen": 8354904,
+      "step": 115
+    },
+    {
+      "epoch": 1.9138110072689511,
+      "grad_norm": 0.09673939645290375,
+      "learning_rate": 4.8856008212906925e-05,
+      "loss": 0.3152,
+      "num_input_tokens_seen": 8442584,
+      "step": 116
+    },
+    {
+      "epoch": 1.9304257528556594,
+      "grad_norm": 0.10827789455652237,
+      "learning_rate": 4.8836354307975026e-05,
+      "loss": 0.2759,
+      "num_input_tokens_seen": 8506688,
+      "step": 117
+    },
+    {
+      "epoch": 1.9470404984423677,
+      "grad_norm": 0.08390220254659653,
+      "learning_rate": 4.881653703133966e-05,
+      "loss": 0.2192,
+      "num_input_tokens_seen": 8610712,
+      "step": 118
+    },
+    {
+      "epoch": 1.9636552440290758,
+      "grad_norm": 0.09252211451530457,
+      "learning_rate": 4.87965565188262e-05,
+      "loss": 0.2618,
+      "num_input_tokens_seen": 8692624,
+      "step": 119
+    },
+    {
+      "epoch": 1.980269989615784,
+      "grad_norm": 0.1107102632522583,
+      "learning_rate": 4.877641290737884e-05,
+      "loss": 0.2666,
+      "num_input_tokens_seen": 8772208,
+      "step": 120
+    },
+    {
+      "epoch": 1.9968847352024923,
+      "grad_norm": 0.0917077362537384,
+      "learning_rate": 4.8756106335059646e-05,
+      "loss": 0.253,
+      "num_input_tokens_seen": 8854904,
+      "step": 121
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2606711685657501,
+      "learning_rate": 4.87356369410476e-05,
+      "loss": 0.235,
+      "num_input_tokens_seen": 8872656,
+      "step": 122
+    },
+    {
+      "epoch": 2.016614745586708,
+      "grad_norm": 0.10363993793725967,
+      "learning_rate": 4.8715004865637614e-05,
+      "loss": 0.266,
+      "num_input_tokens_seen": 8946480,
+      "step": 123
+    },
+    {
+      "epoch": 2.0332294911734166,
+      "grad_norm": 0.09997844696044922,
+      "learning_rate": 4.869421025023965e-05,
+      "loss": 0.2696,
+      "num_input_tokens_seen": 9023328,
+      "step": 124
+    },
+    {
+      "epoch": 2.0498442367601246,
+      "grad_norm": 0.13349319994449615,
+      "learning_rate": 4.867325323737765e-05,
+      "loss": 0.2552,
+      "num_input_tokens_seen": 9074320,
+      "step": 125
+    },
+    {
+      "epoch": 2.0664589823468327,
+      "grad_norm": 0.11201464384794235,
+      "learning_rate": 4.8652133970688636e-05,
+      "loss": 0.2486,
+      "num_input_tokens_seen": 9148784,
+      "step": 126
+    },
+    {
+      "epoch": 2.083073727933541,
+      "grad_norm": 0.10193142294883728,
+      "learning_rate": 4.8630852594921706e-05,
+      "loss": 0.2814,
+      "num_input_tokens_seen": 9246624,
+      "step": 127
+    },
+    {
+      "epoch": 2.0996884735202492,
+      "grad_norm": 0.1305130422115326,
+      "learning_rate": 4.860940925593703e-05,
+      "loss": 0.304,
+      "num_input_tokens_seen": 9328176,
+      "step": 128
+    },
+    {
+      "epoch": 2.1163032191069573,
+      "grad_norm": 0.1137692779302597,
+      "learning_rate": 4.8587804100704845e-05,
+      "loss": 0.2427,
+      "num_input_tokens_seen": 9388936,
+      "step": 129
+    },
+    {
+      "epoch": 2.132917964693666,
+      "grad_norm": 0.12126237154006958,
+      "learning_rate": 4.856603727730447e-05,
+      "loss": 0.2485,
+      "num_input_tokens_seen": 9461664,
+      "step": 130
+    },
+    {
+      "epoch": 2.149532710280374,
+      "grad_norm": 0.11567176878452301,
+      "learning_rate": 4.854410893492326e-05,
+      "loss": 0.2628,
+      "num_input_tokens_seen": 9535000,
+      "step": 131
+    },
+    {
+      "epoch": 2.166147455867082,
+      "grad_norm": 0.1399552971124649,
+      "learning_rate": 4.852201922385564e-05,
+      "loss": 0.2518,
+      "num_input_tokens_seen": 9600296,
+      "step": 132
+    },
+    {
+      "epoch": 2.1827622014537904,
+      "grad_norm": 0.13912151753902435,
+      "learning_rate": 4.8499768295502004e-05,
+      "loss": 0.2429,
+      "num_input_tokens_seen": 9686784,
+      "step": 133
+    },
+    {
+      "epoch": 2.1993769470404985,
+      "grad_norm": 0.11130474507808685,
+      "learning_rate": 4.847735630236773e-05,
+      "loss": 0.2775,
+      "num_input_tokens_seen": 9781112,
+      "step": 134
+    },
+    {
+      "epoch": 2.2159916926272065,
+      "grad_norm": 0.12169156968593597,
+      "learning_rate": 4.8454783398062106e-05,
+      "loss": 0.2439,
+      "num_input_tokens_seen": 9849528,
+      "step": 135
+    },
+    {
+      "epoch": 2.232606438213915,
+      "grad_norm": 0.11766713112592697,
+      "learning_rate": 4.843204973729729e-05,
+      "loss": 0.2538,
+      "num_input_tokens_seen": 9931080,
+      "step": 136
+    },
+    {
+      "epoch": 2.249221183800623,
+      "grad_norm": 0.11854218691587448,
+      "learning_rate": 4.840915547588725e-05,
+      "loss": 0.2782,
+      "num_input_tokens_seen": 10011176,
+      "step": 137
+    },
+    {
+      "epoch": 2.265835929387331,
+      "grad_norm": 0.1340581178665161,
+      "learning_rate": 4.838610077074669e-05,
+      "loss": 0.248,
+      "num_input_tokens_seen": 10084128,
+      "step": 138
+    },
+    {
+      "epoch": 2.2824506749740396,
+      "grad_norm": 0.12075436115264893,
+      "learning_rate": 4.836288577988996e-05,
+      "loss": 0.2582,
+      "num_input_tokens_seen": 10155536,
+      "step": 139
+    },
+    {
+      "epoch": 2.2990654205607477,
+      "grad_norm": 0.10599923878908157,
+      "learning_rate": 4.8339510662430046e-05,
+      "loss": 0.2199,
+      "num_input_tokens_seen": 10251160,
+      "step": 140
+    },
+    {
+      "epoch": 2.3156801661474558,
+      "grad_norm": 0.1117846742272377,
+      "learning_rate": 4.8315975578577355e-05,
+      "loss": 0.2324,
+      "num_input_tokens_seen": 10345864,
+      "step": 141
+    },
+    {
+      "epoch": 2.3322949117341643,
+      "grad_norm": 0.13972057402133942,
+      "learning_rate": 4.8292280689638725e-05,
+      "loss": 0.4072,
+      "num_input_tokens_seen": 10417616,
+      "step": 142
+    },
+    {
+      "epoch": 2.3489096573208723,
+      "grad_norm": 0.13837860524654388,
+      "learning_rate": 4.826842615801628e-05,
+      "loss": 0.2607,
+      "num_input_tokens_seen": 10481816,
+      "step": 143
+    },
+    {
+      "epoch": 2.3655244029075804,
+      "grad_norm": 0.14040137827396393,
+      "learning_rate": 4.8244412147206284e-05,
+      "loss": 0.3094,
+      "num_input_tokens_seen": 10562056,
+      "step": 144
+    },
+    {
+      "epoch": 2.382139148494289,
+      "grad_norm": 0.1393299251794815,
+      "learning_rate": 4.822023882179811e-05,
+      "loss": 0.2407,
+      "num_input_tokens_seen": 10612808,
+      "step": 145
+    },
+    {
+      "epoch": 2.398753894080997,
+      "grad_norm": 0.13878698647022247,
+      "learning_rate": 4.8195906347473e-05,
+      "loss": 0.2481,
+      "num_input_tokens_seen": 10682328,
+      "step": 146
+    },
+    {
+      "epoch": 2.415368639667705,
+      "grad_norm": 0.10430227965116501,
+      "learning_rate": 4.817141489100302e-05,
+      "loss": 0.2528,
+      "num_input_tokens_seen": 10771912,
+      "step": 147
+    },
+    {
+      "epoch": 2.431983385254413,
+      "grad_norm": 0.12963703274726868,
+      "learning_rate": 4.814676462024988e-05,
+      "loss": 0.2739,
+      "num_input_tokens_seen": 10842232,
+      "step": 148
+    },
+    {
+      "epoch": 2.4485981308411215,
+      "grad_norm": 0.13274963200092316,
+      "learning_rate": 4.8121955704163745e-05,
+      "loss": 0.2407,
+      "num_input_tokens_seen": 10902264,
+      "step": 149
+    },
+    {
+      "epoch": 2.4652128764278296,
+      "grad_norm": 0.11079717427492142,
+      "learning_rate": 4.8096988312782174e-05,
+      "loss": 0.2142,
+      "num_input_tokens_seen": 10992744,
+      "step": 150
+    },
+    {
+      "epoch": 2.4818276220145377,
+      "grad_norm": 0.08429212868213654,
+      "learning_rate": 4.8071862617228855e-05,
+      "loss": 0.1428,
+      "num_input_tokens_seen": 11090064,
+      "step": 151
+    },
+    {
+      "epoch": 2.498442367601246,
+      "grad_norm": 0.12903761863708496,
+      "learning_rate": 4.8046578789712515e-05,
+      "loss": 0.2268,
+      "num_input_tokens_seen": 11162864,
+      "step": 152
+    },
+    {
+      "epoch": 2.515057113187954,
+      "grad_norm": 0.14638672769069672,
+      "learning_rate": 4.8021137003525664e-05,
+      "loss": 0.2388,
+      "num_input_tokens_seen": 11224368,
+      "step": 153
+    },
+    {
+      "epoch": 2.5316718587746623,
+      "grad_norm": 0.1372838169336319,
+      "learning_rate": 4.7995537433043446e-05,
+      "loss": 0.2588,
+      "num_input_tokens_seen": 11291056,
+      "step": 154
+    },
+    {
+      "epoch": 2.5482866043613708,
+      "grad_norm": 0.15665481984615326,
+      "learning_rate": 4.796978025372246e-05,
+      "loss": 0.2225,
+      "num_input_tokens_seen": 11345464,
+      "step": 155
+    },
+    {
+      "epoch": 2.564901349948079,
+      "grad_norm": 0.13234855234622955,
+      "learning_rate": 4.794386564209953e-05,
+      "loss": 0.275,
+      "num_input_tokens_seen": 11418912,
+      "step": 156
+    },
+    {
+      "epoch": 2.581516095534787,
+      "grad_norm": 0.13585953414440155,
+      "learning_rate": 4.79177937757905e-05,
+      "loss": 0.2407,
+      "num_input_tokens_seen": 11491216,
+      "step": 157
+    },
+    {
+      "epoch": 2.5981308411214954,
+      "grad_norm": 0.1423913538455963,
+      "learning_rate": 4.7891564833489035e-05,
+      "loss": 0.1971,
+      "num_input_tokens_seen": 11558016,
+      "step": 158
+    },
+    {
+      "epoch": 2.6147455867082035,
+      "grad_norm": 0.13013511896133423,
+      "learning_rate": 4.7865178994965344e-05,
+      "loss": 0.2362,
+      "num_input_tokens_seen": 11630432,
+      "step": 159
+    },
+    {
+      "epoch": 2.6313603322949115,
+      "grad_norm": 0.1587141752243042,
+      "learning_rate": 4.783863644106502e-05,
+      "loss": 0.2252,
+      "num_input_tokens_seen": 11684624,
+      "step": 160
+    },
+    {
+      "epoch": 2.64797507788162,
+      "grad_norm": 0.12592960894107819,
+      "learning_rate": 4.781193735370777e-05,
+      "loss": 0.2506,
+      "num_input_tokens_seen": 11770232,
+      "step": 161
+    },
+    {
+      "epoch": 2.664589823468328,
+      "grad_norm": 0.1583249419927597,
+      "learning_rate": 4.7785081915886134e-05,
+      "loss": 0.2352,
+      "num_input_tokens_seen": 11828360,
+      "step": 162
+    },
+    {
+      "epoch": 2.681204569055036,
+      "grad_norm": 0.14881783723831177,
+      "learning_rate": 4.775807031166428e-05,
+      "loss": 0.2308,
+      "num_input_tokens_seen": 11915944,
+      "step": 163
+    },
+    {
+      "epoch": 2.6978193146417446,
+      "grad_norm": 0.1607823222875595,
+      "learning_rate": 4.773090272617672e-05,
+      "loss": 0.2238,
+      "num_input_tokens_seen": 11981792,
+      "step": 164
+    },
+    {
+      "epoch": 2.7144340602284527,
+      "grad_norm": 0.13583113253116608,
+      "learning_rate": 4.7703579345627035e-05,
+      "loss": 0.3196,
+      "num_input_tokens_seen": 12044024,
+      "step": 165
+    },
+    {
+      "epoch": 2.7310488058151607,
+      "grad_norm": 0.19167298078536987,
+      "learning_rate": 4.7676100357286624e-05,
+      "loss": 0.2745,
+      "num_input_tokens_seen": 12093424,
+      "step": 166
+    },
+    {
+      "epoch": 2.7476635514018692,
+      "grad_norm": 0.130703404545784,
+      "learning_rate": 4.76484659494934e-05,
+      "loss": 0.2285,
+      "num_input_tokens_seen": 12167792,
+      "step": 167
+    },
+    {
+      "epoch": 2.7642782969885773,
+      "grad_norm": 0.14331185817718506,
+      "learning_rate": 4.762067631165049e-05,
+      "loss": 0.2506,
+      "num_input_tokens_seen": 12233712,
+      "step": 168
+    },
+    {
+      "epoch": 2.7808930425752854,
+      "grad_norm": 0.12700341641902924,
+      "learning_rate": 4.7592731634224966e-05,
+      "loss": 0.2052,
+      "num_input_tokens_seen": 12310544,
+      "step": 169
+    },
+    {
+      "epoch": 2.797507788161994,
+      "grad_norm": 0.15118420124053955,
+      "learning_rate": 4.756463210874652e-05,
+      "loss": 0.2309,
+      "num_input_tokens_seen": 12400160,
+      "step": 170
+    },
+    {
+      "epoch": 2.814122533748702,
+      "grad_norm": 0.14001020789146423,
+      "learning_rate": 4.753637792780614e-05,
+      "loss": 0.2544,
+      "num_input_tokens_seen": 12480432,
+      "step": 171
+    },
+    {
+      "epoch": 2.83073727933541,
+      "grad_norm": 0.12076311558485031,
+      "learning_rate": 4.7507969285054845e-05,
+      "loss": 0.2434,
+      "num_input_tokens_seen": 12568064,
+      "step": 172
+    },
+    {
+      "epoch": 2.8473520249221185,
+      "grad_norm": 0.16462342441082,
+      "learning_rate": 4.7479406375202264e-05,
+      "loss": 0.2417,
+      "num_input_tokens_seen": 12647400,
+      "step": 173
+    },
+    {
+      "epoch": 2.8639667705088265,
+      "grad_norm": 0.17294971644878387,
+      "learning_rate": 4.745068939401539e-05,
+      "loss": 0.2121,
+      "num_input_tokens_seen": 12698208,
+      "step": 174
+    },
+    {
+      "epoch": 2.8805815160955346,
+      "grad_norm": 0.16743803024291992,
+      "learning_rate": 4.742181853831721e-05,
+      "loss": 0.2238,
+      "num_input_tokens_seen": 12758528,
+      "step": 175
+    },
+    {
+      "epoch": 2.897196261682243,
+      "grad_norm": 0.14583320915699005,
+      "learning_rate": 4.7392794005985326e-05,
+      "loss": 0.2333,
+      "num_input_tokens_seen": 12837264,
+      "step": 176
+    },
+    {
+      "epoch": 2.913811007268951,
+      "grad_norm": 0.1509270817041397,
+      "learning_rate": 4.7363615995950626e-05,
+      "loss": 0.2179,
+      "num_input_tokens_seen": 12902368,
+      "step": 177
+    },
+    {
+      "epoch": 2.930425752855659,
+      "grad_norm": 0.12910738587379456,
+      "learning_rate": 4.733428470819594e-05,
+      "loss": 0.2144,
+      "num_input_tokens_seen": 12974296,
+      "step": 178
+    },
+    {
+      "epoch": 2.9470404984423677,
+      "grad_norm": 0.142000213265419,
+      "learning_rate": 4.730480034375462e-05,
+      "loss": 0.2413,
+      "num_input_tokens_seen": 13057280,
+      "step": 179
+    },
+    {
+      "epoch": 2.9636552440290758,
+      "grad_norm": 0.131468266248703,
+      "learning_rate": 4.72751631047092e-05,
+      "loss": 0.294,
+      "num_input_tokens_seen": 13158232,
+      "step": 180
+    },
+    {
+      "epoch": 2.980269989615784,
+      "grad_norm": 0.1529342085123062,
+      "learning_rate": 4.7245373194189994e-05,
+      "loss": 0.216,
+      "num_input_tokens_seen": 13229840,
+      "step": 181
+    },
+    {
+      "epoch": 2.9968847352024923,
+      "grad_norm": 0.1573815941810608,
+      "learning_rate": 4.7215430816373726e-05,
+      "loss": 0.2384,
+      "num_input_tokens_seen": 13296520,
+      "step": 182
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.2532118558883667,
+      "learning_rate": 4.718533617648209e-05,
+      "loss": 0.1459,
+      "num_input_tokens_seen": 13309672,
+      "step": 183
+    },
+    {
+      "epoch": 3.016614745586708,
+      "grad_norm": 0.16963432729244232,
+      "learning_rate": 4.715508948078037e-05,
+      "loss": 0.1985,
+      "num_input_tokens_seen": 13371544,
+      "step": 184
+    },
+    {
+      "epoch": 3.0332294911734166,
+      "grad_norm": 0.18877384066581726,
+      "learning_rate": 4.712469093657605e-05,
+      "loss": 0.1856,
+      "num_input_tokens_seen": 13432984,
+      "step": 185
+    },
+    {
+      "epoch": 3.0498442367601246,
+      "grad_norm": 0.14922884106636047,
+      "learning_rate": 4.709414075221734e-05,
+      "loss": 0.2385,
+      "num_input_tokens_seen": 13500016,
+      "step": 186
+    },
+    {
+      "epoch": 3.0664589823468327,
+      "grad_norm": 0.2028326541185379,
+      "learning_rate": 4.706343913709178e-05,
+      "loss": 0.2227,
+      "num_input_tokens_seen": 13579672,
+      "step": 187
+    },
+    {
+      "epoch": 3.083073727933541,
+      "grad_norm": 0.19964616000652313,
+      "learning_rate": 4.70325863016248e-05,
+      "loss": 0.2045,
+      "num_input_tokens_seen": 13630704,
+      "step": 188
+    },
+    {
+      "epoch": 3.0996884735202492,
+      "grad_norm": 0.1594657599925995,
+      "learning_rate": 4.7001582457278304e-05,
+      "loss": 0.2648,
+      "num_input_tokens_seen": 13695472,
+      "step": 189
+    },
+    {
+      "epoch": 3.1163032191069573,
+      "grad_norm": 0.16952532529830933,
+      "learning_rate": 4.697042781654913e-05,
+      "loss": 0.22,
+      "num_input_tokens_seen": 13767792,
+      "step": 190
+    },
+    {
+      "epoch": 3.132917964693666,
+      "grad_norm": 0.16775831580162048,
+      "learning_rate": 4.693912259296773e-05,
+      "loss": 0.2667,
+      "num_input_tokens_seen": 13857352,
+      "step": 191
+    },
+    {
+      "epoch": 3.149532710280374,
+      "grad_norm": 0.15529580414295197,
+      "learning_rate": 4.690766700109659e-05,
+      "loss": 0.2154,
+      "num_input_tokens_seen": 13939928,
+      "step": 192
+    },
+    {
+      "epoch": 3.166147455867082,
+      "grad_norm": 0.1619848757982254,
+      "learning_rate": 4.687606125652882e-05,
+      "loss": 0.1963,
+      "num_input_tokens_seen": 14017936,
+      "step": 193
+    },
+    {
+      "epoch": 3.1827622014537904,
+      "grad_norm": 0.18066684901714325,
+      "learning_rate": 4.684430557588664e-05,
+      "loss": 0.1862,
+      "num_input_tokens_seen": 14074176,
+      "step": 194
+    },
+    {
+      "epoch": 3.1993769470404985,
+      "grad_norm": 0.16520777344703674,
+      "learning_rate": 4.681240017681993e-05,
+      "loss": 0.2576,
+      "num_input_tokens_seen": 14167656,
+      "step": 195
+    },
+    {
+      "epoch": 3.2159916926272065,
+      "grad_norm": 0.15385325253009796,
+      "learning_rate": 4.678034527800474e-05,
+      "loss": 0.1813,
+      "num_input_tokens_seen": 14235800,
+      "step": 196
+    },
+    {
+      "epoch": 3.232606438213915,
+      "grad_norm": 0.16897696256637573,
+      "learning_rate": 4.674814109914174e-05,
+      "loss": 0.1741,
+      "num_input_tokens_seen": 14301272,
+      "step": 197
+    },
+    {
+      "epoch": 3.249221183800623,
+      "grad_norm": 0.19556447863578796,
+      "learning_rate": 4.671578786095478e-05,
+      "loss": 0.2186,
+      "num_input_tokens_seen": 14347352,
+      "step": 198
+    },
+    {
+      "epoch": 3.265835929387331,
+      "grad_norm": 0.17333142459392548,
+      "learning_rate": 4.668328578518933e-05,
+      "loss": 0.2892,
+      "num_input_tokens_seen": 14434600,
+      "step": 199
+    },
+    {
+      "epoch": 3.2824506749740396,
+      "grad_norm": 0.20295488834381104,
+      "learning_rate": 4.665063509461097e-05,
+      "loss": 0.2014,
+      "num_input_tokens_seen": 14484104,
+      "step": 200
+    },
+    {
+      "epoch": 3.2990654205607477,
+      "grad_norm": 0.1597638726234436,
+      "learning_rate": 4.661783601300388e-05,
+      "loss": 0.2158,
+      "num_input_tokens_seen": 14567152,
+      "step": 201
+    },
+    {
+      "epoch": 3.3156801661474558,
+      "grad_norm": 0.19849488139152527,
+      "learning_rate": 4.6584888765169296e-05,
+      "loss": 0.2578,
+      "num_input_tokens_seen": 14647040,
+      "step": 202
+    },
+    {
+      "epoch": 3.3322949117341643,
+      "grad_norm": 0.1508200466632843,
+      "learning_rate": 4.6551793576923964e-05,
+      "loss": 0.2213,
+      "num_input_tokens_seen": 14738216,
+      "step": 203
+    },
+    {
+      "epoch": 3.3489096573208723,
+      "grad_norm": 0.1687687337398529,
+      "learning_rate": 4.65185506750986e-05,
+      "loss": 0.1828,
+      "num_input_tokens_seen": 14811216,
+      "step": 204
+    },
+    {
+      "epoch": 3.3655244029075804,
+      "grad_norm": 0.16587376594543457,
+      "learning_rate": 4.648516028753632e-05,
+      "loss": 0.1619,
+      "num_input_tokens_seen": 14885992,
+      "step": 205
+    },
+    {
+      "epoch": 3.382139148494289,
+      "grad_norm": 0.16600169241428375,
+      "learning_rate": 4.645162264309112e-05,
+      "loss": 0.2438,
+      "num_input_tokens_seen": 14961984,
+      "step": 206
+    },
+    {
+      "epoch": 3.398753894080997,
+      "grad_norm": 0.1877703070640564,
+      "learning_rate": 4.6417937971626245e-05,
+      "loss": 0.1771,
+      "num_input_tokens_seen": 15021240,
+      "step": 207
+    },
+    {
+      "epoch": 3.415368639667705,
+      "grad_norm": 0.20105206966400146,
+      "learning_rate": 4.638410650401267e-05,
+      "loss": 0.1742,
+      "num_input_tokens_seen": 15092016,
+      "step": 208
+    },
+    {
+      "epoch": 3.431983385254413,
+      "grad_norm": 0.12934140861034393,
+      "learning_rate": 4.635012847212748e-05,
+      "loss": 0.1725,
+      "num_input_tokens_seen": 15198192,
+      "step": 209
+    },
+    {
+      "epoch": 3.4485981308411215,
+      "grad_norm": 0.18388882279396057,
+      "learning_rate": 4.6316004108852305e-05,
+      "loss": 0.186,
+      "num_input_tokens_seen": 15258432,
+      "step": 210
+    },
+    {
+      "epoch": 3.4652128764278296,
+      "grad_norm": 0.1766858547925949,
+      "learning_rate": 4.628173364807171e-05,
+      "loss": 0.2166,
+      "num_input_tokens_seen": 15329600,
+      "step": 211
+    },
+    {
+      "epoch": 3.4818276220145377,
+      "grad_norm": 0.16214998066425323,
+      "learning_rate": 4.6247317324671605e-05,
+      "loss": 0.2038,
+      "num_input_tokens_seen": 15407920,
+      "step": 212
+    },
+    {
+      "epoch": 3.498442367601246,
+      "grad_norm": 0.16933797299861908,
+      "learning_rate": 4.6212755374537596e-05,
+      "loss": 0.2017,
+      "num_input_tokens_seen": 15479640,
+      "step": 213
+    },
+    {
+      "epoch": 3.515057113187954,
+      "grad_norm": 0.19472749531269073,
+      "learning_rate": 4.617804803455344e-05,
+      "loss": 0.2048,
+      "num_input_tokens_seen": 15561960,
+      "step": 214
+    },
+    {
+      "epoch": 3.5316718587746623,
+      "grad_norm": 0.33335182070732117,
+      "learning_rate": 4.614319554259934e-05,
+      "loss": 0.2358,
+      "num_input_tokens_seen": 15641440,
+      "step": 215
+    },
+    {
+      "epoch": 3.5482866043613708,
+      "grad_norm": 0.19587557017803192,
+      "learning_rate": 4.610819813755038e-05,
+      "loss": 0.2374,
+      "num_input_tokens_seen": 15728872,
+      "step": 216
+    },
+    {
+      "epoch": 3.564901349948079,
+      "grad_norm": 0.19063518941402435,
+      "learning_rate": 4.607305605927487e-05,
+      "loss": 0.1919,
+      "num_input_tokens_seen": 15798112,
+      "step": 217
+    },
+    {
+      "epoch": 3.581516095534787,
+      "grad_norm": 0.19598323106765747,
+      "learning_rate": 4.6037769548632656e-05,
+      "loss": 0.2583,
+      "num_input_tokens_seen": 15865936,
+      "step": 218
+    },
+    {
+      "epoch": 3.5981308411214954,
+      "grad_norm": 0.18066690862178802,
+      "learning_rate": 4.600233884747355e-05,
+      "loss": 0.2337,
+      "num_input_tokens_seen": 15941368,
+      "step": 219
+    },
+    {
+      "epoch": 3.6147455867082035,
+      "grad_norm": 0.16981899738311768,
+      "learning_rate": 4.5966764198635606e-05,
+      "loss": 0.1818,
+      "num_input_tokens_seen": 16028208,
+      "step": 220
+    },
+    {
+      "epoch": 3.6313603322949115,
+      "grad_norm": 0.180410236120224,
+      "learning_rate": 4.5931045845943474e-05,
+      "loss": 0.1646,
+      "num_input_tokens_seen": 16104408,
+      "step": 221
+    },
+    {
+      "epoch": 3.64797507788162,
+      "grad_norm": 0.19180680811405182,
+      "learning_rate": 4.5895184034206765e-05,
+      "loss": 0.3263,
+      "num_input_tokens_seen": 16156800,
+      "step": 222
+    },
+    {
+      "epoch": 3.664589823468328,
+      "grad_norm": 0.16119280457496643,
+      "learning_rate": 4.585917900921829e-05,
+      "loss": 0.2636,
+      "num_input_tokens_seen": 16256712,
+      "step": 223
+    },
+    {
+      "epoch": 3.681204569055036,
+      "grad_norm": 0.18559172749519348,
+      "learning_rate": 4.5823031017752485e-05,
+      "loss": 0.1759,
+      "num_input_tokens_seen": 16330344,
+      "step": 224
+    },
+    {
+      "epoch": 3.6978193146417446,
+      "grad_norm": 0.17767880856990814,
+      "learning_rate": 4.5786740307563636e-05,
+      "loss": 0.196,
+      "num_input_tokens_seen": 16399792,
+      "step": 225
+    },
+    {
+      "epoch": 3.7144340602284527,
+      "grad_norm": 0.17806987464427948,
+      "learning_rate": 4.575030712738419e-05,
+      "loss": 0.186,
+      "num_input_tokens_seen": 16466368,
+      "step": 226
+    },
+    {
+      "epoch": 3.7310488058151607,
+      "grad_norm": 0.1952792853116989,
+      "learning_rate": 4.571373172692309e-05,
+      "loss": 0.1789,
+      "num_input_tokens_seen": 16530976,
+      "step": 227
+    },
+    {
+      "epoch": 3.7476635514018692,
+      "grad_norm": 0.1774374544620514,
+      "learning_rate": 4.567701435686404e-05,
+      "loss": 0.1929,
+      "num_input_tokens_seen": 16600216,
+      "step": 228
+    },
+    {
+      "epoch": 3.7642782969885773,
+      "grad_norm": 0.18798600137233734,
+      "learning_rate": 4.5640155268863796e-05,
+      "loss": 0.2268,
+      "num_input_tokens_seen": 16673192,
+      "step": 229
+    },
+    {
+      "epoch": 3.7808930425752854,
+      "grad_norm": 0.2022520750761032,
+      "learning_rate": 4.5603154715550386e-05,
+      "loss": 0.1716,
+      "num_input_tokens_seen": 16739912,
+      "step": 230
+    },
+    {
+      "epoch": 3.797507788161994,
+      "grad_norm": 0.15170948207378387,
+      "learning_rate": 4.55660129505215e-05,
+      "loss": 0.1844,
+      "num_input_tokens_seen": 16834632,
+      "step": 231
+    },
+    {
+      "epoch": 3.814122533748702,
+      "grad_norm": 0.16655084490776062,
+      "learning_rate": 4.5528730228342605e-05,
+      "loss": 0.1899,
+      "num_input_tokens_seen": 16914728,
+      "step": 232
+    },
+    {
+      "epoch": 3.83073727933541,
+      "grad_norm": 0.19025221467018127,
+      "learning_rate": 4.549130680454532e-05,
+      "loss": 0.2214,
+      "num_input_tokens_seen": 17014304,
+      "step": 233
+    },
+    {
+      "epoch": 3.8473520249221185,
+      "grad_norm": 0.17126557230949402,
+      "learning_rate": 4.545374293562559e-05,
+      "loss": 0.2062,
+      "num_input_tokens_seen": 17106664,
+      "step": 234
+    },
+    {
+      "epoch": 3.8639667705088265,
+      "grad_norm": 0.16162410378456116,
+      "learning_rate": 4.541603887904198e-05,
+      "loss": 0.2016,
+      "num_input_tokens_seen": 17193744,
+      "step": 235
+    },
+    {
+      "epoch": 3.8805815160955346,
+      "grad_norm": 0.2067136913537979,
+      "learning_rate": 4.537819489321386e-05,
+      "loss": 0.1992,
+      "num_input_tokens_seen": 17254656,
+      "step": 236
+    },
+    {
+      "epoch": 3.897196261682243,
+      "grad_norm": 0.200433611869812,
+      "learning_rate": 4.534021123751968e-05,
+      "loss": 0.1961,
+      "num_input_tokens_seen": 17325896,
+      "step": 237
+    },
+    {
+      "epoch": 3.913811007268951,
+      "grad_norm": 0.2062034010887146,
+      "learning_rate": 4.5302088172295156e-05,
+      "loss": 0.2302,
+      "num_input_tokens_seen": 17394424,
+      "step": 238
+    },
+    {
+      "epoch": 3.930425752855659,
+      "grad_norm": 0.1928798407316208,
+      "learning_rate": 4.526382595883152e-05,
+      "loss": 0.1846,
+      "num_input_tokens_seen": 17456352,
+      "step": 239
+    },
+    {
+      "epoch": 3.9470404984423677,
+      "grad_norm": 0.2011859118938446,
+      "learning_rate": 4.522542485937369e-05,
+      "loss": 0.1879,
+      "num_input_tokens_seen": 17519168,
+      "step": 240
+    },
+    {
+      "epoch": 3.9636552440290758,
+      "grad_norm": 0.20441657304763794,
+      "learning_rate": 4.51868851371185e-05,
+      "loss": 0.206,
+      "num_input_tokens_seen": 17585144,
+      "step": 241
+    },
+    {
+      "epoch": 3.980269989615784,
+      "grad_norm": 0.18314018845558167,
+      "learning_rate": 4.5148207056212896e-05,
+      "loss": 0.1676,
+      "num_input_tokens_seen": 17662024,
+      "step": 242
+    },
+    {
+      "epoch": 3.9968847352024923,
+      "grad_norm": 0.21530692279338837,
+      "learning_rate": 4.5109390881752114e-05,
+      "loss": 0.1961,
+      "num_input_tokens_seen": 17724360,
+      "step": 243
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.38912200927734375,
+      "learning_rate": 4.5070436879777865e-05,
+      "loss": 0.185,
+      "num_input_tokens_seen": 17746200,
+      "step": 244
+    },
+    {
+      "epoch": 4.0166147455867085,
+      "grad_norm": 0.15166164934635162,
+      "learning_rate": 4.503134531727652e-05,
+      "loss": 0.1674,
+      "num_input_tokens_seen": 17830760,
+      "step": 245
+    },
+    {
+      "epoch": 4.033229491173416,
+      "grad_norm": 0.1999833583831787,
+      "learning_rate": 4.499211646217727e-05,
+      "loss": 0.1739,
+      "num_input_tokens_seen": 17903840,
+      "step": 246
+    },
+    {
+      "epoch": 4.049844236760125,
+      "grad_norm": 0.2024000585079193,
+      "learning_rate": 4.495275058335029e-05,
+      "loss": 0.1753,
+      "num_input_tokens_seen": 17990448,
+      "step": 247
+    },
+    {
+      "epoch": 4.066458982346833,
+      "grad_norm": 0.22637376189231873,
+      "learning_rate": 4.491324795060491e-05,
+      "loss": 0.1896,
+      "num_input_tokens_seen": 18069520,
+      "step": 248
+    },
+    {
+      "epoch": 4.083073727933541,
+      "grad_norm": 0.24361123144626617,
+      "learning_rate": 4.487360883468775e-05,
+      "loss": 0.1688,
+      "num_input_tokens_seen": 18129128,
+      "step": 249
+    },
+    {
+      "epoch": 4.099688473520249,
+      "grad_norm": 0.21949416399002075,
+      "learning_rate": 4.4833833507280884e-05,
+      "loss": 0.1928,
+      "num_input_tokens_seen": 18202472,
+      "step": 250
+    },
+    {
+      "epoch": 4.116303219106958,
+      "grad_norm": 0.22039519250392914,
+      "learning_rate": 4.4793922240999933e-05,
+      "loss": 0.1737,
+      "num_input_tokens_seen": 18267232,
+      "step": 251
+    },
+    {
+      "epoch": 4.132917964693665,
+      "grad_norm": 0.23173294961452484,
+      "learning_rate": 4.4753875309392266e-05,
+      "loss": 0.1883,
+      "num_input_tokens_seen": 18325216,
+      "step": 252
+    },
+    {
+      "epoch": 4.149532710280374,
+      "grad_norm": 0.24100351333618164,
+      "learning_rate": 4.471369298693505e-05,
+      "loss": 0.2042,
+      "num_input_tokens_seen": 18406184,
+      "step": 253
+    },
+    {
+      "epoch": 4.166147455867082,
+      "grad_norm": 0.1888919323682785,
+      "learning_rate": 4.467337554903344e-05,
+      "loss": 0.1656,
+      "num_input_tokens_seen": 18481056,
+      "step": 254
+    },
+    {
+      "epoch": 4.18276220145379,
+      "grad_norm": 0.17849119007587433,
+      "learning_rate": 4.463292327201862e-05,
+      "loss": 0.1454,
+      "num_input_tokens_seen": 18554864,
+      "step": 255
+    },
+    {
+      "epoch": 4.1993769470404985,
+      "grad_norm": 0.24600732326507568,
+      "learning_rate": 4.4592336433146e-05,
+      "loss": 0.2039,
+      "num_input_tokens_seen": 18612120,
+      "step": 256
+    },
+    {
+      "epoch": 4.215991692627207,
+      "grad_norm": 0.23695628345012665,
+      "learning_rate": 4.4551615310593195e-05,
+      "loss": 0.2112,
+      "num_input_tokens_seen": 18710408,
+      "step": 257
+    },
+    {
+      "epoch": 4.232606438213915,
+      "grad_norm": 0.2511826753616333,
+      "learning_rate": 4.451076018345825e-05,
+      "loss": 0.1831,
+      "num_input_tokens_seen": 18769400,
+      "step": 258
+    },
+    {
+      "epoch": 4.249221183800623,
+      "grad_norm": 0.1971820890903473,
+      "learning_rate": 4.4469771331757604e-05,
+      "loss": 0.1722,
+      "num_input_tokens_seen": 18849704,
+      "step": 259
+    },
+    {
+      "epoch": 4.265835929387332,
+      "grad_norm": 0.23203876614570618,
+      "learning_rate": 4.442864903642428e-05,
+      "loss": 0.1981,
+      "num_input_tokens_seen": 18943328,
+      "step": 260
+    },
+    {
+      "epoch": 4.282450674974039,
+      "grad_norm": 0.23434185981750488,
+      "learning_rate": 4.4387393579305865e-05,
+      "loss": 0.2014,
+      "num_input_tokens_seen": 19022536,
+      "step": 261
+    },
+    {
+      "epoch": 4.299065420560748,
+      "grad_norm": 0.2373885214328766,
+      "learning_rate": 4.434600524316266e-05,
+      "loss": 0.167,
+      "num_input_tokens_seen": 19089200,
+      "step": 262
+    },
+    {
+      "epoch": 4.315680166147456,
+      "grad_norm": 0.20130722224712372,
+      "learning_rate": 4.430448431166567e-05,
+      "loss": 0.2747,
+      "num_input_tokens_seen": 19171216,
+      "step": 263
+    },
+    {
+      "epoch": 4.332294911734164,
+      "grad_norm": 0.17947593331336975,
+      "learning_rate": 4.426283106939474e-05,
+      "loss": 0.1508,
+      "num_input_tokens_seen": 19271872,
+      "step": 264
+    },
+    {
+      "epoch": 4.348909657320872,
+      "grad_norm": 0.23321041464805603,
+      "learning_rate": 4.4221045801836494e-05,
+      "loss": 0.2585,
+      "num_input_tokens_seen": 19342984,
+      "step": 265
+    },
+    {
+      "epoch": 4.365524402907581,
+      "grad_norm": 0.26581740379333496,
+      "learning_rate": 4.41791287953825e-05,
+      "loss": 0.1789,
+      "num_input_tokens_seen": 19391640,
+      "step": 266
+    },
+    {
+      "epoch": 4.382139148494288,
+      "grad_norm": 0.20715415477752686,
+      "learning_rate": 4.4137080337327205e-05,
+      "loss": 0.1953,
+      "num_input_tokens_seen": 19463232,
+      "step": 267
+    },
+    {
+      "epoch": 4.398753894080997,
+      "grad_norm": 0.20641866326332092,
+      "learning_rate": 4.4094900715866064e-05,
+      "loss": 0.1752,
+      "num_input_tokens_seen": 19523728,
+      "step": 268
+    },
+    {
+      "epoch": 4.415368639667705,
+      "grad_norm": 0.23343385756015778,
+      "learning_rate": 4.4052590220093446e-05,
+      "loss": 0.1904,
+      "num_input_tokens_seen": 19598960,
+      "step": 269
+    },
+    {
+      "epoch": 4.431983385254413,
+      "grad_norm": 0.20117436349391937,
+      "learning_rate": 4.401014914000078e-05,
+      "loss": 0.1801,
+      "num_input_tokens_seen": 19666136,
+      "step": 270
+    },
+    {
+      "epoch": 4.4485981308411215,
+      "grad_norm": 0.24009813368320465,
+      "learning_rate": 4.3967577766474455e-05,
+      "loss": 0.1798,
+      "num_input_tokens_seen": 19728600,
+      "step": 271
+    },
+    {
+      "epoch": 4.46521287642783,
+      "grad_norm": 0.2242031991481781,
+      "learning_rate": 4.3924876391293915e-05,
+      "loss": 0.2221,
+      "num_input_tokens_seen": 19801032,
+      "step": 272
+    },
+    {
+      "epoch": 4.481827622014538,
+      "grad_norm": 0.22890391945838928,
+      "learning_rate": 4.3882045307129594e-05,
+      "loss": 0.1906,
+      "num_input_tokens_seen": 19885496,
+      "step": 273
+    },
+    {
+      "epoch": 4.498442367601246,
+      "grad_norm": 0.21996937692165375,
+      "learning_rate": 4.383908480754095e-05,
+      "loss": 0.1775,
+      "num_input_tokens_seen": 19952072,
+      "step": 274
+    },
+    {
+      "epoch": 4.515057113187955,
+      "grad_norm": 0.1860388070344925,
+      "learning_rate": 4.379599518697444e-05,
+      "loss": 0.1593,
+      "num_input_tokens_seen": 20026536,
+      "step": 275
+    },
+    {
+      "epoch": 4.531671858774662,
+      "grad_norm": 0.20987707376480103,
+      "learning_rate": 4.375277674076149e-05,
+      "loss": 0.1409,
+      "num_input_tokens_seen": 20079112,
+      "step": 276
+    },
+    {
+      "epoch": 4.548286604361371,
+      "grad_norm": 0.21347324550151825,
+      "learning_rate": 4.3709429765116504e-05,
+      "loss": 0.2701,
+      "num_input_tokens_seen": 20144264,
+      "step": 277
+    },
+    {
+      "epoch": 4.564901349948079,
+      "grad_norm": 0.27563896775245667,
+      "learning_rate": 4.366595455713479e-05,
+      "loss": 0.1856,
+      "num_input_tokens_seen": 20207568,
+      "step": 278
+    },
+    {
+      "epoch": 4.581516095534787,
+      "grad_norm": 0.21850791573524475,
+      "learning_rate": 4.3622351414790554e-05,
+      "loss": 0.2204,
+      "num_input_tokens_seen": 20292376,
+      "step": 279
+    },
+    {
+      "epoch": 4.598130841121495,
+      "grad_norm": 0.19672711193561554,
+      "learning_rate": 4.357862063693486e-05,
+      "loss": 0.1397,
+      "num_input_tokens_seen": 20383048,
+      "step": 280
+    },
+    {
+      "epoch": 4.614745586708204,
+      "grad_norm": 0.23507343232631683,
+      "learning_rate": 4.353476252329356e-05,
+      "loss": 0.1655,
+      "num_input_tokens_seen": 20463376,
+      "step": 281
+    },
+    {
+      "epoch": 4.6313603322949115,
+      "grad_norm": 0.23900464177131653,
+      "learning_rate": 4.349077737446525e-05,
+      "loss": 0.1511,
+      "num_input_tokens_seen": 20537808,
+      "step": 282
+    },
+    {
+      "epoch": 4.64797507788162,
+      "grad_norm": 0.2231033891439438,
+      "learning_rate": 4.344666549191921e-05,
+      "loss": 0.1783,
+      "num_input_tokens_seen": 20605496,
+      "step": 283
+    },
+    {
+      "epoch": 4.6645898234683285,
+      "grad_norm": 0.20772908627986908,
+      "learning_rate": 4.3402427177993366e-05,
+      "loss": 0.2135,
+      "num_input_tokens_seen": 20692096,
+      "step": 284
+    },
+    {
+      "epoch": 4.681204569055036,
+      "grad_norm": 0.24927115440368652,
+      "learning_rate": 4.335806273589214e-05,
+      "loss": 0.1919,
+      "num_input_tokens_seen": 20762800,
+      "step": 285
+    },
+    {
+      "epoch": 4.697819314641745,
+      "grad_norm": 0.2112305462360382,
+      "learning_rate": 4.3313572469684474e-05,
+      "loss": 0.1546,
+      "num_input_tokens_seen": 20831584,
+      "step": 286
+    },
+    {
+      "epoch": 4.714434060228453,
+      "grad_norm": 0.22639551758766174,
+      "learning_rate": 4.326895668430166e-05,
+      "loss": 0.124,
+      "num_input_tokens_seen": 20897320,
+      "step": 287
+    },
+    {
+      "epoch": 4.731048805815161,
+      "grad_norm": 0.2295934110879898,
+      "learning_rate": 4.3224215685535294e-05,
+      "loss": 0.1639,
+      "num_input_tokens_seen": 20966136,
+      "step": 288
+    },
+    {
+      "epoch": 4.747663551401869,
+      "grad_norm": 0.2341577112674713,
+      "learning_rate": 4.317934978003517e-05,
+      "loss": 0.1584,
+      "num_input_tokens_seen": 21034800,
+      "step": 289
+    },
+    {
+      "epoch": 4.764278296988578,
+      "grad_norm": 0.2542404234409332,
+      "learning_rate": 4.313435927530719e-05,
+      "loss": 0.1918,
+      "num_input_tokens_seen": 21098672,
+      "step": 290
+    },
+    {
+      "epoch": 4.780893042575285,
+      "grad_norm": 0.23311223089694977,
+      "learning_rate": 4.3089244479711236e-05,
+      "loss": 0.1597,
+      "num_input_tokens_seen": 21177632,
+      "step": 291
+    },
+    {
+      "epoch": 4.797507788161994,
+      "grad_norm": 0.2642923593521118,
+      "learning_rate": 4.304400570245906e-05,
+      "loss": 0.1847,
+      "num_input_tokens_seen": 21240896,
+      "step": 292
+    },
+    {
+      "epoch": 4.814122533748702,
+      "grad_norm": 0.18841278553009033,
+      "learning_rate": 4.299864325361217e-05,
+      "loss": 0.1472,
+      "num_input_tokens_seen": 21322984,
+      "step": 293
+    },
+    {
+      "epoch": 4.83073727933541,
+      "grad_norm": 0.22440434992313385,
+      "learning_rate": 4.295315744407972e-05,
+      "loss": 0.1607,
+      "num_input_tokens_seen": 21389128,
+      "step": 294
+    },
+    {
+      "epoch": 4.8473520249221185,
+      "grad_norm": 0.22145289182662964,
+      "learning_rate": 4.290754858561637e-05,
+      "loss": 0.1851,
+      "num_input_tokens_seen": 21469912,
+      "step": 295
+    },
+    {
+      "epoch": 4.863966770508826,
+      "grad_norm": 0.22817087173461914,
+      "learning_rate": 4.2861816990820084e-05,
+      "loss": 0.1531,
+      "num_input_tokens_seen": 21540320,
+      "step": 296
+    },
+    {
+      "epoch": 4.880581516095535,
+      "grad_norm": 0.22014038264751434,
+      "learning_rate": 4.281596297313013e-05,
+      "loss": 0.1815,
+      "num_input_tokens_seen": 21626312,
+      "step": 297
+    },
+    {
+      "epoch": 4.897196261682243,
+      "grad_norm": 0.2234148383140564,
+      "learning_rate": 4.2769986846824815e-05,
+      "loss": 0.1667,
+      "num_input_tokens_seen": 21702792,
+      "step": 298
+    },
+    {
+      "epoch": 4.913811007268951,
+      "grad_norm": 0.2851375341415405,
+      "learning_rate": 4.272388892701934e-05,
+      "loss": 0.1805,
+      "num_input_tokens_seen": 21771880,
+      "step": 299
+    },
+    {
+      "epoch": 4.930425752855659,
+      "grad_norm": 0.2221265286207199,
+      "learning_rate": 4.267766952966369e-05,
+      "loss": 0.1653,
+      "num_input_tokens_seen": 21844024,
+      "step": 300
+    },
+    {
+      "epoch": 4.947040498442368,
+      "grad_norm": 0.20688939094543457,
+      "learning_rate": 4.2631328971540444e-05,
+      "loss": 0.1654,
+      "num_input_tokens_seen": 21925632,
+      "step": 301
+    },
+    {
+      "epoch": 4.963655244029075,
+      "grad_norm": 0.2270977795124054,
+      "learning_rate": 4.2584867570262597e-05,
+      "loss": 0.1774,
+      "num_input_tokens_seen": 21981952,
+      "step": 302
+    },
+    {
+      "epoch": 4.980269989615784,
+      "grad_norm": 0.18652501702308655,
+      "learning_rate": 4.25382856442714e-05,
+      "loss": 0.1452,
+      "num_input_tokens_seen": 22070440,
+      "step": 303
+    },
+    {
+      "epoch": 4.996884735202492,
+      "grad_norm": 0.19792407751083374,
+      "learning_rate": 4.249158351283414e-05,
+      "loss": 0.1806,
+      "num_input_tokens_seen": 22170184,
+      "step": 304
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.5933757424354553,
+      "learning_rate": 4.244476149604201e-05,
+      "loss": 0.2115,
+      "num_input_tokens_seen": 22181856,
+      "step": 305
+    },
+    {
+      "epoch": 5.0166147455867085,
+      "grad_norm": 0.22399979829788208,
+      "learning_rate": 4.2397819914807856e-05,
+      "loss": 0.1614,
+      "num_input_tokens_seen": 22256808,
+      "step": 306
+    },
+    {
+      "epoch": 5.033229491173416,
+      "grad_norm": 0.2595834732055664,
+      "learning_rate": 4.2350759090864046e-05,
+      "loss": 0.1838,
+      "num_input_tokens_seen": 22325224,
+      "step": 307
+    },
+    {
+      "epoch": 5.049844236760125,
+      "grad_norm": 0.188430517911911,
+      "learning_rate": 4.230357934676017e-05,
+      "loss": 0.1421,
+      "num_input_tokens_seen": 22389624,
+      "step": 308
+    },
+    {
+      "epoch": 5.066458982346833,
+      "grad_norm": 0.325431764125824,
+      "learning_rate": 4.225628100586093e-05,
+      "loss": 0.1852,
+      "num_input_tokens_seen": 22463872,
+      "step": 309
+    },
+    {
+      "epoch": 5.083073727933541,
+      "grad_norm": 0.27097174525260925,
+      "learning_rate": 4.220886439234385e-05,
+      "loss": 0.1526,
+      "num_input_tokens_seen": 22515824,
+      "step": 310
+    },
+    {
+      "epoch": 5.099688473520249,
+      "grad_norm": 0.21379193663597107,
+      "learning_rate": 4.2161329831197095e-05,
+      "loss": 0.151,
+      "num_input_tokens_seen": 22602336,
+      "step": 311
+    },
+    {
+      "epoch": 5.116303219106958,
+      "grad_norm": 0.24827998876571655,
+      "learning_rate": 4.211367764821722e-05,
+      "loss": 0.142,
+      "num_input_tokens_seen": 22655176,
+      "step": 312
+    },
+    {
+      "epoch": 5.132917964693665,
+      "grad_norm": 0.26246964931488037,
+      "learning_rate": 4.2065908170006955e-05,
+      "loss": 0.1589,
+      "num_input_tokens_seen": 22728680,
+      "step": 313
+    },
+    {
+      "epoch": 5.149532710280374,
+      "grad_norm": 0.24459198117256165,
+      "learning_rate": 4.201802172397295e-05,
+      "loss": 0.1435,
+      "num_input_tokens_seen": 22806784,
+      "step": 314
+    },
+    {
+      "epoch": 5.166147455867082,
+      "grad_norm": 0.26540517807006836,
+      "learning_rate": 4.197001863832355e-05,
+      "loss": 0.1447,
+      "num_input_tokens_seen": 22880648,
+      "step": 315
+    },
+    {
+      "epoch": 5.18276220145379,
+      "grad_norm": 0.25646644830703735,
+      "learning_rate": 4.192189924206652e-05,
+      "loss": 0.1418,
+      "num_input_tokens_seen": 22953184,
+      "step": 316
+    },
+    {
+      "epoch": 5.1993769470404985,
+      "grad_norm": 0.2358384132385254,
+      "learning_rate": 4.187366386500683e-05,
+      "loss": 0.1845,
+      "num_input_tokens_seen": 23037392,
+      "step": 317
+    },
+    {
+      "epoch": 5.215991692627207,
+      "grad_norm": 0.2270258218050003,
+      "learning_rate": 4.182531283774434e-05,
+      "loss": 0.2668,
+      "num_input_tokens_seen": 23086552,
+      "step": 318
+    },
+    {
+      "epoch": 5.232606438213915,
+      "grad_norm": 0.24396558105945587,
+      "learning_rate": 4.177684649167158e-05,
+      "loss": 0.1567,
+      "num_input_tokens_seen": 23153152,
+      "step": 319
+    },
+    {
+      "epoch": 5.249221183800623,
+      "grad_norm": 0.2542375922203064,
+      "learning_rate": 4.172826515897146e-05,
+      "loss": 0.1617,
+      "num_input_tokens_seen": 23240928,
+      "step": 320
+    },
+    {
+      "epoch": 5.265835929387332,
+      "grad_norm": 0.2268146276473999,
+      "learning_rate": 4.1679569172614996e-05,
+      "loss": 0.1573,
+      "num_input_tokens_seen": 23325912,
+      "step": 321
+    },
+    {
+      "epoch": 5.282450674974039,
+      "grad_norm": 0.26405712962150574,
+      "learning_rate": 4.163075886635902e-05,
+      "loss": 0.1738,
+      "num_input_tokens_seen": 23401952,
+      "step": 322
+    },
+    {
+      "epoch": 5.299065420560748,
+      "grad_norm": 0.2599943280220032,
+      "learning_rate": 4.1581834574743915e-05,
+      "loss": 0.1433,
+      "num_input_tokens_seen": 23463760,
+      "step": 323
+    },
+    {
+      "epoch": 5.315680166147456,
+      "grad_norm": 0.24650180339813232,
+      "learning_rate": 4.1532796633091296e-05,
+      "loss": 0.1511,
+      "num_input_tokens_seen": 23535272,
+      "step": 324
+    },
+    {
+      "epoch": 5.332294911734164,
+      "grad_norm": 0.22445832192897797,
+      "learning_rate": 4.148364537750172e-05,
+      "loss": 0.1296,
+      "num_input_tokens_seen": 23607752,
+      "step": 325
+    },
+    {
+      "epoch": 5.348909657320872,
+      "grad_norm": 0.20715995132923126,
+      "learning_rate": 4.14343811448524e-05,
+      "loss": 0.142,
+      "num_input_tokens_seen": 23674872,
+      "step": 326
+    },
+    {
+      "epoch": 5.365524402907581,
+      "grad_norm": 0.2707969546318054,
+      "learning_rate": 4.138500427279485e-05,
+      "loss": 0.1679,
+      "num_input_tokens_seen": 23736384,
+      "step": 327
+    },
+    {
+      "epoch": 5.382139148494288,
+      "grad_norm": 0.2017841637134552,
+      "learning_rate": 4.133551509975264e-05,
+      "loss": 0.1346,
+      "num_input_tokens_seen": 23835000,
+      "step": 328
+    },
+    {
+      "epoch": 5.398753894080997,
+      "grad_norm": 0.21116195619106293,
+      "learning_rate": 4.128591396491901e-05,
+      "loss": 0.1364,
+      "num_input_tokens_seen": 23912552,
+      "step": 329
+    },
+    {
+      "epoch": 5.415368639667705,
+      "grad_norm": 0.2331131547689438,
+      "learning_rate": 4.123620120825459e-05,
+      "loss": 0.1719,
+      "num_input_tokens_seen": 23987368,
+      "step": 330
+    },
+    {
+      "epoch": 5.431983385254413,
+      "grad_norm": 0.27115845680236816,
+      "learning_rate": 4.118637717048506e-05,
+      "loss": 0.1468,
+      "num_input_tokens_seen": 24050848,
+      "step": 331
+    },
+    {
+      "epoch": 5.4485981308411215,
+      "grad_norm": 0.21654783189296722,
+      "learning_rate": 4.113644219309877e-05,
+      "loss": 0.1418,
+      "num_input_tokens_seen": 24146104,
+      "step": 332
+    },
+    {
+      "epoch": 5.46521287642783,
+      "grad_norm": 0.2783348560333252,
+      "learning_rate": 4.1086396618344476e-05,
+      "loss": 0.1502,
+      "num_input_tokens_seen": 24194184,
+      "step": 333
+    },
+    {
+      "epoch": 5.481827622014538,
+      "grad_norm": 0.23255467414855957,
+      "learning_rate": 4.1036240789228954e-05,
+      "loss": 0.1571,
+      "num_input_tokens_seen": 24275368,
+      "step": 334
+    },
+    {
+      "epoch": 5.498442367601246,
+      "grad_norm": 0.2655453681945801,
+      "learning_rate": 4.098597504951462e-05,
+      "loss": 0.1607,
+      "num_input_tokens_seen": 24329192,
+      "step": 335
+    },
+    {
+      "epoch": 5.515057113187955,
+      "grad_norm": 0.23245719075202942,
+      "learning_rate": 4.093559974371725e-05,
+      "loss": 0.1453,
+      "num_input_tokens_seen": 24426696,
+      "step": 336
+    },
+    {
+      "epoch": 5.531671858774662,
+      "grad_norm": 0.20036327838897705,
+      "learning_rate": 4.088511521710352e-05,
+      "loss": 0.1411,
+      "num_input_tokens_seen": 24514344,
+      "step": 337
+    },
+    {
+      "epoch": 5.548286604361371,
+      "grad_norm": 0.32294195890426636,
+      "learning_rate": 4.083452181568875e-05,
+      "loss": 0.1467,
+      "num_input_tokens_seen": 24584464,
+      "step": 338
+    },
+    {
+      "epoch": 5.564901349948079,
+      "grad_norm": 0.23969624936580658,
+      "learning_rate": 4.0783819886234445e-05,
+      "loss": 0.1304,
+      "num_input_tokens_seen": 24660600,
+      "step": 339
+    },
+    {
+      "epoch": 5.581516095534787,
+      "grad_norm": 0.2539989948272705,
+      "learning_rate": 4.073300977624594e-05,
+      "loss": 0.1374,
+      "num_input_tokens_seen": 24717088,
+      "step": 340
+    },
+    {
+      "epoch": 5.598130841121495,
+      "grad_norm": 0.26608580350875854,
+      "learning_rate": 4.068209183397004e-05,
+      "loss": 0.1519,
+      "num_input_tokens_seen": 24775352,
+      "step": 341
+    },
+    {
+      "epoch": 5.614745586708204,
+      "grad_norm": 0.2161550372838974,
+      "learning_rate": 4.063106640839264e-05,
+      "loss": 0.1409,
+      "num_input_tokens_seen": 24860072,
+      "step": 342
+    },
+    {
+      "epoch": 5.6313603322949115,
+      "grad_norm": 0.22482600808143616,
+      "learning_rate": 4.057993384923626e-05,
+      "loss": 0.1393,
+      "num_input_tokens_seen": 24947856,
+      "step": 343
+    },
+    {
+      "epoch": 5.64797507788162,
+      "grad_norm": 0.2367829829454422,
+      "learning_rate": 4.052869450695776e-05,
+      "loss": 0.1539,
+      "num_input_tokens_seen": 25024992,
+      "step": 344
+    },
+    {
+      "epoch": 5.6645898234683285,
+      "grad_norm": 0.29229775071144104,
+      "learning_rate": 4.047734873274586e-05,
+      "loss": 0.1522,
+      "num_input_tokens_seen": 25092248,
+      "step": 345
+    },
+    {
+      "epoch": 5.681204569055036,
+      "grad_norm": 0.2589828670024872,
+      "learning_rate": 4.042589687851872e-05,
+      "loss": 0.1493,
+      "num_input_tokens_seen": 25170496,
+      "step": 346
+    },
+    {
+      "epoch": 5.697819314641745,
+      "grad_norm": 0.23003339767456055,
+      "learning_rate": 4.037433929692161e-05,
+      "loss": 0.1529,
+      "num_input_tokens_seen": 25268720,
+      "step": 347
+    },
+    {
+      "epoch": 5.714434060228453,
+      "grad_norm": 0.26932114362716675,
+      "learning_rate": 4.0322676341324415e-05,
+      "loss": 0.1499,
+      "num_input_tokens_seen": 25332688,
+      "step": 348
+    },
+    {
+      "epoch": 5.731048805815161,
+      "grad_norm": 0.27059391140937805,
+      "learning_rate": 4.027090836581925e-05,
+      "loss": 0.173,
+      "num_input_tokens_seen": 25413904,
+      "step": 349
+    },
+    {
+      "epoch": 5.747663551401869,
+      "grad_norm": 0.24265804886817932,
+      "learning_rate": 4.021903572521802e-05,
+      "loss": 0.1531,
+      "num_input_tokens_seen": 25503720,
+      "step": 350
+    },
+    {
+      "epoch": 5.764278296988578,
+      "grad_norm": 0.28688696026802063,
+      "learning_rate": 4.0167058775049996e-05,
+      "loss": 0.1615,
+      "num_input_tokens_seen": 25568560,
+      "step": 351
+    },
+    {
+      "epoch": 5.780893042575285,
+      "grad_norm": 0.26201075315475464,
+      "learning_rate": 4.011497787155938e-05,
+      "loss": 0.1452,
+      "num_input_tokens_seen": 25635184,
+      "step": 352
+    },
+    {
+      "epoch": 5.797507788161994,
+      "grad_norm": 0.22841767966747284,
+      "learning_rate": 4.006279337170283e-05,
+      "loss": 0.148,
+      "num_input_tokens_seen": 25719768,
+      "step": 353
+    },
+    {
+      "epoch": 5.814122533748702,
+      "grad_norm": 0.28246188163757324,
+      "learning_rate": 4.0010505633147106e-05,
+      "loss": 0.1446,
+      "num_input_tokens_seen": 25795016,
+      "step": 354
+    },
+    {
+      "epoch": 5.83073727933541,
+      "grad_norm": 0.2533949315547943,
+      "learning_rate": 3.995811501426648e-05,
+      "loss": 0.1297,
+      "num_input_tokens_seen": 25863184,
+      "step": 355
+    },
+    {
+      "epoch": 5.8473520249221185,
+      "grad_norm": 0.25117793679237366,
+      "learning_rate": 3.99056218741404e-05,
+      "loss": 0.1479,
+      "num_input_tokens_seen": 25935752,
+      "step": 356
+    },
+    {
+      "epoch": 5.863966770508826,
+      "grad_norm": 0.2759612202644348,
+      "learning_rate": 3.985302657255097e-05,
+      "loss": 0.1466,
+      "num_input_tokens_seen": 25995760,
+      "step": 357
+    },
+    {
+      "epoch": 5.880581516095535,
+      "grad_norm": 0.24218028783798218,
+      "learning_rate": 3.980032946998049e-05,
+      "loss": 0.1297,
+      "num_input_tokens_seen": 26061240,
+      "step": 358
+    },
+    {
+      "epoch": 5.897196261682243,
+      "grad_norm": 0.2628185749053955,
+      "learning_rate": 3.974753092760901e-05,
+      "loss": 0.1421,
+      "num_input_tokens_seen": 26131024,
+      "step": 359
+    },
+    {
+      "epoch": 5.913811007268951,
+      "grad_norm": 0.18742328882217407,
+      "learning_rate": 3.969463130731183e-05,
+      "loss": 0.1737,
+      "num_input_tokens_seen": 26233672,
+      "step": 360
+    },
+    {
+      "epoch": 5.930425752855659,
+      "grad_norm": 0.22187161445617676,
+      "learning_rate": 3.964163097165702e-05,
+      "loss": 0.1153,
+      "num_input_tokens_seen": 26303488,
+      "step": 361
+    },
+    {
+      "epoch": 5.947040498442368,
+      "grad_norm": 0.25965312123298645,
+      "learning_rate": 3.958853028390294e-05,
+      "loss": 0.1347,
+      "num_input_tokens_seen": 26377768,
+      "step": 362
+    },
+    {
+      "epoch": 5.963655244029075,
+      "grad_norm": 0.2957039475440979,
+      "learning_rate": 3.953532960799577e-05,
+      "loss": 0.2736,
+      "num_input_tokens_seen": 26435984,
+      "step": 363
+    },
+    {
+      "epoch": 5.980269989615784,
+      "grad_norm": 0.3063158392906189,
+      "learning_rate": 3.948202930856697e-05,
+      "loss": 0.1563,
+      "num_input_tokens_seen": 26513960,
+      "step": 364
+    },
+    {
+      "epoch": 5.996884735202492,
+      "grad_norm": 0.21576139330863953,
+      "learning_rate": 3.942862975093085e-05,
+      "loss": 0.1361,
+      "num_input_tokens_seen": 26599064,
+      "step": 365
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.49579551815986633,
+      "learning_rate": 3.937513130108197e-05,
+      "loss": 0.1849,
+      "num_input_tokens_seen": 26617264,
+      "step": 366
+    },
+    {
+      "epoch": 6.0166147455867085,
+      "grad_norm": 0.25051239132881165,
+      "learning_rate": 3.9321534325692726e-05,
+      "loss": 0.1452,
+      "num_input_tokens_seen": 26717024,
+      "step": 367
+    },
+    {
+      "epoch": 6.033229491173416,
+      "grad_norm": 0.2842366099357605,
+      "learning_rate": 3.92678391921108e-05,
+      "loss": 0.1361,
+      "num_input_tokens_seen": 26788120,
+      "step": 368
+    },
+    {
+      "epoch": 6.049844236760125,
+      "grad_norm": 0.32458341121673584,
+      "learning_rate": 3.92140462683566e-05,
+      "loss": 0.1537,
+      "num_input_tokens_seen": 26853760,
+      "step": 369
+    },
+    {
+      "epoch": 6.066458982346833,
+      "grad_norm": 0.27452871203422546,
+      "learning_rate": 3.916015592312082e-05,
+      "loss": 0.1354,
+      "num_input_tokens_seen": 26923848,
+      "step": 370
+    },
+    {
+      "epoch": 6.083073727933541,
+      "grad_norm": 0.2542634904384613,
+      "learning_rate": 3.9106168525761855e-05,
+      "loss": 0.1233,
+      "num_input_tokens_seen": 26976184,
+      "step": 371
+    },
+    {
+      "epoch": 6.099688473520249,
+      "grad_norm": 0.2786627411842346,
+      "learning_rate": 3.905208444630327e-05,
+      "loss": 0.125,
+      "num_input_tokens_seen": 27065712,
+      "step": 372
+    },
+    {
+      "epoch": 6.116303219106958,
+      "grad_norm": 0.19119420647621155,
+      "learning_rate": 3.899790405543129e-05,
+      "loss": 0.1475,
+      "num_input_tokens_seen": 27145472,
+      "step": 373
+    },
+    {
+      "epoch": 6.132917964693665,
+      "grad_norm": 0.25163477659225464,
+      "learning_rate": 3.894362772449226e-05,
+      "loss": 0.1235,
+      "num_input_tokens_seen": 27233904,
+      "step": 374
+    },
+    {
+      "epoch": 6.149532710280374,
+      "grad_norm": 0.2418321669101715,
+      "learning_rate": 3.888925582549006e-05,
+      "loss": 0.1117,
+      "num_input_tokens_seen": 27311512,
+      "step": 375
+    },
+    {
+      "epoch": 6.166147455867082,
+      "grad_norm": 0.26851242780685425,
+      "learning_rate": 3.883478873108361e-05,
+      "loss": 0.1226,
+      "num_input_tokens_seen": 27387400,
+      "step": 376
+    },
+    {
+      "epoch": 6.18276220145379,
+      "grad_norm": 0.280127614736557,
+      "learning_rate": 3.878022681458426e-05,
+      "loss": 0.1295,
+      "num_input_tokens_seen": 27461280,
+      "step": 377
+    },
+    {
+      "epoch": 6.1993769470404985,
+      "grad_norm": 0.255588173866272,
+      "learning_rate": 3.87255704499533e-05,
+      "loss": 0.1197,
+      "num_input_tokens_seen": 27556400,
+      "step": 378
+    },
+    {
+      "epoch": 6.215991692627207,
+      "grad_norm": 0.2888232171535492,
+      "learning_rate": 3.8670820011799315e-05,
+      "loss": 0.1417,
+      "num_input_tokens_seen": 27613664,
+      "step": 379
+    },
+    {
+      "epoch": 6.232606438213915,
+      "grad_norm": 0.28014957904815674,
+      "learning_rate": 3.861597587537568e-05,
+      "loss": 0.1565,
+      "num_input_tokens_seen": 27681024,
+      "step": 380
+    },
+    {
+      "epoch": 6.249221183800623,
+      "grad_norm": 0.2911868095397949,
+      "learning_rate": 3.856103841657797e-05,
+      "loss": 0.1293,
+      "num_input_tokens_seen": 27759536,
+      "step": 381
+    },
+    {
+      "epoch": 6.265835929387332,
+      "grad_norm": 0.19534167647361755,
+      "learning_rate": 3.850600801194138e-05,
+      "loss": 0.0967,
+      "num_input_tokens_seen": 27857288,
+      "step": 382
+    },
+    {
+      "epoch": 6.282450674974039,
+      "grad_norm": 0.2868412733078003,
+      "learning_rate": 3.8450885038638127e-05,
+      "loss": 0.1395,
+      "num_input_tokens_seen": 27940528,
+      "step": 383
+    },
+    {
+      "epoch": 6.299065420560748,
+      "grad_norm": 0.3421030342578888,
+      "learning_rate": 3.8395669874474915e-05,
+      "loss": 0.1402,
+      "num_input_tokens_seen": 28033824,
+      "step": 384
+    },
+    {
+      "epoch": 6.315680166147456,
+      "grad_norm": 0.2819141447544098,
+      "learning_rate": 3.834036289789029e-05,
+      "loss": 0.1137,
+      "num_input_tokens_seen": 28096192,
+      "step": 385
+    },
+    {
+      "epoch": 6.332294911734164,
+      "grad_norm": 0.22595854103565216,
+      "learning_rate": 3.828496448795207e-05,
+      "loss": 0.1143,
+      "num_input_tokens_seen": 28181256,
+      "step": 386
+    },
+    {
+      "epoch": 6.348909657320872,
+      "grad_norm": 0.2450927495956421,
+      "learning_rate": 3.822947502435477e-05,
+      "loss": 0.121,
+      "num_input_tokens_seen": 28245480,
+      "step": 387
+    },
+    {
+      "epoch": 6.365524402907581,
+      "grad_norm": 0.2552040219306946,
+      "learning_rate": 3.8173894887416945e-05,
+      "loss": 0.1183,
+      "num_input_tokens_seen": 28307200,
+      "step": 388
+    },
+    {
+      "epoch": 6.382139148494288,
+      "grad_norm": 0.3974437415599823,
+      "learning_rate": 3.811822445807863e-05,
+      "loss": 0.1423,
+      "num_input_tokens_seen": 28384640,
+      "step": 389
+    },
+    {
+      "epoch": 6.398753894080997,
+      "grad_norm": 0.2750483453273773,
+      "learning_rate": 3.8062464117898724e-05,
+      "loss": 0.1068,
+      "num_input_tokens_seen": 28447992,
+      "step": 390
+    },
+    {
+      "epoch": 6.415368639667705,
+      "grad_norm": 0.3021928071975708,
+      "learning_rate": 3.800661424905235e-05,
+      "loss": 0.1233,
+      "num_input_tokens_seen": 28513856,
+      "step": 391
+    },
+    {
+      "epoch": 6.431983385254413,
+      "grad_norm": 0.4087376296520233,
+      "learning_rate": 3.795067523432826e-05,
+      "loss": 0.1354,
+      "num_input_tokens_seen": 28596584,
+      "step": 392
+    },
+    {
+      "epoch": 6.4485981308411215,
+      "grad_norm": 0.32668039202690125,
+      "learning_rate": 3.789464745712619e-05,
+      "loss": 0.1483,
+      "num_input_tokens_seen": 28664560,
+      "step": 393
+    },
+    {
+      "epoch": 6.46521287642783,
+      "grad_norm": 0.24187126755714417,
+      "learning_rate": 3.7838531301454254e-05,
+      "loss": 0.1225,
+      "num_input_tokens_seen": 28739512,
+      "step": 394
+    },
+    {
+      "epoch": 6.481827622014538,
+      "grad_norm": 0.2861412465572357,
+      "learning_rate": 3.77823271519263e-05,
+      "loss": 0.1318,
+      "num_input_tokens_seen": 28831848,
+      "step": 395
+    },
+    {
+      "epoch": 6.498442367601246,
+      "grad_norm": 0.30458521842956543,
+      "learning_rate": 3.7726035393759285e-05,
+      "loss": 0.1337,
+      "num_input_tokens_seen": 28898408,
+      "step": 396
+    },
+    {
+      "epoch": 6.515057113187955,
+      "grad_norm": 0.2644833028316498,
+      "learning_rate": 3.76696564127706e-05,
+      "loss": 0.1283,
+      "num_input_tokens_seen": 28960224,
+      "step": 397
+    },
+    {
+      "epoch": 6.531671858774662,
+      "grad_norm": 0.3309299349784851,
+      "learning_rate": 3.761319059537548e-05,
+      "loss": 0.1548,
+      "num_input_tokens_seen": 29020568,
+      "step": 398
+    },
+    {
+      "epoch": 6.548286604361371,
+      "grad_norm": 0.2788560092449188,
+      "learning_rate": 3.755663832858432e-05,
+      "loss": 0.1154,
+      "num_input_tokens_seen": 29095448,
+      "step": 399
+    },
+    {
+      "epoch": 6.564901349948079,
+      "grad_norm": 0.26450732350349426,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.1276,
+      "num_input_tokens_seen": 29186600,
+      "step": 400
+    },
+    {
+      "epoch": 6.581516095534787,
+      "grad_norm": 0.2650800049304962,
+      "learning_rate": 3.744327599781531e-05,
+      "loss": 0.2184,
+      "num_input_tokens_seen": 29258552,
+      "step": 401
+    },
+    {
+      "epoch": 6.598130841121495,
+      "grad_norm": 0.2814129590988159,
+      "learning_rate": 3.7386466710810194e-05,
+      "loss": 0.1508,
+      "num_input_tokens_seen": 29344848,
+      "step": 402
+    },
+    {
+      "epoch": 6.614745586708204,
+      "grad_norm": 0.2940891981124878,
+      "learning_rate": 3.7329572528349146e-05,
+      "loss": 0.2035,
+      "num_input_tokens_seen": 29410184,
+      "step": 403
+    },
+    {
+      "epoch": 6.6313603322949115,
+      "grad_norm": 0.25937849283218384,
+      "learning_rate": 3.727259384037852e-05,
+      "loss": 0.1308,
+      "num_input_tokens_seen": 29484928,
+      "step": 404
+    },
+    {
+      "epoch": 6.64797507788162,
+      "grad_norm": 0.26631370186805725,
+      "learning_rate": 3.721553103742388e-05,
+      "loss": 0.1349,
+      "num_input_tokens_seen": 29566432,
+      "step": 405
+    },
+    {
+      "epoch": 6.6645898234683285,
+      "grad_norm": 0.2721930146217346,
+      "learning_rate": 3.715838451058726e-05,
+      "loss": 0.1246,
+      "num_input_tokens_seen": 29634032,
+      "step": 406
+    },
+    {
+      "epoch": 6.681204569055036,
+      "grad_norm": 0.3131358325481415,
+      "learning_rate": 3.7101154651544584e-05,
+      "loss": 0.142,
+      "num_input_tokens_seen": 29681424,
+      "step": 407
+    },
+    {
+      "epoch": 6.697819314641745,
+      "grad_norm": 0.266184002161026,
+      "learning_rate": 3.704384185254288e-05,
+      "loss": 0.118,
+      "num_input_tokens_seen": 29762208,
+      "step": 408
+    },
+    {
+      "epoch": 6.714434060228453,
+      "grad_norm": 0.29790574312210083,
+      "learning_rate": 3.6986446506397666e-05,
+      "loss": 0.1286,
+      "num_input_tokens_seen": 29816280,
+      "step": 409
+    },
+    {
+      "epoch": 6.731048805815161,
+      "grad_norm": 0.2710191309452057,
+      "learning_rate": 3.692896900649021e-05,
+      "loss": 0.1211,
+      "num_input_tokens_seen": 29893040,
+      "step": 410
+    },
+    {
+      "epoch": 6.747663551401869,
+      "grad_norm": 0.46238377690315247,
+      "learning_rate": 3.6871409746764865e-05,
+      "loss": 0.1324,
+      "num_input_tokens_seen": 29971688,
+      "step": 411
+    },
+    {
+      "epoch": 6.764278296988578,
+      "grad_norm": 0.24678252637386322,
+      "learning_rate": 3.681376912172636e-05,
+      "loss": 0.1175,
+      "num_input_tokens_seen": 30051784,
+      "step": 412
+    },
+    {
+      "epoch": 6.780893042575285,
+      "grad_norm": 0.2699359655380249,
+      "learning_rate": 3.675604752643706e-05,
+      "loss": 0.1279,
+      "num_input_tokens_seen": 30146048,
+      "step": 413
+    },
+    {
+      "epoch": 6.797507788161994,
+      "grad_norm": 0.3923390805721283,
+      "learning_rate": 3.6698245356514335e-05,
+      "loss": 0.1204,
+      "num_input_tokens_seen": 30221296,
+      "step": 414
+    },
+    {
+      "epoch": 6.814122533748702,
+      "grad_norm": 0.26230940222740173,
+      "learning_rate": 3.6640363008127784e-05,
+      "loss": 0.1311,
+      "num_input_tokens_seen": 30287664,
+      "step": 415
+    },
+    {
+      "epoch": 6.83073727933541,
+      "grad_norm": 0.31548890471458435,
+      "learning_rate": 3.6582400877996546e-05,
+      "loss": 0.1417,
+      "num_input_tokens_seen": 30352816,
+      "step": 416
+    },
+    {
+      "epoch": 6.8473520249221185,
+      "grad_norm": 0.25795799493789673,
+      "learning_rate": 3.652435936338656e-05,
+      "loss": 0.1316,
+      "num_input_tokens_seen": 30439688,
+      "step": 417
+    },
+    {
+      "epoch": 6.863966770508826,
+      "grad_norm": 0.3696773648262024,
+      "learning_rate": 3.646623886210788e-05,
+      "loss": 0.1422,
+      "num_input_tokens_seen": 30506856,
+      "step": 418
+    },
+    {
+      "epoch": 6.880581516095535,
+      "grad_norm": 0.28359490633010864,
+      "learning_rate": 3.64080397725119e-05,
+      "loss": 0.1148,
+      "num_input_tokens_seen": 30565848,
+      "step": 419
+    },
+    {
+      "epoch": 6.897196261682243,
+      "grad_norm": 0.3696165978908539,
+      "learning_rate": 3.634976249348867e-05,
+      "loss": 0.1472,
+      "num_input_tokens_seen": 30633944,
+      "step": 420
+    },
+    {
+      "epoch": 6.913811007268951,
+      "grad_norm": 0.2775883078575134,
+      "learning_rate": 3.629140742446414e-05,
+      "loss": 0.1281,
+      "num_input_tokens_seen": 30704760,
+      "step": 421
+    },
+    {
+      "epoch": 6.930425752855659,
+      "grad_norm": 0.3016800582408905,
+      "learning_rate": 3.623297496539741e-05,
+      "loss": 0.1207,
+      "num_input_tokens_seen": 30773792,
+      "step": 422
+    },
+    {
+      "epoch": 6.947040498442368,
+      "grad_norm": 0.2855510711669922,
+      "learning_rate": 3.6174465516778035e-05,
+      "loss": 0.1274,
+      "num_input_tokens_seen": 30848672,
+      "step": 423
+    },
+    {
+      "epoch": 6.963655244029075,
+      "grad_norm": 0.27131325006484985,
+      "learning_rate": 3.611587947962319e-05,
+      "loss": 0.1234,
+      "num_input_tokens_seen": 30906064,
+      "step": 424
+    },
+    {
+      "epoch": 6.980269989615784,
+      "grad_norm": 0.35402408242225647,
+      "learning_rate": 3.6057217255475034e-05,
+      "loss": 0.1465,
+      "num_input_tokens_seen": 30964720,
+      "step": 425
+    },
+    {
+      "epoch": 6.996884735202492,
+      "grad_norm": 0.282308965921402,
+      "learning_rate": 3.599847924639788e-05,
+      "loss": 0.133,
+      "num_input_tokens_seen": 31043152,
+      "step": 426
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.5601735711097717,
+      "learning_rate": 3.593966585497547e-05,
+      "loss": 0.106,
+      "num_input_tokens_seen": 31056056,
+      "step": 427
+    },
+    {
+      "epoch": 7.0166147455867085,
+      "grad_norm": 0.26449844241142273,
+      "learning_rate": 3.588077748430819e-05,
+      "loss": 0.1082,
+      "num_input_tokens_seen": 31135304,
+      "step": 428
+    },
+    {
+      "epoch": 7.033229491173416,
+      "grad_norm": 0.2989669442176819,
+      "learning_rate": 3.582181453801036e-05,
+      "loss": 0.1217,
+      "num_input_tokens_seen": 31185600,
+      "step": 429
+    },
+    {
+      "epoch": 7.049844236760125,
+      "grad_norm": 0.3569459021091461,
+      "learning_rate": 3.576277742020738e-05,
+      "loss": 0.119,
+      "num_input_tokens_seen": 31254312,
+      "step": 430
+    },
+    {
+      "epoch": 7.066458982346833,
+      "grad_norm": 0.3184427320957184,
+      "learning_rate": 3.570366653553307e-05,
+      "loss": 0.1131,
+      "num_input_tokens_seen": 31339112,
+      "step": 431
+    },
+    {
+      "epoch": 7.083073727933541,
+      "grad_norm": 0.2780788540840149,
+      "learning_rate": 3.564448228912682e-05,
+      "loss": 0.0967,
+      "num_input_tokens_seen": 31424024,
+      "step": 432
+    },
+    {
+      "epoch": 7.099688473520249,
+      "grad_norm": 0.29954010248184204,
+      "learning_rate": 3.558522508663081e-05,
+      "loss": 0.124,
+      "num_input_tokens_seen": 31494656,
+      "step": 433
+    },
+    {
+      "epoch": 7.116303219106958,
+      "grad_norm": 0.235326886177063,
+      "learning_rate": 3.552589533418728e-05,
+      "loss": 0.1102,
+      "num_input_tokens_seen": 31588536,
+      "step": 434
+    },
+    {
+      "epoch": 7.132917964693665,
+      "grad_norm": 0.31165215373039246,
+      "learning_rate": 3.54664934384357e-05,
+      "loss": 0.2184,
+      "num_input_tokens_seen": 31657560,
+      "step": 435
+    },
+    {
+      "epoch": 7.149532710280374,
+      "grad_norm": 0.30129772424697876,
+      "learning_rate": 3.540701980651003e-05,
+      "loss": 0.1208,
+      "num_input_tokens_seen": 31743992,
+      "step": 436
+    },
+    {
+      "epoch": 7.166147455867082,
+      "grad_norm": 0.2609243094921112,
+      "learning_rate": 3.534747484603587e-05,
+      "loss": 0.108,
+      "num_input_tokens_seen": 31806520,
+      "step": 437
+    },
+    {
+      "epoch": 7.18276220145379,
+      "grad_norm": 0.3314545452594757,
+      "learning_rate": 3.528785896512772e-05,
+      "loss": 0.1154,
+      "num_input_tokens_seen": 31860464,
+      "step": 438
+    },
+    {
+      "epoch": 7.1993769470404985,
+      "grad_norm": 0.29827994108200073,
+      "learning_rate": 3.5228172572386146e-05,
+      "loss": 0.2637,
+      "num_input_tokens_seen": 31921424,
+      "step": 439
+    },
+    {
+      "epoch": 7.215991692627207,
+      "grad_norm": 0.3889339566230774,
+      "learning_rate": 3.516841607689501e-05,
+      "loss": 0.1164,
+      "num_input_tokens_seen": 31981064,
+      "step": 440
+    },
+    {
+      "epoch": 7.232606438213915,
+      "grad_norm": 0.3092401325702667,
+      "learning_rate": 3.510858988821863e-05,
+      "loss": 0.1085,
+      "num_input_tokens_seen": 32050648,
+      "step": 441
+    },
+    {
+      "epoch": 7.249221183800623,
+      "grad_norm": 0.28353771567344666,
+      "learning_rate": 3.504869441639901e-05,
+      "loss": 0.1026,
+      "num_input_tokens_seen": 32118584,
+      "step": 442
+    },
+    {
+      "epoch": 7.265835929387332,
+      "grad_norm": 0.26136070489883423,
+      "learning_rate": 3.4988730071953004e-05,
+      "loss": 0.1035,
+      "num_input_tokens_seen": 32206384,
+      "step": 443
+    },
+    {
+      "epoch": 7.282450674974039,
+      "grad_norm": 0.26213783025741577,
+      "learning_rate": 3.4928697265869515e-05,
+      "loss": 0.0904,
+      "num_input_tokens_seen": 32299040,
+      "step": 444
+    },
+    {
+      "epoch": 7.299065420560748,
+      "grad_norm": 0.2811000347137451,
+      "learning_rate": 3.486859640960668e-05,
+      "loss": 0.1053,
+      "num_input_tokens_seen": 32355624,
+      "step": 445
+    },
+    {
+      "epoch": 7.315680166147456,
+      "grad_norm": 0.280838280916214,
+      "learning_rate": 3.480842791508904e-05,
+      "loss": 0.1227,
+      "num_input_tokens_seen": 32427792,
+      "step": 446
+    },
+    {
+      "epoch": 7.332294911734164,
+      "grad_norm": 0.3092939555644989,
+      "learning_rate": 3.474819219470471e-05,
+      "loss": 0.1104,
+      "num_input_tokens_seen": 32508696,
+      "step": 447
+    },
+    {
+      "epoch": 7.348909657320872,
+      "grad_norm": 0.25746166706085205,
+      "learning_rate": 3.4687889661302576e-05,
+      "loss": 0.1099,
+      "num_input_tokens_seen": 32601312,
+      "step": 448
+    },
+    {
+      "epoch": 7.365524402907581,
+      "grad_norm": 0.2721652090549469,
+      "learning_rate": 3.4627520728189456e-05,
+      "loss": 0.0966,
+      "num_input_tokens_seen": 32680256,
+      "step": 449
+    },
+    {
+      "epoch": 7.382139148494288,
+      "grad_norm": 0.3046250641345978,
+      "learning_rate": 3.456708580912725e-05,
+      "loss": 0.1138,
+      "num_input_tokens_seen": 32738816,
+      "step": 450
+    },
+    {
+      "epoch": 7.398753894080997,
+      "grad_norm": 0.28120365738868713,
+      "learning_rate": 3.4506585318330125e-05,
+      "loss": 0.1026,
+      "num_input_tokens_seen": 32813240,
+      "step": 451
+    },
+    {
+      "epoch": 7.415368639667705,
+      "grad_norm": 0.29472339153289795,
+      "learning_rate": 3.444601967046168e-05,
+      "loss": 0.1183,
+      "num_input_tokens_seen": 32889680,
+      "step": 452
+    },
+    {
+      "epoch": 7.431983385254413,
+      "grad_norm": 0.32324308156967163,
+      "learning_rate": 3.438538928063208e-05,
+      "loss": 0.1218,
+      "num_input_tokens_seen": 32964760,
+      "step": 453
+    },
+    {
+      "epoch": 7.4485981308411215,
+      "grad_norm": 0.2858980894088745,
+      "learning_rate": 3.432469456439523e-05,
+      "loss": 0.1226,
+      "num_input_tokens_seen": 33048992,
+      "step": 454
+    },
+    {
+      "epoch": 7.46521287642783,
+      "grad_norm": 0.30722907185554504,
+      "learning_rate": 3.426393593774591e-05,
+      "loss": 0.1127,
+      "num_input_tokens_seen": 33130200,
+      "step": 455
+    },
+    {
+      "epoch": 7.481827622014538,
+      "grad_norm": 0.2962627112865448,
+      "learning_rate": 3.4203113817116957e-05,
+      "loss": 0.1057,
+      "num_input_tokens_seen": 33223024,
+      "step": 456
+    },
+    {
+      "epoch": 7.498442367601246,
+      "grad_norm": 0.29225122928619385,
+      "learning_rate": 3.414222861937636e-05,
+      "loss": 0.1127,
+      "num_input_tokens_seen": 33303120,
+      "step": 457
+    },
+    {
+      "epoch": 7.515057113187955,
+      "grad_norm": 0.2806551456451416,
+      "learning_rate": 3.408128076182446e-05,
+      "loss": 0.1201,
+      "num_input_tokens_seen": 33364984,
+      "step": 458
+    },
+    {
+      "epoch": 7.531671858774662,
+      "grad_norm": 0.30782198905944824,
+      "learning_rate": 3.402027066219105e-05,
+      "loss": 0.1318,
+      "num_input_tokens_seen": 33427352,
+      "step": 459
+    },
+    {
+      "epoch": 7.548286604361371,
+      "grad_norm": 0.27443554997444153,
+      "learning_rate": 3.39591987386325e-05,
+      "loss": 0.1045,
+      "num_input_tokens_seen": 33481272,
+      "step": 460
+    },
+    {
+      "epoch": 7.564901349948079,
+      "grad_norm": 0.29877278208732605,
+      "learning_rate": 3.389806540972898e-05,
+      "loss": 0.1,
+      "num_input_tokens_seen": 33538904,
+      "step": 461
+    },
+    {
+      "epoch": 7.581516095534787,
+      "grad_norm": 0.24680256843566895,
+      "learning_rate": 3.383687109448143e-05,
+      "loss": 0.1111,
+      "num_input_tokens_seen": 33635976,
+      "step": 462
+    },
+    {
+      "epoch": 7.598130841121495,
+      "grad_norm": 0.2937908470630646,
+      "learning_rate": 3.377561621230887e-05,
+      "loss": 0.1123,
+      "num_input_tokens_seen": 33711184,
+      "step": 463
+    },
+    {
+      "epoch": 7.614745586708204,
+      "grad_norm": 0.2607901096343994,
+      "learning_rate": 3.3714301183045385e-05,
+      "loss": 0.0937,
+      "num_input_tokens_seen": 33778848,
+      "step": 464
+    },
+    {
+      "epoch": 7.6313603322949115,
+      "grad_norm": 0.28291332721710205,
+      "learning_rate": 3.365292642693732e-05,
+      "loss": 0.1023,
+      "num_input_tokens_seen": 33866024,
+      "step": 465
+    },
+    {
+      "epoch": 7.64797507788162,
+      "grad_norm": 0.2455737590789795,
+      "learning_rate": 3.359149236464041e-05,
+      "loss": 0.1179,
+      "num_input_tokens_seen": 33978144,
+      "step": 466
+    },
+    {
+      "epoch": 7.6645898234683285,
+      "grad_norm": 0.32317623496055603,
+      "learning_rate": 3.35299994172168e-05,
+      "loss": 0.1326,
+      "num_input_tokens_seen": 34047480,
+      "step": 467
+    },
+    {
+      "epoch": 7.681204569055036,
+      "grad_norm": 0.23927746713161469,
+      "learning_rate": 3.346844800613229e-05,
+      "loss": 0.1058,
+      "num_input_tokens_seen": 34134480,
+      "step": 468
+    },
+    {
+      "epoch": 7.697819314641745,
+      "grad_norm": 0.28762567043304443,
+      "learning_rate": 3.340683855325335e-05,
+      "loss": 0.1077,
+      "num_input_tokens_seen": 34190176,
+      "step": 469
+    },
+    {
+      "epoch": 7.714434060228453,
+      "grad_norm": 0.32403019070625305,
+      "learning_rate": 3.3345171480844275e-05,
+      "loss": 0.1224,
+      "num_input_tokens_seen": 34267336,
+      "step": 470
+    },
+    {
+      "epoch": 7.731048805815161,
+      "grad_norm": 0.2981953024864197,
+      "learning_rate": 3.3283447211564276e-05,
+      "loss": 0.116,
+      "num_input_tokens_seen": 34333616,
+      "step": 471
+    },
+    {
+      "epoch": 7.747663551401869,
+      "grad_norm": 0.24373090267181396,
+      "learning_rate": 3.322166616846458e-05,
+      "loss": 0.1153,
+      "num_input_tokens_seen": 34404000,
+      "step": 472
+    },
+    {
+      "epoch": 7.764278296988578,
+      "grad_norm": 0.3025456666946411,
+      "learning_rate": 3.315982877498555e-05,
+      "loss": 0.111,
+      "num_input_tokens_seen": 34466048,
+      "step": 473
+    },
+    {
+      "epoch": 7.780893042575285,
+      "grad_norm": 0.2972700893878937,
+      "learning_rate": 3.309793545495374e-05,
+      "loss": 0.1099,
+      "num_input_tokens_seen": 34547312,
+      "step": 474
+    },
+    {
+      "epoch": 7.797507788161994,
+      "grad_norm": 0.32208284735679626,
+      "learning_rate": 3.303598663257904e-05,
+      "loss": 0.1029,
+      "num_input_tokens_seen": 34600544,
+      "step": 475
+    },
+    {
+      "epoch": 7.814122533748702,
+      "grad_norm": 0.2841314971446991,
+      "learning_rate": 3.2973982732451755e-05,
+      "loss": 0.1111,
+      "num_input_tokens_seen": 34660792,
+      "step": 476
+    },
+    {
+      "epoch": 7.83073727933541,
+      "grad_norm": 0.25448325276374817,
+      "learning_rate": 3.2911924179539656e-05,
+      "loss": 0.162,
+      "num_input_tokens_seen": 34778440,
+      "step": 477
+    },
+    {
+      "epoch": 7.8473520249221185,
+      "grad_norm": 0.30900058150291443,
+      "learning_rate": 3.284981139918513e-05,
+      "loss": 0.1161,
+      "num_input_tokens_seen": 34849760,
+      "step": 478
+    },
+    {
+      "epoch": 7.863966770508826,
+      "grad_norm": 0.26905959844589233,
+      "learning_rate": 3.278764481710221e-05,
+      "loss": 0.0973,
+      "num_input_tokens_seen": 34940776,
+      "step": 479
+    },
+    {
+      "epoch": 7.880581516095535,
+      "grad_norm": 0.2753785252571106,
+      "learning_rate": 3.272542485937369e-05,
+      "loss": 0.1131,
+      "num_input_tokens_seen": 35018104,
+      "step": 480
+    },
+    {
+      "epoch": 7.897196261682243,
+      "grad_norm": 0.27028700709342957,
+      "learning_rate": 3.26631519524482e-05,
+      "loss": 0.1012,
+      "num_input_tokens_seen": 35079744,
+      "step": 481
+    },
+    {
+      "epoch": 7.913811007268951,
+      "grad_norm": 0.2808922231197357,
+      "learning_rate": 3.260082652313726e-05,
+      "loss": 0.1,
+      "num_input_tokens_seen": 35132808,
+      "step": 482
+    },
+    {
+      "epoch": 7.930425752855659,
+      "grad_norm": 0.2821935713291168,
+      "learning_rate": 3.253844899861239e-05,
+      "loss": 0.1005,
+      "num_input_tokens_seen": 35197816,
+      "step": 483
+    },
+    {
+      "epoch": 7.947040498442368,
+      "grad_norm": 0.2640717029571533,
+      "learning_rate": 3.247601980640217e-05,
+      "loss": 0.0914,
+      "num_input_tokens_seen": 35275528,
+      "step": 484
+    },
+    {
+      "epoch": 7.963655244029075,
+      "grad_norm": 0.31811895966529846,
+      "learning_rate": 3.241353937438927e-05,
+      "loss": 0.119,
+      "num_input_tokens_seen": 35333280,
+      "step": 485
+    },
+    {
+      "epoch": 7.980269989615784,
+      "grad_norm": 0.30607783794403076,
+      "learning_rate": 3.23510081308076e-05,
+      "loss": 0.1156,
+      "num_input_tokens_seen": 35412944,
+      "step": 486
+    },
+    {
+      "epoch": 7.996884735202492,
+      "grad_norm": 0.29681020975112915,
+      "learning_rate": 3.228842650423929e-05,
+      "loss": 0.1217,
+      "num_input_tokens_seen": 35485056,
+      "step": 487
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.6129417419433594,
+      "learning_rate": 3.222579492361179e-05,
+      "loss": 0.1188,
+      "num_input_tokens_seen": 35494824,
+      "step": 488
+    },
+    {
+      "epoch": 8.016614745586708,
+      "grad_norm": 0.26822784543037415,
+      "learning_rate": 3.2163113818194964e-05,
+      "loss": 0.0967,
+      "num_input_tokens_seen": 35557768,
+      "step": 489
+    },
+    {
+      "epoch": 8.033229491173417,
+      "grad_norm": 0.261436402797699,
+      "learning_rate": 3.210038361759807e-05,
+      "loss": 0.0989,
+      "num_input_tokens_seen": 35613120,
+      "step": 490
+    },
+    {
+      "epoch": 8.049844236760125,
+      "grad_norm": 0.26140278577804565,
+      "learning_rate": 3.2037604751766885e-05,
+      "loss": 0.096,
+      "num_input_tokens_seen": 35674176,
+      "step": 491
+    },
+    {
+      "epoch": 8.066458982346832,
+      "grad_norm": 0.285133957862854,
+      "learning_rate": 3.1974777650980735e-05,
+      "loss": 0.1049,
+      "num_input_tokens_seen": 35786664,
+      "step": 492
+    },
+    {
+      "epoch": 8.083073727933542,
+      "grad_norm": 0.34326228499412537,
+      "learning_rate": 3.191190274584952e-05,
+      "loss": 0.1102,
+      "num_input_tokens_seen": 35840720,
+      "step": 493
+    },
+    {
+      "epoch": 8.09968847352025,
+      "grad_norm": 0.23448359966278076,
+      "learning_rate": 3.184898046731082e-05,
+      "loss": 0.0864,
+      "num_input_tokens_seen": 35936736,
+      "step": 494
+    },
+    {
+      "epoch": 8.116303219106957,
+      "grad_norm": 0.2525218427181244,
+      "learning_rate": 3.178601124662686e-05,
+      "loss": 0.1,
+      "num_input_tokens_seen": 36013800,
+      "step": 495
+    },
+    {
+      "epoch": 8.132917964693666,
+      "grad_norm": 0.27955323457717896,
+      "learning_rate": 3.172299551538164e-05,
+      "loss": 0.0957,
+      "num_input_tokens_seen": 36097904,
+      "step": 496
+    },
+    {
+      "epoch": 8.149532710280374,
+      "grad_norm": 0.2300575226545334,
+      "learning_rate": 3.165993370547794e-05,
+      "loss": 0.077,
+      "num_input_tokens_seen": 36195544,
+      "step": 497
+    },
+    {
+      "epoch": 8.166147455867081,
+      "grad_norm": 0.3093201816082001,
+      "learning_rate": 3.1596826249134324e-05,
+      "loss": 0.1195,
+      "num_input_tokens_seen": 36261256,
+      "step": 498
+    },
+    {
+      "epoch": 8.18276220145379,
+      "grad_norm": 0.36045631766319275,
+      "learning_rate": 3.153367357888224e-05,
+      "loss": 0.1199,
+      "num_input_tokens_seen": 36325024,
+      "step": 499
+    },
+    {
+      "epoch": 8.199376947040498,
+      "grad_norm": 0.28502553701400757,
+      "learning_rate": 3.147047612756302e-05,
+      "loss": 0.1066,
+      "num_input_tokens_seen": 36377368,
+      "step": 500
+    },
+    {
+      "epoch": 8.215991692627206,
+      "grad_norm": 0.2474561631679535,
+      "learning_rate": 3.140723432832492e-05,
+      "loss": 0.0908,
+      "num_input_tokens_seen": 36459240,
+      "step": 501
+    },
+    {
+      "epoch": 8.232606438213915,
+      "grad_norm": 0.2631191313266754,
+      "learning_rate": 3.1343948614620145e-05,
+      "loss": 0.1022,
+      "num_input_tokens_seen": 36553088,
+      "step": 502
+    },
+    {
+      "epoch": 8.249221183800623,
+      "grad_norm": 0.34360605478286743,
+      "learning_rate": 3.128061942020189e-05,
+      "loss": 0.2542,
+      "num_input_tokens_seen": 36611464,
+      "step": 503
+    },
+    {
+      "epoch": 8.26583592938733,
+      "grad_norm": 0.21943052113056183,
+      "learning_rate": 3.121724717912138e-05,
+      "loss": 0.0817,
+      "num_input_tokens_seen": 36705696,
+      "step": 504
+    },
+    {
+      "epoch": 8.28245067497404,
+      "grad_norm": 0.27961695194244385,
+      "learning_rate": 3.115383232572483e-05,
+      "loss": 0.0879,
+      "num_input_tokens_seen": 36762744,
+      "step": 505
+    },
+    {
+      "epoch": 8.299065420560748,
+      "grad_norm": 0.3083556294441223,
+      "learning_rate": 3.109037529465056e-05,
+      "loss": 0.1057,
+      "num_input_tokens_seen": 36827816,
+      "step": 506
+    },
+    {
+      "epoch": 8.315680166147455,
+      "grad_norm": 0.24268627166748047,
+      "learning_rate": 3.102687652082597e-05,
+      "loss": 0.0924,
+      "num_input_tokens_seen": 36931424,
+      "step": 507
+    },
+    {
+      "epoch": 8.332294911734165,
+      "grad_norm": 0.4235898554325104,
+      "learning_rate": 3.0963336439464526e-05,
+      "loss": 0.093,
+      "num_input_tokens_seen": 36991464,
+      "step": 508
+    },
+    {
+      "epoch": 8.348909657320872,
+      "grad_norm": 0.2394886463880539,
+      "learning_rate": 3.089975548606283e-05,
+      "loss": 0.0784,
+      "num_input_tokens_seen": 37092928,
+      "step": 509
+    },
+    {
+      "epoch": 8.36552440290758,
+      "grad_norm": 0.2834029495716095,
+      "learning_rate": 3.083613409639764e-05,
+      "loss": 0.0953,
+      "num_input_tokens_seen": 37168792,
+      "step": 510
+    },
+    {
+      "epoch": 8.38213914849429,
+      "grad_norm": 0.3311532735824585,
+      "learning_rate": 3.0772472706522806e-05,
+      "loss": 0.0964,
+      "num_input_tokens_seen": 37258864,
+      "step": 511
+    },
+    {
+      "epoch": 8.398753894080997,
+      "grad_norm": 0.28146132826805115,
+      "learning_rate": 3.0708771752766394e-05,
+      "loss": 0.1084,
+      "num_input_tokens_seen": 37343224,
+      "step": 512
+    },
+    {
+      "epoch": 8.415368639667705,
+      "grad_norm": 0.35688018798828125,
+      "learning_rate": 3.06450316717276e-05,
+      "loss": 0.1068,
+      "num_input_tokens_seen": 37395488,
+      "step": 513
+    },
+    {
+      "epoch": 8.431983385254414,
+      "grad_norm": 0.24343422055244446,
+      "learning_rate": 3.0581252900273786e-05,
+      "loss": 0.0849,
+      "num_input_tokens_seen": 37473248,
+      "step": 514
+    },
+    {
+      "epoch": 8.448598130841122,
+      "grad_norm": 0.30105647444725037,
+      "learning_rate": 3.0517435875537536e-05,
+      "loss": 0.0883,
+      "num_input_tokens_seen": 37532096,
+      "step": 515
+    },
+    {
+      "epoch": 8.46521287642783,
+      "grad_norm": 0.24181726574897766,
+      "learning_rate": 3.045358103491357e-05,
+      "loss": 0.0839,
+      "num_input_tokens_seen": 37622328,
+      "step": 516
+    },
+    {
+      "epoch": 8.481827622014539,
+      "grad_norm": 0.3248032331466675,
+      "learning_rate": 3.038968881605583e-05,
+      "loss": 0.0999,
+      "num_input_tokens_seen": 37686304,
+      "step": 517
+    },
+    {
+      "epoch": 8.498442367601246,
+      "grad_norm": 0.28919899463653564,
+      "learning_rate": 3.0325759656874418e-05,
+      "loss": 0.1023,
+      "num_input_tokens_seen": 37770856,
+      "step": 518
+    },
+    {
+      "epoch": 8.515057113187954,
+      "grad_norm": 0.32156485319137573,
+      "learning_rate": 3.026179399553264e-05,
+      "loss": 0.0907,
+      "num_input_tokens_seen": 37834072,
+      "step": 519
+    },
+    {
+      "epoch": 8.531671858774663,
+      "grad_norm": 0.36424997448921204,
+      "learning_rate": 3.0197792270443982e-05,
+      "loss": 0.0899,
+      "num_input_tokens_seen": 37889928,
+      "step": 520
+    },
+    {
+      "epoch": 8.54828660436137,
+      "grad_norm": 0.3248436450958252,
+      "learning_rate": 3.0133754920269103e-05,
+      "loss": 0.2102,
+      "num_input_tokens_seen": 37971296,
+      "step": 521
+    },
+    {
+      "epoch": 8.564901349948078,
+      "grad_norm": 0.3297094702720642,
+      "learning_rate": 3.0069682383912813e-05,
+      "loss": 0.0978,
+      "num_input_tokens_seen": 38049288,
+      "step": 522
+    },
+    {
+      "epoch": 8.581516095534788,
+      "grad_norm": 0.3013320565223694,
+      "learning_rate": 3.0005575100521118e-05,
+      "loss": 0.1069,
+      "num_input_tokens_seen": 38123392,
+      "step": 523
+    },
+    {
+      "epoch": 8.598130841121495,
+      "grad_norm": 0.2589408755302429,
+      "learning_rate": 2.9941433509478156e-05,
+      "loss": 0.0973,
+      "num_input_tokens_seen": 38208264,
+      "step": 524
+    },
+    {
+      "epoch": 8.614745586708203,
+      "grad_norm": 0.30129629373550415,
+      "learning_rate": 2.9877258050403212e-05,
+      "loss": 0.1045,
+      "num_input_tokens_seen": 38258192,
+      "step": 525
+    },
+    {
+      "epoch": 8.631360332294912,
+      "grad_norm": 0.28428325057029724,
+      "learning_rate": 2.9813049163147688e-05,
+      "loss": 0.1014,
+      "num_input_tokens_seen": 38332408,
+      "step": 526
+    },
+    {
+      "epoch": 8.64797507788162,
+      "grad_norm": 0.23376743495464325,
+      "learning_rate": 2.974880728779212e-05,
+      "loss": 0.0816,
+      "num_input_tokens_seen": 38404960,
+      "step": 527
+    },
+    {
+      "epoch": 8.664589823468328,
+      "grad_norm": 0.26018524169921875,
+      "learning_rate": 2.9684532864643122e-05,
+      "loss": 0.1103,
+      "num_input_tokens_seen": 38481704,
+      "step": 528
+    },
+    {
+      "epoch": 8.681204569055037,
+      "grad_norm": 0.3135533332824707,
+      "learning_rate": 2.9620226334230388e-05,
+      "loss": 0.0833,
+      "num_input_tokens_seen": 38546304,
+      "step": 529
+    },
+    {
+      "epoch": 8.697819314641745,
+      "grad_norm": 0.318968266248703,
+      "learning_rate": 2.9555888137303695e-05,
+      "loss": 0.1225,
+      "num_input_tokens_seen": 38621024,
+      "step": 530
+    },
+    {
+      "epoch": 8.714434060228452,
+      "grad_norm": 0.2814541161060333,
+      "learning_rate": 2.949151871482982e-05,
+      "loss": 0.0874,
+      "num_input_tokens_seen": 38679368,
+      "step": 531
+    },
+    {
+      "epoch": 8.731048805815162,
+      "grad_norm": 0.304311603307724,
+      "learning_rate": 2.9427118507989586e-05,
+      "loss": 0.1045,
+      "num_input_tokens_seen": 38753984,
+      "step": 532
+    },
+    {
+      "epoch": 8.74766355140187,
+      "grad_norm": 0.28034818172454834,
+      "learning_rate": 2.93626879581748e-05,
+      "loss": 0.0939,
+      "num_input_tokens_seen": 38808336,
+      "step": 533
+    },
+    {
+      "epoch": 8.764278296988577,
+      "grad_norm": 0.29433172941207886,
+      "learning_rate": 2.929822750698524e-05,
+      "loss": 0.2031,
+      "num_input_tokens_seen": 38876624,
+      "step": 534
+    },
+    {
+      "epoch": 8.780893042575286,
+      "grad_norm": 0.2822960913181305,
+      "learning_rate": 2.9233737596225613e-05,
+      "loss": 0.0906,
+      "num_input_tokens_seen": 38933576,
+      "step": 535
+    },
+    {
+      "epoch": 8.797507788161994,
+      "grad_norm": 0.26343122124671936,
+      "learning_rate": 2.916921866790256e-05,
+      "loss": 0.0876,
+      "num_input_tokens_seen": 39050816,
+      "step": 536
+    },
+    {
+      "epoch": 8.814122533748701,
+      "grad_norm": 0.2758285403251648,
+      "learning_rate": 2.9104671164221576e-05,
+      "loss": 0.0899,
+      "num_input_tokens_seen": 39101856,
+      "step": 537
+    },
+    {
+      "epoch": 8.83073727933541,
+      "grad_norm": 0.2846848964691162,
+      "learning_rate": 2.9040095527584032e-05,
+      "loss": 0.094,
+      "num_input_tokens_seen": 39161928,
+      "step": 538
+    },
+    {
+      "epoch": 8.847352024922118,
+      "grad_norm": 0.32669830322265625,
+      "learning_rate": 2.897549220058411e-05,
+      "loss": 0.1077,
+      "num_input_tokens_seen": 39216048,
+      "step": 539
+    },
+    {
+      "epoch": 8.863966770508826,
+      "grad_norm": 0.28703227639198303,
+      "learning_rate": 2.8910861626005776e-05,
+      "loss": 0.0857,
+      "num_input_tokens_seen": 39317320,
+      "step": 540
+    },
+    {
+      "epoch": 8.880581516095535,
+      "grad_norm": 0.32840797305107117,
+      "learning_rate": 2.884620424681976e-05,
+      "loss": 0.0981,
+      "num_input_tokens_seen": 39383120,
+      "step": 541
+    },
+    {
+      "epoch": 8.897196261682243,
+      "grad_norm": 0.2913094758987427,
+      "learning_rate": 2.8781520506180486e-05,
+      "loss": 0.0969,
+      "num_input_tokens_seen": 39458584,
+      "step": 542
+    },
+    {
+      "epoch": 8.91381100726895,
+      "grad_norm": 0.23351797461509705,
+      "learning_rate": 2.871681084742308e-05,
+      "loss": 0.0809,
+      "num_input_tokens_seen": 39535152,
+      "step": 543
+    },
+    {
+      "epoch": 8.93042575285566,
+      "grad_norm": 0.32754406332969666,
+      "learning_rate": 2.8652075714060295e-05,
+      "loss": 0.1064,
+      "num_input_tokens_seen": 39590360,
+      "step": 544
+    },
+    {
+      "epoch": 8.947040498442368,
+      "grad_norm": 0.28002288937568665,
+      "learning_rate": 2.858731554977948e-05,
+      "loss": 0.0897,
+      "num_input_tokens_seen": 39669984,
+      "step": 545
+    },
+    {
+      "epoch": 8.963655244029075,
+      "grad_norm": 0.2599540054798126,
+      "learning_rate": 2.8522530798439567e-05,
+      "loss": 0.0883,
+      "num_input_tokens_seen": 39757392,
+      "step": 546
+    },
+    {
+      "epoch": 8.980269989615785,
+      "grad_norm": 0.29728278517723083,
+      "learning_rate": 2.845772190406798e-05,
+      "loss": 0.1024,
+      "num_input_tokens_seen": 39848064,
+      "step": 547
+    },
+    {
+      "epoch": 8.996884735202492,
+      "grad_norm": 0.27901116013526917,
+      "learning_rate": 2.8392889310857612e-05,
+      "loss": 0.093,
+      "num_input_tokens_seen": 39922288,
+      "step": 548
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.7538627982139587,
+      "learning_rate": 2.832803346316381e-05,
+      "loss": 0.0997,
+      "num_input_tokens_seen": 39932640,
+      "step": 549
+    },
+    {
+      "epoch": 9.016614745586708,
+      "grad_norm": 0.259941965341568,
+      "learning_rate": 2.8263154805501297e-05,
+      "loss": 0.0756,
+      "num_input_tokens_seen": 40005688,
+      "step": 550
+    },
+    {
+      "epoch": 9.033229491173417,
+      "grad_norm": 0.27237728238105774,
+      "learning_rate": 2.819825378254111e-05,
+      "loss": 0.0851,
+      "num_input_tokens_seen": 40057120,
+      "step": 551
+    },
+    {
+      "epoch": 9.049844236760125,
+      "grad_norm": 0.28233349323272705,
+      "learning_rate": 2.8133330839107608e-05,
+      "loss": 0.0841,
+      "num_input_tokens_seen": 40135992,
+      "step": 552
+    },
+    {
+      "epoch": 9.066458982346832,
+      "grad_norm": 0.278773695230484,
+      "learning_rate": 2.8068386420175375e-05,
+      "loss": 0.1069,
+      "num_input_tokens_seen": 40196928,
+      "step": 553
+    },
+    {
+      "epoch": 9.083073727933542,
+      "grad_norm": 0.28890740871429443,
+      "learning_rate": 2.8003420970866177e-05,
+      "loss": 0.0851,
+      "num_input_tokens_seen": 40269392,
+      "step": 554
+    },
+    {
+      "epoch": 9.09968847352025,
+      "grad_norm": 0.33677053451538086,
+      "learning_rate": 2.7938434936445945e-05,
+      "loss": 0.0899,
+      "num_input_tokens_seen": 40350080,
+      "step": 555
+    },
+    {
+      "epoch": 9.116303219106957,
+      "grad_norm": 0.33169543743133545,
+      "learning_rate": 2.787342876232167e-05,
+      "loss": 0.0991,
+      "num_input_tokens_seen": 40416360,
+      "step": 556
+    },
+    {
+      "epoch": 9.132917964693666,
+      "grad_norm": 0.2758564054965973,
+      "learning_rate": 2.780840289403839e-05,
+      "loss": 0.0733,
+      "num_input_tokens_seen": 40490432,
+      "step": 557
+    },
+    {
+      "epoch": 9.149532710280374,
+      "grad_norm": 0.29123055934906006,
+      "learning_rate": 2.774335777727613e-05,
+      "loss": 0.0927,
+      "num_input_tokens_seen": 40561784,
+      "step": 558
+    },
+    {
+      "epoch": 9.166147455867081,
+      "grad_norm": 0.26718494296073914,
+      "learning_rate": 2.7678293857846844e-05,
+      "loss": 0.083,
+      "num_input_tokens_seen": 40641728,
+      "step": 559
+    },
+    {
+      "epoch": 9.18276220145379,
+      "grad_norm": 0.30949828028678894,
+      "learning_rate": 2.761321158169134e-05,
+      "loss": 0.0975,
+      "num_input_tokens_seen": 40700744,
+      "step": 560
+    },
+    {
+      "epoch": 9.199376947040498,
+      "grad_norm": 0.3187688887119293,
+      "learning_rate": 2.754811139487625e-05,
+      "loss": 0.0995,
+      "num_input_tokens_seen": 40748048,
+      "step": 561
+    },
+    {
+      "epoch": 9.215991692627206,
+      "grad_norm": 0.2677261531352997,
+      "learning_rate": 2.7482993743590978e-05,
+      "loss": 0.073,
+      "num_input_tokens_seen": 40810104,
+      "step": 562
+    },
+    {
+      "epoch": 9.232606438213915,
+      "grad_norm": 0.3078905940055847,
+      "learning_rate": 2.7417859074144604e-05,
+      "loss": 0.0952,
+      "num_input_tokens_seen": 40899480,
+      "step": 563
+    },
+    {
+      "epoch": 9.249221183800623,
+      "grad_norm": 0.25825080275535583,
+      "learning_rate": 2.7352707832962865e-05,
+      "loss": 0.0808,
+      "num_input_tokens_seen": 40993536,
+      "step": 564
+    },
+    {
+      "epoch": 9.26583592938733,
+      "grad_norm": 0.23915870487689972,
+      "learning_rate": 2.7287540466585065e-05,
+      "loss": 0.0698,
+      "num_input_tokens_seen": 41060848,
+      "step": 565
+    },
+    {
+      "epoch": 9.28245067497404,
+      "grad_norm": 0.28298041224479675,
+      "learning_rate": 2.7222357421661042e-05,
+      "loss": 0.0931,
+      "num_input_tokens_seen": 41138352,
+      "step": 566
+    },
+    {
+      "epoch": 9.299065420560748,
+      "grad_norm": 0.3111128509044647,
+      "learning_rate": 2.7157159144948092e-05,
+      "loss": 0.0996,
+      "num_input_tokens_seen": 41212624,
+      "step": 567
+    },
+    {
+      "epoch": 9.315680166147455,
+      "grad_norm": 0.2831686735153198,
+      "learning_rate": 2.7091946083307896e-05,
+      "loss": 0.0764,
+      "num_input_tokens_seen": 41279472,
+      "step": 568
+    },
+    {
+      "epoch": 9.332294911734165,
+      "grad_norm": 0.27329790592193604,
+      "learning_rate": 2.7026718683703473e-05,
+      "loss": 0.0942,
+      "num_input_tokens_seen": 41353544,
+      "step": 569
+    },
+    {
+      "epoch": 9.348909657320872,
+      "grad_norm": 0.23970037698745728,
+      "learning_rate": 2.6961477393196126e-05,
+      "loss": 0.0718,
+      "num_input_tokens_seen": 41444896,
+      "step": 570
+    },
+    {
+      "epoch": 9.36552440290758,
+      "grad_norm": 0.2980060279369354,
+      "learning_rate": 2.6896222658942348e-05,
+      "loss": 0.0906,
+      "num_input_tokens_seen": 41505152,
+      "step": 571
+    },
+    {
+      "epoch": 9.38213914849429,
+      "grad_norm": 0.32479342818260193,
+      "learning_rate": 2.6830954928190794e-05,
+      "loss": 0.1081,
+      "num_input_tokens_seen": 41566696,
+      "step": 572
+    },
+    {
+      "epoch": 9.398753894080997,
+      "grad_norm": 0.2624221444129944,
+      "learning_rate": 2.6765674648279172e-05,
+      "loss": 0.0922,
+      "num_input_tokens_seen": 41641736,
+      "step": 573
+    },
+    {
+      "epoch": 9.415368639667705,
+      "grad_norm": 0.22994311153888702,
+      "learning_rate": 2.6700382266631206e-05,
+      "loss": 0.0748,
+      "num_input_tokens_seen": 41740008,
+      "step": 574
+    },
+    {
+      "epoch": 9.431983385254414,
+      "grad_norm": 0.264123797416687,
+      "learning_rate": 2.663507823075358e-05,
+      "loss": 0.0813,
+      "num_input_tokens_seen": 41842808,
+      "step": 575
+    },
+    {
+      "epoch": 9.448598130841122,
+      "grad_norm": 0.2524157464504242,
+      "learning_rate": 2.656976298823284e-05,
+      "loss": 0.0779,
+      "num_input_tokens_seen": 41917128,
+      "step": 576
+    },
+    {
+      "epoch": 9.46521287642783,
+      "grad_norm": 0.2805820405483246,
+      "learning_rate": 2.6504436986732338e-05,
+      "loss": 0.0882,
+      "num_input_tokens_seen": 41984232,
+      "step": 577
+    },
+    {
+      "epoch": 9.481827622014539,
+      "grad_norm": 0.2692512273788452,
+      "learning_rate": 2.6439100673989187e-05,
+      "loss": 0.0905,
+      "num_input_tokens_seen": 42047216,
+      "step": 578
+    },
+    {
+      "epoch": 9.498442367601246,
+      "grad_norm": 0.26741349697113037,
+      "learning_rate": 2.637375449781115e-05,
+      "loss": 0.081,
+      "num_input_tokens_seen": 42122072,
+      "step": 579
+    },
+    {
+      "epoch": 9.515057113187954,
+      "grad_norm": 0.24825426936149597,
+      "learning_rate": 2.63083989060736e-05,
+      "loss": 0.0777,
+      "num_input_tokens_seen": 42198480,
+      "step": 580
+    },
+    {
+      "epoch": 9.531671858774663,
+      "grad_norm": 0.26209551095962524,
+      "learning_rate": 2.624303434671645e-05,
+      "loss": 0.0817,
+      "num_input_tokens_seen": 42289336,
+      "step": 581
+    },
+    {
+      "epoch": 9.54828660436137,
+      "grad_norm": 0.3296469748020172,
+      "learning_rate": 2.6177661267741065e-05,
+      "loss": 0.0931,
+      "num_input_tokens_seen": 42352288,
+      "step": 582
+    },
+    {
+      "epoch": 9.564901349948078,
+      "grad_norm": 0.29804670810699463,
+      "learning_rate": 2.611228011720722e-05,
+      "loss": 0.0852,
+      "num_input_tokens_seen": 42453832,
+      "step": 583
+    },
+    {
+      "epoch": 9.581516095534788,
+      "grad_norm": 0.30008113384246826,
+      "learning_rate": 2.604689134322999e-05,
+      "loss": 0.0917,
+      "num_input_tokens_seen": 42514704,
+      "step": 584
+    },
+    {
+      "epoch": 9.598130841121495,
+      "grad_norm": 0.3668535649776459,
+      "learning_rate": 2.598149539397672e-05,
+      "loss": 0.184,
+      "num_input_tokens_seen": 42576344,
+      "step": 585
+    },
+    {
+      "epoch": 9.614745586708203,
+      "grad_norm": 0.26044410467147827,
+      "learning_rate": 2.591609271766391e-05,
+      "loss": 0.0799,
+      "num_input_tokens_seen": 42662824,
+      "step": 586
+    },
+    {
+      "epoch": 9.631360332294912,
+      "grad_norm": 0.23652903735637665,
+      "learning_rate": 2.5850683762554184e-05,
+      "loss": 0.0681,
+      "num_input_tokens_seen": 42752496,
+      "step": 587
+    },
+    {
+      "epoch": 9.64797507788162,
+      "grad_norm": 0.2591301202774048,
+      "learning_rate": 2.578526897695321e-05,
+      "loss": 0.07,
+      "num_input_tokens_seen": 42826064,
+      "step": 588
+    },
+    {
+      "epoch": 9.664589823468328,
+      "grad_norm": 0.27966633439064026,
+      "learning_rate": 2.5719848809206586e-05,
+      "loss": 0.1774,
+      "num_input_tokens_seen": 42895808,
+      "step": 589
+    },
+    {
+      "epoch": 9.681204569055037,
+      "grad_norm": 0.2826574146747589,
+      "learning_rate": 2.5654423707696833e-05,
+      "loss": 0.0925,
+      "num_input_tokens_seen": 42952408,
+      "step": 590
+    },
+    {
+      "epoch": 9.697819314641745,
+      "grad_norm": 0.29687875509262085,
+      "learning_rate": 2.558899412084026e-05,
+      "loss": 0.0824,
+      "num_input_tokens_seen": 43025536,
+      "step": 591
+    },
+    {
+      "epoch": 9.714434060228452,
+      "grad_norm": 0.2201530486345291,
+      "learning_rate": 2.5523560497083926e-05,
+      "loss": 0.0634,
+      "num_input_tokens_seen": 43118024,
+      "step": 592
+    },
+    {
+      "epoch": 9.731048805815162,
+      "grad_norm": 0.25466641783714294,
+      "learning_rate": 2.5458123284902573e-05,
+      "loss": 0.0796,
+      "num_input_tokens_seen": 43198360,
+      "step": 593
+    },
+    {
+      "epoch": 9.74766355140187,
+      "grad_norm": 0.2543363571166992,
+      "learning_rate": 2.539268293279552e-05,
+      "loss": 0.0681,
+      "num_input_tokens_seen": 43264072,
+      "step": 594
+    },
+    {
+      "epoch": 9.764278296988577,
+      "grad_norm": 0.35776007175445557,
+      "learning_rate": 2.5327239889283612e-05,
+      "loss": 0.098,
+      "num_input_tokens_seen": 43339600,
+      "step": 595
+    },
+    {
+      "epoch": 9.780893042575286,
+      "grad_norm": 0.2951027452945709,
+      "learning_rate": 2.5261794602906145e-05,
+      "loss": 0.0857,
+      "num_input_tokens_seen": 43401136,
+      "step": 596
+    },
+    {
+      "epoch": 9.797507788161994,
+      "grad_norm": 0.2961483299732208,
+      "learning_rate": 2.5196347522217784e-05,
+      "loss": 0.0853,
+      "num_input_tokens_seen": 43477528,
+      "step": 597
+    },
+    {
+      "epoch": 9.814122533748701,
+      "grad_norm": 0.23898787796497345,
+      "learning_rate": 2.513089909578549e-05,
+      "loss": 0.0825,
+      "num_input_tokens_seen": 43557352,
+      "step": 598
+    },
+    {
+      "epoch": 9.83073727933541,
+      "grad_norm": 0.3080408275127411,
+      "learning_rate": 2.5065449772185456e-05,
+      "loss": 0.0985,
+      "num_input_tokens_seen": 43643104,
+      "step": 599
+    },
+    {
+      "epoch": 9.847352024922118,
+      "grad_norm": 0.3034234941005707,
+      "learning_rate": 2.5e-05,
+      "loss": 0.0859,
+      "num_input_tokens_seen": 43721408,
+      "step": 600
+    },
+    {
+      "epoch": 9.863966770508826,
+      "grad_norm": 0.23991604149341583,
+      "learning_rate": 2.4934550227814553e-05,
+      "loss": 0.0745,
+      "num_input_tokens_seen": 43802136,
+      "step": 601
+    },
+    {
+      "epoch": 9.880581516095535,
+      "grad_norm": 0.30657002329826355,
+      "learning_rate": 2.486910090421451e-05,
+      "loss": 0.0909,
+      "num_input_tokens_seen": 43862904,
+      "step": 602
+    },
+    {
+      "epoch": 9.897196261682243,
+      "grad_norm": 0.27239277958869934,
+      "learning_rate": 2.480365247778223e-05,
+      "loss": 0.0946,
+      "num_input_tokens_seen": 43942056,
+      "step": 603
+    },
+    {
+      "epoch": 9.91381100726895,
+      "grad_norm": 0.2195497751235962,
+      "learning_rate": 2.4738205397093864e-05,
+      "loss": 0.163,
+      "num_input_tokens_seen": 44016096,
+      "step": 604
+    },
+    {
+      "epoch": 9.93042575285566,
+      "grad_norm": 0.33310699462890625,
+      "learning_rate": 2.4672760110716394e-05,
+      "loss": 0.1059,
+      "num_input_tokens_seen": 44065504,
+      "step": 605
+    },
+    {
+      "epoch": 9.947040498442368,
+      "grad_norm": 0.2776658833026886,
+      "learning_rate": 2.460731706720449e-05,
+      "loss": 0.0852,
+      "num_input_tokens_seen": 44114776,
+      "step": 606
+    },
+    {
+      "epoch": 9.963655244029075,
+      "grad_norm": 0.32274240255355835,
+      "learning_rate": 2.4541876715097432e-05,
+      "loss": 0.1065,
+      "num_input_tokens_seen": 44175184,
+      "step": 607
+    },
+    {
+      "epoch": 9.980269989615785,
+      "grad_norm": 0.2639608085155487,
+      "learning_rate": 2.447643950291608e-05,
+      "loss": 0.0708,
+      "num_input_tokens_seen": 44265024,
+      "step": 608
+    },
+    {
+      "epoch": 9.996884735202492,
+      "grad_norm": 0.26623424887657166,
+      "learning_rate": 2.4411005879159753e-05,
+      "loss": 0.0803,
+      "num_input_tokens_seen": 44355400,
+      "step": 609
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.6041810512542725,
+      "learning_rate": 2.4345576292303176e-05,
+      "loss": 0.0686,
+      "num_input_tokens_seen": 44370360,
+      "step": 610
+    },
+    {
+      "epoch": 10.016614745586708,
+      "grad_norm": 0.2492121458053589,
+      "learning_rate": 2.4280151190793417e-05,
+      "loss": 0.0685,
+      "num_input_tokens_seen": 44446816,
+      "step": 611
+    },
+    {
+      "epoch": 10.033229491173417,
+      "grad_norm": 0.307436466217041,
+      "learning_rate": 2.4214731023046793e-05,
+      "loss": 0.0814,
+      "num_input_tokens_seen": 44503632,
+      "step": 612
+    },
+    {
+      "epoch": 10.049844236760125,
+      "grad_norm": 0.2768141031265259,
+      "learning_rate": 2.4149316237445812e-05,
+      "loss": 0.0782,
+      "num_input_tokens_seen": 44590320,
+      "step": 613
+    },
+    {
+      "epoch": 10.066458982346832,
+      "grad_norm": 0.29155638813972473,
+      "learning_rate": 2.408390728233609e-05,
+      "loss": 0.0793,
+      "num_input_tokens_seen": 44655224,
+      "step": 614
+    },
+    {
+      "epoch": 10.083073727933542,
+      "grad_norm": 0.24246808886528015,
+      "learning_rate": 2.4018504606023293e-05,
+      "loss": 0.1559,
+      "num_input_tokens_seen": 44736200,
+      "step": 615
+    },
+    {
+      "epoch": 10.09968847352025,
+      "grad_norm": 0.2551959455013275,
+      "learning_rate": 2.3953108656770016e-05,
+      "loss": 0.069,
+      "num_input_tokens_seen": 44804416,
+      "step": 616
+    },
+    {
+      "epoch": 10.116303219106957,
+      "grad_norm": 0.28595054149627686,
+      "learning_rate": 2.3887719882792785e-05,
+      "loss": 0.0694,
+      "num_input_tokens_seen": 44873864,
+      "step": 617
+    },
+    {
+      "epoch": 10.132917964693666,
+      "grad_norm": 0.2920657992362976,
+      "learning_rate": 2.3822338732258937e-05,
+      "loss": 0.081,
+      "num_input_tokens_seen": 44936736,
+      "step": 618
+    },
+    {
+      "epoch": 10.149532710280374,
+      "grad_norm": 0.23846709728240967,
+      "learning_rate": 2.3756965653283557e-05,
+      "loss": 0.0612,
+      "num_input_tokens_seen": 45026952,
+      "step": 619
+    },
+    {
+      "epoch": 10.166147455867081,
+      "grad_norm": 0.258646160364151,
+      "learning_rate": 2.3691601093926404e-05,
+      "loss": 0.0776,
+      "num_input_tokens_seen": 45104816,
+      "step": 620
+    },
+    {
+      "epoch": 10.18276220145379,
+      "grad_norm": 0.2821391522884369,
+      "learning_rate": 2.3626245502188864e-05,
+      "loss": 0.0707,
+      "num_input_tokens_seen": 45177392,
+      "step": 621
+    },
+    {
+      "epoch": 10.199376947040498,
+      "grad_norm": 0.34291672706604004,
+      "learning_rate": 2.3560899326010822e-05,
+      "loss": 0.0849,
+      "num_input_tokens_seen": 45237200,
+      "step": 622
+    },
+    {
+      "epoch": 10.215991692627206,
+      "grad_norm": 0.32930925488471985,
+      "learning_rate": 2.3495563013267664e-05,
+      "loss": 0.0856,
+      "num_input_tokens_seen": 45293376,
+      "step": 623
+    },
+    {
+      "epoch": 10.232606438213915,
+      "grad_norm": 0.2715514004230499,
+      "learning_rate": 2.3430237011767167e-05,
+      "loss": 0.081,
+      "num_input_tokens_seen": 45369376,
+      "step": 624
+    },
+    {
+      "epoch": 10.249221183800623,
+      "grad_norm": 0.3008752167224884,
+      "learning_rate": 2.3364921769246423e-05,
+      "loss": 0.0707,
+      "num_input_tokens_seen": 45439920,
+      "step": 625
+    },
+    {
+      "epoch": 10.26583592938733,
+      "grad_norm": 0.3019203841686249,
+      "learning_rate": 2.3299617733368806e-05,
+      "loss": 0.0791,
+      "num_input_tokens_seen": 45497992,
+      "step": 626
+    },
+    {
+      "epoch": 10.28245067497404,
+      "grad_norm": 0.2555345892906189,
+      "learning_rate": 2.323432535172084e-05,
+      "loss": 0.0649,
+      "num_input_tokens_seen": 45562832,
+      "step": 627
+    },
+    {
+      "epoch": 10.299065420560748,
+      "grad_norm": 0.2813156843185425,
+      "learning_rate": 2.3169045071809215e-05,
+      "loss": 0.0835,
+      "num_input_tokens_seen": 45638712,
+      "step": 628
+    },
+    {
+      "epoch": 10.315680166147455,
+      "grad_norm": 0.30063870549201965,
+      "learning_rate": 2.3103777341057655e-05,
+      "loss": 0.0797,
+      "num_input_tokens_seen": 45697336,
+      "step": 629
+    },
+    {
+      "epoch": 10.332294911734165,
+      "grad_norm": 0.26722076535224915,
+      "learning_rate": 2.303852260680388e-05,
+      "loss": 0.0892,
+      "num_input_tokens_seen": 45765680,
+      "step": 630
+    },
+    {
+      "epoch": 10.348909657320872,
+      "grad_norm": 0.28536635637283325,
+      "learning_rate": 2.2973281316296533e-05,
+      "loss": 0.081,
+      "num_input_tokens_seen": 45829592,
+      "step": 631
+    },
+    {
+      "epoch": 10.36552440290758,
+      "grad_norm": 0.2656691074371338,
+      "learning_rate": 2.2908053916692117e-05,
+      "loss": 0.0816,
+      "num_input_tokens_seen": 45903304,
+      "step": 632
+    },
+    {
+      "epoch": 10.38213914849429,
+      "grad_norm": 0.23176686465740204,
+      "learning_rate": 2.284284085505192e-05,
+      "loss": 0.0773,
+      "num_input_tokens_seen": 45999840,
+      "step": 633
+    },
+    {
+      "epoch": 10.398753894080997,
+      "grad_norm": 0.32047316431999207,
+      "learning_rate": 2.2777642578338963e-05,
+      "loss": 0.0773,
+      "num_input_tokens_seen": 46107264,
+      "step": 634
+    },
+    {
+      "epoch": 10.415368639667705,
+      "grad_norm": 0.2911181151866913,
+      "learning_rate": 2.2712459533414944e-05,
+      "loss": 0.0696,
+      "num_input_tokens_seen": 46174720,
+      "step": 635
+    },
+    {
+      "epoch": 10.431983385254414,
+      "grad_norm": 0.22214826941490173,
+      "learning_rate": 2.2647292167037144e-05,
+      "loss": 0.063,
+      "num_input_tokens_seen": 46269984,
+      "step": 636
+    },
+    {
+      "epoch": 10.448598130841122,
+      "grad_norm": 0.2614048719406128,
+      "learning_rate": 2.2582140925855395e-05,
+      "loss": 0.0672,
+      "num_input_tokens_seen": 46354792,
+      "step": 637
+    },
+    {
+      "epoch": 10.46521287642783,
+      "grad_norm": 0.3499446213245392,
+      "learning_rate": 2.251700625640903e-05,
+      "loss": 0.0755,
+      "num_input_tokens_seen": 46420480,
+      "step": 638
+    },
+    {
+      "epoch": 10.481827622014539,
+      "grad_norm": 0.30158787965774536,
+      "learning_rate": 2.2451888605123754e-05,
+      "loss": 0.0773,
+      "num_input_tokens_seen": 46487288,
+      "step": 639
+    },
+    {
+      "epoch": 10.498442367601246,
+      "grad_norm": 0.2722439169883728,
+      "learning_rate": 2.238678841830867e-05,
+      "loss": 0.0769,
+      "num_input_tokens_seen": 46560016,
+      "step": 640
+    },
+    {
+      "epoch": 10.515057113187954,
+      "grad_norm": 0.27054470777511597,
+      "learning_rate": 2.2321706142153162e-05,
+      "loss": 0.079,
+      "num_input_tokens_seen": 46640864,
+      "step": 641
+    },
+    {
+      "epoch": 10.531671858774663,
+      "grad_norm": 0.3000243306159973,
+      "learning_rate": 2.225664222272387e-05,
+      "loss": 0.0798,
+      "num_input_tokens_seen": 46712072,
+      "step": 642
+    },
+    {
+      "epoch": 10.54828660436137,
+      "grad_norm": 0.37168607115745544,
+      "learning_rate": 2.219159710596161e-05,
+      "loss": 0.075,
+      "num_input_tokens_seen": 46787552,
+      "step": 643
+    },
+    {
+      "epoch": 10.564901349948078,
+      "grad_norm": 0.28578755259513855,
+      "learning_rate": 2.212657123767834e-05,
+      "loss": 0.0724,
+      "num_input_tokens_seen": 46862880,
+      "step": 644
+    },
+    {
+      "epoch": 10.581516095534788,
+      "grad_norm": 0.24598972499370575,
+      "learning_rate": 2.2061565063554064e-05,
+      "loss": 0.0711,
+      "num_input_tokens_seen": 46944632,
+      "step": 645
+    },
+    {
+      "epoch": 10.598130841121495,
+      "grad_norm": 0.24780628085136414,
+      "learning_rate": 2.1996579029133825e-05,
+      "loss": 0.0699,
+      "num_input_tokens_seen": 47029152,
+      "step": 646
+    },
+    {
+      "epoch": 10.614745586708203,
+      "grad_norm": 0.2675541341304779,
+      "learning_rate": 2.1931613579824628e-05,
+      "loss": 0.0717,
+      "num_input_tokens_seen": 47133448,
+      "step": 647
+    },
+    {
+      "epoch": 10.631360332294912,
+      "grad_norm": 0.28133469820022583,
+      "learning_rate": 2.186666916089239e-05,
+      "loss": 0.0698,
+      "num_input_tokens_seen": 47219888,
+      "step": 648
+    },
+    {
+      "epoch": 10.64797507788162,
+      "grad_norm": 0.2919703722000122,
+      "learning_rate": 2.180174621745889e-05,
+      "loss": 0.07,
+      "num_input_tokens_seen": 47286344,
+      "step": 649
+    },
+    {
+      "epoch": 10.664589823468328,
+      "grad_norm": 0.26447877287864685,
+      "learning_rate": 2.173684519449872e-05,
+      "loss": 0.0619,
+      "num_input_tokens_seen": 47392832,
+      "step": 650
+    },
+    {
+      "epoch": 10.681204569055037,
+      "grad_norm": 0.32841357588768005,
+      "learning_rate": 2.1671966536836196e-05,
+      "loss": 0.0796,
+      "num_input_tokens_seen": 47456448,
+      "step": 651
+    },
+    {
+      "epoch": 10.697819314641745,
+      "grad_norm": 0.29451075196266174,
+      "learning_rate": 2.1607110689142393e-05,
+      "loss": 0.0696,
+      "num_input_tokens_seen": 47544608,
+      "step": 652
+    },
+    {
+      "epoch": 10.714434060228452,
+      "grad_norm": 0.2618974447250366,
+      "learning_rate": 2.154227809593203e-05,
+      "loss": 0.0804,
+      "num_input_tokens_seen": 47605504,
+      "step": 653
+    },
+    {
+      "epoch": 10.731048805815162,
+      "grad_norm": 0.2926179766654968,
+      "learning_rate": 2.1477469201560435e-05,
+      "loss": 0.0675,
+      "num_input_tokens_seen": 47668856,
+      "step": 654
+    },
+    {
+      "epoch": 10.74766355140187,
+      "grad_norm": 0.26044753193855286,
+      "learning_rate": 2.141268445022052e-05,
+      "loss": 0.0779,
+      "num_input_tokens_seen": 47732072,
+      "step": 655
+    },
+    {
+      "epoch": 10.764278296988577,
+      "grad_norm": 0.26107361912727356,
+      "learning_rate": 2.1347924285939714e-05,
+      "loss": 0.0645,
+      "num_input_tokens_seen": 47821128,
+      "step": 656
+    },
+    {
+      "epoch": 10.780893042575286,
+      "grad_norm": 0.334241658449173,
+      "learning_rate": 2.1283189152576925e-05,
+      "loss": 0.0975,
+      "num_input_tokens_seen": 47888432,
+      "step": 657
+    },
+    {
+      "epoch": 10.797507788161994,
+      "grad_norm": 0.2968047857284546,
+      "learning_rate": 2.121847949381952e-05,
+      "loss": 0.0789,
+      "num_input_tokens_seen": 47961872,
+      "step": 658
+    },
+    {
+      "epoch": 10.814122533748701,
+      "grad_norm": 0.2523380219936371,
+      "learning_rate": 2.1153795753180247e-05,
+      "loss": 0.0607,
+      "num_input_tokens_seen": 48049624,
+      "step": 659
+    },
+    {
+      "epoch": 10.83073727933541,
+      "grad_norm": 0.32714295387268066,
+      "learning_rate": 2.1089138373994223e-05,
+      "loss": 0.0868,
+      "num_input_tokens_seen": 48119288,
+      "step": 660
+    },
+    {
+      "epoch": 10.847352024922118,
+      "grad_norm": 0.3273003101348877,
+      "learning_rate": 2.1024507799415903e-05,
+      "loss": 0.0888,
+      "num_input_tokens_seen": 48176432,
+      "step": 661
+    },
+    {
+      "epoch": 10.863966770508826,
+      "grad_norm": 0.27770254015922546,
+      "learning_rate": 2.0959904472415977e-05,
+      "loss": 0.1672,
+      "num_input_tokens_seen": 48252000,
+      "step": 662
+    },
+    {
+      "epoch": 10.880581516095535,
+      "grad_norm": 0.2864398956298828,
+      "learning_rate": 2.089532883577843e-05,
+      "loss": 0.0815,
+      "num_input_tokens_seen": 48317320,
+      "step": 663
+    },
+    {
+      "epoch": 10.897196261682243,
+      "grad_norm": 0.30251842737197876,
+      "learning_rate": 2.0830781332097446e-05,
+      "loss": 0.0764,
+      "num_input_tokens_seen": 48384728,
+      "step": 664
+    },
+    {
+      "epoch": 10.91381100726895,
+      "grad_norm": 0.29925107955932617,
+      "learning_rate": 2.0766262403774386e-05,
+      "loss": 0.181,
+      "num_input_tokens_seen": 48451128,
+      "step": 665
+    },
+    {
+      "epoch": 10.93042575285566,
+      "grad_norm": 0.24659354984760284,
+      "learning_rate": 2.070177249301476e-05,
+      "loss": 0.0658,
+      "num_input_tokens_seen": 48535696,
+      "step": 666
+    },
+    {
+      "epoch": 10.947040498442368,
+      "grad_norm": 0.27392905950546265,
+      "learning_rate": 2.0637312041825205e-05,
+      "loss": 0.0701,
+      "num_input_tokens_seen": 48606496,
+      "step": 667
+    },
+    {
+      "epoch": 10.963655244029075,
+      "grad_norm": 0.33744704723358154,
+      "learning_rate": 2.057288149201042e-05,
+      "loss": 0.091,
+      "num_input_tokens_seen": 48671728,
+      "step": 668
+    },
+    {
+      "epoch": 10.980269989615785,
+      "grad_norm": 0.28141456842422485,
+      "learning_rate": 2.0508481285170186e-05,
+      "loss": 0.0652,
+      "num_input_tokens_seen": 48735408,
+      "step": 669
+    },
+    {
+      "epoch": 10.996884735202492,
+      "grad_norm": 0.35630133748054504,
+      "learning_rate": 2.0444111862696314e-05,
+      "loss": 0.0916,
+      "num_input_tokens_seen": 48794360,
+      "step": 670
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 0.9144778847694397,
+      "learning_rate": 2.037977366576961e-05,
+      "loss": 0.1137,
+      "num_input_tokens_seen": 48806360,
+      "step": 671
+    },
+    {
+      "epoch": 11.016614745586708,
+      "grad_norm": 0.2536274790763855,
+      "learning_rate": 2.031546713535688e-05,
+      "loss": 0.0732,
+      "num_input_tokens_seen": 48873568,
+      "step": 672
+    },
+    {
+      "epoch": 11.033229491173417,
+      "grad_norm": 0.22841258347034454,
+      "learning_rate": 2.025119271220789e-05,
+      "loss": 0.0561,
+      "num_input_tokens_seen": 48959880,
+      "step": 673
+    },
+    {
+      "epoch": 11.049844236760125,
+      "grad_norm": 0.30224987864494324,
+      "learning_rate": 2.018695083685232e-05,
+      "loss": 0.0598,
+      "num_input_tokens_seen": 49048936,
+      "step": 674
+    },
+    {
+      "epoch": 11.066458982346832,
+      "grad_norm": 0.24764464795589447,
+      "learning_rate": 2.0122741949596797e-05,
+      "loss": 0.0586,
+      "num_input_tokens_seen": 49119216,
+      "step": 675
+    },
+    {
+      "epoch": 11.083073727933542,
+      "grad_norm": 0.30028969049453735,
+      "learning_rate": 2.0058566490521847e-05,
+      "loss": 0.0718,
+      "num_input_tokens_seen": 49183408,
+      "step": 676
+    },
+    {
+      "epoch": 11.09968847352025,
+      "grad_norm": 0.2609259784221649,
+      "learning_rate": 1.9994424899478885e-05,
+      "loss": 0.062,
+      "num_input_tokens_seen": 49275352,
+      "step": 677
+    },
+    {
+      "epoch": 11.116303219106957,
+      "grad_norm": 0.34472572803497314,
+      "learning_rate": 1.9930317616087196e-05,
+      "loss": 0.0794,
+      "num_input_tokens_seen": 49326536,
+      "step": 678
+    },
+    {
+      "epoch": 11.132917964693666,
+      "grad_norm": 0.28830835223197937,
+      "learning_rate": 1.986624507973091e-05,
+      "loss": 0.066,
+      "num_input_tokens_seen": 49390840,
+      "step": 679
+    },
+    {
+      "epoch": 11.149532710280374,
+      "grad_norm": 0.38434937596321106,
+      "learning_rate": 1.980220772955602e-05,
+      "loss": 0.0612,
+      "num_input_tokens_seen": 49474128,
+      "step": 680
+    },
+    {
+      "epoch": 11.166147455867081,
+      "grad_norm": 0.31176745891571045,
+      "learning_rate": 1.9738206004467363e-05,
+      "loss": 0.0715,
+      "num_input_tokens_seen": 49534248,
+      "step": 681
+    },
+    {
+      "epoch": 11.18276220145379,
+      "grad_norm": 0.23110291361808777,
+      "learning_rate": 1.9674240343125588e-05,
+      "loss": 0.0535,
+      "num_input_tokens_seen": 49621320,
+      "step": 682
+    },
+    {
+      "epoch": 11.199376947040498,
+      "grad_norm": 0.3060101270675659,
+      "learning_rate": 1.961031118394418e-05,
+      "loss": 0.0729,
+      "num_input_tokens_seen": 49689760,
+      "step": 683
+    },
+    {
+      "epoch": 11.215991692627206,
+      "grad_norm": 0.2873547077178955,
+      "learning_rate": 1.9546418965086442e-05,
+      "loss": 0.0635,
+      "num_input_tokens_seen": 49771640,
+      "step": 684
+    },
+    {
+      "epoch": 11.232606438213915,
+      "grad_norm": 0.2780902087688446,
+      "learning_rate": 1.9482564124462476e-05,
+      "loss": 0.0644,
+      "num_input_tokens_seen": 49838736,
+      "step": 685
+    },
+    {
+      "epoch": 11.249221183800623,
+      "grad_norm": 0.3456045091152191,
+      "learning_rate": 1.941874709972622e-05,
+      "loss": 0.079,
+      "num_input_tokens_seen": 49908376,
+      "step": 686
+    },
+    {
+      "epoch": 11.26583592938733,
+      "grad_norm": 0.22289884090423584,
+      "learning_rate": 1.935496832827241e-05,
+      "loss": 0.0518,
+      "num_input_tokens_seen": 50006464,
+      "step": 687
+    },
+    {
+      "epoch": 11.28245067497404,
+      "grad_norm": 0.330370157957077,
+      "learning_rate": 1.9291228247233605e-05,
+      "loss": 0.075,
+      "num_input_tokens_seen": 50081920,
+      "step": 688
+    },
+    {
+      "epoch": 11.299065420560748,
+      "grad_norm": 0.2642322778701782,
+      "learning_rate": 1.9227527293477186e-05,
+      "loss": 0.1651,
+      "num_input_tokens_seen": 50145832,
+      "step": 689
+    },
+    {
+      "epoch": 11.315680166147455,
+      "grad_norm": 0.32244759798049927,
+      "learning_rate": 1.9163865903602374e-05,
+      "loss": 0.0777,
+      "num_input_tokens_seen": 50193752,
+      "step": 690
+    },
+    {
+      "epoch": 11.332294911734165,
+      "grad_norm": 0.2895197570323944,
+      "learning_rate": 1.9100244513937174e-05,
+      "loss": 0.0765,
+      "num_input_tokens_seen": 50252792,
+      "step": 691
+    },
+    {
+      "epoch": 11.348909657320872,
+      "grad_norm": 0.31584399938583374,
+      "learning_rate": 1.9036663560535483e-05,
+      "loss": 0.0693,
+      "num_input_tokens_seen": 50315264,
+      "step": 692
+    },
+    {
+      "epoch": 11.36552440290758,
+      "grad_norm": 0.2692103087902069,
+      "learning_rate": 1.897312347917404e-05,
+      "loss": 0.0658,
+      "num_input_tokens_seen": 50379280,
+      "step": 693
+    },
+    {
+      "epoch": 11.38213914849429,
+      "grad_norm": 0.2717333436012268,
+      "learning_rate": 1.890962470534944e-05,
+      "loss": 0.0709,
+      "num_input_tokens_seen": 50459248,
+      "step": 694
+    },
+    {
+      "epoch": 11.398753894080997,
+      "grad_norm": 0.23477673530578613,
+      "learning_rate": 1.8846167674275176e-05,
+      "loss": 0.0649,
+      "num_input_tokens_seen": 50552672,
+      "step": 695
+    },
+    {
+      "epoch": 11.415368639667705,
+      "grad_norm": 0.30369827151298523,
+      "learning_rate": 1.8782752820878634e-05,
+      "loss": 0.062,
+      "num_input_tokens_seen": 50613744,
+      "step": 696
+    },
+    {
+      "epoch": 11.431983385254414,
+      "grad_norm": 0.2870631814002991,
+      "learning_rate": 1.8719380579798112e-05,
+      "loss": 0.0667,
+      "num_input_tokens_seen": 50690248,
+      "step": 697
+    },
+    {
+      "epoch": 11.448598130841122,
+      "grad_norm": 0.2412702739238739,
+      "learning_rate": 1.865605138537986e-05,
+      "loss": 0.0638,
+      "num_input_tokens_seen": 50782136,
+      "step": 698
+    },
+    {
+      "epoch": 11.46521287642783,
+      "grad_norm": 0.3324171304702759,
+      "learning_rate": 1.8592765671675084e-05,
+      "loss": 0.0806,
+      "num_input_tokens_seen": 50843280,
+      "step": 699
+    },
+    {
+      "epoch": 11.481827622014539,
+      "grad_norm": 0.25983986258506775,
+      "learning_rate": 1.852952387243698e-05,
+      "loss": 0.0738,
+      "num_input_tokens_seen": 50909176,
+      "step": 700
+    },
+    {
+      "epoch": 11.498442367601246,
+      "grad_norm": 0.2748391628265381,
+      "learning_rate": 1.846632642111777e-05,
+      "loss": 0.0823,
+      "num_input_tokens_seen": 50979512,
+      "step": 701
+    },
+    {
+      "epoch": 11.515057113187954,
+      "grad_norm": 0.2953915297985077,
+      "learning_rate": 1.8403173750865685e-05,
+      "loss": 0.0726,
+      "num_input_tokens_seen": 51043064,
+      "step": 702
+    },
+    {
+      "epoch": 11.531671858774663,
+      "grad_norm": 0.27582335472106934,
+      "learning_rate": 1.8340066294522068e-05,
+      "loss": 0.0812,
+      "num_input_tokens_seen": 51117784,
+      "step": 703
+    },
+    {
+      "epoch": 11.54828660436137,
+      "grad_norm": 0.27300143241882324,
+      "learning_rate": 1.827700448461836e-05,
+      "loss": 0.0661,
+      "num_input_tokens_seen": 51190232,
+      "step": 704
+    },
+    {
+      "epoch": 11.564901349948078,
+      "grad_norm": 0.27855339646339417,
+      "learning_rate": 1.8213988753373146e-05,
+      "loss": 0.0714,
+      "num_input_tokens_seen": 51261080,
+      "step": 705
+    },
+    {
+      "epoch": 11.581516095534788,
+      "grad_norm": 0.21793155372142792,
+      "learning_rate": 1.815101953268919e-05,
+      "loss": 0.0573,
+      "num_input_tokens_seen": 51341592,
+      "step": 706
+    },
+    {
+      "epoch": 11.598130841121495,
+      "grad_norm": 0.29607975482940674,
+      "learning_rate": 1.8088097254150486e-05,
+      "loss": 0.0625,
+      "num_input_tokens_seen": 51411168,
+      "step": 707
+    },
+    {
+      "epoch": 11.614745586708203,
+      "grad_norm": 0.3065016269683838,
+      "learning_rate": 1.802522234901927e-05,
+      "loss": 0.0729,
+      "num_input_tokens_seen": 51467184,
+      "step": 708
+    },
+    {
+      "epoch": 11.631360332294912,
+      "grad_norm": 0.2603945732116699,
+      "learning_rate": 1.7962395248233114e-05,
+      "loss": 0.1445,
+      "num_input_tokens_seen": 51567896,
+      "step": 709
+    },
+    {
+      "epoch": 11.64797507788162,
+      "grad_norm": 0.298199862241745,
+      "learning_rate": 1.7899616382401936e-05,
+      "loss": 0.0814,
+      "num_input_tokens_seen": 51629240,
+      "step": 710
+    },
+    {
+      "epoch": 11.664589823468328,
+      "grad_norm": 0.33891791105270386,
+      "learning_rate": 1.783688618180504e-05,
+      "loss": 0.087,
+      "num_input_tokens_seen": 51687552,
+      "step": 711
+    },
+    {
+      "epoch": 11.681204569055037,
+      "grad_norm": 0.2577643394470215,
+      "learning_rate": 1.7774205076388206e-05,
+      "loss": 0.0658,
+      "num_input_tokens_seen": 51784232,
+      "step": 712
+    },
+    {
+      "epoch": 11.697819314641745,
+      "grad_norm": 0.27919164299964905,
+      "learning_rate": 1.7711573495760725e-05,
+      "loss": 0.0708,
+      "num_input_tokens_seen": 51852288,
+      "step": 713
+    },
+    {
+      "epoch": 11.714434060228452,
+      "grad_norm": 0.22918739914894104,
+      "learning_rate": 1.7648991869192405e-05,
+      "loss": 0.0526,
+      "num_input_tokens_seen": 51948408,
+      "step": 714
+    },
+    {
+      "epoch": 11.731048805815162,
+      "grad_norm": 0.2642481327056885,
+      "learning_rate": 1.7586460625610728e-05,
+      "loss": 0.0648,
+      "num_input_tokens_seen": 52029512,
+      "step": 715
+    },
+    {
+      "epoch": 11.74766355140187,
+      "grad_norm": 0.2955224812030792,
+      "learning_rate": 1.7523980193597836e-05,
+      "loss": 0.073,
+      "num_input_tokens_seen": 52095864,
+      "step": 716
+    },
+    {
+      "epoch": 11.764278296988577,
+      "grad_norm": 0.2940084934234619,
+      "learning_rate": 1.746155100138761e-05,
+      "loss": 0.0764,
+      "num_input_tokens_seen": 52154880,
+      "step": 717
+    },
+    {
+      "epoch": 11.780893042575286,
+      "grad_norm": 0.3002694547176361,
+      "learning_rate": 1.739917347686274e-05,
+      "loss": 0.0807,
+      "num_input_tokens_seen": 52209696,
+      "step": 718
+    },
+    {
+      "epoch": 11.797507788161994,
+      "grad_norm": 0.22952137887477875,
+      "learning_rate": 1.7336848047551814e-05,
+      "loss": 0.0552,
+      "num_input_tokens_seen": 52290608,
+      "step": 719
+    },
+    {
+      "epoch": 11.814122533748701,
+      "grad_norm": 0.256875216960907,
+      "learning_rate": 1.7274575140626318e-05,
+      "loss": 0.0628,
+      "num_input_tokens_seen": 52364616,
+      "step": 720
+    },
+    {
+      "epoch": 11.83073727933541,
+      "grad_norm": 0.4035111367702484,
+      "learning_rate": 1.72123551828978e-05,
+      "loss": 0.0687,
+      "num_input_tokens_seen": 52446432,
+      "step": 721
+    },
+    {
+      "epoch": 11.847352024922118,
+      "grad_norm": 0.34936726093292236,
+      "learning_rate": 1.7150188600814877e-05,
+      "loss": 0.1985,
+      "num_input_tokens_seen": 52500320,
+      "step": 722
+    },
+    {
+      "epoch": 11.863966770508826,
+      "grad_norm": 0.2872942090034485,
+      "learning_rate": 1.7088075820460346e-05,
+      "loss": 0.0639,
+      "num_input_tokens_seen": 52572888,
+      "step": 723
+    },
+    {
+      "epoch": 11.880581516095535,
+      "grad_norm": 0.2703973650932312,
+      "learning_rate": 1.702601726754825e-05,
+      "loss": 0.0715,
+      "num_input_tokens_seen": 52649240,
+      "step": 724
+    },
+    {
+      "epoch": 11.897196261682243,
+      "grad_norm": 0.2533208727836609,
+      "learning_rate": 1.6964013367420966e-05,
+      "loss": 0.0638,
+      "num_input_tokens_seen": 52731312,
+      "step": 725
+    },
+    {
+      "epoch": 11.91381100726895,
+      "grad_norm": 0.24604761600494385,
+      "learning_rate": 1.690206454504627e-05,
+      "loss": 0.0668,
+      "num_input_tokens_seen": 52829968,
+      "step": 726
+    },
+    {
+      "epoch": 11.93042575285566,
+      "grad_norm": 0.21317115426063538,
+      "learning_rate": 1.6840171225014457e-05,
+      "loss": 0.0567,
+      "num_input_tokens_seen": 52938304,
+      "step": 727
+    },
+    {
+      "epoch": 11.947040498442368,
+      "grad_norm": 0.3644583821296692,
+      "learning_rate": 1.677833383153542e-05,
+      "loss": 0.0798,
+      "num_input_tokens_seen": 53003736,
+      "step": 728
+    },
+    {
+      "epoch": 11.963655244029075,
+      "grad_norm": 0.293142169713974,
+      "learning_rate": 1.6716552788435724e-05,
+      "loss": 0.0621,
+      "num_input_tokens_seen": 53071872,
+      "step": 729
+    },
+    {
+      "epoch": 11.980269989615785,
+      "grad_norm": 0.2816755473613739,
+      "learning_rate": 1.665482851915573e-05,
+      "loss": 0.06,
+      "num_input_tokens_seen": 53145192,
+      "step": 730
+    },
+    {
+      "epoch": 11.996884735202492,
+      "grad_norm": 0.21882176399230957,
+      "learning_rate": 1.659316144674666e-05,
+      "loss": 0.0551,
+      "num_input_tokens_seen": 53229240,
+      "step": 731
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.5393900275230408,
+      "learning_rate": 1.6531551993867717e-05,
+      "loss": 0.0551,
+      "num_input_tokens_seen": 53245592,
+      "step": 732
+    },
+    {
+      "epoch": 12.016614745586708,
+      "grad_norm": 0.2326558381319046,
+      "learning_rate": 1.6470000582783204e-05,
+      "loss": 0.0617,
+      "num_input_tokens_seen": 53327016,
+      "step": 733
+    },
+    {
+      "epoch": 12.033229491173417,
+      "grad_norm": 0.24347102642059326,
+      "learning_rate": 1.64085076353596e-05,
+      "loss": 0.062,
+      "num_input_tokens_seen": 53396464,
+      "step": 734
+    },
+    {
+      "epoch": 12.049844236760125,
+      "grad_norm": 0.2822597622871399,
+      "learning_rate": 1.6347073573062672e-05,
+      "loss": 0.0568,
+      "num_input_tokens_seen": 53485056,
+      "step": 735
+    },
+    {
+      "epoch": 12.066458982346832,
+      "grad_norm": 0.27358195185661316,
+      "learning_rate": 1.6285698816954624e-05,
+      "loss": 0.0666,
+      "num_input_tokens_seen": 53553376,
+      "step": 736
+    },
+    {
+      "epoch": 12.083073727933542,
+      "grad_norm": 0.2802799344062805,
+      "learning_rate": 1.622438378769114e-05,
+      "loss": 0.0675,
+      "num_input_tokens_seen": 53615616,
+      "step": 737
+    },
+    {
+      "epoch": 12.09968847352025,
+      "grad_norm": 0.26887986063957214,
+      "learning_rate": 1.6163128905518578e-05,
+      "loss": 0.0631,
+      "num_input_tokens_seen": 53689512,
+      "step": 738
+    },
+    {
+      "epoch": 12.116303219106957,
+      "grad_norm": 0.33475375175476074,
+      "learning_rate": 1.6101934590271032e-05,
+      "loss": 0.0775,
+      "num_input_tokens_seen": 53743080,
+      "step": 739
+    },
+    {
+      "epoch": 12.132917964693666,
+      "grad_norm": 0.28297415375709534,
+      "learning_rate": 1.6040801261367493e-05,
+      "loss": 0.0568,
+      "num_input_tokens_seen": 53830432,
+      "step": 740
+    },
+    {
+      "epoch": 12.149532710280374,
+      "grad_norm": 0.2604190707206726,
+      "learning_rate": 1.5979729337808955e-05,
+      "loss": 0.0545,
+      "num_input_tokens_seen": 53914536,
+      "step": 741
+    },
+    {
+      "epoch": 12.166147455867081,
+      "grad_norm": 0.36683493852615356,
+      "learning_rate": 1.5918719238175544e-05,
+      "loss": 0.0553,
+      "num_input_tokens_seen": 53977072,
+      "step": 742
+    },
+    {
+      "epoch": 12.18276220145379,
+      "grad_norm": 0.27039340138435364,
+      "learning_rate": 1.5857771380623643e-05,
+      "loss": 0.0609,
+      "num_input_tokens_seen": 54051576,
+      "step": 743
+    },
+    {
+      "epoch": 12.199376947040498,
+      "grad_norm": 0.2632916271686554,
+      "learning_rate": 1.5796886182883053e-05,
+      "loss": 0.0629,
+      "num_input_tokens_seen": 54126832,
+      "step": 744
+    },
+    {
+      "epoch": 12.215991692627206,
+      "grad_norm": 0.2716232240200043,
+      "learning_rate": 1.5736064062254094e-05,
+      "loss": 0.0579,
+      "num_input_tokens_seen": 54203832,
+      "step": 745
+    },
+    {
+      "epoch": 12.232606438213915,
+      "grad_norm": 0.29032984375953674,
+      "learning_rate": 1.5675305435604775e-05,
+      "loss": 0.0652,
+      "num_input_tokens_seen": 54268712,
+      "step": 746
+    },
+    {
+      "epoch": 12.249221183800623,
+      "grad_norm": 0.2442651391029358,
+      "learning_rate": 1.561461071936792e-05,
+      "loss": 0.1332,
+      "num_input_tokens_seen": 54342024,
+      "step": 747
+    },
+    {
+      "epoch": 12.26583592938733,
+      "grad_norm": 0.25280654430389404,
+      "learning_rate": 1.5553980329538326e-05,
+      "loss": 0.0549,
+      "num_input_tokens_seen": 54407352,
+      "step": 748
+    },
+    {
+      "epoch": 12.28245067497404,
+      "grad_norm": 0.3102169930934906,
+      "learning_rate": 1.549341468166988e-05,
+      "loss": 0.0639,
+      "num_input_tokens_seen": 54466560,
+      "step": 749
+    },
+    {
+      "epoch": 12.299065420560748,
+      "grad_norm": 0.2568191885948181,
+      "learning_rate": 1.5432914190872757e-05,
+      "loss": 0.0611,
+      "num_input_tokens_seen": 54559704,
+      "step": 750
+    },
+    {
+      "epoch": 12.315680166147455,
+      "grad_norm": 0.3201739192008972,
+      "learning_rate": 1.537247927181055e-05,
+      "loss": 0.0735,
+      "num_input_tokens_seen": 54623736,
+      "step": 751
+    },
+    {
+      "epoch": 12.332294911734165,
+      "grad_norm": 0.25409963726997375,
+      "learning_rate": 1.5312110338697426e-05,
+      "loss": 0.0655,
+      "num_input_tokens_seen": 54699928,
+      "step": 752
+    },
+    {
+      "epoch": 12.348909657320872,
+      "grad_norm": 0.2382766753435135,
+      "learning_rate": 1.5251807805295302e-05,
+      "loss": 0.0529,
+      "num_input_tokens_seen": 54782808,
+      "step": 753
+    },
+    {
+      "epoch": 12.36552440290758,
+      "grad_norm": 0.2533799111843109,
+      "learning_rate": 1.519157208491097e-05,
+      "loss": 0.0607,
+      "num_input_tokens_seen": 54868720,
+      "step": 754
+    },
+    {
+      "epoch": 12.38213914849429,
+      "grad_norm": 0.19882524013519287,
+      "learning_rate": 1.5131403590393323e-05,
+      "loss": 0.0496,
+      "num_input_tokens_seen": 54960384,
+      "step": 755
+    },
+    {
+      "epoch": 12.398753894080997,
+      "grad_norm": 0.2437567114830017,
+      "learning_rate": 1.5071302734130489e-05,
+      "loss": 0.0553,
+      "num_input_tokens_seen": 55044232,
+      "step": 756
+    },
+    {
+      "epoch": 12.415368639667705,
+      "grad_norm": 0.24854236841201782,
+      "learning_rate": 1.5011269928047003e-05,
+      "loss": 0.1655,
+      "num_input_tokens_seen": 55125632,
+      "step": 757
+    },
+    {
+      "epoch": 12.431983385254414,
+      "grad_norm": 0.2629762887954712,
+      "learning_rate": 1.4951305583601e-05,
+      "loss": 0.0588,
+      "num_input_tokens_seen": 55201720,
+      "step": 758
+    },
+    {
+      "epoch": 12.448598130841122,
+      "grad_norm": 0.4379749000072479,
+      "learning_rate": 1.4891410111781378e-05,
+      "loss": 0.0522,
+      "num_input_tokens_seen": 55265832,
+      "step": 759
+    },
+    {
+      "epoch": 12.46521287642783,
+      "grad_norm": 0.3276868462562561,
+      "learning_rate": 1.4831583923104999e-05,
+      "loss": 0.0776,
+      "num_input_tokens_seen": 55316928,
+      "step": 760
+    },
+    {
+      "epoch": 12.481827622014539,
+      "grad_norm": 0.34918782114982605,
+      "learning_rate": 1.4771827427613855e-05,
+      "loss": 0.059,
+      "num_input_tokens_seen": 55391232,
+      "step": 761
+    },
+    {
+      "epoch": 12.498442367601246,
+      "grad_norm": 0.28458765149116516,
+      "learning_rate": 1.4712141034872282e-05,
+      "loss": 0.0604,
+      "num_input_tokens_seen": 55460984,
+      "step": 762
+    },
+    {
+      "epoch": 12.515057113187954,
+      "grad_norm": 0.27223336696624756,
+      "learning_rate": 1.4652525153964131e-05,
+      "loss": 0.0556,
+      "num_input_tokens_seen": 55537056,
+      "step": 763
+    },
+    {
+      "epoch": 12.531671858774663,
+      "grad_norm": 0.27431365847587585,
+      "learning_rate": 1.4592980193489975e-05,
+      "loss": 0.0598,
+      "num_input_tokens_seen": 55625272,
+      "step": 764
+    },
+    {
+      "epoch": 12.54828660436137,
+      "grad_norm": 0.2332375943660736,
+      "learning_rate": 1.4533506561564306e-05,
+      "loss": 0.05,
+      "num_input_tokens_seen": 55701992,
+      "step": 765
+    },
+    {
+      "epoch": 12.564901349948078,
+      "grad_norm": 0.3134738802909851,
+      "learning_rate": 1.4474104665812727e-05,
+      "loss": 0.066,
+      "num_input_tokens_seen": 55759312,
+      "step": 766
+    },
+    {
+      "epoch": 12.581516095534788,
+      "grad_norm": 0.26774415373802185,
+      "learning_rate": 1.44147749133692e-05,
+      "loss": 0.0541,
+      "num_input_tokens_seen": 55840400,
+      "step": 767
+    },
+    {
+      "epoch": 12.598130841121495,
+      "grad_norm": 0.28324708342552185,
+      "learning_rate": 1.4355517710873184e-05,
+      "loss": 0.0623,
+      "num_input_tokens_seen": 55912176,
+      "step": 768
+    },
+    {
+      "epoch": 12.614745586708203,
+      "grad_norm": 0.22127535939216614,
+      "learning_rate": 1.429633346446693e-05,
+      "loss": 0.0542,
+      "num_input_tokens_seen": 55989272,
+      "step": 769
+    },
+    {
+      "epoch": 12.631360332294912,
+      "grad_norm": 0.30930501222610474,
+      "learning_rate": 1.4237222579792618e-05,
+      "loss": 0.0719,
+      "num_input_tokens_seen": 56046720,
+      "step": 770
+    },
+    {
+      "epoch": 12.64797507788162,
+      "grad_norm": 0.27857473492622375,
+      "learning_rate": 1.4178185461989662e-05,
+      "loss": 0.0572,
+      "num_input_tokens_seen": 56132720,
+      "step": 771
+    },
+    {
+      "epoch": 12.664589823468328,
+      "grad_norm": 0.30680912733078003,
+      "learning_rate": 1.4119222515691816e-05,
+      "loss": 0.0648,
+      "num_input_tokens_seen": 56187824,
+      "step": 772
+    },
+    {
+      "epoch": 12.681204569055037,
+      "grad_norm": 0.2681683301925659,
+      "learning_rate": 1.4060334145024542e-05,
+      "loss": 0.0576,
+      "num_input_tokens_seen": 56259248,
+      "step": 773
+    },
+    {
+      "epoch": 12.697819314641745,
+      "grad_norm": 0.2639155983924866,
+      "learning_rate": 1.4001520753602121e-05,
+      "loss": 0.0637,
+      "num_input_tokens_seen": 56347584,
+      "step": 774
+    },
+    {
+      "epoch": 12.714434060228452,
+      "grad_norm": 0.25560155510902405,
+      "learning_rate": 1.3942782744524973e-05,
+      "loss": 0.0584,
+      "num_input_tokens_seen": 56433184,
+      "step": 775
+    },
+    {
+      "epoch": 12.731048805815162,
+      "grad_norm": 0.2899349629878998,
+      "learning_rate": 1.388412052037682e-05,
+      "loss": 0.0866,
+      "num_input_tokens_seen": 56491176,
+      "step": 776
+    },
+    {
+      "epoch": 12.74766355140187,
+      "grad_norm": 0.26568692922592163,
+      "learning_rate": 1.3825534483221974e-05,
+      "loss": 0.0536,
+      "num_input_tokens_seen": 56560952,
+      "step": 777
+    },
+    {
+      "epoch": 12.764278296988577,
+      "grad_norm": 0.27220121026039124,
+      "learning_rate": 1.376702503460259e-05,
+      "loss": 0.0638,
+      "num_input_tokens_seen": 56622384,
+      "step": 778
+    },
+    {
+      "epoch": 12.780893042575286,
+      "grad_norm": 0.24911001324653625,
+      "learning_rate": 1.3708592575535858e-05,
+      "loss": 0.0606,
+      "num_input_tokens_seen": 56709344,
+      "step": 779
+    },
+    {
+      "epoch": 12.797507788161994,
+      "grad_norm": 0.3394135534763336,
+      "learning_rate": 1.3650237506511331e-05,
+      "loss": 0.0752,
+      "num_input_tokens_seen": 56780664,
+      "step": 780
+    },
+    {
+      "epoch": 12.814122533748701,
+      "grad_norm": 0.26311007142066956,
+      "learning_rate": 1.3591960227488098e-05,
+      "loss": 0.0537,
+      "num_input_tokens_seen": 56862528,
+      "step": 781
+    },
+    {
+      "epoch": 12.83073727933541,
+      "grad_norm": 0.2802240550518036,
+      "learning_rate": 1.3533761137892136e-05,
+      "loss": 0.0575,
+      "num_input_tokens_seen": 56933384,
+      "step": 782
+    },
+    {
+      "epoch": 12.847352024922118,
+      "grad_norm": 0.2789711058139801,
+      "learning_rate": 1.3475640636613446e-05,
+      "loss": 0.0555,
+      "num_input_tokens_seen": 57023008,
+      "step": 783
+    },
+    {
+      "epoch": 12.863966770508826,
+      "grad_norm": 0.30197107791900635,
+      "learning_rate": 1.3417599122003464e-05,
+      "loss": 0.0743,
+      "num_input_tokens_seen": 57095032,
+      "step": 784
+    },
+    {
+      "epoch": 12.880581516095535,
+      "grad_norm": 0.2779630422592163,
+      "learning_rate": 1.3359636991872215e-05,
+      "loss": 0.1914,
+      "num_input_tokens_seen": 57162840,
+      "step": 785
+    },
+    {
+      "epoch": 12.897196261682243,
+      "grad_norm": 0.21620029211044312,
+      "learning_rate": 1.330175464348567e-05,
+      "loss": 0.0483,
+      "num_input_tokens_seen": 57247992,
+      "step": 786
+    },
+    {
+      "epoch": 12.91381100726895,
+      "grad_norm": 0.28678637742996216,
+      "learning_rate": 1.3243952473562942e-05,
+      "loss": 0.0709,
+      "num_input_tokens_seen": 57312216,
+      "step": 787
+    },
+    {
+      "epoch": 12.93042575285566,
+      "grad_norm": 0.3376156687736511,
+      "learning_rate": 1.3186230878273653e-05,
+      "loss": 0.0697,
+      "num_input_tokens_seen": 57367104,
+      "step": 788
+    },
+    {
+      "epoch": 12.947040498442368,
+      "grad_norm": 0.2537081837654114,
+      "learning_rate": 1.312859025323514e-05,
+      "loss": 0.0562,
+      "num_input_tokens_seen": 57437176,
+      "step": 789
+    },
+    {
+      "epoch": 12.963655244029075,
+      "grad_norm": 0.3236291706562042,
+      "learning_rate": 1.3071030993509788e-05,
+      "loss": 0.0653,
+      "num_input_tokens_seen": 57495304,
+      "step": 790
+    },
+    {
+      "epoch": 12.980269989615785,
+      "grad_norm": 0.2506027817726135,
+      "learning_rate": 1.3013553493602338e-05,
+      "loss": 0.0566,
+      "num_input_tokens_seen": 57582400,
+      "step": 791
+    },
+    {
+      "epoch": 12.996884735202492,
+      "grad_norm": 0.24630357325077057,
+      "learning_rate": 1.2956158147457115e-05,
+      "loss": 0.0577,
+      "num_input_tokens_seen": 57666136,
+      "step": 792
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 0.40804192423820496,
+      "learning_rate": 1.2898845348455418e-05,
+      "loss": 0.0425,
+      "num_input_tokens_seen": 57682352,
+      "step": 793
+    },
+    {
+      "epoch": 13.016614745586708,
+      "grad_norm": 0.24909114837646484,
+      "learning_rate": 1.2841615489412739e-05,
+      "loss": 0.0538,
+      "num_input_tokens_seen": 57745080,
+      "step": 794
+    },
+    {
+      "epoch": 13.033229491173417,
+      "grad_norm": 0.26190048456192017,
+      "learning_rate": 1.2784468962576136e-05,
+      "loss": 0.0569,
+      "num_input_tokens_seen": 57808680,
+      "step": 795
+    },
+    {
+      "epoch": 13.049844236760125,
+      "grad_norm": 0.2653025686740875,
+      "learning_rate": 1.272740615962148e-05,
+      "loss": 0.067,
+      "num_input_tokens_seen": 57878016,
+      "step": 796
+    },
+    {
+      "epoch": 13.066458982346832,
+      "grad_norm": 0.2509848177433014,
+      "learning_rate": 1.2670427471650864e-05,
+      "loss": 0.0593,
+      "num_input_tokens_seen": 57937640,
+      "step": 797
+    },
+    {
+      "epoch": 13.083073727933542,
+      "grad_norm": 0.2545263171195984,
+      "learning_rate": 1.261353328918981e-05,
+      "loss": 0.0554,
+      "num_input_tokens_seen": 58009448,
+      "step": 798
+    },
+    {
+      "epoch": 13.09968847352025,
+      "grad_norm": 0.2856726348400116,
+      "learning_rate": 1.2556724002184696e-05,
+      "loss": 0.2005,
+      "num_input_tokens_seen": 58068936,
+      "step": 799
+    },
+    {
+      "epoch": 13.116303219106957,
+      "grad_norm": 0.2398708462715149,
+      "learning_rate": 1.2500000000000006e-05,
+      "loss": 0.0501,
+      "num_input_tokens_seen": 58148824,
+      "step": 800
+    },
+    {
+      "epoch": 13.132917964693666,
+      "grad_norm": 0.24167364835739136,
+      "learning_rate": 1.2443361671415687e-05,
+      "loss": 0.048,
+      "num_input_tokens_seen": 58220720,
+      "step": 801
+    },
+    {
+      "epoch": 13.149532710280374,
+      "grad_norm": 0.2955470681190491,
+      "learning_rate": 1.2386809404624521e-05,
+      "loss": 0.05,
+      "num_input_tokens_seen": 58290944,
+      "step": 802
+    },
+    {
+      "epoch": 13.166147455867081,
+      "grad_norm": 0.24610291421413422,
+      "learning_rate": 1.2330343587229396e-05,
+      "loss": 0.0517,
+      "num_input_tokens_seen": 58382480,
+      "step": 803
+    },
+    {
+      "epoch": 13.18276220145379,
+      "grad_norm": 0.31220707297325134,
+      "learning_rate": 1.2273964606240718e-05,
+      "loss": 0.0558,
+      "num_input_tokens_seen": 58443424,
+      "step": 804
+    },
+    {
+      "epoch": 13.199376947040498,
+      "grad_norm": 0.27322131395339966,
+      "learning_rate": 1.2217672848073702e-05,
+      "loss": 0.0602,
+      "num_input_tokens_seen": 58534424,
+      "step": 805
+    },
+    {
+      "epoch": 13.215991692627206,
+      "grad_norm": 0.29824891686439514,
+      "learning_rate": 1.2161468698545755e-05,
+      "loss": 0.0585,
+      "num_input_tokens_seen": 58597336,
+      "step": 806
+    },
+    {
+      "epoch": 13.232606438213915,
+      "grad_norm": 0.3074946701526642,
+      "learning_rate": 1.2105352542873815e-05,
+      "loss": 0.0642,
+      "num_input_tokens_seen": 58647872,
+      "step": 807
+    },
+    {
+      "epoch": 13.249221183800623,
+      "grad_norm": 0.27672338485717773,
+      "learning_rate": 1.2049324765671749e-05,
+      "loss": 0.0544,
+      "num_input_tokens_seen": 58731544,
+      "step": 808
+    },
+    {
+      "epoch": 13.26583592938733,
+      "grad_norm": 0.2907312512397766,
+      "learning_rate": 1.199338575094765e-05,
+      "loss": 0.0639,
+      "num_input_tokens_seen": 58788488,
+      "step": 809
+    },
+    {
+      "epoch": 13.28245067497404,
+      "grad_norm": 0.17112018167972565,
+      "learning_rate": 1.1937535882101281e-05,
+      "loss": 0.0392,
+      "num_input_tokens_seen": 58886640,
+      "step": 810
+    },
+    {
+      "epoch": 13.299065420560748,
+      "grad_norm": 0.3354332149028778,
+      "learning_rate": 1.1881775541921378e-05,
+      "loss": 0.0644,
+      "num_input_tokens_seen": 58943304,
+      "step": 811
+    },
+    {
+      "epoch": 13.315680166147455,
+      "grad_norm": 0.25639235973358154,
+      "learning_rate": 1.1826105112583061e-05,
+      "loss": 0.0632,
+      "num_input_tokens_seen": 59020992,
+      "step": 812
+    },
+    {
+      "epoch": 13.332294911734165,
+      "grad_norm": 0.21590979397296906,
+      "learning_rate": 1.1770524975645238e-05,
+      "loss": 0.0424,
+      "num_input_tokens_seen": 59111368,
+      "step": 813
+    },
+    {
+      "epoch": 13.348909657320872,
+      "grad_norm": 0.28309866786003113,
+      "learning_rate": 1.1715035512047925e-05,
+      "loss": 0.0574,
+      "num_input_tokens_seen": 59194880,
+      "step": 814
+    },
+    {
+      "epoch": 13.36552440290758,
+      "grad_norm": 0.2805883586406708,
+      "learning_rate": 1.1659637102109714e-05,
+      "loss": 0.0521,
+      "num_input_tokens_seen": 59250648,
+      "step": 815
+    },
+    {
+      "epoch": 13.38213914849429,
+      "grad_norm": 0.25733205676078796,
+      "learning_rate": 1.1604330125525079e-05,
+      "loss": 0.0515,
+      "num_input_tokens_seen": 59324888,
+      "step": 816
+    },
+    {
+      "epoch": 13.398753894080997,
+      "grad_norm": 0.28011178970336914,
+      "learning_rate": 1.154911496136188e-05,
+      "loss": 0.0585,
+      "num_input_tokens_seen": 59395120,
+      "step": 817
+    },
+    {
+      "epoch": 13.415368639667705,
+      "grad_norm": 0.2550284266471863,
+      "learning_rate": 1.1493991988058625e-05,
+      "loss": 0.0526,
+      "num_input_tokens_seen": 59475152,
+      "step": 818
+    },
+    {
+      "epoch": 13.431983385254414,
+      "grad_norm": 0.2467280924320221,
+      "learning_rate": 1.1438961583422037e-05,
+      "loss": 0.0484,
+      "num_input_tokens_seen": 59538328,
+      "step": 819
+    },
+    {
+      "epoch": 13.448598130841122,
+      "grad_norm": 0.31466010212898254,
+      "learning_rate": 1.1384024124624324e-05,
+      "loss": 0.0599,
+      "num_input_tokens_seen": 59606280,
+      "step": 820
+    },
+    {
+      "epoch": 13.46521287642783,
+      "grad_norm": 0.2625578045845032,
+      "learning_rate": 1.1329179988200694e-05,
+      "loss": 0.0585,
+      "num_input_tokens_seen": 59694032,
+      "step": 821
+    },
+    {
+      "epoch": 13.481827622014539,
+      "grad_norm": 0.24019013345241547,
+      "learning_rate": 1.1274429550046704e-05,
+      "loss": 0.0534,
+      "num_input_tokens_seen": 59764904,
+      "step": 822
+    },
+    {
+      "epoch": 13.498442367601246,
+      "grad_norm": 0.23081430792808533,
+      "learning_rate": 1.121977318541575e-05,
+      "loss": 0.0486,
+      "num_input_tokens_seen": 59839232,
+      "step": 823
+    },
+    {
+      "epoch": 13.515057113187954,
+      "grad_norm": 0.2902800738811493,
+      "learning_rate": 1.11652112689164e-05,
+      "loss": 0.0608,
+      "num_input_tokens_seen": 59907248,
+      "step": 824
+    },
+    {
+      "epoch": 13.531671858774663,
+      "grad_norm": 0.27917391061782837,
+      "learning_rate": 1.1110744174509952e-05,
+      "loss": 0.062,
+      "num_input_tokens_seen": 59971912,
+      "step": 825
+    },
+    {
+      "epoch": 13.54828660436137,
+      "grad_norm": 0.2244357317686081,
+      "learning_rate": 1.1056372275507749e-05,
+      "loss": 0.0495,
+      "num_input_tokens_seen": 60054728,
+      "step": 826
+    },
+    {
+      "epoch": 13.564901349948078,
+      "grad_norm": 0.24689778685569763,
+      "learning_rate": 1.1002095944568707e-05,
+      "loss": 0.0506,
+      "num_input_tokens_seen": 60133824,
+      "step": 827
+    },
+    {
+      "epoch": 13.581516095534788,
+      "grad_norm": 0.2756742537021637,
+      "learning_rate": 1.0947915553696742e-05,
+      "loss": 0.058,
+      "num_input_tokens_seen": 60194832,
+      "step": 828
+    },
+    {
+      "epoch": 13.598130841121495,
+      "grad_norm": 0.2537204623222351,
+      "learning_rate": 1.089383147423815e-05,
+      "loss": 0.0592,
+      "num_input_tokens_seen": 60280224,
+      "step": 829
+    },
+    {
+      "epoch": 13.614745586708203,
+      "grad_norm": 0.3149183392524719,
+      "learning_rate": 1.0839844076879185e-05,
+      "loss": 0.0658,
+      "num_input_tokens_seen": 60341144,
+      "step": 830
+    },
+    {
+      "epoch": 13.631360332294912,
+      "grad_norm": 0.26036155223846436,
+      "learning_rate": 1.07859537316434e-05,
+      "loss": 0.0504,
+      "num_input_tokens_seen": 60425128,
+      "step": 831
+    },
+    {
+      "epoch": 13.64797507788162,
+      "grad_norm": 0.20459908246994019,
+      "learning_rate": 1.0732160807889211e-05,
+      "loss": 0.0523,
+      "num_input_tokens_seen": 60520456,
+      "step": 832
+    },
+    {
+      "epoch": 13.664589823468328,
+      "grad_norm": 0.24871429800987244,
+      "learning_rate": 1.0678465674307273e-05,
+      "loss": 0.0495,
+      "num_input_tokens_seen": 60617640,
+      "step": 833
+    },
+    {
+      "epoch": 13.681204569055037,
+      "grad_norm": 0.30655166506767273,
+      "learning_rate": 1.0624868698918045e-05,
+      "loss": 0.0632,
+      "num_input_tokens_seen": 60676584,
+      "step": 834
+    },
+    {
+      "epoch": 13.697819314641745,
+      "grad_norm": 0.25585272908210754,
+      "learning_rate": 1.0571370249069162e-05,
+      "loss": 0.1454,
+      "num_input_tokens_seen": 60750760,
+      "step": 835
+    },
+    {
+      "epoch": 13.714434060228452,
+      "grad_norm": 0.3102145791053772,
+      "learning_rate": 1.0517970691433035e-05,
+      "loss": 0.0637,
+      "num_input_tokens_seen": 60823520,
+      "step": 836
+    },
+    {
+      "epoch": 13.731048805815162,
+      "grad_norm": 0.21035313606262207,
+      "learning_rate": 1.0464670392004235e-05,
+      "loss": 0.0587,
+      "num_input_tokens_seen": 60917712,
+      "step": 837
+    },
+    {
+      "epoch": 13.74766355140187,
+      "grad_norm": 0.2799111306667328,
+      "learning_rate": 1.0411469716097067e-05,
+      "loss": 0.0646,
+      "num_input_tokens_seen": 60979928,
+      "step": 838
+    },
+    {
+      "epoch": 13.764278296988577,
+      "grad_norm": 0.2967042624950409,
+      "learning_rate": 1.0358369028342985e-05,
+      "loss": 0.0609,
+      "num_input_tokens_seen": 61050592,
+      "step": 839
+    },
+    {
+      "epoch": 13.780893042575286,
+      "grad_norm": 0.24081780016422272,
+      "learning_rate": 1.0305368692688174e-05,
+      "loss": 0.0457,
+      "num_input_tokens_seen": 61126704,
+      "step": 840
+    },
+    {
+      "epoch": 13.797507788161994,
+      "grad_norm": 0.28168758749961853,
+      "learning_rate": 1.0252469072390994e-05,
+      "loss": 0.0592,
+      "num_input_tokens_seen": 61192016,
+      "step": 841
+    },
+    {
+      "epoch": 13.814122533748701,
+      "grad_norm": 0.3008675277233124,
+      "learning_rate": 1.0199670530019511e-05,
+      "loss": 0.0687,
+      "num_input_tokens_seen": 61264256,
+      "step": 842
+    },
+    {
+      "epoch": 13.83073727933541,
+      "grad_norm": 0.28137174248695374,
+      "learning_rate": 1.0146973427449038e-05,
+      "loss": 0.0559,
+      "num_input_tokens_seen": 61330712,
+      "step": 843
+    },
+    {
+      "epoch": 13.847352024922118,
+      "grad_norm": 0.29582953453063965,
+      "learning_rate": 1.0094378125859602e-05,
+      "loss": 0.0539,
+      "num_input_tokens_seen": 61403992,
+      "step": 844
+    },
+    {
+      "epoch": 13.863966770508826,
+      "grad_norm": 0.2721082866191864,
+      "learning_rate": 1.0041884985733524e-05,
+      "loss": 0.0551,
+      "num_input_tokens_seen": 61505072,
+      "step": 845
+    },
+    {
+      "epoch": 13.880581516095535,
+      "grad_norm": 0.29855334758758545,
+      "learning_rate": 9.989494366852904e-06,
+      "loss": 0.058,
+      "num_input_tokens_seen": 61574704,
+      "step": 846
+    },
+    {
+      "epoch": 13.897196261682243,
+      "grad_norm": 0.2921924591064453,
+      "learning_rate": 9.937206628297172e-06,
+      "loss": 0.0584,
+      "num_input_tokens_seen": 61647640,
+      "step": 847
+    },
+    {
+      "epoch": 13.91381100726895,
+      "grad_norm": 0.32202112674713135,
+      "learning_rate": 9.88502212844063e-06,
+      "loss": 0.0682,
+      "num_input_tokens_seen": 61708696,
+      "step": 848
+    },
+    {
+      "epoch": 13.93042575285566,
+      "grad_norm": 0.2174832671880722,
+      "learning_rate": 9.832941224950012e-06,
+      "loss": 0.0436,
+      "num_input_tokens_seen": 61801464,
+      "step": 849
+    },
+    {
+      "epoch": 13.947040498442368,
+      "grad_norm": 0.255197674036026,
+      "learning_rate": 9.780964274781984e-06,
+      "loss": 0.0593,
+      "num_input_tokens_seen": 61876376,
+      "step": 850
+    },
+    {
+      "epoch": 13.963655244029075,
+      "grad_norm": 0.25164854526519775,
+      "learning_rate": 9.729091634180756e-06,
+      "loss": 0.0562,
+      "num_input_tokens_seen": 61947232,
+      "step": 851
+    },
+    {
+      "epoch": 13.980269989615785,
+      "grad_norm": 0.2508520781993866,
+      "learning_rate": 9.677323658675594e-06,
+      "loss": 0.0485,
+      "num_input_tokens_seen": 62026520,
+      "step": 852
+    },
+    {
+      "epoch": 13.996884735202492,
+      "grad_norm": 0.24607762694358826,
+      "learning_rate": 9.625660703078392e-06,
+      "loss": 0.1437,
+      "num_input_tokens_seen": 62107152,
+      "step": 853
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 0.6298788189888,
+      "learning_rate": 9.574103121481288e-06,
+      "loss": 0.054,
+      "num_input_tokens_seen": 62117944,
+      "step": 854
+    },
+    {
+      "epoch": 14.016614745586708,
+      "grad_norm": 0.2503069341182709,
+      "learning_rate": 9.522651267254149e-06,
+      "loss": 0.0527,
+      "num_input_tokens_seen": 62192136,
+      "step": 855
+    },
+    {
+      "epoch": 14.033229491173417,
+      "grad_norm": 0.22820697724819183,
+      "learning_rate": 9.471305493042243e-06,
+      "loss": 0.0479,
+      "num_input_tokens_seen": 62290128,
+      "step": 856
+    },
+    {
+      "epoch": 14.049844236760125,
+      "grad_norm": 0.2834872901439667,
+      "learning_rate": 9.420066150763748e-06,
+      "loss": 0.0542,
+      "num_input_tokens_seen": 62346640,
+      "step": 857
+    },
+    {
+      "epoch": 14.066458982346832,
+      "grad_norm": 0.21925635635852814,
+      "learning_rate": 9.368933591607378e-06,
+      "loss": 0.0412,
+      "num_input_tokens_seen": 62430464,
+      "step": 858
+    },
+    {
+      "epoch": 14.083073727933542,
+      "grad_norm": 0.27315613627433777,
+      "learning_rate": 9.317908166029962e-06,
+      "loss": 0.055,
+      "num_input_tokens_seen": 62487224,
+      "step": 859
+    },
+    {
+      "epoch": 14.09968847352025,
+      "grad_norm": 0.2407708764076233,
+      "learning_rate": 9.266990223754069e-06,
+      "loss": 0.0644,
+      "num_input_tokens_seen": 62570352,
+      "step": 860
+    },
+    {
+      "epoch": 14.116303219106957,
+      "grad_norm": 0.23930609226226807,
+      "learning_rate": 9.216180113765558e-06,
+      "loss": 0.0509,
+      "num_input_tokens_seen": 62677432,
+      "step": 861
+    },
+    {
+      "epoch": 14.132917964693666,
+      "grad_norm": 0.2706838548183441,
+      "learning_rate": 9.16547818431125e-06,
+      "loss": 0.0538,
+      "num_input_tokens_seen": 62750304,
+      "step": 862
+    },
+    {
+      "epoch": 14.149532710280374,
+      "grad_norm": 0.2537670433521271,
+      "learning_rate": 9.114884782896483e-06,
+      "loss": 0.0495,
+      "num_input_tokens_seen": 62817736,
+      "step": 863
+    },
+    {
+      "epoch": 14.166147455867081,
+      "grad_norm": 0.2521739602088928,
+      "learning_rate": 9.064400256282757e-06,
+      "loss": 0.0521,
+      "num_input_tokens_seen": 62896016,
+      "step": 864
+    },
+    {
+      "epoch": 14.18276220145379,
+      "grad_norm": 0.2947329580783844,
+      "learning_rate": 9.014024950485383e-06,
+      "loss": 0.0483,
+      "num_input_tokens_seen": 62957488,
+      "step": 865
+    },
+    {
+      "epoch": 14.199376947040498,
+      "grad_norm": 0.2857937514781952,
+      "learning_rate": 8.963759210771052e-06,
+      "loss": 0.0526,
+      "num_input_tokens_seen": 63013344,
+      "step": 866
+    },
+    {
+      "epoch": 14.215991692627206,
+      "grad_norm": 0.27009105682373047,
+      "learning_rate": 8.913603381655528e-06,
+      "loss": 0.0543,
+      "num_input_tokens_seen": 63073504,
+      "step": 867
+    },
+    {
+      "epoch": 14.232606438213915,
+      "grad_norm": 0.24146534502506256,
+      "learning_rate": 8.863557806901233e-06,
+      "loss": 0.0438,
+      "num_input_tokens_seen": 63143344,
+      "step": 868
+    },
+    {
+      "epoch": 14.249221183800623,
+      "grad_norm": 0.19567841291427612,
+      "learning_rate": 8.813622829514956e-06,
+      "loss": 0.041,
+      "num_input_tokens_seen": 63245488,
+      "step": 869
+    },
+    {
+      "epoch": 14.26583592938733,
+      "grad_norm": 0.2820407748222351,
+      "learning_rate": 8.763798791745411e-06,
+      "loss": 0.0523,
+      "num_input_tokens_seen": 63297208,
+      "step": 870
+    },
+    {
+      "epoch": 14.28245067497404,
+      "grad_norm": 0.28351321816444397,
+      "learning_rate": 8.714086035080996e-06,
+      "loss": 0.055,
+      "num_input_tokens_seen": 63357392,
+      "step": 871
+    },
+    {
+      "epoch": 14.299065420560748,
+      "grad_norm": 0.23567432165145874,
+      "learning_rate": 8.664484900247363e-06,
+      "loss": 0.0499,
+      "num_input_tokens_seen": 63429064,
+      "step": 872
+    },
+    {
+      "epoch": 14.315680166147455,
+      "grad_norm": 0.32714203000068665,
+      "learning_rate": 8.614995727205156e-06,
+      "loss": 0.0698,
+      "num_input_tokens_seen": 63478112,
+      "step": 873
+    },
+    {
+      "epoch": 14.332294911734165,
+      "grad_norm": 0.2806774377822876,
+      "learning_rate": 8.565618855147603e-06,
+      "loss": 0.0464,
+      "num_input_tokens_seen": 63558472,
+      "step": 874
+    },
+    {
+      "epoch": 14.348909657320872,
+      "grad_norm": 0.22814752161502838,
+      "learning_rate": 8.51635462249828e-06,
+      "loss": 0.0474,
+      "num_input_tokens_seen": 63641776,
+      "step": 875
+    },
+    {
+      "epoch": 14.36552440290758,
+      "grad_norm": 0.20048512518405914,
+      "learning_rate": 8.467203366908707e-06,
+      "loss": 0.0458,
+      "num_input_tokens_seen": 63746080,
+      "step": 876
+    },
+    {
+      "epoch": 14.38213914849429,
+      "grad_norm": 0.17419511079788208,
+      "learning_rate": 8.41816542525608e-06,
+      "loss": 0.1133,
+      "num_input_tokens_seen": 63846320,
+      "step": 877
+    },
+    {
+      "epoch": 14.398753894080997,
+      "grad_norm": 0.26787328720092773,
+      "learning_rate": 8.369241133640982e-06,
+      "loss": 0.0462,
+      "num_input_tokens_seen": 63917736,
+      "step": 878
+    },
+    {
+      "epoch": 14.415368639667705,
+      "grad_norm": 0.2565169334411621,
+      "learning_rate": 8.320430827385003e-06,
+      "loss": 0.133,
+      "num_input_tokens_seen": 63981576,
+      "step": 879
+    },
+    {
+      "epoch": 14.431983385254414,
+      "grad_norm": 0.31951773166656494,
+      "learning_rate": 8.271734841028553e-06,
+      "loss": 0.0649,
+      "num_input_tokens_seen": 64032848,
+      "step": 880
+    },
+    {
+      "epoch": 14.448598130841122,
+      "grad_norm": 0.261004775762558,
+      "learning_rate": 8.22315350832843e-06,
+      "loss": 0.0533,
+      "num_input_tokens_seen": 64098072,
+      "step": 881
+    },
+    {
+      "epoch": 14.46521287642783,
+      "grad_norm": 0.2179432213306427,
+      "learning_rate": 8.174687162255672e-06,
+      "loss": 0.0484,
+      "num_input_tokens_seen": 64186088,
+      "step": 882
+    },
+    {
+      "epoch": 14.481827622014539,
+      "grad_norm": 0.31436631083488464,
+      "learning_rate": 8.126336134993176e-06,
+      "loss": 0.05,
+      "num_input_tokens_seen": 64277520,
+      "step": 883
+    },
+    {
+      "epoch": 14.498442367601246,
+      "grad_norm": 0.3128837049007416,
+      "learning_rate": 8.078100757933485e-06,
+      "loss": 0.0603,
+      "num_input_tokens_seen": 64340728,
+      "step": 884
+    },
+    {
+      "epoch": 14.515057113187954,
+      "grad_norm": 0.2509966492652893,
+      "learning_rate": 8.029981361676456e-06,
+      "loss": 0.0497,
+      "num_input_tokens_seen": 64417080,
+      "step": 885
+    },
+    {
+      "epoch": 14.531671858774663,
+      "grad_norm": 0.2274176925420761,
+      "learning_rate": 7.981978276027055e-06,
+      "loss": 0.0396,
+      "num_input_tokens_seen": 64503560,
+      "step": 886
+    },
+    {
+      "epoch": 14.54828660436137,
+      "grad_norm": 0.23473070561885834,
+      "learning_rate": 7.934091829993055e-06,
+      "loss": 0.0505,
+      "num_input_tokens_seen": 64575432,
+      "step": 887
+    },
+    {
+      "epoch": 14.564901349948078,
+      "grad_norm": 0.2230038344860077,
+      "learning_rate": 7.886322351782783e-06,
+      "loss": 0.0455,
+      "num_input_tokens_seen": 64669216,
+      "step": 888
+    },
+    {
+      "epoch": 14.581516095534788,
+      "grad_norm": 0.27560481429100037,
+      "learning_rate": 7.838670168802909e-06,
+      "loss": 0.0633,
+      "num_input_tokens_seen": 64728640,
+      "step": 889
+    },
+    {
+      "epoch": 14.598130841121495,
+      "grad_norm": 0.30317774415016174,
+      "learning_rate": 7.791135607656147e-06,
+      "loss": 0.0598,
+      "num_input_tokens_seen": 64780032,
+      "step": 890
+    },
+    {
+      "epoch": 14.614745586708203,
+      "grad_norm": 0.2672622799873352,
+      "learning_rate": 7.743718994139071e-06,
+      "loss": 0.0536,
+      "num_input_tokens_seen": 64863760,
+      "step": 891
+    },
+    {
+      "epoch": 14.631360332294912,
+      "grad_norm": 0.29224610328674316,
+      "learning_rate": 7.696420653239833e-06,
+      "loss": 0.0515,
+      "num_input_tokens_seen": 64928808,
+      "step": 892
+    },
+    {
+      "epoch": 14.64797507788162,
+      "grad_norm": 0.2772779166698456,
+      "learning_rate": 7.649240909135965e-06,
+      "loss": 0.0579,
+      "num_input_tokens_seen": 64999256,
+      "step": 893
+    },
+    {
+      "epoch": 14.664589823468328,
+      "grad_norm": 0.22769714891910553,
+      "learning_rate": 7.602180085192143e-06,
+      "loss": 0.0462,
+      "num_input_tokens_seen": 65081544,
+      "step": 894
+    },
+    {
+      "epoch": 14.681204569055037,
+      "grad_norm": 0.28400924801826477,
+      "learning_rate": 7.555238503958001e-06,
+      "loss": 0.0545,
+      "num_input_tokens_seen": 65145096,
+      "step": 895
+    },
+    {
+      "epoch": 14.697819314641745,
+      "grad_norm": 0.2113635390996933,
+      "learning_rate": 7.508416487165862e-06,
+      "loss": 0.0446,
+      "num_input_tokens_seen": 65229504,
+      "step": 896
+    },
+    {
+      "epoch": 14.714434060228452,
+      "grad_norm": 0.2663005590438843,
+      "learning_rate": 7.461714355728608e-06,
+      "loss": 0.0511,
+      "num_input_tokens_seen": 65299408,
+      "step": 897
+    },
+    {
+      "epoch": 14.731048805815162,
+      "grad_norm": 0.2649020850658417,
+      "learning_rate": 7.415132429737407e-06,
+      "loss": 0.061,
+      "num_input_tokens_seen": 65376216,
+      "step": 898
+    },
+    {
+      "epoch": 14.74766355140187,
+      "grad_norm": 0.22136181592941284,
+      "learning_rate": 7.368671028459564e-06,
+      "loss": 0.0453,
+      "num_input_tokens_seen": 65457240,
+      "step": 899
+    },
+    {
+      "epoch": 14.764278296988577,
+      "grad_norm": 0.2594226896762848,
+      "learning_rate": 7.3223304703363135e-06,
+      "loss": 0.0585,
+      "num_input_tokens_seen": 65531064,
+      "step": 900
+    },
+    {
+      "epoch": 14.780893042575286,
+      "grad_norm": 0.24974443018436432,
+      "learning_rate": 7.276111072980663e-06,
+      "loss": 0.1328,
+      "num_input_tokens_seen": 65600688,
+      "step": 901
+    },
+    {
+      "epoch": 14.797507788161994,
+      "grad_norm": 0.2837437689304352,
+      "learning_rate": 7.230013153175188e-06,
+      "loss": 0.0515,
+      "num_input_tokens_seen": 65678640,
+      "step": 902
+    },
+    {
+      "epoch": 14.814122533748701,
+      "grad_norm": 0.25543513894081116,
+      "learning_rate": 7.184037026869867e-06,
+      "loss": 0.0496,
+      "num_input_tokens_seen": 65755816,
+      "step": 903
+    },
+    {
+      "epoch": 14.83073727933541,
+      "grad_norm": 0.24205391108989716,
+      "learning_rate": 7.138183009179922e-06,
+      "loss": 0.0499,
+      "num_input_tokens_seen": 65837696,
+      "step": 904
+    },
+    {
+      "epoch": 14.847352024922118,
+      "grad_norm": 0.325641393661499,
+      "learning_rate": 7.092451414383644e-06,
+      "loss": 0.0545,
+      "num_input_tokens_seen": 65901792,
+      "step": 905
+    },
+    {
+      "epoch": 14.863966770508826,
+      "grad_norm": 0.28845149278640747,
+      "learning_rate": 7.046842555920283e-06,
+      "loss": 0.0614,
+      "num_input_tokens_seen": 65969960,
+      "step": 906
+    },
+    {
+      "epoch": 14.880581516095535,
+      "grad_norm": 0.2403915971517563,
+      "learning_rate": 7.00135674638783e-06,
+      "loss": 0.053,
+      "num_input_tokens_seen": 66043752,
+      "step": 907
+    },
+    {
+      "epoch": 14.897196261682243,
+      "grad_norm": 0.35168591141700745,
+      "learning_rate": 6.9559942975409465e-06,
+      "loss": 0.0562,
+      "num_input_tokens_seen": 66100456,
+      "step": 908
+    },
+    {
+      "epoch": 14.91381100726895,
+      "grad_norm": 0.2577241361141205,
+      "learning_rate": 6.91075552028877e-06,
+      "loss": 0.0582,
+      "num_input_tokens_seen": 66177320,
+      "step": 909
+    },
+    {
+      "epoch": 14.93042575285566,
+      "grad_norm": 0.28989753127098083,
+      "learning_rate": 6.865640724692815e-06,
+      "loss": 0.0582,
+      "num_input_tokens_seen": 66231656,
+      "step": 910
+    },
+    {
+      "epoch": 14.947040498442368,
+      "grad_norm": 0.25177815556526184,
+      "learning_rate": 6.820650219964833e-06,
+      "loss": 0.0477,
+      "num_input_tokens_seen": 66315768,
+      "step": 911
+    },
+    {
+      "epoch": 14.963655244029075,
+      "grad_norm": 0.2509918808937073,
+      "learning_rate": 6.775784314464717e-06,
+      "loss": 0.0522,
+      "num_input_tokens_seen": 66391256,
+      "step": 912
+    },
+    {
+      "epoch": 14.980269989615785,
+      "grad_norm": 0.30490386486053467,
+      "learning_rate": 6.731043315698346e-06,
+      "loss": 0.0617,
+      "num_input_tokens_seen": 66469184,
+      "step": 913
+    },
+    {
+      "epoch": 14.996884735202492,
+      "grad_norm": 0.2597614526748657,
+      "learning_rate": 6.686427530315534e-06,
+      "loss": 0.0553,
+      "num_input_tokens_seen": 66540392,
+      "step": 914
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 0.3879867196083069,
+      "learning_rate": 6.641937264107867e-06,
+      "loss": 0.038,
+      "num_input_tokens_seen": 66561792,
+      "step": 915
+    },
+    {
+      "epoch": 15.016614745586708,
+      "grad_norm": 0.26023438572883606,
+      "learning_rate": 6.5975728220066425e-06,
+      "loss": 0.0557,
+      "num_input_tokens_seen": 66635072,
+      "step": 916
+    },
+    {
+      "epoch": 15.033229491173417,
+      "grad_norm": 0.22681663930416107,
+      "learning_rate": 6.553334508080794e-06,
+      "loss": 0.0475,
+      "num_input_tokens_seen": 66708200,
+      "step": 917
+    },
+    {
+      "epoch": 15.049844236760125,
+      "grad_norm": 0.19712482392787933,
+      "learning_rate": 6.509222625534755e-06,
+      "loss": 0.0401,
+      "num_input_tokens_seen": 66796600,
+      "step": 918
+    },
+    {
+      "epoch": 15.066458982346832,
+      "grad_norm": 0.2641635835170746,
+      "learning_rate": 6.465237476706449e-06,
+      "loss": 0.0624,
+      "num_input_tokens_seen": 66878560,
+      "step": 919
+    },
+    {
+      "epoch": 15.083073727933542,
+      "grad_norm": 0.24981576204299927,
+      "learning_rate": 6.421379363065142e-06,
+      "loss": 0.0502,
+      "num_input_tokens_seen": 66944600,
+      "step": 920
+    },
+    {
+      "epoch": 15.09968847352025,
+      "grad_norm": 0.19411137700080872,
+      "learning_rate": 6.377648585209456e-06,
+      "loss": 0.0384,
+      "num_input_tokens_seen": 67026488,
+      "step": 921
+    },
+    {
+      "epoch": 15.116303219106957,
+      "grad_norm": 0.2732497751712799,
+      "learning_rate": 6.334045442865219e-06,
+      "loss": 0.0525,
+      "num_input_tokens_seen": 67095344,
+      "step": 922
+    },
+    {
+      "epoch": 15.132917964693666,
+      "grad_norm": 0.26258111000061035,
+      "learning_rate": 6.290570234883505e-06,
+      "loss": 0.0567,
+      "num_input_tokens_seen": 67171352,
+      "step": 923
+    },
+    {
+      "epoch": 15.149532710280374,
+      "grad_norm": 0.3083919286727905,
+      "learning_rate": 6.247223259238511e-06,
+      "loss": 0.0611,
+      "num_input_tokens_seen": 67232288,
+      "step": 924
+    },
+    {
+      "epoch": 15.166147455867081,
+      "grad_norm": 0.3361482322216034,
+      "learning_rate": 6.204004813025568e-06,
+      "loss": 0.0583,
+      "num_input_tokens_seen": 67287128,
+      "step": 925
+    },
+    {
+      "epoch": 15.18276220145379,
+      "grad_norm": 0.21773198246955872,
+      "learning_rate": 6.160915192459058e-06,
+      "loss": 0.0419,
+      "num_input_tokens_seen": 67362248,
+      "step": 926
+    },
+    {
+      "epoch": 15.199376947040498,
+      "grad_norm": 0.27043187618255615,
+      "learning_rate": 6.117954692870412e-06,
+      "loss": 0.0484,
+      "num_input_tokens_seen": 67427448,
+      "step": 927
+    },
+    {
+      "epoch": 15.215991692627206,
+      "grad_norm": 0.22789040207862854,
+      "learning_rate": 6.075123608706093e-06,
+      "loss": 0.0477,
+      "num_input_tokens_seen": 67506648,
+      "step": 928
+    },
+    {
+      "epoch": 15.232606438213915,
+      "grad_norm": 0.21521471440792084,
+      "learning_rate": 6.032422233525545e-06,
+      "loss": 0.0411,
+      "num_input_tokens_seen": 67589024,
+      "step": 929
+    },
+    {
+      "epoch": 15.249221183800623,
+      "grad_norm": 0.2916419804096222,
+      "learning_rate": 5.989850859999227e-06,
+      "loss": 0.0511,
+      "num_input_tokens_seen": 67649592,
+      "step": 930
+    },
+    {
+      "epoch": 15.26583592938733,
+      "grad_norm": 0.2598723769187927,
+      "learning_rate": 5.947409779906554e-06,
+      "loss": 0.048,
+      "num_input_tokens_seen": 67717080,
+      "step": 931
+    },
+    {
+      "epoch": 15.28245067497404,
+      "grad_norm": 0.22704635560512543,
+      "learning_rate": 5.905099284133952e-06,
+      "loss": 0.0511,
+      "num_input_tokens_seen": 67791560,
+      "step": 932
+    },
+    {
+      "epoch": 15.299065420560748,
+      "grad_norm": 0.23701129853725433,
+      "learning_rate": 5.8629196626728e-06,
+      "loss": 0.0452,
+      "num_input_tokens_seen": 67852376,
+      "step": 933
+    },
+    {
+      "epoch": 15.315680166147455,
+      "grad_norm": 0.18171566724777222,
+      "learning_rate": 5.820871204617515e-06,
+      "loss": 0.0384,
+      "num_input_tokens_seen": 67946192,
+      "step": 934
+    },
+    {
+      "epoch": 15.332294911734165,
+      "grad_norm": 0.21201346814632416,
+      "learning_rate": 5.778954198163514e-06,
+      "loss": 0.0404,
+      "num_input_tokens_seen": 68025248,
+      "step": 935
+    },
+    {
+      "epoch": 15.348909657320872,
+      "grad_norm": 0.24997848272323608,
+      "learning_rate": 5.737168930605272e-06,
+      "loss": 0.0517,
+      "num_input_tokens_seen": 68109776,
+      "step": 936
+    },
+    {
+      "epoch": 15.36552440290758,
+      "grad_norm": 0.22962713241577148,
+      "learning_rate": 5.6955156883343265e-06,
+      "loss": 0.0481,
+      "num_input_tokens_seen": 68187496,
+      "step": 937
+    },
+    {
+      "epoch": 15.38213914849429,
+      "grad_norm": 0.21411846578121185,
+      "learning_rate": 5.653994756837347e-06,
+      "loss": 0.0429,
+      "num_input_tokens_seen": 68268560,
+      "step": 938
+    },
+    {
+      "epoch": 15.398753894080997,
+      "grad_norm": 0.25815847516059875,
+      "learning_rate": 5.612606420694141e-06,
+      "loss": 0.058,
+      "num_input_tokens_seen": 68331208,
+      "step": 939
+    },
+    {
+      "epoch": 15.415368639667705,
+      "grad_norm": 0.25215816497802734,
+      "learning_rate": 5.571350963575728e-06,
+      "loss": 0.0561,
+      "num_input_tokens_seen": 68401712,
+      "step": 940
+    },
+    {
+      "epoch": 15.431983385254414,
+      "grad_norm": 0.19544276595115662,
+      "learning_rate": 5.530228668242402e-06,
+      "loss": 0.0411,
+      "num_input_tokens_seen": 68487824,
+      "step": 941
+    },
+    {
+      "epoch": 15.448598130841122,
+      "grad_norm": 0.2418648898601532,
+      "learning_rate": 5.489239816541755e-06,
+      "loss": 0.05,
+      "num_input_tokens_seen": 68558256,
+      "step": 942
+    },
+    {
+      "epoch": 15.46521287642783,
+      "grad_norm": 0.27279072999954224,
+      "learning_rate": 5.4483846894068046e-06,
+      "loss": 0.0493,
+      "num_input_tokens_seen": 68652768,
+      "step": 943
+    },
+    {
+      "epoch": 15.481827622014539,
+      "grad_norm": 0.21187296509742737,
+      "learning_rate": 5.4076635668540075e-06,
+      "loss": 0.0433,
+      "num_input_tokens_seen": 68745352,
+      "step": 944
+    },
+    {
+      "epoch": 15.498442367601246,
+      "grad_norm": 0.20128142833709717,
+      "learning_rate": 5.367076727981382e-06,
+      "loss": 0.0411,
+      "num_input_tokens_seen": 68830520,
+      "step": 945
+    },
+    {
+      "epoch": 15.515057113187954,
+      "grad_norm": 0.30053532123565674,
+      "learning_rate": 5.326624450966569e-06,
+      "loss": 0.2515,
+      "num_input_tokens_seen": 68894064,
+      "step": 946
+    },
+    {
+      "epoch": 15.531671858774663,
+      "grad_norm": 0.23420806229114532,
+      "learning_rate": 5.286307013064956e-06,
+      "loss": 0.0441,
+      "num_input_tokens_seen": 68963888,
+      "step": 947
+    },
+    {
+      "epoch": 15.54828660436137,
+      "grad_norm": 0.2615622282028198,
+      "learning_rate": 5.24612469060774e-06,
+      "loss": 0.048,
+      "num_input_tokens_seen": 69025032,
+      "step": 948
+    },
+    {
+      "epoch": 15.564901349948078,
+      "grad_norm": 0.2581160366535187,
+      "learning_rate": 5.206077759000069e-06,
+      "loss": 0.0477,
+      "num_input_tokens_seen": 69087736,
+      "step": 949
+    },
+    {
+      "epoch": 15.581516095534788,
+      "grad_norm": 0.26595795154571533,
+      "learning_rate": 5.166166492719124e-06,
+      "loss": 0.0533,
+      "num_input_tokens_seen": 69168720,
+      "step": 950
+    },
+    {
+      "epoch": 15.598130841121495,
+      "grad_norm": 0.30843034386634827,
+      "learning_rate": 5.12639116531225e-06,
+      "loss": 0.0589,
+      "num_input_tokens_seen": 69234496,
+      "step": 951
+    },
+    {
+      "epoch": 15.614745586708203,
+      "grad_norm": 0.2536829113960266,
+      "learning_rate": 5.086752049395094e-06,
+      "loss": 0.0476,
+      "num_input_tokens_seen": 69303200,
+      "step": 952
+    },
+    {
+      "epoch": 15.631360332294912,
+      "grad_norm": 0.2525365650653839,
+      "learning_rate": 5.0472494166497135e-06,
+      "loss": 0.0486,
+      "num_input_tokens_seen": 69387968,
+      "step": 953
+    },
+    {
+      "epoch": 15.64797507788162,
+      "grad_norm": 0.21798045933246613,
+      "learning_rate": 5.007883537822736e-06,
+      "loss": 0.0428,
+      "num_input_tokens_seen": 69468496,
+      "step": 954
+    },
+    {
+      "epoch": 15.664589823468328,
+      "grad_norm": 0.25125470757484436,
+      "learning_rate": 4.9686546827234865e-06,
+      "loss": 0.0485,
+      "num_input_tokens_seen": 69545408,
+      "step": 955
+    },
+    {
+      "epoch": 15.681204569055037,
+      "grad_norm": 0.24528349936008453,
+      "learning_rate": 4.929563120222141e-06,
+      "loss": 0.0468,
+      "num_input_tokens_seen": 69604664,
+      "step": 956
+    },
+    {
+      "epoch": 15.697819314641745,
+      "grad_norm": 0.3065119683742523,
+      "learning_rate": 4.890609118247888e-06,
+      "loss": 0.061,
+      "num_input_tokens_seen": 69663456,
+      "step": 957
+    },
+    {
+      "epoch": 15.714434060228452,
+      "grad_norm": 0.32543885707855225,
+      "learning_rate": 4.851792943787109e-06,
+      "loss": 0.0588,
+      "num_input_tokens_seen": 69726392,
+      "step": 958
+    },
+    {
+      "epoch": 15.731048805815162,
+      "grad_norm": 0.2984660565853119,
+      "learning_rate": 4.813114862881502e-06,
+      "loss": 0.0527,
+      "num_input_tokens_seen": 69774792,
+      "step": 959
+    },
+    {
+      "epoch": 15.74766355140187,
+      "grad_norm": 0.21434898674488068,
+      "learning_rate": 4.7745751406263165e-06,
+      "loss": 0.0377,
+      "num_input_tokens_seen": 69864152,
+      "step": 960
+    },
+    {
+      "epoch": 15.764278296988577,
+      "grad_norm": 0.2650623321533203,
+      "learning_rate": 4.7361740411684865e-06,
+      "loss": 0.0498,
+      "num_input_tokens_seen": 69932112,
+      "step": 961
+    },
+    {
+      "epoch": 15.780893042575286,
+      "grad_norm": 0.24544315040111542,
+      "learning_rate": 4.6979118277048426e-06,
+      "loss": 0.0467,
+      "num_input_tokens_seen": 69997512,
+      "step": 962
+    },
+    {
+      "epoch": 15.797507788161994,
+      "grad_norm": 0.2633085250854492,
+      "learning_rate": 4.659788762480327e-06,
+      "loss": 0.047,
+      "num_input_tokens_seen": 70063808,
+      "step": 963
+    },
+    {
+      "epoch": 15.814122533748701,
+      "grad_norm": 0.24816827476024628,
+      "learning_rate": 4.621805106786142e-06,
+      "loss": 0.0486,
+      "num_input_tokens_seen": 70140760,
+      "step": 964
+    },
+    {
+      "epoch": 15.83073727933541,
+      "grad_norm": 0.24450010061264038,
+      "learning_rate": 4.583961120958027e-06,
+      "loss": 0.0468,
+      "num_input_tokens_seen": 70218960,
+      "step": 965
+    },
+    {
+      "epoch": 15.847352024922118,
+      "grad_norm": 0.22891941666603088,
+      "learning_rate": 4.54625706437441e-06,
+      "loss": 0.0467,
+      "num_input_tokens_seen": 70293824,
+      "step": 966
+    },
+    {
+      "epoch": 15.863966770508826,
+      "grad_norm": 0.2394842803478241,
+      "learning_rate": 4.508693195454694e-06,
+      "loss": 0.0515,
+      "num_input_tokens_seen": 70384464,
+      "step": 967
+    },
+    {
+      "epoch": 15.880581516095535,
+      "grad_norm": 0.23781423270702362,
+      "learning_rate": 4.4712697716574e-06,
+      "loss": 0.042,
+      "num_input_tokens_seen": 70455440,
+      "step": 968
+    },
+    {
+      "epoch": 15.897196261682243,
+      "grad_norm": 0.2540624737739563,
+      "learning_rate": 4.433987049478508e-06,
+      "loss": 0.0487,
+      "num_input_tokens_seen": 70513984,
+      "step": 969
+    },
+    {
+      "epoch": 15.91381100726895,
+      "grad_norm": 0.2405335158109665,
+      "learning_rate": 4.396845284449608e-06,
+      "loss": 0.0435,
+      "num_input_tokens_seen": 70588896,
+      "step": 970
+    },
+    {
+      "epoch": 15.93042575285566,
+      "grad_norm": 0.2510586977005005,
+      "learning_rate": 4.359844731136209e-06,
+      "loss": 0.1434,
+      "num_input_tokens_seen": 70673296,
+      "step": 971
+    },
+    {
+      "epoch": 15.947040498442368,
+      "grad_norm": 0.2762887179851532,
+      "learning_rate": 4.322985643135952e-06,
+      "loss": 0.0477,
+      "num_input_tokens_seen": 70742760,
+      "step": 972
+    },
+    {
+      "epoch": 15.963655244029075,
+      "grad_norm": 0.2639402449131012,
+      "learning_rate": 4.286268273076915e-06,
+      "loss": 0.0543,
+      "num_input_tokens_seen": 70806728,
+      "step": 973
+    },
+    {
+      "epoch": 15.980269989615785,
+      "grad_norm": 0.24937745928764343,
+      "learning_rate": 4.2496928726158155e-06,
+      "loss": 0.0496,
+      "num_input_tokens_seen": 70909320,
+      "step": 974
+    },
+    {
+      "epoch": 15.996884735202492,
+      "grad_norm": 0.2835952937602997,
+      "learning_rate": 4.213259692436367e-06,
+      "loss": 0.0566,
+      "num_input_tokens_seen": 70989640,
+      "step": 975
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 0.7339544296264648,
+      "learning_rate": 4.176968982247514e-06,
+      "loss": 0.049,
+      "num_input_tokens_seen": 70997848,
+      "step": 976
+    },
+    {
+      "epoch": 16.016614745586708,
+      "grad_norm": 0.28989917039871216,
+      "learning_rate": 4.140820990781705e-06,
+      "loss": 0.0534,
+      "num_input_tokens_seen": 71051960,
+      "step": 977
+    },
+    {
+      "epoch": 16.033229491173415,
+      "grad_norm": 0.22893474996089935,
+      "learning_rate": 4.104815965793249e-06,
+      "loss": 0.0468,
+      "num_input_tokens_seen": 71132536,
+      "step": 978
+    },
+    {
+      "epoch": 16.049844236760123,
+      "grad_norm": 0.23161743581295013,
+      "learning_rate": 4.068954154056528e-06,
+      "loss": 0.0448,
+      "num_input_tokens_seen": 71202432,
+      "step": 979
+    },
+    {
+      "epoch": 16.066458982346834,
+      "grad_norm": 0.2329898178577423,
+      "learning_rate": 4.0332358013644016e-06,
+      "loss": 0.0415,
+      "num_input_tokens_seen": 71287576,
+      "step": 980
+    },
+    {
+      "epoch": 16.08307372793354,
+      "grad_norm": 0.2544691860675812,
+      "learning_rate": 3.9976611525264525e-06,
+      "loss": 0.1111,
+      "num_input_tokens_seen": 71367248,
+      "step": 981
+    },
+    {
+      "epoch": 16.09968847352025,
+      "grad_norm": 0.20354589819908142,
+      "learning_rate": 3.962230451367349e-06,
+      "loss": 0.0432,
+      "num_input_tokens_seen": 71447928,
+      "step": 982
+    },
+    {
+      "epoch": 16.116303219106957,
+      "grad_norm": 0.28865981101989746,
+      "learning_rate": 3.926943940725137e-06,
+      "loss": 0.0499,
+      "num_input_tokens_seen": 71501472,
+      "step": 983
+    },
+    {
+      "epoch": 16.132917964693664,
+      "grad_norm": 0.30155521631240845,
+      "learning_rate": 3.891801862449629e-06,
+      "loss": 0.0511,
+      "num_input_tokens_seen": 71568048,
+      "step": 984
+    },
+    {
+      "epoch": 16.149532710280372,
+      "grad_norm": 0.21292831003665924,
+      "learning_rate": 3.85680445740067e-06,
+      "loss": 0.0476,
+      "num_input_tokens_seen": 71641112,
+      "step": 985
+    },
+    {
+      "epoch": 16.166147455867083,
+      "grad_norm": 0.22878766059875488,
+      "learning_rate": 3.821951965446577e-06,
+      "loss": 0.0456,
+      "num_input_tokens_seen": 71716800,
+      "step": 986
+    },
+    {
+      "epoch": 16.18276220145379,
+      "grad_norm": 0.18548932671546936,
+      "learning_rate": 3.7872446254624104e-06,
+      "loss": 0.0341,
+      "num_input_tokens_seen": 71799616,
+      "step": 987
+    },
+    {
+      "epoch": 16.1993769470405,
+      "grad_norm": 0.26984840631484985,
+      "learning_rate": 3.752682675328406e-06,
+      "loss": 0.0496,
+      "num_input_tokens_seen": 71863392,
+      "step": 988
+    },
+    {
+      "epoch": 16.215991692627206,
+      "grad_norm": 0.28214019536972046,
+      "learning_rate": 3.718266351928287e-06,
+      "loss": 0.0458,
+      "num_input_tokens_seen": 71927704,
+      "step": 989
+    },
+    {
+      "epoch": 16.232606438213914,
+      "grad_norm": 0.2738666534423828,
+      "learning_rate": 3.6839958911476957e-06,
+      "loss": 0.0473,
+      "num_input_tokens_seen": 71984208,
+      "step": 990
+    },
+    {
+      "epoch": 16.24922118380062,
+      "grad_norm": 0.22236013412475586,
+      "learning_rate": 3.6498715278725254e-06,
+      "loss": 0.0378,
+      "num_input_tokens_seen": 72053448,
+      "step": 991
+    },
+    {
+      "epoch": 16.265835929387332,
+      "grad_norm": 0.2531529664993286,
+      "learning_rate": 3.6158934959873353e-06,
+      "loss": 0.0422,
+      "num_input_tokens_seen": 72123464,
+      "step": 992
+    },
+    {
+      "epoch": 16.28245067497404,
+      "grad_norm": 0.2313385158777237,
+      "learning_rate": 3.5820620283737616e-06,
+      "loss": 0.1494,
+      "num_input_tokens_seen": 72197920,
+      "step": 993
+    },
+    {
+      "epoch": 16.299065420560748,
+      "grad_norm": 0.2416224330663681,
+      "learning_rate": 3.5483773569088856e-06,
+      "loss": 0.044,
+      "num_input_tokens_seen": 72295824,
+      "step": 994
+    },
+    {
+      "epoch": 16.315680166147455,
+      "grad_norm": 0.2035658210515976,
+      "learning_rate": 3.5148397124636826e-06,
+      "loss": 0.0428,
+      "num_input_tokens_seen": 72389296,
+      "step": 995
+    },
+    {
+      "epoch": 16.332294911734163,
+      "grad_norm": 0.29375237226486206,
+      "learning_rate": 3.4814493249014116e-06,
+      "loss": 0.0546,
+      "num_input_tokens_seen": 72451080,
+      "step": 996
+    },
+    {
+      "epoch": 16.34890965732087,
+      "grad_norm": 0.2521064579486847,
+      "learning_rate": 3.4482064230760474e-06,
+      "loss": 0.0457,
+      "num_input_tokens_seen": 72523168,
+      "step": 997
+    },
+    {
+      "epoch": 16.36552440290758,
+      "grad_norm": 0.32425233721733093,
+      "learning_rate": 3.415111234830709e-06,
+      "loss": 0.0573,
+      "num_input_tokens_seen": 72585384,
+      "step": 998
+    },
+    {
+      "epoch": 16.38213914849429,
+      "grad_norm": 0.27826961874961853,
+      "learning_rate": 3.382163986996126e-06,
+      "loss": 0.0484,
+      "num_input_tokens_seen": 72645544,
+      "step": 999
+    },
+    {
+      "epoch": 16.398753894080997,
+      "grad_norm": 0.223323792219162,
+      "learning_rate": 3.3493649053890326e-06,
+      "loss": 0.0412,
+      "num_input_tokens_seen": 72722408,
+      "step": 1000
+    },
+    {
+      "epoch": 16.415368639667705,
+      "grad_norm": 0.28299546241760254,
+      "learning_rate": 3.3167142148106762e-06,
+      "loss": 0.0498,
+      "num_input_tokens_seen": 72798192,
+      "step": 1001
+    },
+    {
+      "epoch": 16.431983385254412,
+      "grad_norm": 0.26504456996917725,
+      "learning_rate": 3.284212139045223e-06,
+      "loss": 0.0492,
+      "num_input_tokens_seen": 72862440,
+      "step": 1002
+    },
+    {
+      "epoch": 16.44859813084112,
+      "grad_norm": 0.2836116552352905,
+      "learning_rate": 3.2518589008582597e-06,
+      "loss": 0.0507,
+      "num_input_tokens_seen": 72911448,
+      "step": 1003
+    },
+    {
+      "epoch": 16.46521287642783,
+      "grad_norm": 0.25678548216819763,
+      "learning_rate": 3.219654721995266e-06,
+      "loss": 0.0501,
+      "num_input_tokens_seen": 72980368,
+      "step": 1004
+    },
+    {
+      "epoch": 16.48182762201454,
+      "grad_norm": 0.29687047004699707,
+      "learning_rate": 3.187599823180071e-06,
+      "loss": 0.0548,
+      "num_input_tokens_seen": 73044888,
+      "step": 1005
+    },
+    {
+      "epoch": 16.498442367601246,
+      "grad_norm": 0.23092421889305115,
+      "learning_rate": 3.15569442411337e-06,
+      "loss": 0.0427,
+      "num_input_tokens_seen": 73119968,
+      "step": 1006
+    },
+    {
+      "epoch": 16.515057113187954,
+      "grad_norm": 0.2751709222793579,
+      "learning_rate": 3.1239387434711916e-06,
+      "loss": 0.0475,
+      "num_input_tokens_seen": 73184104,
+      "step": 1007
+    },
+    {
+      "epoch": 16.53167185877466,
+      "grad_norm": 0.29244229197502136,
+      "learning_rate": 3.092332998903416e-06,
+      "loss": 0.0517,
+      "num_input_tokens_seen": 73239464,
+      "step": 1008
+    },
+    {
+      "epoch": 16.54828660436137,
+      "grad_norm": 0.25361648201942444,
+      "learning_rate": 3.06087740703227e-06,
+      "loss": 0.0548,
+      "num_input_tokens_seen": 73312808,
+      "step": 1009
+    },
+    {
+      "epoch": 16.56490134994808,
+      "grad_norm": 0.23709113895893097,
+      "learning_rate": 3.029572183450868e-06,
+      "loss": 0.1552,
+      "num_input_tokens_seen": 73381528,
+      "step": 1010
+    },
+    {
+      "epoch": 16.581516095534788,
+      "grad_norm": 0.2809208035469055,
+      "learning_rate": 2.9984175427217016e-06,
+      "loss": 0.0461,
+      "num_input_tokens_seen": 73455200,
+      "step": 1011
+    },
+    {
+      "epoch": 16.598130841121495,
+      "grad_norm": 0.28413939476013184,
+      "learning_rate": 2.967413698375196e-06,
+      "loss": 0.0539,
+      "num_input_tokens_seen": 73527448,
+      "step": 1012
+    },
+    {
+      "epoch": 16.614745586708203,
+      "grad_norm": 0.23016762733459473,
+      "learning_rate": 2.936560862908225e-06,
+      "loss": 0.0475,
+      "num_input_tokens_seen": 73615192,
+      "step": 1013
+    },
+    {
+      "epoch": 16.63136033229491,
+      "grad_norm": 0.2616693675518036,
+      "learning_rate": 2.9058592477826636e-06,
+      "loss": 0.0471,
+      "num_input_tokens_seen": 73692976,
+      "step": 1014
+    },
+    {
+      "epoch": 16.64797507788162,
+      "grad_norm": 0.25557368993759155,
+      "learning_rate": 2.875309063423956e-06,
+      "loss": 0.0443,
+      "num_input_tokens_seen": 73753968,
+      "step": 1015
+    },
+    {
+      "epoch": 16.66458982346833,
+      "grad_norm": 0.2636294960975647,
+      "learning_rate": 2.8449105192196316e-06,
+      "loss": 0.0624,
+      "num_input_tokens_seen": 73816936,
+      "step": 1016
+    },
+    {
+      "epoch": 16.681204569055037,
+      "grad_norm": 0.25584691762924194,
+      "learning_rate": 2.8146638235179213e-06,
+      "loss": 0.0516,
+      "num_input_tokens_seen": 73895184,
+      "step": 1017
+    },
+    {
+      "epoch": 16.697819314641745,
+      "grad_norm": 0.23321789503097534,
+      "learning_rate": 2.784569183626276e-06,
+      "loss": 0.0471,
+      "num_input_tokens_seen": 74005280,
+      "step": 1018
+    },
+    {
+      "epoch": 16.714434060228452,
+      "grad_norm": 0.2339867651462555,
+      "learning_rate": 2.7546268058100096e-06,
+      "loss": 0.046,
+      "num_input_tokens_seen": 74078960,
+      "step": 1019
+    },
+    {
+      "epoch": 16.73104880581516,
+      "grad_norm": 0.24815727770328522,
+      "learning_rate": 2.7248368952908053e-06,
+      "loss": 0.0442,
+      "num_input_tokens_seen": 74143592,
+      "step": 1020
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1200,
+  "num_input_tokens_seen": 74143592,
+  "num_train_epochs": 20,
+  "save_steps": 60,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.254877437100818e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}