diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,19524 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 9742,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0010264832683227264,
+      "grad_norm": 1.7093752482690165,
+      "learning_rate": 9.99999584028644e-06,
+      "loss": 0.6885,
+      "num_input_tokens_seen": 288880,
+      "step": 5,
+      "train_runtime": 31.0607,
+      "train_tokens_per_second": 9300.503
+    },
+    {
+      "epoch": 0.0020529665366454527,
+      "grad_norm": 1.140044192832091,
+      "learning_rate": 9.999978941461961e-06,
+      "loss": 0.6061,
+      "num_input_tokens_seen": 594024,
+      "step": 10,
+      "train_runtime": 65.2987,
+      "train_tokens_per_second": 9097.027
+    },
+    {
+      "epoch": 0.003079449804968179,
+      "grad_norm": 1.098428867833084,
+      "learning_rate": 9.999949043588364e-06,
+      "loss": 0.5583,
+      "num_input_tokens_seen": 883616,
+      "step": 15,
+      "train_runtime": 92.9473,
+      "train_tokens_per_second": 9506.634
+    },
+    {
+      "epoch": 0.0041059330732909054,
+      "grad_norm": 0.870157129956912,
+      "learning_rate": 9.999906146743383e-06,
+      "loss": 0.5532,
+      "num_input_tokens_seen": 1140208,
+      "step": 20,
+      "train_runtime": 118.581,
+      "train_tokens_per_second": 9615.435
+    },
+    {
+      "epoch": 0.005132416341613631,
+      "grad_norm": 0.9470635609095528,
+      "learning_rate": 9.999850251038536e-06,
+      "loss": 0.5169,
+      "num_input_tokens_seen": 1395832,
+      "step": 25,
+      "train_runtime": 149.8219,
+      "train_tokens_per_second": 9316.61
+    },
+    {
+      "epoch": 0.006158899609936358,
+      "grad_norm": 0.9013044956495757,
+      "learning_rate": 9.999781356619146e-06,
+      "loss": 0.5023,
+      "num_input_tokens_seen": 1674296,
+      "step": 30,
+      "train_runtime": 181.1625,
+      "train_tokens_per_second": 9241.956
+    },
+    {
+      "epoch": 0.007185382878259084,
+      "grad_norm": 0.9111357572393852,
+      "learning_rate": 9.999699463664327e-06,
+      "loss": 0.4917,
+      "num_input_tokens_seen": 1962496,
+      "step": 35,
+      "train_runtime": 205.3882,
+      "train_tokens_per_second": 9555.056
+    },
+    {
+      "epoch": 0.008211866146581811,
+      "grad_norm": 0.8518491598760526,
+      "learning_rate": 9.999604572386982e-06,
+      "loss": 0.4884,
+      "num_input_tokens_seen": 2264416,
+      "step": 40,
+      "train_runtime": 234.117,
+      "train_tokens_per_second": 9672.156
+    },
+    {
+      "epoch": 0.009238349414904538,
+      "grad_norm": 0.9919497354784894,
+      "learning_rate": 9.999496683033817e-06,
+      "loss": 0.4867,
+      "num_input_tokens_seen": 2549952,
+      "step": 45,
+      "train_runtime": 263.7014,
+      "train_tokens_per_second": 9669.847
+    },
+    {
+      "epoch": 0.010264832683227263,
+      "grad_norm": 1.0099575770745444,
+      "learning_rate": 9.999375795885318e-06,
+      "loss": 0.4769,
+      "num_input_tokens_seen": 2809896,
+      "step": 50,
+      "train_runtime": 297.301,
+      "train_tokens_per_second": 9451.351
+    },
+    {
+      "epoch": 0.01129131595154999,
+      "grad_norm": 0.9263167594265198,
+      "learning_rate": 9.999241911255774e-06,
+      "loss": 0.496,
+      "num_input_tokens_seen": 3114560,
+      "step": 55,
+      "train_runtime": 333.1167,
+      "train_tokens_per_second": 9349.755
+    },
+    {
+      "epoch": 0.012317799219872716,
+      "grad_norm": 1.039487820167108,
+      "learning_rate": 9.99909502949326e-06,
+      "loss": 0.4818,
+      "num_input_tokens_seen": 3383920,
+      "step": 60,
+      "train_runtime": 362.6056,
+      "train_tokens_per_second": 9332.234
+    },
+    {
+      "epoch": 0.013344282488195443,
+      "grad_norm": 0.8347687730273666,
+      "learning_rate": 9.998935150979644e-06,
+      "loss": 0.4822,
+      "num_input_tokens_seen": 3638840,
+      "step": 65,
+      "train_runtime": 389.9966,
+      "train_tokens_per_second": 9330.44
+    },
+    {
+      "epoch": 0.014370765756518168,
+      "grad_norm": 1.0463764499991062,
+      "learning_rate": 9.99876227613058e-06,
+      "loss": 0.4848,
+      "num_input_tokens_seen": 3896512,
+      "step": 70,
+      "train_runtime": 420.3149,
+      "train_tokens_per_second": 9270.458
+    },
+    {
+      "epoch": 0.015397249024840895,
+      "grad_norm": 0.9138533313514547,
+      "learning_rate": 9.998576405395512e-06,
+      "loss": 0.4626,
+      "num_input_tokens_seen": 4179824,
+      "step": 75,
+      "train_runtime": 453.0576,
+      "train_tokens_per_second": 9225.812
+    },
+    {
+      "epoch": 0.016423732293163622,
+      "grad_norm": 1.0993684463890951,
+      "learning_rate": 9.998377539257669e-06,
+      "loss": 0.4428,
+      "num_input_tokens_seen": 4437944,
+      "step": 80,
+      "train_runtime": 479.8491,
+      "train_tokens_per_second": 9248.625
+    },
+    {
+      "epoch": 0.01745021556148635,
+      "grad_norm": 1.0424752877552625,
+      "learning_rate": 9.998165678234071e-06,
+      "loss": 0.4362,
+      "num_input_tokens_seen": 4700480,
+      "step": 85,
+      "train_runtime": 508.331,
+      "train_tokens_per_second": 9246.888
+    },
+    {
+      "epoch": 0.018476698829809075,
+      "grad_norm": 1.0224224365419694,
+      "learning_rate": 9.997940822875514e-06,
+      "loss": 0.4782,
+      "num_input_tokens_seen": 4979472,
+      "step": 90,
+      "train_runtime": 544.1995,
+      "train_tokens_per_second": 9150.085
+    },
+    {
+      "epoch": 0.019503182098131802,
+      "grad_norm": 0.8075645114393942,
+      "learning_rate": 9.997702973766586e-06,
+      "loss": 0.4564,
+      "num_input_tokens_seen": 5274512,
+      "step": 95,
+      "train_runtime": 574.05,
+      "train_tokens_per_second": 9188.245
+    },
+    {
+      "epoch": 0.020529665366454525,
+      "grad_norm": 0.8465301111221318,
+      "learning_rate": 9.997452131525647e-06,
+      "loss": 0.4263,
+      "num_input_tokens_seen": 5547736,
+      "step": 100,
+      "train_runtime": 610.1886,
+      "train_tokens_per_second": 9091.838
+    },
+    {
+      "epoch": 0.021556148634777252,
+      "grad_norm": 0.776841575458339,
+      "learning_rate": 9.997188296804848e-06,
+      "loss": 0.4807,
+      "num_input_tokens_seen": 5824416,
+      "step": 105,
+      "train_runtime": 641.2867,
+      "train_tokens_per_second": 9082.391
+    },
+    {
+      "epoch": 0.02258263190309998,
+      "grad_norm": 0.8220225245529158,
+      "learning_rate": 9.996911470290107e-06,
+      "loss": 0.4666,
+      "num_input_tokens_seen": 6099384,
+      "step": 110,
+      "train_runtime": 669.5557,
+      "train_tokens_per_second": 9109.599
+    },
+    {
+      "epoch": 0.023609115171422706,
+      "grad_norm": 0.803090866887073,
+      "learning_rate": 9.996621652701127e-06,
+      "loss": 0.4539,
+      "num_input_tokens_seen": 6375920,
+      "step": 115,
+      "train_runtime": 703.7587,
+      "train_tokens_per_second": 9059.81
+    },
+    {
+      "epoch": 0.024635598439745433,
+      "grad_norm": 0.886756926269955,
+      "learning_rate": 9.99631884479138e-06,
+      "loss": 0.4803,
+      "num_input_tokens_seen": 6661424,
+      "step": 120,
+      "train_runtime": 734.3139,
+      "train_tokens_per_second": 9071.63
+    },
+    {
+      "epoch": 0.02566208170806816,
+      "grad_norm": 0.8609144168268476,
+      "learning_rate": 9.99600304734811e-06,
+      "loss": 0.4897,
+      "num_input_tokens_seen": 6941384,
+      "step": 125,
+      "train_runtime": 766.2757,
+      "train_tokens_per_second": 9058.598
+    },
+    {
+      "epoch": 0.026688564976390886,
+      "grad_norm": 0.9644263543668842,
+      "learning_rate": 9.995674261192339e-06,
+      "loss": 0.4609,
+      "num_input_tokens_seen": 7215808,
+      "step": 130,
+      "train_runtime": 799.7705,
+      "train_tokens_per_second": 9022.348
+    },
+    {
+      "epoch": 0.02771504824471361,
+      "grad_norm": 0.8299591491756749,
+      "learning_rate": 9.995332487178848e-06,
+      "loss": 0.4521,
+      "num_input_tokens_seen": 7496440,
+      "step": 135,
+      "train_runtime": 834.3146,
+      "train_tokens_per_second": 8985.148
+    },
+    {
+      "epoch": 0.028741531513036336,
+      "grad_norm": 1.002369715042943,
+      "learning_rate": 9.99497772619619e-06,
+      "loss": 0.4183,
+      "num_input_tokens_seen": 7755872,
+      "step": 140,
+      "train_runtime": 863.6696,
+      "train_tokens_per_second": 8980.138
+    },
+    {
+      "epoch": 0.029768014781359063,
+      "grad_norm": 0.7723939918742386,
+      "learning_rate": 9.994609979166677e-06,
+      "loss": 0.411,
+      "num_input_tokens_seen": 8035704,
+      "step": 145,
+      "train_runtime": 890.5168,
+      "train_tokens_per_second": 9023.641
+    },
+    {
+      "epoch": 0.03079449804968179,
+      "grad_norm": 0.8067357125528577,
+      "learning_rate": 9.99422924704639e-06,
+      "loss": 0.4551,
+      "num_input_tokens_seen": 8332528,
+      "step": 150,
+      "train_runtime": 921.5655,
+      "train_tokens_per_second": 9041.71
+    },
+    {
+      "epoch": 0.03182098131800452,
+      "grad_norm": 0.8877951833773499,
+      "learning_rate": 9.99383553082516e-06,
+      "loss": 0.4441,
+      "num_input_tokens_seen": 8600584,
+      "step": 155,
+      "train_runtime": 955.0269,
+      "train_tokens_per_second": 9005.594
+    },
+    {
+      "epoch": 0.032847464586327244,
+      "grad_norm": 0.8609696359321181,
+      "learning_rate": 9.993428831526583e-06,
+      "loss": 0.442,
+      "num_input_tokens_seen": 8874504,
+      "step": 160,
+      "train_runtime": 987.1518,
+      "train_tokens_per_second": 8990.01
+    },
+    {
+      "epoch": 0.03387394785464997,
+      "grad_norm": 1.0124004319938071,
+      "learning_rate": 9.993009150208e-06,
+      "loss": 0.4264,
+      "num_input_tokens_seen": 9174808,
+      "step": 165,
+      "train_runtime": 1022.5472,
+      "train_tokens_per_second": 8972.503
+    },
+    {
+      "epoch": 0.0349004311229727,
+      "grad_norm": 0.9241603684381633,
+      "learning_rate": 9.99257648796051e-06,
+      "loss": 0.4516,
+      "num_input_tokens_seen": 9463536,
+      "step": 170,
+      "train_runtime": 1051.3271,
+      "train_tokens_per_second": 9001.515
+    },
+    {
+      "epoch": 0.035926914391295424,
+      "grad_norm": 0.8876117074199815,
+      "learning_rate": 9.992130845908956e-06,
+      "loss": 0.452,
+      "num_input_tokens_seen": 9737080,
+      "step": 175,
+      "train_runtime": 1084.3615,
+      "train_tokens_per_second": 8979.551
+    },
+    {
+      "epoch": 0.03695339765961815,
+      "grad_norm": 0.884873374009993,
+      "learning_rate": 9.991672225211928e-06,
+      "loss": 0.4833,
+      "num_input_tokens_seen": 10018904,
+      "step": 180,
+      "train_runtime": 1112.1854,
+      "train_tokens_per_second": 9008.304
+    },
+    {
+      "epoch": 0.03797988092794088,
+      "grad_norm": 1.0714561310857418,
+      "learning_rate": 9.991200627061757e-06,
+      "loss": 0.4474,
+      "num_input_tokens_seen": 10299080,
+      "step": 185,
+      "train_runtime": 1145.4655,
+      "train_tokens_per_second": 8991.174
+    },
+    {
+      "epoch": 0.039006364196263604,
+      "grad_norm": 0.807157976161116,
+      "learning_rate": 9.990716052684515e-06,
+      "loss": 0.4477,
+      "num_input_tokens_seen": 10587232,
+      "step": 190,
+      "train_runtime": 1177.8197,
+      "train_tokens_per_second": 8988.839
+    },
+    {
+      "epoch": 0.040032847464586324,
+      "grad_norm": 0.9356509727853985,
+      "learning_rate": 9.990218503340007e-06,
+      "loss": 0.436,
+      "num_input_tokens_seen": 10857048,
+      "step": 195,
+      "train_runtime": 1202.9703,
+      "train_tokens_per_second": 9025.2
+    },
+    {
+      "epoch": 0.04105933073290905,
+      "grad_norm": 0.753940542574059,
+      "learning_rate": 9.989707980321772e-06,
+      "loss": 0.4474,
+      "num_input_tokens_seen": 11149960,
+      "step": 200,
+      "train_runtime": 1225.2813,
+      "train_tokens_per_second": 9099.919
+    },
+    {
+      "epoch": 0.04208581400123178,
+      "grad_norm": 0.8018935929130898,
+      "learning_rate": 9.989184484957081e-06,
+      "loss": 0.4411,
+      "num_input_tokens_seen": 11433584,
+      "step": 205,
+      "train_runtime": 1254.3529,
+      "train_tokens_per_second": 9115.125
+    },
+    {
+      "epoch": 0.043112297269554505,
+      "grad_norm": 0.8585332552421142,
+      "learning_rate": 9.988648018606926e-06,
+      "loss": 0.3998,
+      "num_input_tokens_seen": 11710848,
+      "step": 210,
+      "train_runtime": 1286.1308,
+      "train_tokens_per_second": 9105.487
+    },
+    {
+      "epoch": 0.04413878053787723,
+      "grad_norm": 0.8889894600684649,
+      "learning_rate": 9.988098582666022e-06,
+      "loss": 0.4286,
+      "num_input_tokens_seen": 11977560,
+      "step": 215,
+      "train_runtime": 1313.1455,
+      "train_tokens_per_second": 9121.274
+    },
+    {
+      "epoch": 0.04516526380619996,
+      "grad_norm": 0.788862758060421,
+      "learning_rate": 9.987536178562806e-06,
+      "loss": 0.4306,
+      "num_input_tokens_seen": 12261280,
+      "step": 220,
+      "train_runtime": 1345.688,
+      "train_tokens_per_second": 9111.532
+    },
+    {
+      "epoch": 0.046191747074522685,
+      "grad_norm": 0.9675288090164625,
+      "learning_rate": 9.986960807759428e-06,
+      "loss": 0.4373,
+      "num_input_tokens_seen": 12528344,
+      "step": 225,
+      "train_runtime": 1377.4008,
+      "train_tokens_per_second": 9095.642
+    },
+    {
+      "epoch": 0.04721823034284541,
+      "grad_norm": 0.9077608255052017,
+      "learning_rate": 9.986372471751749e-06,
+      "loss": 0.4402,
+      "num_input_tokens_seen": 12816320,
+      "step": 230,
+      "train_runtime": 1406.8966,
+      "train_tokens_per_second": 9109.639
+    },
+    {
+      "epoch": 0.04824471361116814,
+      "grad_norm": 0.7792004529065182,
+      "learning_rate": 9.985771172069337e-06,
+      "loss": 0.466,
+      "num_input_tokens_seen": 13113376,
+      "step": 235,
+      "train_runtime": 1437.6249,
+      "train_tokens_per_second": 9121.556
+    },
+    {
+      "epoch": 0.049271196879490865,
+      "grad_norm": 0.8305428943576649,
+      "learning_rate": 9.98515691027546e-06,
+      "loss": 0.4307,
+      "num_input_tokens_seen": 13376944,
+      "step": 240,
+      "train_runtime": 1464.3193,
+      "train_tokens_per_second": 9135.264
+    },
+    {
+      "epoch": 0.05029768014781359,
+      "grad_norm": 0.8688733247479157,
+      "learning_rate": 9.984529687967094e-06,
+      "loss": 0.4355,
+      "num_input_tokens_seen": 13664848,
+      "step": 245,
+      "train_runtime": 1499.1362,
+      "train_tokens_per_second": 9115.147
+    },
+    {
+      "epoch": 0.05132416341613632,
+      "grad_norm": 0.9050789743379986,
+      "learning_rate": 9.983889506774902e-06,
+      "loss": 0.4257,
+      "num_input_tokens_seen": 13916624,
+      "step": 250,
+      "train_runtime": 1532.4725,
+      "train_tokens_per_second": 9081.157
+    },
+    {
+      "epoch": 0.052350646684459046,
+      "grad_norm": 0.682459769341108,
+      "learning_rate": 9.98323636836324e-06,
+      "loss": 0.402,
+      "num_input_tokens_seen": 14213848,
+      "step": 255,
+      "train_runtime": 1565.7275,
+      "train_tokens_per_second": 9078.111
+    },
+    {
+      "epoch": 0.05337712995278177,
+      "grad_norm": 0.8178388766923528,
+      "learning_rate": 9.982570274430152e-06,
+      "loss": 0.3991,
+      "num_input_tokens_seen": 14475360,
+      "step": 260,
+      "train_runtime": 1596.5564,
+      "train_tokens_per_second": 9066.614
+    },
+    {
+      "epoch": 0.0544036132211045,
+      "grad_norm": 1.3056261036505794,
+      "learning_rate": 9.98189122670736e-06,
+      "loss": 0.4315,
+      "num_input_tokens_seen": 14765488,
+      "step": 265,
+      "train_runtime": 1629.0985,
+      "train_tokens_per_second": 9063.595
+    },
+    {
+      "epoch": 0.05543009648942722,
+      "grad_norm": 0.8251242242215888,
+      "learning_rate": 9.981199226960272e-06,
+      "loss": 0.4272,
+      "num_input_tokens_seen": 15035456,
+      "step": 270,
+      "train_runtime": 1657.1371,
+      "train_tokens_per_second": 9073.151
+    },
+    {
+      "epoch": 0.056456579757749946,
+      "grad_norm": 0.956870601187203,
+      "learning_rate": 9.980494276987957e-06,
+      "loss": 0.4489,
+      "num_input_tokens_seen": 15304240,
+      "step": 275,
+      "train_runtime": 1687.2437,
+      "train_tokens_per_second": 9070.557
+    },
+    {
+      "epoch": 0.05748306302607267,
+      "grad_norm": 0.9076638961219162,
+      "learning_rate": 9.979776378623163e-06,
+      "loss": 0.4397,
+      "num_input_tokens_seen": 15587400,
+      "step": 280,
+      "train_runtime": 1712.0718,
+      "train_tokens_per_second": 9104.408
+    },
+    {
+      "epoch": 0.0585095462943954,
+      "grad_norm": 0.7219805059717627,
+      "learning_rate": 9.979045533732294e-06,
+      "loss": 0.4099,
+      "num_input_tokens_seen": 15866456,
+      "step": 285,
+      "train_runtime": 1741.1174,
+      "train_tokens_per_second": 9112.801
+    },
+    {
+      "epoch": 0.059536029562718126,
+      "grad_norm": 0.8849031210741065,
+      "learning_rate": 9.978301744215418e-06,
+      "loss": 0.4569,
+      "num_input_tokens_seen": 16135824,
+      "step": 290,
+      "train_runtime": 1768.5055,
+      "train_tokens_per_second": 9123.988
+    },
+    {
+      "epoch": 0.06056251283104085,
+      "grad_norm": 0.8429681725465481,
+      "learning_rate": 9.977545012006254e-06,
+      "loss": 0.4223,
+      "num_input_tokens_seen": 16425752,
+      "step": 295,
+      "train_runtime": 1807.3891,
+      "train_tokens_per_second": 9088.111
+    },
+    {
+      "epoch": 0.06158899609936358,
+      "grad_norm": 0.8758981175371494,
+      "learning_rate": 9.97677533907217e-06,
+      "loss": 0.4077,
+      "num_input_tokens_seen": 16690656,
+      "step": 300,
+      "train_runtime": 1838.1256,
+      "train_tokens_per_second": 9080.259
+    },
+    {
+      "epoch": 0.06261547936768631,
+      "grad_norm": 0.9765039298560739,
+      "learning_rate": 9.975992727414177e-06,
+      "loss": 0.4447,
+      "num_input_tokens_seen": 16950272,
+      "step": 305,
+      "train_runtime": 1868.8159,
+      "train_tokens_per_second": 9070.06
+    },
+    {
+      "epoch": 0.06364196263600903,
+      "grad_norm": 0.7758495288578562,
+      "learning_rate": 9.975197179066925e-06,
+      "loss": 0.4192,
+      "num_input_tokens_seen": 17222032,
+      "step": 310,
+      "train_runtime": 1898.1063,
+      "train_tokens_per_second": 9073.27
+    },
+    {
+      "epoch": 0.06466844590433175,
+      "grad_norm": 0.8097501620622239,
+      "learning_rate": 9.974388696098699e-06,
+      "loss": 0.43,
+      "num_input_tokens_seen": 17516576,
+      "step": 315,
+      "train_runtime": 1926.6219,
+      "train_tokens_per_second": 9091.86
+    },
+    {
+      "epoch": 0.06569492917265449,
+      "grad_norm": 0.7307111178314502,
+      "learning_rate": 9.973567280611407e-06,
+      "loss": 0.4351,
+      "num_input_tokens_seen": 17798968,
+      "step": 320,
+      "train_runtime": 1961.226,
+      "train_tokens_per_second": 9075.429
+    },
+    {
+      "epoch": 0.0667214124409772,
+      "grad_norm": 0.8936759621526961,
+      "learning_rate": 9.972732934740586e-06,
+      "loss": 0.4248,
+      "num_input_tokens_seen": 18068920,
+      "step": 325,
+      "train_runtime": 1986.894,
+      "train_tokens_per_second": 9094.053
+    },
+    {
+      "epoch": 0.06774789570929994,
+      "grad_norm": 0.8589147987401641,
+      "learning_rate": 9.971885660655381e-06,
+      "loss": 0.423,
+      "num_input_tokens_seen": 18331136,
+      "step": 330,
+      "train_runtime": 2010.4936,
+      "train_tokens_per_second": 9117.729
+    },
+    {
+      "epoch": 0.06877437897762266,
+      "grad_norm": 0.8917404954492028,
+      "learning_rate": 9.971025460558556e-06,
+      "loss": 0.4236,
+      "num_input_tokens_seen": 18595112,
+      "step": 335,
+      "train_runtime": 2042.6408,
+      "train_tokens_per_second": 9103.466
+    },
+    {
+      "epoch": 0.0698008622459454,
+      "grad_norm": 0.8859170158353316,
+      "learning_rate": 9.970152336686476e-06,
+      "loss": 0.4037,
+      "num_input_tokens_seen": 18875496,
+      "step": 340,
+      "train_runtime": 2069.5262,
+      "train_tokens_per_second": 9120.685
+    },
+    {
+      "epoch": 0.07082734551426811,
+      "grad_norm": 0.7972717428427775,
+      "learning_rate": 9.969266291309105e-06,
+      "loss": 0.4363,
+      "num_input_tokens_seen": 19154760,
+      "step": 345,
+      "train_runtime": 2105.0862,
+      "train_tokens_per_second": 9099.276
+    },
+    {
+      "epoch": 0.07185382878259085,
+      "grad_norm": 0.7751956045409947,
+      "learning_rate": 9.968367326730006e-06,
+      "loss": 0.411,
+      "num_input_tokens_seen": 19420800,
+      "step": 350,
+      "train_runtime": 2135.4315,
+      "train_tokens_per_second": 9094.555
+    },
+    {
+      "epoch": 0.07288031205091357,
+      "grad_norm": 0.8629942470015681,
+      "learning_rate": 9.967455445286322e-06,
+      "loss": 0.4022,
+      "num_input_tokens_seen": 19717864,
+      "step": 355,
+      "train_runtime": 2169.9768,
+      "train_tokens_per_second": 9086.67
+    },
+    {
+      "epoch": 0.0739067953192363,
+      "grad_norm": 0.788580913592726,
+      "learning_rate": 9.966530649348785e-06,
+      "loss": 0.4041,
+      "num_input_tokens_seen": 19970288,
+      "step": 360,
+      "train_runtime": 2201.0647,
+      "train_tokens_per_second": 9073.013
+    },
+    {
+      "epoch": 0.07493327858755902,
+      "grad_norm": 0.7698116817872513,
+      "learning_rate": 9.965592941321695e-06,
+      "loss": 0.4342,
+      "num_input_tokens_seen": 20240480,
+      "step": 365,
+      "train_runtime": 2231.5127,
+      "train_tokens_per_second": 9070.296
+    },
+    {
+      "epoch": 0.07595976185588175,
+      "grad_norm": 1.5643799711689113,
+      "learning_rate": 9.964642323642927e-06,
+      "loss": 0.4351,
+      "num_input_tokens_seen": 20538768,
+      "step": 370,
+      "train_runtime": 2260.1015,
+      "train_tokens_per_second": 9087.542
+    },
+    {
+      "epoch": 0.07698624512420447,
+      "grad_norm": 0.7497406074876252,
+      "learning_rate": 9.963678798783915e-06,
+      "loss": 0.4334,
+      "num_input_tokens_seen": 20829744,
+      "step": 375,
+      "train_runtime": 2294.8437,
+      "train_tokens_per_second": 9076.759
+    },
+    {
+      "epoch": 0.07801272839252721,
+      "grad_norm": 0.9607863961521076,
+      "learning_rate": 9.962702369249654e-06,
+      "loss": 0.4069,
+      "num_input_tokens_seen": 21121776,
+      "step": 380,
+      "train_runtime": 2330.2637,
+      "train_tokens_per_second": 9064.114
+    },
+    {
+      "epoch": 0.07903921166084993,
+      "grad_norm": 0.9086486828770094,
+      "learning_rate": 9.961713037578681e-06,
+      "loss": 0.3852,
+      "num_input_tokens_seen": 21391480,
+      "step": 385,
+      "train_runtime": 2366.8768,
+      "train_tokens_per_second": 9037.851
+    },
+    {
+      "epoch": 0.08006569492917265,
+      "grad_norm": 0.8304539530306914,
+      "learning_rate": 9.960710806343087e-06,
+      "loss": 0.4256,
+      "num_input_tokens_seen": 21646720,
+      "step": 390,
+      "train_runtime": 2398.2468,
+      "train_tokens_per_second": 9026.06
+    },
+    {
+      "epoch": 0.08109217819749538,
+      "grad_norm": 0.8551558434742199,
+      "learning_rate": 9.95969567814849e-06,
+      "loss": 0.4055,
+      "num_input_tokens_seen": 21941880,
+      "step": 395,
+      "train_runtime": 2425.6434,
+      "train_tokens_per_second": 9045.798
+    },
+    {
+      "epoch": 0.0821186614658181,
+      "grad_norm": 0.8550407384372883,
+      "learning_rate": 9.95866765563404e-06,
+      "loss": 0.4025,
+      "num_input_tokens_seen": 22240600,
+      "step": 400,
+      "train_runtime": 2459.1312,
+      "train_tokens_per_second": 9044.088
+    },
+    {
+      "epoch": 0.08314514473414084,
+      "grad_norm": 0.8673296175902743,
+      "learning_rate": 9.957626741472412e-06,
+      "loss": 0.4349,
+      "num_input_tokens_seen": 22502152,
+      "step": 405,
+      "train_runtime": 2490.6198,
+      "train_tokens_per_second": 9034.76
+    },
+    {
+      "epoch": 0.08417162800246356,
+      "grad_norm": 0.8403672782992784,
+      "learning_rate": 9.9565729383698e-06,
+      "loss": 0.3863,
+      "num_input_tokens_seen": 22775944,
+      "step": 410,
+      "train_runtime": 2524.9001,
+      "train_tokens_per_second": 9020.533
+    },
+    {
+      "epoch": 0.08519811127078629,
+      "grad_norm": 0.8341234279331601,
+      "learning_rate": 9.955506249065899e-06,
+      "loss": 0.4178,
+      "num_input_tokens_seen": 23078536,
+      "step": 415,
+      "train_runtime": 2555.7889,
+      "train_tokens_per_second": 9029.907
+    },
+    {
+      "epoch": 0.08622459453910901,
+      "grad_norm": 1.023106779298129,
+      "learning_rate": 9.954426676333913e-06,
+      "loss": 0.4287,
+      "num_input_tokens_seen": 23370864,
+      "step": 420,
+      "train_runtime": 2586.4737,
+      "train_tokens_per_second": 9035.802
+    },
+    {
+      "epoch": 0.08725107780743174,
+      "grad_norm": 0.7249186412405721,
+      "learning_rate": 9.953334222980534e-06,
+      "loss": 0.3814,
+      "num_input_tokens_seen": 23627032,
+      "step": 425,
+      "train_runtime": 2615.7423,
+      "train_tokens_per_second": 9032.63
+    },
+    {
+      "epoch": 0.08827756107575446,
+      "grad_norm": 0.8819813429563228,
+      "learning_rate": 9.952228891845948e-06,
+      "loss": 0.4337,
+      "num_input_tokens_seen": 23895208,
+      "step": 430,
+      "train_runtime": 2644.6516,
+      "train_tokens_per_second": 9035.295
+    },
+    {
+      "epoch": 0.0893040443440772,
+      "grad_norm": 0.7683347690302661,
+      "learning_rate": 9.951110685803818e-06,
+      "loss": 0.3777,
+      "num_input_tokens_seen": 24173656,
+      "step": 435,
+      "train_runtime": 2674.2574,
+      "train_tokens_per_second": 9039.39
+    },
+    {
+      "epoch": 0.09033052761239992,
+      "grad_norm": 0.9096722254579354,
+      "learning_rate": 9.949979607761277e-06,
+      "loss": 0.429,
+      "num_input_tokens_seen": 24436936,
+      "step": 440,
+      "train_runtime": 2711.2228,
+      "train_tokens_per_second": 9013.253
+    },
+    {
+      "epoch": 0.09135701088072265,
+      "grad_norm": 0.699143379328376,
+      "learning_rate": 9.948835660658927e-06,
+      "loss": 0.4235,
+      "num_input_tokens_seen": 24703048,
+      "step": 445,
+      "train_runtime": 2742.9066,
+      "train_tokens_per_second": 9006.157
+    },
+    {
+      "epoch": 0.09238349414904537,
+      "grad_norm": 1.001133044256508,
+      "learning_rate": 9.947678847470824e-06,
+      "loss": 0.3906,
+      "num_input_tokens_seen": 24989080,
+      "step": 450,
+      "train_runtime": 2769.6753,
+      "train_tokens_per_second": 9022.386
+    },
+    {
+      "epoch": 0.0934099774173681,
+      "grad_norm": 0.8969947217777525,
+      "learning_rate": 9.946509171204474e-06,
+      "loss": 0.4507,
+      "num_input_tokens_seen": 25255168,
+      "step": 455,
+      "train_runtime": 2798.0101,
+      "train_tokens_per_second": 9026.117
+    },
+    {
+      "epoch": 0.09443646068569082,
+      "grad_norm": 0.7803528354856185,
+      "learning_rate": 9.94532663490083e-06,
+      "loss": 0.4001,
+      "num_input_tokens_seen": 25551208,
+      "step": 460,
+      "train_runtime": 2830.1231,
+      "train_tokens_per_second": 9028.303
+    },
+    {
+      "epoch": 0.09546294395401354,
+      "grad_norm": 0.7990471170349084,
+      "learning_rate": 9.944131241634271e-06,
+      "loss": 0.4023,
+      "num_input_tokens_seen": 25806800,
+      "step": 465,
+      "train_runtime": 2853.4162,
+      "train_tokens_per_second": 9044.176
+    },
+    {
+      "epoch": 0.09648942722233628,
+      "grad_norm": 0.6492898537418841,
+      "learning_rate": 9.942922994512608e-06,
+      "loss": 0.3859,
+      "num_input_tokens_seen": 26086312,
+      "step": 470,
+      "train_runtime": 2885.1601,
+      "train_tokens_per_second": 9041.547
+    },
+    {
+      "epoch": 0.097515910490659,
+      "grad_norm": 0.9446880449459889,
+      "learning_rate": 9.941701896677063e-06,
+      "loss": 0.3876,
+      "num_input_tokens_seen": 26359192,
+      "step": 475,
+      "train_runtime": 2917.9075,
+      "train_tokens_per_second": 9033.594
+    },
+    {
+      "epoch": 0.09854239375898173,
+      "grad_norm": 0.7877115378096643,
+      "learning_rate": 9.940467951302275e-06,
+      "loss": 0.3967,
+      "num_input_tokens_seen": 26611464,
+      "step": 480,
+      "train_runtime": 2944.865,
+      "train_tokens_per_second": 9036.565
+    },
+    {
+      "epoch": 0.09956887702730445,
+      "grad_norm": 0.7305896917056838,
+      "learning_rate": 9.939221161596279e-06,
+      "loss": 0.4038,
+      "num_input_tokens_seen": 26887208,
+      "step": 485,
+      "train_runtime": 2968.9688,
+      "train_tokens_per_second": 9056.076
+    },
+    {
+      "epoch": 0.10059536029562718,
+      "grad_norm": 0.918455805865829,
+      "learning_rate": 9.937961530800503e-06,
+      "loss": 0.4222,
+      "num_input_tokens_seen": 27151856,
+      "step": 490,
+      "train_runtime": 2997.3075,
+      "train_tokens_per_second": 9058.749
+    },
+    {
+      "epoch": 0.1016218435639499,
+      "grad_norm": 0.7035912232509952,
+      "learning_rate": 9.936689062189766e-06,
+      "loss": 0.4178,
+      "num_input_tokens_seen": 27432352,
+      "step": 495,
+      "train_runtime": 3029.6388,
+      "train_tokens_per_second": 9054.661
+    },
+    {
+      "epoch": 0.10264832683227264,
+      "grad_norm": 0.726145188915723,
+      "learning_rate": 9.935403759072255e-06,
+      "loss": 0.4136,
+      "num_input_tokens_seen": 27689784,
+      "step": 500,
+      "train_runtime": 3058.5617,
+      "train_tokens_per_second": 9053.204
+    },
+    {
+      "epoch": 0.10367481010059536,
+      "grad_norm": 0.9808260401089935,
+      "learning_rate": 9.934105624789529e-06,
+      "loss": 0.4181,
+      "num_input_tokens_seen": 27989320,
+      "step": 505,
+      "train_runtime": 3093.2973,
+      "train_tokens_per_second": 9048.377
+    },
+    {
+      "epoch": 0.10470129336891809,
+      "grad_norm": 0.9024630958289642,
+      "learning_rate": 9.932794662716506e-06,
+      "loss": 0.3975,
+      "num_input_tokens_seen": 28267480,
+      "step": 510,
+      "train_runtime": 3125.3978,
+      "train_tokens_per_second": 9044.442
+    },
+    {
+      "epoch": 0.10572777663724081,
+      "grad_norm": 0.7773816594911963,
+      "learning_rate": 9.931470876261446e-06,
+      "loss": 0.3981,
+      "num_input_tokens_seen": 28550488,
+      "step": 515,
+      "train_runtime": 3152.6089,
+      "train_tokens_per_second": 9056.146
+    },
+    {
+      "epoch": 0.10675425990556354,
+      "grad_norm": 0.7358142375641107,
+      "learning_rate": 9.930134268865964e-06,
+      "loss": 0.3935,
+      "num_input_tokens_seen": 28824160,
+      "step": 520,
+      "train_runtime": 3179.0552,
+      "train_tokens_per_second": 9066.895
+    },
+    {
+      "epoch": 0.10778074317388626,
+      "grad_norm": 0.7817110514397422,
+      "learning_rate": 9.928784844004996e-06,
+      "loss": 0.4142,
+      "num_input_tokens_seen": 29108080,
+      "step": 525,
+      "train_runtime": 3207.1625,
+      "train_tokens_per_second": 9075.96
+    },
+    {
+      "epoch": 0.108807226442209,
+      "grad_norm": 0.800665159373056,
+      "learning_rate": 9.927422605186805e-06,
+      "loss": 0.4082,
+      "num_input_tokens_seen": 29374424,
+      "step": 530,
+      "train_runtime": 3236.5815,
+      "train_tokens_per_second": 9075.756
+    },
+    {
+      "epoch": 0.10983370971053172,
+      "grad_norm": 0.8464047799045803,
+      "learning_rate": 9.92604755595297e-06,
+      "loss": 0.4339,
+      "num_input_tokens_seen": 29643072,
+      "step": 535,
+      "train_runtime": 3263.4396,
+      "train_tokens_per_second": 9083.383
+    },
+    {
+      "epoch": 0.11086019297885444,
+      "grad_norm": 0.8733334150843193,
+      "learning_rate": 9.92465969987837e-06,
+      "loss": 0.4168,
+      "num_input_tokens_seen": 29918824,
+      "step": 540,
+      "train_runtime": 3291.1583,
+      "train_tokens_per_second": 9090.667
+    },
+    {
+      "epoch": 0.11188667624717717,
+      "grad_norm": 0.7772255500945142,
+      "learning_rate": 9.92325904057118e-06,
+      "loss": 0.3992,
+      "num_input_tokens_seen": 30199936,
+      "step": 545,
+      "train_runtime": 3317.0403,
+      "train_tokens_per_second": 9104.483
+    },
+    {
+      "epoch": 0.11291315951549989,
+      "grad_norm": 0.7743680316095446,
+      "learning_rate": 9.921845581672868e-06,
+      "loss": 0.4203,
+      "num_input_tokens_seen": 30483936,
+      "step": 550,
+      "train_runtime": 3353.3823,
+      "train_tokens_per_second": 9090.504
+    },
+    {
+      "epoch": 0.11393964278382263,
+      "grad_norm": 0.8553401620109244,
+      "learning_rate": 9.920419326858174e-06,
+      "loss": 0.4241,
+      "num_input_tokens_seen": 30749936,
+      "step": 555,
+      "train_runtime": 3386.8123,
+      "train_tokens_per_second": 9079.315
+    },
+    {
+      "epoch": 0.11496612605214535,
+      "grad_norm": 0.7982543579755981,
+      "learning_rate": 9.9189802798351e-06,
+      "loss": 0.41,
+      "num_input_tokens_seen": 31014472,
+      "step": 560,
+      "train_runtime": 3420.4368,
+      "train_tokens_per_second": 9067.401
+    },
+    {
+      "epoch": 0.11599260932046808,
+      "grad_norm": 0.7896000719681316,
+      "learning_rate": 9.917528444344915e-06,
+      "loss": 0.4018,
+      "num_input_tokens_seen": 31279368,
+      "step": 565,
+      "train_runtime": 3451.3558,
+      "train_tokens_per_second": 9062.922
+    },
+    {
+      "epoch": 0.1170190925887908,
+      "grad_norm": 0.7273168484517984,
+      "learning_rate": 9.916063824162128e-06,
+      "loss": 0.4069,
+      "num_input_tokens_seen": 31571872,
+      "step": 570,
+      "train_runtime": 3484.4347,
+      "train_tokens_per_second": 9060.83
+    },
+    {
+      "epoch": 0.11804557585711353,
+      "grad_norm": 0.75831208821697,
+      "learning_rate": 9.91458642309449e-06,
+      "loss": 0.4259,
+      "num_input_tokens_seen": 31841432,
+      "step": 575,
+      "train_runtime": 3518.1982,
+      "train_tokens_per_second": 9050.494
+    },
+    {
+      "epoch": 0.11907205912543625,
+      "grad_norm": 0.7393015504713972,
+      "learning_rate": 9.91309624498298e-06,
+      "loss": 0.4027,
+      "num_input_tokens_seen": 32123264,
+      "step": 580,
+      "train_runtime": 3550.369,
+      "train_tokens_per_second": 9047.866
+    },
+    {
+      "epoch": 0.12009854239375899,
+      "grad_norm": 0.7271366965696996,
+      "learning_rate": 9.91159329370179e-06,
+      "loss": 0.4196,
+      "num_input_tokens_seen": 32384448,
+      "step": 585,
+      "train_runtime": 3576.0792,
+      "train_tokens_per_second": 9055.853
+    },
+    {
+      "epoch": 0.1211250256620817,
+      "grad_norm": 0.7883028504117854,
+      "learning_rate": 9.910077573158331e-06,
+      "loss": 0.4162,
+      "num_input_tokens_seen": 32668032,
+      "step": 590,
+      "train_runtime": 3606.0431,
+      "train_tokens_per_second": 9059.246
+    },
+    {
+      "epoch": 0.12215150893040444,
+      "grad_norm": 0.8959463101469392,
+      "learning_rate": 9.9085490872932e-06,
+      "loss": 0.4161,
+      "num_input_tokens_seen": 32948160,
+      "step": 595,
+      "train_runtime": 3635.2186,
+      "train_tokens_per_second": 9063.598
+    },
+    {
+      "epoch": 0.12317799219872716,
+      "grad_norm": 0.7117335650767859,
+      "learning_rate": 9.907007840080185e-06,
+      "loss": 0.4063,
+      "num_input_tokens_seen": 33227592,
+      "step": 600,
+      "train_runtime": 3661.0108,
+      "train_tokens_per_second": 9076.07
+    },
+    {
+      "epoch": 0.1242044754670499,
+      "grad_norm": 0.8600828061732327,
+      "learning_rate": 9.905453835526257e-06,
+      "loss": 0.3885,
+      "num_input_tokens_seen": 33512568,
+      "step": 605,
+      "train_runtime": 3690.0434,
+      "train_tokens_per_second": 9081.89
+    },
+    {
+      "epoch": 0.12523095873537263,
+      "grad_norm": 0.8471122092861801,
+      "learning_rate": 9.903887077671545e-06,
+      "loss": 0.4029,
+      "num_input_tokens_seen": 33796112,
+      "step": 610,
+      "train_runtime": 3722.7358,
+      "train_tokens_per_second": 9078.3
+    },
+    {
+      "epoch": 0.12625744200369535,
+      "grad_norm": 0.8090676648308716,
+      "learning_rate": 9.902307570589343e-06,
+      "loss": 0.4292,
+      "num_input_tokens_seen": 34092904,
+      "step": 615,
+      "train_runtime": 3753.9036,
+      "train_tokens_per_second": 9081.987
+    },
+    {
+      "epoch": 0.12728392527201807,
+      "grad_norm": 0.8533413175902423,
+      "learning_rate": 9.90071531838608e-06,
+      "loss": 0.3906,
+      "num_input_tokens_seen": 34358664,
+      "step": 620,
+      "train_runtime": 3791.3814,
+      "train_tokens_per_second": 9062.307
+    },
+    {
+      "epoch": 0.1283104085403408,
+      "grad_norm": 0.8356664351117207,
+      "learning_rate": 9.899110325201335e-06,
+      "loss": 0.4295,
+      "num_input_tokens_seen": 34654488,
+      "step": 625,
+      "train_runtime": 3825.7878,
+      "train_tokens_per_second": 9058.131
+    },
+    {
+      "epoch": 0.1293368918086635,
+      "grad_norm": 0.8034465102750565,
+      "learning_rate": 9.897492595207795e-06,
+      "loss": 0.4312,
+      "num_input_tokens_seen": 34915200,
+      "step": 630,
+      "train_runtime": 3853.6553,
+      "train_tokens_per_second": 9060.281
+    },
+    {
+      "epoch": 0.13036337507698625,
+      "grad_norm": 0.8701516666203803,
+      "learning_rate": 9.895862132611272e-06,
+      "loss": 0.3987,
+      "num_input_tokens_seen": 35174168,
+      "step": 635,
+      "train_runtime": 3884.6038,
+      "train_tokens_per_second": 9054.763
+    },
+    {
+      "epoch": 0.13138985834530897,
+      "grad_norm": 0.6687006844635585,
+      "learning_rate": 9.894218941650675e-06,
+      "loss": 0.3868,
+      "num_input_tokens_seen": 35468232,
+      "step": 640,
+      "train_runtime": 3917.5209,
+      "train_tokens_per_second": 9053.744
+    },
+    {
+      "epoch": 0.1324163416136317,
+      "grad_norm": 0.7155154916982882,
+      "learning_rate": 9.892563026598005e-06,
+      "loss": 0.4125,
+      "num_input_tokens_seen": 35754968,
+      "step": 645,
+      "train_runtime": 3948.506,
+      "train_tokens_per_second": 9055.316
+    },
+    {
+      "epoch": 0.1334428248819544,
+      "grad_norm": 0.8185685908046004,
+      "learning_rate": 9.890894391758348e-06,
+      "loss": 0.4038,
+      "num_input_tokens_seen": 36018760,
+      "step": 650,
+      "train_runtime": 3979.7277,
+      "train_tokens_per_second": 9050.559
+    },
+    {
+      "epoch": 0.13446930815027716,
+      "grad_norm": 0.7641533764651858,
+      "learning_rate": 9.889213041469851e-06,
+      "loss": 0.4156,
+      "num_input_tokens_seen": 36304872,
+      "step": 655,
+      "train_runtime": 4016.1803,
+      "train_tokens_per_second": 9039.652
+    },
+    {
+      "epoch": 0.13549579141859988,
+      "grad_norm": 0.774803054395058,
+      "learning_rate": 9.887518980103728e-06,
+      "loss": 0.4439,
+      "num_input_tokens_seen": 36586512,
+      "step": 660,
+      "train_runtime": 4045.8897,
+      "train_tokens_per_second": 9042.884
+    },
+    {
+      "epoch": 0.1365222746869226,
+      "grad_norm": 0.7401160539857111,
+      "learning_rate": 9.885812212064231e-06,
+      "loss": 0.4078,
+      "num_input_tokens_seen": 36860904,
+      "step": 665,
+      "train_runtime": 4070.6824,
+      "train_tokens_per_second": 9055.215
+    },
+    {
+      "epoch": 0.13754875795524532,
+      "grad_norm": 0.8024385212485208,
+      "learning_rate": 9.884092741788655e-06,
+      "loss": 0.4144,
+      "num_input_tokens_seen": 37136944,
+      "step": 670,
+      "train_runtime": 4099.805,
+      "train_tokens_per_second": 9058.222
+    },
+    {
+      "epoch": 0.13857524122356807,
+      "grad_norm": 0.8348166575669559,
+      "learning_rate": 9.88236057374731e-06,
+      "loss": 0.3993,
+      "num_input_tokens_seen": 37407120,
+      "step": 675,
+      "train_runtime": 4126.2391,
+      "train_tokens_per_second": 9065.67
+    },
+    {
+      "epoch": 0.1396017244918908,
+      "grad_norm": 0.8943884825192004,
+      "learning_rate": 9.880615712443526e-06,
+      "loss": 0.4006,
+      "num_input_tokens_seen": 37677512,
+      "step": 680,
+      "train_runtime": 4157.7534,
+      "train_tokens_per_second": 9061.988
+    },
+    {
+      "epoch": 0.1406282077602135,
+      "grad_norm": 0.8949007453586688,
+      "learning_rate": 9.878858162413627e-06,
+      "loss": 0.4039,
+      "num_input_tokens_seen": 37930968,
+      "step": 685,
+      "train_runtime": 4188.8457,
+      "train_tokens_per_second": 9055.232
+    },
+    {
+      "epoch": 0.14165469102853623,
+      "grad_norm": 0.9512519304437591,
+      "learning_rate": 9.87708792822693e-06,
+      "loss": 0.441,
+      "num_input_tokens_seen": 38239280,
+      "step": 690,
+      "train_runtime": 4221.205,
+      "train_tokens_per_second": 9058.854
+    },
+    {
+      "epoch": 0.14268117429685895,
+      "grad_norm": 0.7574463389297761,
+      "learning_rate": 9.875305014485726e-06,
+      "loss": 0.3925,
+      "num_input_tokens_seen": 38528968,
+      "step": 695,
+      "train_runtime": 4246.3244,
+      "train_tokens_per_second": 9073.487
+    },
+    {
+      "epoch": 0.1437076575651817,
+      "grad_norm": 0.794754946614291,
+      "learning_rate": 9.873509425825271e-06,
+      "loss": 0.4303,
+      "num_input_tokens_seen": 38814808,
+      "step": 700,
+      "train_runtime": 4272.7859,
+      "train_tokens_per_second": 9084.192
+    },
+    {
+      "epoch": 0.14473414083350442,
+      "grad_norm": 0.7794034930035713,
+      "learning_rate": 9.871701166913775e-06,
+      "loss": 0.4035,
+      "num_input_tokens_seen": 39078904,
+      "step": 705,
+      "train_runtime": 4302.8853,
+      "train_tokens_per_second": 9082.023
+    },
+    {
+      "epoch": 0.14576062410182714,
+      "grad_norm": 0.8224562803844636,
+      "learning_rate": 9.869880242452385e-06,
+      "loss": 0.392,
+      "num_input_tokens_seen": 39381704,
+      "step": 710,
+      "train_runtime": 4335.245,
+      "train_tokens_per_second": 9084.078
+    },
+    {
+      "epoch": 0.14678710737014986,
+      "grad_norm": 0.7123813412774904,
+      "learning_rate": 9.86804665717518e-06,
+      "loss": 0.4275,
+      "num_input_tokens_seen": 39659264,
+      "step": 715,
+      "train_runtime": 4361.6892,
+      "train_tokens_per_second": 9092.639
+    },
+    {
+      "epoch": 0.1478135906384726,
+      "grad_norm": 0.8543662997434422,
+      "learning_rate": 9.86620041584915e-06,
+      "loss": 0.4388,
+      "num_input_tokens_seen": 39913000,
+      "step": 720,
+      "train_runtime": 4387.3283,
+      "train_tokens_per_second": 9097.336
+    },
+    {
+      "epoch": 0.14884007390679532,
+      "grad_norm": 0.8299599597764751,
+      "learning_rate": 9.864341523274197e-06,
+      "loss": 0.3994,
+      "num_input_tokens_seen": 40162128,
+      "step": 725,
+      "train_runtime": 4418.1846,
+      "train_tokens_per_second": 9090.188
+    },
+    {
+      "epoch": 0.14986655717511804,
+      "grad_norm": 0.8012105199308094,
+      "learning_rate": 9.862469984283103e-06,
+      "loss": 0.4216,
+      "num_input_tokens_seen": 40438584,
+      "step": 730,
+      "train_runtime": 4448.9187,
+      "train_tokens_per_second": 9089.531
+    },
+    {
+      "epoch": 0.15089304044344076,
+      "grad_norm": 0.9236589763979595,
+      "learning_rate": 9.860585803741536e-06,
+      "loss": 0.4096,
+      "num_input_tokens_seen": 40706728,
+      "step": 735,
+      "train_runtime": 4475.9638,
+      "train_tokens_per_second": 9094.517
+    },
+    {
+      "epoch": 0.1519195237117635,
+      "grad_norm": 0.7034651036046181,
+      "learning_rate": 9.85868898654803e-06,
+      "loss": 0.4079,
+      "num_input_tokens_seen": 40971152,
+      "step": 740,
+      "train_runtime": 4500.6079,
+      "train_tokens_per_second": 9103.471
+    },
+    {
+      "epoch": 0.15294600698008623,
+      "grad_norm": 1.6181784720089798,
+      "learning_rate": 9.856779537633968e-06,
+      "loss": 0.3692,
+      "num_input_tokens_seen": 41253392,
+      "step": 745,
+      "train_runtime": 4528.1292,
+      "train_tokens_per_second": 9110.471
+    },
+    {
+      "epoch": 0.15397249024840895,
+      "grad_norm": 0.8004059440701587,
+      "learning_rate": 9.854857461963572e-06,
+      "loss": 0.3953,
+      "num_input_tokens_seen": 41519168,
+      "step": 750,
+      "train_runtime": 4557.2264,
+      "train_tokens_per_second": 9110.622
+    },
+    {
+      "epoch": 0.15499897351673167,
+      "grad_norm": 0.7528244056765105,
+      "learning_rate": 9.852922764533899e-06,
+      "loss": 0.385,
+      "num_input_tokens_seen": 41797328,
+      "step": 755,
+      "train_runtime": 4581.2562,
+      "train_tokens_per_second": 9123.552
+    },
+    {
+      "epoch": 0.15602545678505442,
+      "grad_norm": 0.7908962723294183,
+      "learning_rate": 9.850975450374814e-06,
+      "loss": 0.3368,
+      "num_input_tokens_seen": 42092360,
+      "step": 760,
+      "train_runtime": 4616.7669,
+      "train_tokens_per_second": 9117.281
+    },
+    {
+      "epoch": 0.15705194005337714,
+      "grad_norm": 0.8363021642365176,
+      "learning_rate": 9.849015524548982e-06,
+      "loss": 0.3918,
+      "num_input_tokens_seen": 42387680,
+      "step": 765,
+      "train_runtime": 4651.852,
+      "train_tokens_per_second": 9112.001
+    },
+    {
+      "epoch": 0.15807842332169986,
+      "grad_norm": 0.8589830405896183,
+      "learning_rate": 9.847042992151864e-06,
+      "loss": 0.3917,
+      "num_input_tokens_seen": 42664952,
+      "step": 770,
+      "train_runtime": 4676.6877,
+      "train_tokens_per_second": 9122.9
+    },
+    {
+      "epoch": 0.15910490659002258,
+      "grad_norm": 0.6444728006733647,
+      "learning_rate": 9.845057858311686e-06,
+      "loss": 0.3954,
+      "num_input_tokens_seen": 42967576,
+      "step": 775,
+      "train_runtime": 4709.6527,
+      "train_tokens_per_second": 9123.3
+    },
+    {
+      "epoch": 0.1601313898583453,
+      "grad_norm": 0.8962188971560138,
+      "learning_rate": 9.843060128189444e-06,
+      "loss": 0.4286,
+      "num_input_tokens_seen": 43226816,
+      "step": 780,
+      "train_runtime": 4743.0418,
+      "train_tokens_per_second": 9113.733
+    },
+    {
+      "epoch": 0.16115787312666804,
+      "grad_norm": 0.8120360628375487,
+      "learning_rate": 9.841049806978874e-06,
+      "loss": 0.3849,
+      "num_input_tokens_seen": 43504208,
+      "step": 785,
+      "train_runtime": 4771.8766,
+      "train_tokens_per_second": 9116.792
+    },
+    {
+      "epoch": 0.16218435639499076,
+      "grad_norm": 0.7365829651621817,
+      "learning_rate": 9.839026899906454e-06,
+      "loss": 0.3994,
+      "num_input_tokens_seen": 43781288,
+      "step": 790,
+      "train_runtime": 4803.8057,
+      "train_tokens_per_second": 9113.876
+    },
+    {
+      "epoch": 0.16321083966331348,
+      "grad_norm": 0.9021919855046685,
+      "learning_rate": 9.83699141223138e-06,
+      "loss": 0.3811,
+      "num_input_tokens_seen": 44057600,
+      "step": 795,
+      "train_runtime": 4839.277,
+      "train_tokens_per_second": 9104.17
+    },
+    {
+      "epoch": 0.1642373229316362,
+      "grad_norm": 0.7148451077064707,
+      "learning_rate": 9.834943349245554e-06,
+      "loss": 0.3975,
+      "num_input_tokens_seen": 44315800,
+      "step": 800,
+      "train_runtime": 4868.6034,
+      "train_tokens_per_second": 9102.364
+    },
+    {
+      "epoch": 0.16526380619995895,
+      "grad_norm": 0.9229643721311263,
+      "learning_rate": 9.832882716273572e-06,
+      "loss": 0.4086,
+      "num_input_tokens_seen": 44574728,
+      "step": 805,
+      "train_runtime": 4891.7499,
+      "train_tokens_per_second": 9112.225
+    },
+    {
+      "epoch": 0.16629028946828167,
+      "grad_norm": 0.81441203646081,
+      "learning_rate": 9.830809518672713e-06,
+      "loss": 0.4374,
+      "num_input_tokens_seen": 44841696,
+      "step": 810,
+      "train_runtime": 4922.2313,
+      "train_tokens_per_second": 9110.034
+    },
+    {
+      "epoch": 0.1673167727366044,
+      "grad_norm": 1.0201274726818677,
+      "learning_rate": 9.828723761832914e-06,
+      "loss": 0.4077,
+      "num_input_tokens_seen": 45107512,
+      "step": 815,
+      "train_runtime": 4953.9817,
+      "train_tokens_per_second": 9105.304
+    },
+    {
+      "epoch": 0.1683432560049271,
+      "grad_norm": 0.7407391883278915,
+      "learning_rate": 9.826625451176777e-06,
+      "loss": 0.4018,
+      "num_input_tokens_seen": 45397560,
+      "step": 820,
+      "train_runtime": 4988.7259,
+      "train_tokens_per_second": 9100.031
+    },
+    {
+      "epoch": 0.16936973927324986,
+      "grad_norm": 0.7787411197741427,
+      "learning_rate": 9.824514592159526e-06,
+      "loss": 0.4106,
+      "num_input_tokens_seen": 45673792,
+      "step": 825,
+      "train_runtime": 5021.2954,
+      "train_tokens_per_second": 9096.018
+    },
+    {
+      "epoch": 0.17039622254157258,
+      "grad_norm": 0.7806451599186509,
+      "learning_rate": 9.82239119026902e-06,
+      "loss": 0.4032,
+      "num_input_tokens_seen": 45948088,
+      "step": 830,
+      "train_runtime": 5050.8736,
+      "train_tokens_per_second": 9097.058
+    },
+    {
+      "epoch": 0.1714227058098953,
+      "grad_norm": 0.7211129637718052,
+      "learning_rate": 9.820255251025724e-06,
+      "loss": 0.4133,
+      "num_input_tokens_seen": 46202368,
+      "step": 835,
+      "train_runtime": 5073.9941,
+      "train_tokens_per_second": 9105.72
+    },
+    {
+      "epoch": 0.17244918907821802,
+      "grad_norm": 0.8395271808194775,
+      "learning_rate": 9.818106779982693e-06,
+      "loss": 0.4139,
+      "num_input_tokens_seen": 46502728,
+      "step": 840,
+      "train_runtime": 5107.7595,
+      "train_tokens_per_second": 9104.33
+    },
+    {
+      "epoch": 0.17347567234654074,
+      "grad_norm": 0.6497742081145332,
+      "learning_rate": 9.81594578272557e-06,
+      "loss": 0.3952,
+      "num_input_tokens_seen": 46769456,
+      "step": 845,
+      "train_runtime": 5135.4706,
+      "train_tokens_per_second": 9107.141
+    },
+    {
+      "epoch": 0.17450215561486349,
+      "grad_norm": 0.8066409845504592,
+      "learning_rate": 9.81377226487256e-06,
+      "loss": 0.3809,
+      "num_input_tokens_seen": 47035008,
+      "step": 850,
+      "train_runtime": 5157.855,
+      "train_tokens_per_second": 9119.102
+    },
+    {
+      "epoch": 0.1755286388831862,
+      "grad_norm": 0.7326407537453494,
+      "learning_rate": 9.811586232074421e-06,
+      "loss": 0.4097,
+      "num_input_tokens_seen": 47297160,
+      "step": 855,
+      "train_runtime": 5187.2167,
+      "train_tokens_per_second": 9118.023
+    },
+    {
+      "epoch": 0.17655512215150893,
+      "grad_norm": 0.7658091343428134,
+      "learning_rate": 9.809387690014446e-06,
+      "loss": 0.3973,
+      "num_input_tokens_seen": 47551656,
+      "step": 860,
+      "train_runtime": 5214.4054,
+      "train_tokens_per_second": 9119.286
+    },
+    {
+      "epoch": 0.17758160541983165,
+      "grad_norm": 0.776850575643609,
+      "learning_rate": 9.807176644408449e-06,
+      "loss": 0.4118,
+      "num_input_tokens_seen": 47816256,
+      "step": 865,
+      "train_runtime": 5247.8085,
+      "train_tokens_per_second": 9111.662
+    },
+    {
+      "epoch": 0.1786080886881544,
+      "grad_norm": 0.9894904238450722,
+      "learning_rate": 9.804953101004752e-06,
+      "loss": 0.4037,
+      "num_input_tokens_seen": 48079224,
+      "step": 870,
+      "train_runtime": 5276.1389,
+      "train_tokens_per_second": 9112.577
+    },
+    {
+      "epoch": 0.1796345719564771,
+      "grad_norm": 1.3104978065370896,
+      "learning_rate": 9.80271706558417e-06,
+      "loss": 0.4027,
+      "num_input_tokens_seen": 48375632,
+      "step": 875,
+      "train_runtime": 5309.5805,
+      "train_tokens_per_second": 9111.008
+    },
+    {
+      "epoch": 0.18066105522479983,
+      "grad_norm": 0.8043649013622309,
+      "learning_rate": 9.800468543959996e-06,
+      "loss": 0.399,
+      "num_input_tokens_seen": 48633536,
+      "step": 880,
+      "train_runtime": 5336.9197,
+      "train_tokens_per_second": 9112.66
+    },
+    {
+      "epoch": 0.18168753849312255,
+      "grad_norm": 0.7607699735059696,
+      "learning_rate": 9.798207541977981e-06,
+      "loss": 0.4037,
+      "num_input_tokens_seen": 48906416,
+      "step": 885,
+      "train_runtime": 5366.4507,
+      "train_tokens_per_second": 9113.364
+    },
+    {
+      "epoch": 0.1827140217614453,
+      "grad_norm": 0.8648793687842831,
+      "learning_rate": 9.795934065516327e-06,
+      "loss": 0.39,
+      "num_input_tokens_seen": 49153520,
+      "step": 890,
+      "train_runtime": 5399.1409,
+      "train_tokens_per_second": 9103.952
+    },
+    {
+      "epoch": 0.18374050502976802,
+      "grad_norm": 0.7147097732214664,
+      "learning_rate": 9.793648120485661e-06,
+      "loss": 0.3825,
+      "num_input_tokens_seen": 49452096,
+      "step": 895,
+      "train_runtime": 5422.5997,
+      "train_tokens_per_second": 9119.629
+    },
+    {
+      "epoch": 0.18476698829809074,
+      "grad_norm": 0.6641017721260983,
+      "learning_rate": 9.791349712829037e-06,
+      "loss": 0.3786,
+      "num_input_tokens_seen": 49741408,
+      "step": 900,
+      "train_runtime": 5451.3621,
+      "train_tokens_per_second": 9124.583
+    },
+    {
+      "epoch": 0.18579347156641346,
+      "grad_norm": 0.7456366591497028,
+      "learning_rate": 9.789038848521897e-06,
+      "loss": 0.3624,
+      "num_input_tokens_seen": 50029976,
+      "step": 905,
+      "train_runtime": 5481.0733,
+      "train_tokens_per_second": 9127.77
+    },
+    {
+      "epoch": 0.1868199548347362,
+      "grad_norm": 0.9175117130912842,
+      "learning_rate": 9.786715533572078e-06,
+      "loss": 0.4021,
+      "num_input_tokens_seen": 50317320,
+      "step": 910,
+      "train_runtime": 5504.6476,
+      "train_tokens_per_second": 9140.879
+    },
+    {
+      "epoch": 0.18784643810305893,
+      "grad_norm": 0.8352781709569066,
+      "learning_rate": 9.78437977401978e-06,
+      "loss": 0.3904,
+      "num_input_tokens_seen": 50594504,
+      "step": 915,
+      "train_runtime": 5529.6908,
+      "train_tokens_per_second": 9149.608
+    },
+    {
+      "epoch": 0.18887292137138165,
+      "grad_norm": 0.8703462014623028,
+      "learning_rate": 9.78203157593756e-06,
+      "loss": 0.4011,
+      "num_input_tokens_seen": 50866192,
+      "step": 920,
+      "train_runtime": 5554.9583,
+      "train_tokens_per_second": 9156.899
+    },
+    {
+      "epoch": 0.18989940463970437,
+      "grad_norm": 0.7467924053746212,
+      "learning_rate": 9.779670945430314e-06,
+      "loss": 0.4268,
+      "num_input_tokens_seen": 51164768,
+      "step": 925,
+      "train_runtime": 5588.5892,
+      "train_tokens_per_second": 9155.221
+    },
+    {
+      "epoch": 0.1909258879080271,
+      "grad_norm": 0.783145634544983,
+      "learning_rate": 9.777297888635254e-06,
+      "loss": 0.3648,
+      "num_input_tokens_seen": 51429384,
+      "step": 930,
+      "train_runtime": 5619.2878,
+      "train_tokens_per_second": 9152.296
+    },
+    {
+      "epoch": 0.19195237117634983,
+      "grad_norm": 0.735239218871721,
+      "learning_rate": 9.774912411721907e-06,
+      "loss": 0.4002,
+      "num_input_tokens_seen": 51720872,
+      "step": 935,
+      "train_runtime": 5648.9522,
+      "train_tokens_per_second": 9155.835
+    },
+    {
+      "epoch": 0.19297885444467255,
+      "grad_norm": 0.782687811391192,
+      "learning_rate": 9.772514520892081e-06,
+      "loss": 0.4016,
+      "num_input_tokens_seen": 51978264,
+      "step": 940,
+      "train_runtime": 5681.9622,
+      "train_tokens_per_second": 9147.943
+    },
+    {
+      "epoch": 0.19400533771299527,
+      "grad_norm": 0.9611143558156527,
+      "learning_rate": 9.770104222379866e-06,
+      "loss": 0.4119,
+      "num_input_tokens_seen": 52225240,
+      "step": 945,
+      "train_runtime": 5712.0165,
+      "train_tokens_per_second": 9143.048
+    },
+    {
+      "epoch": 0.195031820981318,
+      "grad_norm": 0.9880073792322263,
+      "learning_rate": 9.767681522451603e-06,
+      "loss": 0.3798,
+      "num_input_tokens_seen": 52520544,
+      "step": 950,
+      "train_runtime": 5741.0745,
+      "train_tokens_per_second": 9148.208
+    },
+    {
+      "epoch": 0.19605830424964074,
+      "grad_norm": 1.0927875314172302,
+      "learning_rate": 9.765246427405883e-06,
+      "loss": 0.3788,
+      "num_input_tokens_seen": 52782048,
+      "step": 955,
+      "train_runtime": 5768.8432,
+      "train_tokens_per_second": 9149.503
+    },
+    {
+      "epoch": 0.19708478751796346,
+      "grad_norm": 0.7276368009349232,
+      "learning_rate": 9.76279894357351e-06,
+      "loss": 0.4148,
+      "num_input_tokens_seen": 53027024,
+      "step": 960,
+      "train_runtime": 5802.1363,
+      "train_tokens_per_second": 9139.224
+    },
+    {
+      "epoch": 0.19811127078628618,
+      "grad_norm": 0.7905953507521015,
+      "learning_rate": 9.760339077317508e-06,
+      "loss": 0.402,
+      "num_input_tokens_seen": 53275432,
+      "step": 965,
+      "train_runtime": 5827.6999,
+      "train_tokens_per_second": 9141.76
+    },
+    {
+      "epoch": 0.1991377540546089,
+      "grad_norm": 0.717058390400047,
+      "learning_rate": 9.757866835033088e-06,
+      "loss": 0.4055,
+      "num_input_tokens_seen": 53560864,
+      "step": 970,
+      "train_runtime": 5857.8292,
+      "train_tokens_per_second": 9143.466
+    },
+    {
+      "epoch": 0.20016423732293165,
+      "grad_norm": 0.8827673609636191,
+      "learning_rate": 9.755382223147634e-06,
+      "loss": 0.4072,
+      "num_input_tokens_seen": 53819904,
+      "step": 975,
+      "train_runtime": 5889.1256,
+      "train_tokens_per_second": 9138.862
+    },
+    {
+      "epoch": 0.20119072059125437,
+      "grad_norm": 0.8419087785248859,
+      "learning_rate": 9.752885248120694e-06,
+      "loss": 0.3843,
+      "num_input_tokens_seen": 54087136,
+      "step": 980,
+      "train_runtime": 5925.7531,
+      "train_tokens_per_second": 9127.47
+    },
+    {
+      "epoch": 0.2022172038595771,
+      "grad_norm": 0.7606995257745054,
+      "learning_rate": 9.750375916443952e-06,
+      "loss": 0.4068,
+      "num_input_tokens_seen": 54366328,
+      "step": 985,
+      "train_runtime": 5958.1519,
+      "train_tokens_per_second": 9124.697
+    },
+    {
+      "epoch": 0.2032436871278998,
+      "grad_norm": 0.818628324139173,
+      "learning_rate": 9.747854234641228e-06,
+      "loss": 0.3879,
+      "num_input_tokens_seen": 54626720,
+      "step": 990,
+      "train_runtime": 5989.3792,
+      "train_tokens_per_second": 9120.598
+    },
+    {
+      "epoch": 0.20427017039622253,
+      "grad_norm": 0.7700534803301051,
+      "learning_rate": 9.745320209268434e-06,
+      "loss": 0.4271,
+      "num_input_tokens_seen": 54907000,
+      "step": 995,
+      "train_runtime": 6017.8507,
+      "train_tokens_per_second": 9124.022
+    },
+    {
+      "epoch": 0.20529665366454528,
+      "grad_norm": 0.7186329121994488,
+      "learning_rate": 9.742773846913587e-06,
+      "loss": 0.4215,
+      "num_input_tokens_seen": 55176896,
+      "step": 1000,
+      "train_runtime": 6043.353,
+      "train_tokens_per_second": 9130.179
+    },
+    {
+      "epoch": 0.206323136932868,
+      "grad_norm": 0.6461686750636164,
+      "learning_rate": 9.740215154196773e-06,
+      "loss": 0.3892,
+      "num_input_tokens_seen": 55467192,
+      "step": 1005,
+      "train_runtime": 6074.081,
+      "train_tokens_per_second": 9131.783
+    },
+    {
+      "epoch": 0.20734962020119072,
+      "grad_norm": 0.7677615192539505,
+      "learning_rate": 9.737644137770133e-06,
+      "loss": 0.42,
+      "num_input_tokens_seen": 55758608,
+      "step": 1010,
+      "train_runtime": 6104.0152,
+      "train_tokens_per_second": 9134.743
+    },
+    {
+      "epoch": 0.20837610346951344,
+      "grad_norm": 0.7466547437248883,
+      "learning_rate": 9.735060804317848e-06,
+      "loss": 0.3819,
+      "num_input_tokens_seen": 56051288,
+      "step": 1015,
+      "train_runtime": 6133.834,
+      "train_tokens_per_second": 9138.051
+    },
+    {
+      "epoch": 0.20940258673783618,
+      "grad_norm": 0.7826862700642607,
+      "learning_rate": 9.732465160556123e-06,
+      "loss": 0.4072,
+      "num_input_tokens_seen": 56332704,
+      "step": 1020,
+      "train_runtime": 6160.5197,
+      "train_tokens_per_second": 9144.148
+    },
+    {
+      "epoch": 0.2104290700061589,
+      "grad_norm": 0.6831111898975775,
+      "learning_rate": 9.729857213233169e-06,
+      "loss": 0.3906,
+      "num_input_tokens_seen": 56617584,
+      "step": 1025,
+      "train_runtime": 6191.132,
+      "train_tokens_per_second": 9144.949
+    },
+    {
+      "epoch": 0.21145555327448162,
+      "grad_norm": 0.7107797995579107,
+      "learning_rate": 9.727236969129176e-06,
+      "loss": 0.3835,
+      "num_input_tokens_seen": 56910048,
+      "step": 1030,
+      "train_runtime": 6225.7954,
+      "train_tokens_per_second": 9141.008
+    },
+    {
+      "epoch": 0.21248203654280434,
+      "grad_norm": 0.7599682786968505,
+      "learning_rate": 9.724604435056316e-06,
+      "loss": 0.3745,
+      "num_input_tokens_seen": 57180336,
+      "step": 1035,
+      "train_runtime": 6256.1327,
+      "train_tokens_per_second": 9139.885
+    },
+    {
+      "epoch": 0.2135085198111271,
+      "grad_norm": 0.9183764053694906,
+      "learning_rate": 9.721959617858699e-06,
+      "loss": 0.3863,
+      "num_input_tokens_seen": 57470840,
+      "step": 1040,
+      "train_runtime": 6292.0799,
+      "train_tokens_per_second": 9133.838
+    },
+    {
+      "epoch": 0.2145350030794498,
+      "grad_norm": 0.7100611448055162,
+      "learning_rate": 9.71930252441238e-06,
+      "loss": 0.3809,
+      "num_input_tokens_seen": 57752368,
+      "step": 1045,
+      "train_runtime": 6322.3037,
+      "train_tokens_per_second": 9134.703
+    },
+    {
+      "epoch": 0.21556148634777253,
+      "grad_norm": 0.693736115543734,
+      "learning_rate": 9.716633161625324e-06,
+      "loss": 0.3808,
+      "num_input_tokens_seen": 58025184,
+      "step": 1050,
+      "train_runtime": 6347.0335,
+      "train_tokens_per_second": 9142.095
+    },
+    {
+      "epoch": 0.21658796961609525,
+      "grad_norm": 0.7414240037799011,
+      "learning_rate": 9.713951536437395e-06,
+      "loss": 0.4014,
+      "num_input_tokens_seen": 58298672,
+      "step": 1055,
+      "train_runtime": 6373.6705,
+      "train_tokens_per_second": 9146.797
+    },
+    {
+      "epoch": 0.217614452884418,
+      "grad_norm": 0.7493171227061323,
+      "learning_rate": 9.71125765582034e-06,
+      "loss": 0.4047,
+      "num_input_tokens_seen": 58601896,
+      "step": 1060,
+      "train_runtime": 6407.5784,
+      "train_tokens_per_second": 9145.717
+    },
+    {
+      "epoch": 0.21864093615274072,
+      "grad_norm": 0.7465571484112782,
+      "learning_rate": 9.708551526777767e-06,
+      "loss": 0.4043,
+      "num_input_tokens_seen": 58883896,
+      "step": 1065,
+      "train_runtime": 6438.3139,
+      "train_tokens_per_second": 9145.857
+    },
+    {
+      "epoch": 0.21966741942106344,
+      "grad_norm": 0.7121161408858652,
+      "learning_rate": 9.705833156345124e-06,
+      "loss": 0.4093,
+      "num_input_tokens_seen": 59162000,
+      "step": 1070,
+      "train_runtime": 6466.6074,
+      "train_tokens_per_second": 9148.847
+    },
+    {
+      "epoch": 0.22069390268938616,
+      "grad_norm": 0.6175487774798406,
+      "learning_rate": 9.703102551589686e-06,
+      "loss": 0.3818,
+      "num_input_tokens_seen": 59463376,
+      "step": 1075,
+      "train_runtime": 6493.9325,
+      "train_tokens_per_second": 9156.759
+    },
+    {
+      "epoch": 0.22172038595770888,
+      "grad_norm": 0.8491725971767244,
+      "learning_rate": 9.700359719610538e-06,
+      "loss": 0.3997,
+      "num_input_tokens_seen": 59709480,
+      "step": 1080,
+      "train_runtime": 6525.6314,
+      "train_tokens_per_second": 9149.993
+    },
+    {
+      "epoch": 0.22274686922603162,
+      "grad_norm": 0.7974983681796582,
+      "learning_rate": 9.697604667538549e-06,
+      "loss": 0.4014,
+      "num_input_tokens_seen": 59990656,
+      "step": 1085,
+      "train_runtime": 6556.2123,
+      "train_tokens_per_second": 9150.2
+    },
+    {
+      "epoch": 0.22377335249435434,
+      "grad_norm": 0.7135161870858717,
+      "learning_rate": 9.694837402536366e-06,
+      "loss": 0.3833,
+      "num_input_tokens_seen": 60257344,
+      "step": 1090,
+      "train_runtime": 6588.3886,
+      "train_tokens_per_second": 9145.991
+    },
+    {
+      "epoch": 0.22479983576267706,
+      "grad_norm": 0.6998408316888066,
+      "learning_rate": 9.692057931798376e-06,
+      "loss": 0.3773,
+      "num_input_tokens_seen": 60557448,
+      "step": 1095,
+      "train_runtime": 6618.3661,
+      "train_tokens_per_second": 9149.909
+    },
+    {
+      "epoch": 0.22582631903099978,
+      "grad_norm": 0.7064428324130719,
+      "learning_rate": 9.689266262550709e-06,
+      "loss": 0.3737,
+      "num_input_tokens_seen": 60823608,
+      "step": 1100,
+      "train_runtime": 6646.6085,
+      "train_tokens_per_second": 9151.074
+    },
+    {
+      "epoch": 0.22685280229932253,
+      "grad_norm": 0.8851867141687565,
+      "learning_rate": 9.686462402051202e-06,
+      "loss": 0.4051,
+      "num_input_tokens_seen": 61071016,
+      "step": 1105,
+      "train_runtime": 6674.0063,
+      "train_tokens_per_second": 9150.578
+    },
+    {
+      "epoch": 0.22787928556764525,
+      "grad_norm": 0.9339515488261513,
+      "learning_rate": 9.68364635758939e-06,
+      "loss": 0.3886,
+      "num_input_tokens_seen": 61323840,
+      "step": 1110,
+      "train_runtime": 6702.4664,
+      "train_tokens_per_second": 9149.444
+    },
+    {
+      "epoch": 0.22890576883596797,
+      "grad_norm": 0.7906557815162372,
+      "learning_rate": 9.680818136486484e-06,
+      "loss": 0.3904,
+      "num_input_tokens_seen": 61575856,
+      "step": 1115,
+      "train_runtime": 6733.5577,
+      "train_tokens_per_second": 9144.624
+    },
+    {
+      "epoch": 0.2299322521042907,
+      "grad_norm": 0.7524731693885143,
+      "learning_rate": 9.677977746095356e-06,
+      "loss": 0.343,
+      "num_input_tokens_seen": 61837064,
+      "step": 1120,
+      "train_runtime": 6761.9497,
+      "train_tokens_per_second": 9144.857
+    },
+    {
+      "epoch": 0.23095873537261344,
+      "grad_norm": 0.6733977797771152,
+      "learning_rate": 9.675125193800508e-06,
+      "loss": 0.3674,
+      "num_input_tokens_seen": 62123856,
+      "step": 1125,
+      "train_runtime": 6798.0094,
+      "train_tokens_per_second": 9138.536
+    },
+    {
+      "epoch": 0.23198521864093616,
+      "grad_norm": 0.700332220176808,
+      "learning_rate": 9.672260487018064e-06,
+      "loss": 0.4015,
+      "num_input_tokens_seen": 62409392,
+      "step": 1130,
+      "train_runtime": 6834.071,
+      "train_tokens_per_second": 9132.096
+    },
+    {
+      "epoch": 0.23301170190925888,
+      "grad_norm": 0.7830907447502897,
+      "learning_rate": 9.669383633195749e-06,
+      "loss": 0.4069,
+      "num_input_tokens_seen": 62669136,
+      "step": 1135,
+      "train_runtime": 6860.1989,
+      "train_tokens_per_second": 9135.178
+    },
+    {
+      "epoch": 0.2340381851775816,
+      "grad_norm": 0.8497889059146713,
+      "learning_rate": 9.666494639812869e-06,
+      "loss": 0.4329,
+      "num_input_tokens_seen": 62961288,
+      "step": 1140,
+      "train_runtime": 6888.5612,
+      "train_tokens_per_second": 9139.977
+    },
+    {
+      "epoch": 0.23506466844590435,
+      "grad_norm": 0.7267336657602529,
+      "learning_rate": 9.663593514380287e-06,
+      "loss": 0.4111,
+      "num_input_tokens_seen": 63237704,
+      "step": 1145,
+      "train_runtime": 6918.5243,
+      "train_tokens_per_second": 9140.346
+    },
+    {
+      "epoch": 0.23609115171422707,
+      "grad_norm": 0.8102920676465493,
+      "learning_rate": 9.660680264440409e-06,
+      "loss": 0.3868,
+      "num_input_tokens_seen": 63520648,
+      "step": 1150,
+      "train_runtime": 6947.6545,
+      "train_tokens_per_second": 9142.747
+    },
+    {
+      "epoch": 0.23711763498254979,
+      "grad_norm": 0.6442175629132532,
+      "learning_rate": 9.657754897567167e-06,
+      "loss": 0.3724,
+      "num_input_tokens_seen": 63777576,
+      "step": 1155,
+      "train_runtime": 6975.4966,
+      "train_tokens_per_second": 9143.088
+    },
+    {
+      "epoch": 0.2381441182508725,
+      "grad_norm": 0.745497106911732,
+      "learning_rate": 9.654817421365985e-06,
+      "loss": 0.3976,
+      "num_input_tokens_seen": 64059128,
+      "step": 1160,
+      "train_runtime": 7004.4383,
+      "train_tokens_per_second": 9145.505
+    },
+    {
+      "epoch": 0.23917060151919523,
+      "grad_norm": 0.7850082152253003,
+      "learning_rate": 9.651867843473777e-06,
+      "loss": 0.4093,
+      "num_input_tokens_seen": 64350424,
+      "step": 1165,
+      "train_runtime": 7034.4548,
+      "train_tokens_per_second": 9147.891
+    },
+    {
+      "epoch": 0.24019708478751797,
+      "grad_norm": 0.7729292299306528,
+      "learning_rate": 9.648906171558917e-06,
+      "loss": 0.4158,
+      "num_input_tokens_seen": 64609232,
+      "step": 1170,
+      "train_runtime": 7068.4168,
+      "train_tokens_per_second": 9140.552
+    },
+    {
+      "epoch": 0.2412235680558407,
+      "grad_norm": 0.79892293037197,
+      "learning_rate": 9.645932413321222e-06,
+      "loss": 0.4281,
+      "num_input_tokens_seen": 64885544,
+      "step": 1175,
+      "train_runtime": 7095.4481,
+      "train_tokens_per_second": 9144.672
+    },
+    {
+      "epoch": 0.2422500513241634,
+      "grad_norm": 0.7386074978396872,
+      "learning_rate": 9.642946576491928e-06,
+      "loss": 0.3876,
+      "num_input_tokens_seen": 65138256,
+      "step": 1180,
+      "train_runtime": 7125.1255,
+      "train_tokens_per_second": 9142.05
+    },
+    {
+      "epoch": 0.24327653459248613,
+      "grad_norm": 1.1879085192245964,
+      "learning_rate": 9.639948668833677e-06,
+      "loss": 0.3415,
+      "num_input_tokens_seen": 65412144,
+      "step": 1185,
+      "train_runtime": 7153.3591,
+      "train_tokens_per_second": 9144.256
+    },
+    {
+      "epoch": 0.24430301786080888,
+      "grad_norm": 0.7016694531805002,
+      "learning_rate": 9.636938698140493e-06,
+      "loss": 0.4069,
+      "num_input_tokens_seen": 65687160,
+      "step": 1190,
+      "train_runtime": 7187.8579,
+      "train_tokens_per_second": 9138.628
+    },
+    {
+      "epoch": 0.2453295011291316,
+      "grad_norm": 0.8529934973281005,
+      "learning_rate": 9.633916672237758e-06,
+      "loss": 0.3792,
+      "num_input_tokens_seen": 65954408,
+      "step": 1195,
+      "train_runtime": 7216.6411,
+      "train_tokens_per_second": 9139.211
+    },
+    {
+      "epoch": 0.24635598439745432,
+      "grad_norm": 0.7443081448839253,
+      "learning_rate": 9.630882598982198e-06,
+      "loss": 0.4269,
+      "num_input_tokens_seen": 66220408,
+      "step": 1200,
+      "train_runtime": 7248.0176,
+      "train_tokens_per_second": 9136.348
+    },
+    {
+      "epoch": 0.24738246766577704,
+      "grad_norm": 0.7944672553322008,
+      "learning_rate": 9.627836486261862e-06,
+      "loss": 0.394,
+      "num_input_tokens_seen": 66490176,
+      "step": 1205,
+      "train_runtime": 7279.3614,
+      "train_tokens_per_second": 9134.067
+    },
+    {
+      "epoch": 0.2484089509340998,
+      "grad_norm": 0.8041485730015021,
+      "learning_rate": 9.624778341996094e-06,
+      "loss": 0.3611,
+      "num_input_tokens_seen": 66765120,
+      "step": 1210,
+      "train_runtime": 7299.498,
+      "train_tokens_per_second": 9146.536
+    },
+    {
+      "epoch": 0.2494354342024225,
+      "grad_norm": 0.7870785232236602,
+      "learning_rate": 9.621708174135525e-06,
+      "loss": 0.414,
+      "num_input_tokens_seen": 67041624,
+      "step": 1215,
+      "train_runtime": 7323.8452,
+      "train_tokens_per_second": 9153.883
+    },
+    {
+      "epoch": 0.25046191747074525,
+      "grad_norm": 0.7575113326338325,
+      "learning_rate": 9.618625990662037e-06,
+      "loss": 0.4116,
+      "num_input_tokens_seen": 67324976,
+      "step": 1220,
+      "train_runtime": 7352.5543,
+      "train_tokens_per_second": 9156.679
+    },
+    {
+      "epoch": 0.251488400739068,
+      "grad_norm": 0.7811475396865072,
+      "learning_rate": 9.615531799588756e-06,
+      "loss": 0.4159,
+      "num_input_tokens_seen": 67598248,
+      "step": 1225,
+      "train_runtime": 7385.5959,
+      "train_tokens_per_second": 9152.714
+    },
+    {
+      "epoch": 0.2525148840073907,
+      "grad_norm": 1.014320066446144,
+      "learning_rate": 9.612425608960024e-06,
+      "loss": 0.4094,
+      "num_input_tokens_seen": 67853912,
+      "step": 1230,
+      "train_runtime": 7415.5142,
+      "train_tokens_per_second": 9150.264
+    },
+    {
+      "epoch": 0.2535413672757134,
+      "grad_norm": 0.6855949570476443,
+      "learning_rate": 9.60930742685138e-06,
+      "loss": 0.3986,
+      "num_input_tokens_seen": 68127224,
+      "step": 1235,
+      "train_runtime": 7445.5787,
+      "train_tokens_per_second": 9150.024
+    },
+    {
+      "epoch": 0.25456785054403613,
+      "grad_norm": 0.737156005884372,
+      "learning_rate": 9.606177261369541e-06,
+      "loss": 0.3954,
+      "num_input_tokens_seen": 68405248,
+      "step": 1240,
+      "train_runtime": 7475.9951,
+      "train_tokens_per_second": 9149.986
+    },
+    {
+      "epoch": 0.25559433381235885,
+      "grad_norm": 0.7708948122590483,
+      "learning_rate": 9.603035120652373e-06,
+      "loss": 0.3992,
+      "num_input_tokens_seen": 68656376,
+      "step": 1245,
+      "train_runtime": 7501.6003,
+      "train_tokens_per_second": 9152.231
+    },
+    {
+      "epoch": 0.2566208170806816,
+      "grad_norm": 2.4045199925385004,
+      "learning_rate": 9.59988101286888e-06,
+      "loss": 0.3987,
+      "num_input_tokens_seen": 68937976,
+      "step": 1250,
+      "train_runtime": 7534.3558,
+      "train_tokens_per_second": 9149.817
+    },
+    {
+      "epoch": 0.2576473003490043,
+      "grad_norm": 0.7883390437706823,
+      "learning_rate": 9.596714946219178e-06,
+      "loss": 0.4021,
+      "num_input_tokens_seen": 69206080,
+      "step": 1255,
+      "train_runtime": 7567.6626,
+      "train_tokens_per_second": 9144.974
+    },
+    {
+      "epoch": 0.258673783617327,
+      "grad_norm": 0.7261379447218289,
+      "learning_rate": 9.593536928934471e-06,
+      "loss": 0.3469,
+      "num_input_tokens_seen": 69465936,
+      "step": 1260,
+      "train_runtime": 7602.0149,
+      "train_tokens_per_second": 9137.832
+    },
+    {
+      "epoch": 0.2597002668856498,
+      "grad_norm": 0.7761096082361607,
+      "learning_rate": 9.590346969277036e-06,
+      "loss": 0.3713,
+      "num_input_tokens_seen": 69748616,
+      "step": 1265,
+      "train_runtime": 7636.2763,
+      "train_tokens_per_second": 9133.852
+    },
+    {
+      "epoch": 0.2607267501539725,
+      "grad_norm": 0.6920431842811126,
+      "learning_rate": 9.587145075540194e-06,
+      "loss": 0.3485,
+      "num_input_tokens_seen": 70010128,
+      "step": 1270,
+      "train_runtime": 7670.2663,
+      "train_tokens_per_second": 9127.47
+    },
+    {
+      "epoch": 0.26175323342229523,
+      "grad_norm": 0.7914894889985296,
+      "learning_rate": 9.583931256048298e-06,
+      "loss": 0.4265,
+      "num_input_tokens_seen": 70309552,
+      "step": 1275,
+      "train_runtime": 7695.4373,
+      "train_tokens_per_second": 9136.525
+    },
+    {
+      "epoch": 0.26277971669061795,
+      "grad_norm": 0.7284993629216492,
+      "learning_rate": 9.580705519156698e-06,
+      "loss": 0.4018,
+      "num_input_tokens_seen": 70599336,
+      "step": 1280,
+      "train_runtime": 7721.8257,
+      "train_tokens_per_second": 9142.829
+    },
+    {
+      "epoch": 0.26380619995894067,
+      "grad_norm": 0.717536058385607,
+      "learning_rate": 9.577467873251738e-06,
+      "loss": 0.3787,
+      "num_input_tokens_seen": 70857296,
+      "step": 1285,
+      "train_runtime": 7748.9153,
+      "train_tokens_per_second": 9144.157
+    },
+    {
+      "epoch": 0.2648326832272634,
+      "grad_norm": 0.6089000307238214,
+      "learning_rate": 9.574218326750713e-06,
+      "loss": 0.3822,
+      "num_input_tokens_seen": 71131832,
+      "step": 1290,
+      "train_runtime": 7780.3722,
+      "train_tokens_per_second": 9142.472
+    },
+    {
+      "epoch": 0.2658591664955861,
+      "grad_norm": 0.6877303330566175,
+      "learning_rate": 9.570956888101861e-06,
+      "loss": 0.3977,
+      "num_input_tokens_seen": 71410840,
+      "step": 1295,
+      "train_runtime": 7810.6007,
+      "train_tokens_per_second": 9142.81
+    },
+    {
+      "epoch": 0.2668856497639088,
+      "grad_norm": 1.3843169466368819,
+      "learning_rate": 9.567683565784343e-06,
+      "loss": 0.4066,
+      "num_input_tokens_seen": 71678072,
+      "step": 1300,
+      "train_runtime": 7841.6233,
+      "train_tokens_per_second": 9140.719
+    },
+    {
+      "epoch": 0.26791213303223155,
+      "grad_norm": 0.7173729520355662,
+      "learning_rate": 9.564398368308203e-06,
+      "loss": 0.3834,
+      "num_input_tokens_seen": 71966536,
+      "step": 1305,
+      "train_runtime": 7867.2302,
+      "train_tokens_per_second": 9147.633
+    },
+    {
+      "epoch": 0.2689386163005543,
+      "grad_norm": 0.6581267135751493,
+      "learning_rate": 9.56110130421437e-06,
+      "loss": 0.3872,
+      "num_input_tokens_seen": 72213720,
+      "step": 1310,
+      "train_runtime": 7898.7056,
+      "train_tokens_per_second": 9142.475
+    },
+    {
+      "epoch": 0.26996509956887704,
+      "grad_norm": 0.7973037242382532,
+      "learning_rate": 9.557792382074621e-06,
+      "loss": 0.3924,
+      "num_input_tokens_seen": 72491216,
+      "step": 1315,
+      "train_runtime": 7930.6727,
+      "train_tokens_per_second": 9140.614
+    },
+    {
+      "epoch": 0.27099158283719976,
+      "grad_norm": 0.6948891254313848,
+      "learning_rate": 9.554471610491557e-06,
+      "loss": 0.3815,
+      "num_input_tokens_seen": 72774936,
+      "step": 1320,
+      "train_runtime": 7965.7194,
+      "train_tokens_per_second": 9136.016
+    },
+    {
+      "epoch": 0.2720180661055225,
+      "grad_norm": 0.7060688458894412,
+      "learning_rate": 9.55113899809859e-06,
+      "loss": 0.3814,
+      "num_input_tokens_seen": 73047616,
+      "step": 1325,
+      "train_runtime": 7992.0354,
+      "train_tokens_per_second": 9140.052
+    },
+    {
+      "epoch": 0.2730445493738452,
+      "grad_norm": 0.8358048066232971,
+      "learning_rate": 9.547794553559916e-06,
+      "loss": 0.3797,
+      "num_input_tokens_seen": 73343128,
+      "step": 1330,
+      "train_runtime": 8019.6481,
+      "train_tokens_per_second": 9145.43
+    },
+    {
+      "epoch": 0.2740710326421679,
+      "grad_norm": 0.7233305009101423,
+      "learning_rate": 9.544438285570492e-06,
+      "loss": 0.3966,
+      "num_input_tokens_seen": 73611840,
+      "step": 1335,
+      "train_runtime": 8050.4993,
+      "train_tokens_per_second": 9143.761
+    },
+    {
+      "epoch": 0.27509751591049064,
+      "grad_norm": 0.8478232719206497,
+      "learning_rate": 9.54107020285601e-06,
+      "loss": 0.399,
+      "num_input_tokens_seen": 73876512,
+      "step": 1340,
+      "train_runtime": 8081.2136,
+      "train_tokens_per_second": 9141.76
+    },
+    {
+      "epoch": 0.27612399917881336,
+      "grad_norm": 0.7561518749859467,
+      "learning_rate": 9.537690314172887e-06,
+      "loss": 0.4061,
+      "num_input_tokens_seen": 74160248,
+      "step": 1345,
+      "train_runtime": 8112.1926,
+      "train_tokens_per_second": 9141.825
+    },
+    {
+      "epoch": 0.27715048244713614,
+      "grad_norm": 0.6757078277190117,
+      "learning_rate": 9.534298628308221e-06,
+      "loss": 0.3756,
+      "num_input_tokens_seen": 74433736,
+      "step": 1350,
+      "train_runtime": 8140.9436,
+      "train_tokens_per_second": 9143.134
+    },
+    {
+      "epoch": 0.27817696571545886,
+      "grad_norm": 0.7158419204370354,
+      "learning_rate": 9.530895154079791e-06,
+      "loss": 0.4074,
+      "num_input_tokens_seen": 74689240,
+      "step": 1355,
+      "train_runtime": 8171.2038,
+      "train_tokens_per_second": 9140.543
+    },
+    {
+      "epoch": 0.2792034489837816,
+      "grad_norm": 0.8501573582623446,
+      "learning_rate": 9.527479900336022e-06,
+      "loss": 0.3769,
+      "num_input_tokens_seen": 74936904,
+      "step": 1360,
+      "train_runtime": 8199.2142,
+      "train_tokens_per_second": 9139.523
+    },
+    {
+      "epoch": 0.2802299322521043,
+      "grad_norm": 0.7799168741049674,
+      "learning_rate": 9.524052875955956e-06,
+      "loss": 0.3765,
+      "num_input_tokens_seen": 75207928,
+      "step": 1365,
+      "train_runtime": 8227.3746,
+      "train_tokens_per_second": 9141.182
+    },
+    {
+      "epoch": 0.281256415520427,
+      "grad_norm": 0.7512961754458702,
+      "learning_rate": 9.520614089849247e-06,
+      "loss": 0.3771,
+      "num_input_tokens_seen": 75464336,
+      "step": 1370,
+      "train_runtime": 8254.119,
+      "train_tokens_per_second": 9142.628
+    },
+    {
+      "epoch": 0.28228289878874974,
+      "grad_norm": 0.7091192129546435,
+      "learning_rate": 9.517163550956121e-06,
+      "loss": 0.3793,
+      "num_input_tokens_seen": 75745936,
+      "step": 1375,
+      "train_runtime": 8278.708,
+      "train_tokens_per_second": 9149.488
+    },
+    {
+      "epoch": 0.28330938205707246,
+      "grad_norm": 0.7289109862949428,
+      "learning_rate": 9.51370126824736e-06,
+      "loss": 0.3833,
+      "num_input_tokens_seen": 76018784,
+      "step": 1380,
+      "train_runtime": 8309.3504,
+      "train_tokens_per_second": 9148.583
+    },
+    {
+      "epoch": 0.2843358653253952,
+      "grad_norm": 0.769640618355137,
+      "learning_rate": 9.51022725072428e-06,
+      "loss": 0.3853,
+      "num_input_tokens_seen": 76312920,
+      "step": 1385,
+      "train_runtime": 8343.5049,
+      "train_tokens_per_second": 9146.386
+    },
+    {
+      "epoch": 0.2853623485937179,
+      "grad_norm": 0.7678726999739379,
+      "learning_rate": 9.506741507418703e-06,
+      "loss": 0.4051,
+      "num_input_tokens_seen": 76593616,
+      "step": 1390,
+      "train_runtime": 8376.5431,
+      "train_tokens_per_second": 9143.822
+    },
+    {
+      "epoch": 0.28638883186204067,
+      "grad_norm": 0.718799563023163,
+      "learning_rate": 9.503244047392936e-06,
+      "loss": 0.3896,
+      "num_input_tokens_seen": 76880568,
+      "step": 1395,
+      "train_runtime": 8405.0723,
+      "train_tokens_per_second": 9146.925
+    },
+    {
+      "epoch": 0.2874153151303634,
+      "grad_norm": 0.7055826760826535,
+      "learning_rate": 9.499734879739752e-06,
+      "loss": 0.3902,
+      "num_input_tokens_seen": 77142664,
+      "step": 1400,
+      "train_runtime": 8438.8837,
+      "train_tokens_per_second": 9141.335
+    },
+    {
+      "epoch": 0.2884417983986861,
+      "grad_norm": 0.7663713747716872,
+      "learning_rate": 9.496214013582352e-06,
+      "loss": 0.3641,
+      "num_input_tokens_seen": 77405616,
+      "step": 1405,
+      "train_runtime": 8467.6757,
+      "train_tokens_per_second": 9141.306
+    },
+    {
+      "epoch": 0.28946828166700883,
+      "grad_norm": 0.7551390060846025,
+      "learning_rate": 9.492681458074362e-06,
+      "loss": 0.3858,
+      "num_input_tokens_seen": 77692176,
+      "step": 1410,
+      "train_runtime": 8493.527,
+      "train_tokens_per_second": 9147.222
+    },
+    {
+      "epoch": 0.29049476493533155,
+      "grad_norm": 0.8573767080247109,
+      "learning_rate": 9.489137222399791e-06,
+      "loss": 0.4173,
+      "num_input_tokens_seen": 77995624,
+      "step": 1415,
+      "train_runtime": 8528.6431,
+      "train_tokens_per_second": 9145.139
+    },
+    {
+      "epoch": 0.29152124820365427,
+      "grad_norm": 0.6429675392689181,
+      "learning_rate": 9.485581315773016e-06,
+      "loss": 0.3632,
+      "num_input_tokens_seen": 78279416,
+      "step": 1420,
+      "train_runtime": 8563.9227,
+      "train_tokens_per_second": 9140.603
+    },
+    {
+      "epoch": 0.292547731471977,
+      "grad_norm": 0.7864810662114388,
+      "learning_rate": 9.482013747438758e-06,
+      "loss": 0.3467,
+      "num_input_tokens_seen": 78551128,
+      "step": 1425,
+      "train_runtime": 8592.789,
+      "train_tokens_per_second": 9141.517
+    },
+    {
+      "epoch": 0.2935742147402997,
+      "grad_norm": 0.8745533170134085,
+      "learning_rate": 9.478434526672056e-06,
+      "loss": 0.3931,
+      "num_input_tokens_seen": 78813976,
+      "step": 1430,
+      "train_runtime": 8626.9967,
+      "train_tokens_per_second": 9135.737
+    },
+    {
+      "epoch": 0.2946006980086225,
+      "grad_norm": 0.7244926930829404,
+      "learning_rate": 9.474843662778242e-06,
+      "loss": 0.3819,
+      "num_input_tokens_seen": 79087344,
+      "step": 1435,
+      "train_runtime": 8652.6771,
+      "train_tokens_per_second": 9140.217
+    },
+    {
+      "epoch": 0.2956271812769452,
+      "grad_norm": 0.7595972243325454,
+      "learning_rate": 9.471241165092917e-06,
+      "loss": 0.3881,
+      "num_input_tokens_seen": 79358856,
+      "step": 1440,
+      "train_runtime": 8686.4208,
+      "train_tokens_per_second": 9135.967
+    },
+    {
+      "epoch": 0.2966536645452679,
+      "grad_norm": 0.745317289970527,
+      "learning_rate": 9.46762704298193e-06,
+      "loss": 0.3656,
+      "num_input_tokens_seen": 79632448,
+      "step": 1445,
+      "train_runtime": 8712.7887,
+      "train_tokens_per_second": 9139.72
+    },
+    {
+      "epoch": 0.29768014781359065,
+      "grad_norm": 0.7515249512115387,
+      "learning_rate": 9.464001305841351e-06,
+      "loss": 0.3737,
+      "num_input_tokens_seen": 79938648,
+      "step": 1450,
+      "train_runtime": 8748.1981,
+      "train_tokens_per_second": 9137.727
+    },
+    {
+      "epoch": 0.29870663108191337,
+      "grad_norm": 0.7698796400163836,
+      "learning_rate": 9.460363963097448e-06,
+      "loss": 0.366,
+      "num_input_tokens_seen": 80207696,
+      "step": 1455,
+      "train_runtime": 8781.3553,
+      "train_tokens_per_second": 9133.863
+    },
+    {
+      "epoch": 0.2997331143502361,
+      "grad_norm": 0.7193003183337814,
+      "learning_rate": 9.45671502420666e-06,
+      "loss": 0.3763,
+      "num_input_tokens_seen": 80465928,
+      "step": 1460,
+      "train_runtime": 8803.3637,
+      "train_tokens_per_second": 9140.362
+    },
+    {
+      "epoch": 0.3007595976185588,
+      "grad_norm": 0.7775212788676693,
+      "learning_rate": 9.45305449865557e-06,
+      "loss": 0.4038,
+      "num_input_tokens_seen": 80743880,
+      "step": 1465,
+      "train_runtime": 8833.0661,
+      "train_tokens_per_second": 9141.093
+    },
+    {
+      "epoch": 0.3017860808868815,
+      "grad_norm": 0.803119285146583,
+      "learning_rate": 9.449382395960895e-06,
+      "loss": 0.3692,
+      "num_input_tokens_seen": 81028600,
+      "step": 1470,
+      "train_runtime": 8859.7157,
+      "train_tokens_per_second": 9145.734
+    },
+    {
+      "epoch": 0.30281256415520424,
+      "grad_norm": 0.6858683304604537,
+      "learning_rate": 9.44569872566944e-06,
+      "loss": 0.3921,
+      "num_input_tokens_seen": 81272960,
+      "step": 1475,
+      "train_runtime": 8893.3707,
+      "train_tokens_per_second": 9138.6
+    },
+    {
+      "epoch": 0.303839047423527,
+      "grad_norm": 0.6728874867575027,
+      "learning_rate": 9.442003497358086e-06,
+      "loss": 0.4097,
+      "num_input_tokens_seen": 81548824,
+      "step": 1480,
+      "train_runtime": 8929.7779,
+      "train_tokens_per_second": 9132.234
+    },
+    {
+      "epoch": 0.30486553069184974,
+      "grad_norm": 0.778829414014511,
+      "learning_rate": 9.438296720633766e-06,
+      "loss": 0.3846,
+      "num_input_tokens_seen": 81798456,
+      "step": 1485,
+      "train_runtime": 8958.879,
+      "train_tokens_per_second": 9130.434
+    },
+    {
+      "epoch": 0.30589201396017246,
+      "grad_norm": 0.6727939859031058,
+      "learning_rate": 9.434578405133437e-06,
+      "loss": 0.4247,
+      "num_input_tokens_seen": 82080512,
+      "step": 1490,
+      "train_runtime": 8990.9115,
+      "train_tokens_per_second": 9129.276
+    },
+    {
+      "epoch": 0.3069184972284952,
+      "grad_norm": 0.754285636505493,
+      "learning_rate": 9.43084856052405e-06,
+      "loss": 0.373,
+      "num_input_tokens_seen": 82363472,
+      "step": 1495,
+      "train_runtime": 9023.896,
+      "train_tokens_per_second": 9127.263
+    },
+    {
+      "epoch": 0.3079449804968179,
+      "grad_norm": 0.6993596026153885,
+      "learning_rate": 9.427107196502536e-06,
+      "loss": 0.3618,
+      "num_input_tokens_seen": 82637104,
+      "step": 1500,
+      "train_runtime": 9052.5805,
+      "train_tokens_per_second": 9128.569
+    },
+    {
+      "epoch": 0.3089714637651406,
+      "grad_norm": 0.7566609912434713,
+      "learning_rate": 9.423354322795768e-06,
+      "loss": 0.388,
+      "num_input_tokens_seen": 82914152,
+      "step": 1505,
+      "train_runtime": 9082.3842,
+      "train_tokens_per_second": 9129.117
+    },
+    {
+      "epoch": 0.30999794703346334,
+      "grad_norm": 0.6417189911237462,
+      "learning_rate": 9.419589949160545e-06,
+      "loss": 0.4166,
+      "num_input_tokens_seen": 83192696,
+      "step": 1510,
+      "train_runtime": 9107.2886,
+      "train_tokens_per_second": 9134.738
+    },
+    {
+      "epoch": 0.31102443030178606,
+      "grad_norm": 0.7209498977224079,
+      "learning_rate": 9.415814085383566e-06,
+      "loss": 0.3529,
+      "num_input_tokens_seen": 83492416,
+      "step": 1515,
+      "train_runtime": 9136.2158,
+      "train_tokens_per_second": 9138.621
+    },
+    {
+      "epoch": 0.31205091357010883,
+      "grad_norm": 0.6759800286820219,
+      "learning_rate": 9.412026741281397e-06,
+      "loss": 0.3625,
+      "num_input_tokens_seen": 83762448,
+      "step": 1520,
+      "train_runtime": 9167.7987,
+      "train_tokens_per_second": 9136.593
+    },
+    {
+      "epoch": 0.31307739683843155,
+      "grad_norm": 0.775475634970074,
+      "learning_rate": 9.408227926700455e-06,
+      "loss": 0.4068,
+      "num_input_tokens_seen": 83987216,
+      "step": 1525,
+      "train_runtime": 9195.3321,
+      "train_tokens_per_second": 9133.679
+    },
+    {
+      "epoch": 0.3141038801067543,
+      "grad_norm": 0.6363160513989018,
+      "learning_rate": 9.40441765151698e-06,
+      "loss": 0.3683,
+      "num_input_tokens_seen": 84261144,
+      "step": 1530,
+      "train_runtime": 9226.1122,
+      "train_tokens_per_second": 9132.898
+    },
+    {
+      "epoch": 0.315130363375077,
+      "grad_norm": 0.7300444637440023,
+      "learning_rate": 9.400595925636999e-06,
+      "loss": 0.4075,
+      "num_input_tokens_seen": 84547976,
+      "step": 1535,
+      "train_runtime": 9251.0431,
+      "train_tokens_per_second": 9139.291
+    },
+    {
+      "epoch": 0.3161568466433997,
+      "grad_norm": 0.6455534246860312,
+      "learning_rate": 9.39676275899632e-06,
+      "loss": 0.4,
+      "num_input_tokens_seen": 84821000,
+      "step": 1540,
+      "train_runtime": 9281.5628,
+      "train_tokens_per_second": 9138.655
+    },
+    {
+      "epoch": 0.31718332991172243,
+      "grad_norm": 0.9033058358758117,
+      "learning_rate": 9.392918161560485e-06,
+      "loss": 0.4078,
+      "num_input_tokens_seen": 85117688,
+      "step": 1545,
+      "train_runtime": 9313.8275,
+      "train_tokens_per_second": 9138.852
+    },
+    {
+      "epoch": 0.31820981318004515,
+      "grad_norm": 0.6731679864479334,
+      "learning_rate": 9.38906214332476e-06,
+      "loss": 0.4009,
+      "num_input_tokens_seen": 85393312,
+      "step": 1550,
+      "train_runtime": 9347.6347,
+      "train_tokens_per_second": 9135.286
+    },
+    {
+      "epoch": 0.3192362964483679,
+      "grad_norm": 0.7159486156061052,
+      "learning_rate": 9.385194714314102e-06,
+      "loss": 0.3819,
+      "num_input_tokens_seen": 85661408,
+      "step": 1555,
+      "train_runtime": 9379.6072,
+      "train_tokens_per_second": 9132.729
+    },
+    {
+      "epoch": 0.3202627797166906,
+      "grad_norm": 0.7372561853201023,
+      "learning_rate": 9.381315884583133e-06,
+      "loss": 0.4371,
+      "num_input_tokens_seen": 85937768,
+      "step": 1560,
+      "train_runtime": 9414.3587,
+      "train_tokens_per_second": 9128.372
+    },
+    {
+      "epoch": 0.32128926298501337,
+      "grad_norm": 0.7213644453208212,
+      "learning_rate": 9.377425664216115e-06,
+      "loss": 0.3568,
+      "num_input_tokens_seen": 86210264,
+      "step": 1565,
+      "train_runtime": 9441.1786,
+      "train_tokens_per_second": 9131.303
+    },
+    {
+      "epoch": 0.3223157462533361,
+      "grad_norm": 0.7147884012172306,
+      "learning_rate": 9.373524063326922e-06,
+      "loss": 0.3871,
+      "num_input_tokens_seen": 86498016,
+      "step": 1570,
+      "train_runtime": 9476.59,
+      "train_tokens_per_second": 9127.547
+    },
+    {
+      "epoch": 0.3233422295216588,
+      "grad_norm": 0.7883056886951613,
+      "learning_rate": 9.36961109205902e-06,
+      "loss": 0.3624,
+      "num_input_tokens_seen": 86779352,
+      "step": 1575,
+      "train_runtime": 9512.3742,
+      "train_tokens_per_second": 9122.786
+    },
+    {
+      "epoch": 0.32436871278998153,
+      "grad_norm": 0.7367925417922598,
+      "learning_rate": 9.365686760585431e-06,
+      "loss": 0.3487,
+      "num_input_tokens_seen": 87051968,
+      "step": 1580,
+      "train_runtime": 9542.721,
+      "train_tokens_per_second": 9122.342
+    },
+    {
+      "epoch": 0.32539519605830425,
+      "grad_norm": 0.6796889180634916,
+      "learning_rate": 9.361751079108713e-06,
+      "loss": 0.3681,
+      "num_input_tokens_seen": 87316832,
+      "step": 1585,
+      "train_runtime": 9576.0988,
+      "train_tokens_per_second": 9118.205
+    },
+    {
+      "epoch": 0.32642167932662697,
+      "grad_norm": 0.646910161395551,
+      "learning_rate": 9.357804057860934e-06,
+      "loss": 0.3944,
+      "num_input_tokens_seen": 87621136,
+      "step": 1590,
+      "train_runtime": 9599.9018,
+      "train_tokens_per_second": 9127.295
+    },
+    {
+      "epoch": 0.3274481625949497,
+      "grad_norm": 0.7737087042798728,
+      "learning_rate": 9.353845707103641e-06,
+      "loss": 0.3944,
+      "num_input_tokens_seen": 87897104,
+      "step": 1595,
+      "train_runtime": 9632.6197,
+      "train_tokens_per_second": 9124.943
+    },
+    {
+      "epoch": 0.3284746458632724,
+      "grad_norm": 0.756554964703144,
+      "learning_rate": 9.349876037127837e-06,
+      "loss": 0.3953,
+      "num_input_tokens_seen": 88180712,
+      "step": 1600,
+      "train_runtime": 9664.5636,
+      "train_tokens_per_second": 9124.128
+    },
+    {
+      "epoch": 0.3295011291315951,
+      "grad_norm": 0.7956259976426554,
+      "learning_rate": 9.345895058253952e-06,
+      "loss": 0.4261,
+      "num_input_tokens_seen": 88466008,
+      "step": 1605,
+      "train_runtime": 9694.7151,
+      "train_tokens_per_second": 9125.179
+    },
+    {
+      "epoch": 0.3305276123999179,
+      "grad_norm": 0.7273514259296426,
+      "learning_rate": 9.341902780831817e-06,
+      "loss": 0.3736,
+      "num_input_tokens_seen": 88765776,
+      "step": 1610,
+      "train_runtime": 9723.0666,
+      "train_tokens_per_second": 9129.401
+    },
+    {
+      "epoch": 0.3315540956682406,
+      "grad_norm": 0.6919641684812229,
+      "learning_rate": 9.33789921524064e-06,
+      "loss": 0.3809,
+      "num_input_tokens_seen": 89054128,
+      "step": 1615,
+      "train_runtime": 9745.0759,
+      "train_tokens_per_second": 9138.372
+    },
+    {
+      "epoch": 0.33258057893656334,
+      "grad_norm": 0.7591883061546578,
+      "learning_rate": 9.333884371888972e-06,
+      "loss": 0.4149,
+      "num_input_tokens_seen": 89323440,
+      "step": 1620,
+      "train_runtime": 9771.8864,
+      "train_tokens_per_second": 9140.859
+    },
+    {
+      "epoch": 0.33360706220488606,
+      "grad_norm": 0.6506880126712047,
+      "learning_rate": 9.329858261214687e-06,
+      "loss": 0.3681,
+      "num_input_tokens_seen": 89619168,
+      "step": 1625,
+      "train_runtime": 9805.3406,
+      "train_tokens_per_second": 9139.832
+    },
+    {
+      "epoch": 0.3346335454732088,
+      "grad_norm": 0.78588655441044,
+      "learning_rate": 9.32582089368495e-06,
+      "loss": 0.3953,
+      "num_input_tokens_seen": 89895968,
+      "step": 1630,
+      "train_runtime": 9839.0242,
+      "train_tokens_per_second": 9136.675
+    },
+    {
+      "epoch": 0.3356600287415315,
+      "grad_norm": 0.755907337427022,
+      "learning_rate": 9.321772279796193e-06,
+      "loss": 0.3997,
+      "num_input_tokens_seen": 90172056,
+      "step": 1635,
+      "train_runtime": 9869.5803,
+      "train_tokens_per_second": 9136.362
+    },
+    {
+      "epoch": 0.3366865120098542,
+      "grad_norm": 0.6349990976517407,
+      "learning_rate": 9.31771243007409e-06,
+      "loss": 0.3724,
+      "num_input_tokens_seen": 90454048,
+      "step": 1640,
+      "train_runtime": 9901.8414,
+      "train_tokens_per_second": 9135.073
+    },
+    {
+      "epoch": 0.33771299527817694,
+      "grad_norm": 0.7188332797819461,
+      "learning_rate": 9.313641355073517e-06,
+      "loss": 0.3843,
+      "num_input_tokens_seen": 90722792,
+      "step": 1645,
+      "train_runtime": 9934.029,
+      "train_tokens_per_second": 9132.527
+    },
+    {
+      "epoch": 0.3387394785464997,
+      "grad_norm": 0.6975325720326289,
+      "learning_rate": 9.309559065378544e-06,
+      "loss": 0.3426,
+      "num_input_tokens_seen": 91006744,
+      "step": 1650,
+      "train_runtime": 9964.8734,
+      "train_tokens_per_second": 9132.755
+    },
+    {
+      "epoch": 0.33976596181482244,
+      "grad_norm": 0.7438484973135873,
+      "learning_rate": 9.30546557160239e-06,
+      "loss": 0.3952,
+      "num_input_tokens_seen": 91271056,
+      "step": 1655,
+      "train_runtime": 9992.6121,
+      "train_tokens_per_second": 9133.854
+    },
+    {
+      "epoch": 0.34079244508314516,
+      "grad_norm": 1.039963636798865,
+      "learning_rate": 9.301360884387408e-06,
+      "loss": 0.3837,
+      "num_input_tokens_seen": 91540880,
+      "step": 1660,
+      "train_runtime": 10024.7734,
+      "train_tokens_per_second": 9131.466
+    },
+    {
+      "epoch": 0.3418189283514679,
+      "grad_norm": 0.6335454657376006,
+      "learning_rate": 9.297245014405045e-06,
+      "loss": 0.3988,
+      "num_input_tokens_seen": 91792464,
+      "step": 1665,
+      "train_runtime": 10058.0913,
+      "train_tokens_per_second": 9126.231
+    },
+    {
+      "epoch": 0.3428454116197906,
+      "grad_norm": 0.8294348660723462,
+      "learning_rate": 9.293117972355828e-06,
+      "loss": 0.3797,
+      "num_input_tokens_seen": 92061096,
+      "step": 1670,
+      "train_runtime": 10089.5712,
+      "train_tokens_per_second": 9124.381
+    },
+    {
+      "epoch": 0.3438718948881133,
+      "grad_norm": 0.7651791719653793,
+      "learning_rate": 9.288979768969327e-06,
+      "loss": 0.3788,
+      "num_input_tokens_seen": 92333200,
+      "step": 1675,
+      "train_runtime": 10117.5439,
+      "train_tokens_per_second": 9126.049
+    },
+    {
+      "epoch": 0.34489837815643604,
+      "grad_norm": 0.699632632585877,
+      "learning_rate": 9.284830415004126e-06,
+      "loss": 0.375,
+      "num_input_tokens_seen": 92606848,
+      "step": 1680,
+      "train_runtime": 10146.5597,
+      "train_tokens_per_second": 9126.921
+    },
+    {
+      "epoch": 0.34592486142475876,
+      "grad_norm": 0.6626066704925232,
+      "learning_rate": 9.280669921247805e-06,
+      "loss": 0.3738,
+      "num_input_tokens_seen": 92888392,
+      "step": 1685,
+      "train_runtime": 10174.0837,
+      "train_tokens_per_second": 9129.903
+    },
+    {
+      "epoch": 0.3469513446930815,
+      "grad_norm": 0.6900520261280432,
+      "learning_rate": 9.2764982985169e-06,
+      "loss": 0.3922,
+      "num_input_tokens_seen": 93172112,
+      "step": 1690,
+      "train_runtime": 10202.4975,
+      "train_tokens_per_second": 9132.285
+    },
+    {
+      "epoch": 0.34797782796140425,
+      "grad_norm": 0.7093711010428975,
+      "learning_rate": 9.272315557656884e-06,
+      "loss": 0.418,
+      "num_input_tokens_seen": 93429968,
+      "step": 1695,
+      "train_runtime": 10232.9952,
+      "train_tokens_per_second": 9130.266
+    },
+    {
+      "epoch": 0.34900431122972697,
+      "grad_norm": 0.7198571602212869,
+      "learning_rate": 9.268121709542133e-06,
+      "loss": 0.3772,
+      "num_input_tokens_seen": 93705920,
+      "step": 1700,
+      "train_runtime": 10262.9081,
+      "train_tokens_per_second": 9130.543
+    },
+    {
+      "epoch": 0.3500307944980497,
+      "grad_norm": 0.769077576171908,
+      "learning_rate": 9.2639167650759e-06,
+      "loss": 0.3931,
+      "num_input_tokens_seen": 94001848,
+      "step": 1705,
+      "train_runtime": 10294.8313,
+      "train_tokens_per_second": 9130.975
+    },
+    {
+      "epoch": 0.3510572777663724,
+      "grad_norm": 0.730683787117151,
+      "learning_rate": 9.259700735190287e-06,
+      "loss": 0.4019,
+      "num_input_tokens_seen": 94272672,
+      "step": 1710,
+      "train_runtime": 10326.8303,
+      "train_tokens_per_second": 9128.907
+    },
+    {
+      "epoch": 0.35208376103469513,
+      "grad_norm": 0.6780810563453115,
+      "learning_rate": 9.255473630846218e-06,
+      "loss": 0.362,
+      "num_input_tokens_seen": 94562880,
+      "step": 1715,
+      "train_runtime": 10358.7999,
+      "train_tokens_per_second": 9128.749
+    },
+    {
+      "epoch": 0.35311024430301785,
+      "grad_norm": 0.6143517037830429,
+      "learning_rate": 9.251235463033403e-06,
+      "loss": 0.374,
+      "num_input_tokens_seen": 94857288,
+      "step": 1720,
+      "train_runtime": 10387.3195,
+      "train_tokens_per_second": 9132.028
+    },
+    {
+      "epoch": 0.35413672757134057,
+      "grad_norm": 0.76596886360452,
+      "learning_rate": 9.246986242770324e-06,
+      "loss": 0.3755,
+      "num_input_tokens_seen": 95122568,
+      "step": 1725,
+      "train_runtime": 10417.2451,
+      "train_tokens_per_second": 9131.26
+    },
+    {
+      "epoch": 0.3551632108396633,
+      "grad_norm": 0.6293394878963015,
+      "learning_rate": 9.242725981104186e-06,
+      "loss": 0.3609,
+      "num_input_tokens_seen": 95407816,
+      "step": 1730,
+      "train_runtime": 10450.9044,
+      "train_tokens_per_second": 9129.144
+    },
+    {
+      "epoch": 0.35618969410798607,
+      "grad_norm": 0.7010186808665287,
+      "learning_rate": 9.238454689110909e-06,
+      "loss": 0.4095,
+      "num_input_tokens_seen": 95702056,
+      "step": 1735,
+      "train_runtime": 10485.094,
+      "train_tokens_per_second": 9127.439
+    },
+    {
+      "epoch": 0.3572161773763088,
+      "grad_norm": 0.6834484863201761,
+      "learning_rate": 9.234172377895085e-06,
+      "loss": 0.4068,
+      "num_input_tokens_seen": 95975016,
+      "step": 1740,
+      "train_runtime": 10513.7406,
+      "train_tokens_per_second": 9128.532
+    },
+    {
+      "epoch": 0.3582426606446315,
+      "grad_norm": 0.7042043668607236,
+      "learning_rate": 9.229879058589959e-06,
+      "loss": 0.3663,
+      "num_input_tokens_seen": 96251688,
+      "step": 1745,
+      "train_runtime": 10537.7187,
+      "train_tokens_per_second": 9134.016
+    },
+    {
+      "epoch": 0.3592691439129542,
+      "grad_norm": 0.7968747578888067,
+      "learning_rate": 9.22557474235739e-06,
+      "loss": 0.357,
+      "num_input_tokens_seen": 96535208,
+      "step": 1750,
+      "train_runtime": 10575.2949,
+      "train_tokens_per_second": 9128.37
+    },
+    {
+      "epoch": 0.36029562718127695,
+      "grad_norm": 0.6866654356823125,
+      "learning_rate": 9.221259440387827e-06,
+      "loss": 0.3711,
+      "num_input_tokens_seen": 96819032,
+      "step": 1755,
+      "train_runtime": 10608.5906,
+      "train_tokens_per_second": 9126.475
+    },
+    {
+      "epoch": 0.36132211044959966,
+      "grad_norm": 0.7452354657905673,
+      "learning_rate": 9.216933163900284e-06,
+      "loss": 0.3872,
+      "num_input_tokens_seen": 97101776,
+      "step": 1760,
+      "train_runtime": 10640.0116,
+      "train_tokens_per_second": 9126.097
+    },
+    {
+      "epoch": 0.3623485937179224,
+      "grad_norm": 1.1208553163271142,
+      "learning_rate": 9.212595924142304e-06,
+      "loss": 0.3871,
+      "num_input_tokens_seen": 97388728,
+      "step": 1765,
+      "train_runtime": 10675.3838,
+      "train_tokens_per_second": 9122.738
+    },
+    {
+      "epoch": 0.3633750769862451,
+      "grad_norm": 0.6768177781212251,
+      "learning_rate": 9.208247732389933e-06,
+      "loss": 0.371,
+      "num_input_tokens_seen": 97656424,
+      "step": 1770,
+      "train_runtime": 10710.7499,
+      "train_tokens_per_second": 9117.608
+    },
+    {
+      "epoch": 0.3644015602545678,
+      "grad_norm": 0.8140729995757358,
+      "learning_rate": 9.203888599947687e-06,
+      "loss": 0.3718,
+      "num_input_tokens_seen": 97927272,
+      "step": 1775,
+      "train_runtime": 10742.4195,
+      "train_tokens_per_second": 9115.942
+    },
+    {
+      "epoch": 0.3654280435228906,
+      "grad_norm": 0.8148694525094718,
+      "learning_rate": 9.199518538148534e-06,
+      "loss": 0.3715,
+      "num_input_tokens_seen": 98195560,
+      "step": 1780,
+      "train_runtime": 10763.1177,
+      "train_tokens_per_second": 9123.338
+    },
+    {
+      "epoch": 0.3664545267912133,
+      "grad_norm": 0.8636781826508881,
+      "learning_rate": 9.19513755835385e-06,
+      "loss": 0.3847,
+      "num_input_tokens_seen": 98455248,
+      "step": 1785,
+      "train_runtime": 10798.6814,
+      "train_tokens_per_second": 9117.34
+    },
+    {
+      "epoch": 0.36748101005953604,
+      "grad_norm": 0.7100805031832539,
+      "learning_rate": 9.190745671953397e-06,
+      "loss": 0.3821,
+      "num_input_tokens_seen": 98720968,
+      "step": 1790,
+      "train_runtime": 10828.5057,
+      "train_tokens_per_second": 9116.767
+    },
+    {
+      "epoch": 0.36850749332785876,
+      "grad_norm": 0.6822598864964778,
+      "learning_rate": 9.18634289036529e-06,
+      "loss": 0.3864,
+      "num_input_tokens_seen": 98995776,
+      "step": 1795,
+      "train_runtime": 10857.5131,
+      "train_tokens_per_second": 9117.721
+    },
+    {
+      "epoch": 0.3695339765961815,
+      "grad_norm": 0.6788257486077849,
+      "learning_rate": 9.181929225035975e-06,
+      "loss": 0.3606,
+      "num_input_tokens_seen": 99250240,
+      "step": 1800,
+      "train_runtime": 10889.5728,
+      "train_tokens_per_second": 9114.245
+    },
+    {
+      "epoch": 0.3705604598645042,
+      "grad_norm": 0.7155616573459661,
+      "learning_rate": 9.17750468744019e-06,
+      "loss": 0.4084,
+      "num_input_tokens_seen": 99534592,
+      "step": 1805,
+      "train_runtime": 10920.5366,
+      "train_tokens_per_second": 9114.441
+    },
+    {
+      "epoch": 0.3715869431328269,
+      "grad_norm": 0.9085457429545923,
+      "learning_rate": 9.173069289080941e-06,
+      "loss": 0.4159,
+      "num_input_tokens_seen": 99808488,
+      "step": 1810,
+      "train_runtime": 10949.7627,
+      "train_tokens_per_second": 9115.128
+    },
+    {
+      "epoch": 0.37261342640114964,
+      "grad_norm": 0.6518870783926743,
+      "learning_rate": 9.168623041489466e-06,
+      "loss": 0.3281,
+      "num_input_tokens_seen": 100092744,
+      "step": 1815,
+      "train_runtime": 10978.2467,
+      "train_tokens_per_second": 9117.371
+    },
+    {
+      "epoch": 0.3736399096694724,
+      "grad_norm": 0.6653225988804985,
+      "learning_rate": 9.164165956225213e-06,
+      "loss": 0.3644,
+      "num_input_tokens_seen": 100369960,
+      "step": 1820,
+      "train_runtime": 11006.8805,
+      "train_tokens_per_second": 9118.838
+    },
+    {
+      "epoch": 0.37466639293779513,
+      "grad_norm": 0.6017606637053066,
+      "learning_rate": 9.159698044875806e-06,
+      "loss": 0.3722,
+      "num_input_tokens_seen": 100654448,
+      "step": 1825,
+      "train_runtime": 11037.7869,
+      "train_tokens_per_second": 9119.079
+    },
+    {
+      "epoch": 0.37569287620611785,
+      "grad_norm": 0.6442533960193247,
+      "learning_rate": 9.155219319057013e-06,
+      "loss": 0.3816,
+      "num_input_tokens_seen": 100901408,
+      "step": 1830,
+      "train_runtime": 11066.4786,
+      "train_tokens_per_second": 9117.752
+    },
+    {
+      "epoch": 0.3767193594744406,
+      "grad_norm": 0.6605147916767188,
+      "learning_rate": 9.15072979041272e-06,
+      "loss": 0.3876,
+      "num_input_tokens_seen": 101177256,
+      "step": 1835,
+      "train_runtime": 11097.1198,
+      "train_tokens_per_second": 9117.434
+    },
+    {
+      "epoch": 0.3777458427427633,
+      "grad_norm": 0.7350131537794723,
+      "learning_rate": 9.146229470614894e-06,
+      "loss": 0.4001,
+      "num_input_tokens_seen": 101437712,
+      "step": 1840,
+      "train_runtime": 11125.3012,
+      "train_tokens_per_second": 9117.75
+    },
+    {
+      "epoch": 0.378772326011086,
+      "grad_norm": 0.7242982831071183,
+      "learning_rate": 9.14171837136356e-06,
+      "loss": 0.3627,
+      "num_input_tokens_seen": 101709136,
+      "step": 1845,
+      "train_runtime": 11156.3307,
+      "train_tokens_per_second": 9116.719
+    },
+    {
+      "epoch": 0.37979880927940873,
+      "grad_norm": 0.6414714603143467,
+      "learning_rate": 9.13719650438677e-06,
+      "loss": 0.3642,
+      "num_input_tokens_seen": 102007080,
+      "step": 1850,
+      "train_runtime": 11181.7088,
+      "train_tokens_per_second": 9122.674
+    },
+    {
+      "epoch": 0.38082529254773145,
+      "grad_norm": 0.6871836305369112,
+      "learning_rate": 9.132663881440566e-06,
+      "loss": 0.3622,
+      "num_input_tokens_seen": 102270264,
+      "step": 1855,
+      "train_runtime": 11210.1314,
+      "train_tokens_per_second": 9123.021
+    },
+    {
+      "epoch": 0.3818517758160542,
+      "grad_norm": 0.650737982206852,
+      "learning_rate": 9.128120514308957e-06,
+      "loss": 0.3514,
+      "num_input_tokens_seen": 102552472,
+      "step": 1860,
+      "train_runtime": 11241.2193,
+      "train_tokens_per_second": 9122.896
+    },
+    {
+      "epoch": 0.38287825908437695,
+      "grad_norm": 0.6688883700163434,
+      "learning_rate": 9.12356641480388e-06,
+      "loss": 0.3878,
+      "num_input_tokens_seen": 102813560,
+      "step": 1865,
+      "train_runtime": 11277.5579,
+      "train_tokens_per_second": 9116.651
+    },
+    {
+      "epoch": 0.38390474235269967,
+      "grad_norm": 0.6511698642020315,
+      "learning_rate": 9.11900159476518e-06,
+      "loss": 0.3678,
+      "num_input_tokens_seen": 103103200,
+      "step": 1870,
+      "train_runtime": 11316.0174,
+      "train_tokens_per_second": 9111.262
+    },
+    {
+      "epoch": 0.3849312256210224,
+      "grad_norm": 0.6598238431992423,
+      "learning_rate": 9.11442606606057e-06,
+      "loss": 0.3519,
+      "num_input_tokens_seen": 103386392,
+      "step": 1875,
+      "train_runtime": 11344.1918,
+      "train_tokens_per_second": 9113.597
+    },
+    {
+      "epoch": 0.3859577088893451,
+      "grad_norm": 0.8097931980347263,
+      "learning_rate": 9.109839840585606e-06,
+      "loss": 0.3841,
+      "num_input_tokens_seen": 103659440,
+      "step": 1880,
+      "train_runtime": 11369.578,
+      "train_tokens_per_second": 9117.264
+    },
+    {
+      "epoch": 0.38698419215766783,
+      "grad_norm": 0.7028203109056594,
+      "learning_rate": 9.105242930263652e-06,
+      "loss": 0.3631,
+      "num_input_tokens_seen": 103923216,
+      "step": 1885,
+      "train_runtime": 11406.1796,
+      "train_tokens_per_second": 9111.133
+    },
+    {
+      "epoch": 0.38801067542599055,
+      "grad_norm": 0.7771112518335765,
+      "learning_rate": 9.100635347045848e-06,
+      "loss": 0.3476,
+      "num_input_tokens_seen": 104184752,
+      "step": 1890,
+      "train_runtime": 11441.4564,
+      "train_tokens_per_second": 9105.899
+    },
+    {
+      "epoch": 0.38903715869431327,
+      "grad_norm": 0.7841024274435517,
+      "learning_rate": 9.096017102911088e-06,
+      "loss": 0.3708,
+      "num_input_tokens_seen": 104462584,
+      "step": 1895,
+      "train_runtime": 11468.0522,
+      "train_tokens_per_second": 9109.008
+    },
+    {
+      "epoch": 0.390063641962636,
+      "grad_norm": 0.6586781518222155,
+      "learning_rate": 9.091388209865977e-06,
+      "loss": 0.3925,
+      "num_input_tokens_seen": 104753208,
+      "step": 1900,
+      "train_runtime": 11493.8659,
+      "train_tokens_per_second": 9113.836
+    },
+    {
+      "epoch": 0.39109012523095876,
+      "grad_norm": 0.6740147829736638,
+      "learning_rate": 9.086748679944807e-06,
+      "loss": 0.3763,
+      "num_input_tokens_seen": 105045664,
+      "step": 1905,
+      "train_runtime": 11523.1558,
+      "train_tokens_per_second": 9116.05
+    },
+    {
+      "epoch": 0.3921166084992815,
+      "grad_norm": 0.7487014795695699,
+      "learning_rate": 9.082098525209527e-06,
+      "loss": 0.3929,
+      "num_input_tokens_seen": 105321368,
+      "step": 1910,
+      "train_runtime": 11556.5724,
+      "train_tokens_per_second": 9113.547
+    },
+    {
+      "epoch": 0.3931430917676042,
+      "grad_norm": 0.6596030329378766,
+      "learning_rate": 9.077437757749703e-06,
+      "loss": 0.3732,
+      "num_input_tokens_seen": 105613880,
+      "step": 1915,
+      "train_runtime": 11581.9078,
+      "train_tokens_per_second": 9118.867
+    },
+    {
+      "epoch": 0.3941695750359269,
+      "grad_norm": 0.7185061809660211,
+      "learning_rate": 9.072766389682497e-06,
+      "loss": 0.3402,
+      "num_input_tokens_seen": 105889896,
+      "step": 1920,
+      "train_runtime": 11609.565,
+      "train_tokens_per_second": 9120.918
+    },
+    {
+      "epoch": 0.39519605830424964,
+      "grad_norm": 0.685601571525062,
+      "learning_rate": 9.068084433152629e-06,
+      "loss": 0.38,
+      "num_input_tokens_seen": 106149496,
+      "step": 1925,
+      "train_runtime": 11633.8111,
+      "train_tokens_per_second": 9124.224
+    },
+    {
+      "epoch": 0.39622254157257236,
+      "grad_norm": 0.6974978218570849,
+      "learning_rate": 9.063391900332345e-06,
+      "loss": 0.4144,
+      "num_input_tokens_seen": 106440488,
+      "step": 1930,
+      "train_runtime": 11669.3611,
+      "train_tokens_per_second": 9121.364
+    },
+    {
+      "epoch": 0.3972490248408951,
+      "grad_norm": 0.6401337672602817,
+      "learning_rate": 9.05868880342139e-06,
+      "loss": 0.3994,
+      "num_input_tokens_seen": 106718288,
+      "step": 1935,
+      "train_runtime": 11706.4258,
+      "train_tokens_per_second": 9116.214
+    },
+    {
+      "epoch": 0.3982755081092178,
+      "grad_norm": 0.7816071161092908,
+      "learning_rate": 9.053975154646973e-06,
+      "loss": 0.3832,
+      "num_input_tokens_seen": 106981680,
+      "step": 1940,
+      "train_runtime": 11734.1368,
+      "train_tokens_per_second": 9117.132
+    },
+    {
+      "epoch": 0.3993019913775405,
+      "grad_norm": 0.6587330261359072,
+      "learning_rate": 9.049250966263736e-06,
+      "loss": 0.3693,
+      "num_input_tokens_seen": 107293864,
+      "step": 1945,
+      "train_runtime": 11764.4615,
+      "train_tokens_per_second": 9120.168
+    },
+    {
+      "epoch": 0.4003284746458633,
+      "grad_norm": 0.7011459665202934,
+      "learning_rate": 9.044516250553723e-06,
+      "loss": 0.3761,
+      "num_input_tokens_seen": 107556880,
+      "step": 1950,
+      "train_runtime": 11792.7217,
+      "train_tokens_per_second": 9120.615
+    },
+    {
+      "epoch": 0.401354957914186,
+      "grad_norm": 0.6300752980009519,
+      "learning_rate": 9.039771019826344e-06,
+      "loss": 0.3785,
+      "num_input_tokens_seen": 107841080,
+      "step": 1955,
+      "train_runtime": 11819.7424,
+      "train_tokens_per_second": 9123.81
+    },
+    {
+      "epoch": 0.40238144118250874,
+      "grad_norm": 0.684637099301222,
+      "learning_rate": 9.035015286418352e-06,
+      "loss": 0.3536,
+      "num_input_tokens_seen": 108126376,
+      "step": 1960,
+      "train_runtime": 11853.5621,
+      "train_tokens_per_second": 9121.847
+    },
+    {
+      "epoch": 0.40340792445083146,
+      "grad_norm": 0.568639844611764,
+      "learning_rate": 9.030249062693799e-06,
+      "loss": 0.3593,
+      "num_input_tokens_seen": 108410568,
+      "step": 1965,
+      "train_runtime": 11876.7,
+      "train_tokens_per_second": 9128.004
+    },
+    {
+      "epoch": 0.4044344077191542,
+      "grad_norm": 0.6536315878664977,
+      "learning_rate": 9.025472361044011e-06,
+      "loss": 0.329,
+      "num_input_tokens_seen": 108697000,
+      "step": 1970,
+      "train_runtime": 11911.2842,
+      "train_tokens_per_second": 9125.548
+    },
+    {
+      "epoch": 0.4054608909874769,
+      "grad_norm": 0.7102246225015607,
+      "learning_rate": 9.02068519388756e-06,
+      "loss": 0.3962,
+      "num_input_tokens_seen": 108981792,
+      "step": 1975,
+      "train_runtime": 11942.7537,
+      "train_tokens_per_second": 9125.349
+    },
+    {
+      "epoch": 0.4064873742557996,
+      "grad_norm": 0.6196101434547192,
+      "learning_rate": 9.015887573670222e-06,
+      "loss": 0.3612,
+      "num_input_tokens_seen": 109281464,
+      "step": 1980,
+      "train_runtime": 11975.5601,
+      "train_tokens_per_second": 9125.374
+    },
+    {
+      "epoch": 0.40751385752412234,
+      "grad_norm": 0.8077423576282892,
+      "learning_rate": 9.011079512864945e-06,
+      "loss": 0.4048,
+      "num_input_tokens_seen": 109566360,
+      "step": 1985,
+      "train_runtime": 12007.2715,
+      "train_tokens_per_second": 9125.001
+    },
+    {
+      "epoch": 0.40854034079244506,
+      "grad_norm": 0.7259244923572262,
+      "learning_rate": 9.006261023971832e-06,
+      "loss": 0.3988,
+      "num_input_tokens_seen": 109851168,
+      "step": 1990,
+      "train_runtime": 12039.667,
+      "train_tokens_per_second": 9124.104
+    },
+    {
+      "epoch": 0.40956682406076783,
+      "grad_norm": 0.6883871482222504,
+      "learning_rate": 9.001432119518086e-06,
+      "loss": 0.3597,
+      "num_input_tokens_seen": 110109216,
+      "step": 1995,
+      "train_runtime": 12067.659,
+      "train_tokens_per_second": 9124.323
+    },
+    {
+      "epoch": 0.41059330732909055,
+      "grad_norm": 0.6010279472760849,
+      "learning_rate": 8.996592812057992e-06,
+      "loss": 0.378,
+      "num_input_tokens_seen": 110391344,
+      "step": 2000,
+      "train_runtime": 12101.8381,
+      "train_tokens_per_second": 9121.866
+    },
+    {
+      "epoch": 0.41161979059741327,
+      "grad_norm": 0.671743852461452,
+      "learning_rate": 8.991743114172885e-06,
+      "loss": 0.3424,
+      "num_input_tokens_seen": 110667896,
+      "step": 2005,
+      "train_runtime": 12132.1531,
+      "train_tokens_per_second": 9121.868
+    },
+    {
+      "epoch": 0.412646273865736,
+      "grad_norm": 0.6260613893696378,
+      "learning_rate": 8.986883038471112e-06,
+      "loss": 0.384,
+      "num_input_tokens_seen": 110923032,
+      "step": 2010,
+      "train_runtime": 12168.9889,
+      "train_tokens_per_second": 9115.222
+    },
+    {
+      "epoch": 0.4136727571340587,
+      "grad_norm": 0.6287572804342813,
+      "learning_rate": 8.982012597587996e-06,
+      "loss": 0.3459,
+      "num_input_tokens_seen": 111230128,
+      "step": 2015,
+      "train_runtime": 12191.9513,
+      "train_tokens_per_second": 9123.242
+    },
+    {
+      "epoch": 0.41469924040238143,
+      "grad_norm": 0.7911887140036147,
+      "learning_rate": 8.977131804185812e-06,
+      "loss": 0.3821,
+      "num_input_tokens_seen": 111487816,
+      "step": 2020,
+      "train_runtime": 12222.869,
+      "train_tokens_per_second": 9121.248
+    },
+    {
+      "epoch": 0.41572572367070415,
+      "grad_norm": 0.6192594074093732,
+      "learning_rate": 8.972240670953745e-06,
+      "loss": 0.3523,
+      "num_input_tokens_seen": 111778264,
+      "step": 2025,
+      "train_runtime": 12249.9785,
+      "train_tokens_per_second": 9124.772
+    },
+    {
+      "epoch": 0.41675220693902687,
+      "grad_norm": 0.6492223795269265,
+      "learning_rate": 8.967339210607868e-06,
+      "loss": 0.3754,
+      "num_input_tokens_seen": 112038640,
+      "step": 2030,
+      "train_runtime": 12282.3561,
+      "train_tokens_per_second": 9121.918
+    },
+    {
+      "epoch": 0.41777869020734965,
+      "grad_norm": 0.6432229124904583,
+      "learning_rate": 8.962427435891096e-06,
+      "loss": 0.3533,
+      "num_input_tokens_seen": 112327240,
+      "step": 2035,
+      "train_runtime": 12314.7522,
+      "train_tokens_per_second": 9121.356
+    },
+    {
+      "epoch": 0.41880517347567237,
+      "grad_norm": 0.6492594106581043,
+      "learning_rate": 8.957505359573168e-06,
+      "loss": 0.359,
+      "num_input_tokens_seen": 112629680,
+      "step": 2040,
+      "train_runtime": 12344.7826,
+      "train_tokens_per_second": 9123.667
+    },
+    {
+      "epoch": 0.4198316567439951,
+      "grad_norm": 0.7394374455644706,
+      "learning_rate": 8.952572994450595e-06,
+      "loss": 0.4055,
+      "num_input_tokens_seen": 112886768,
+      "step": 2045,
+      "train_runtime": 12377.1642,
+      "train_tokens_per_second": 9120.568
+    },
+    {
+      "epoch": 0.4208581400123178,
+      "grad_norm": 0.8118040446062235,
+      "learning_rate": 8.947630353346646e-06,
+      "loss": 0.3695,
+      "num_input_tokens_seen": 113165448,
+      "step": 2050,
+      "train_runtime": 12408.0385,
+      "train_tokens_per_second": 9120.333
+    },
+    {
+      "epoch": 0.4218846232806405,
+      "grad_norm": 0.6095877840113083,
+      "learning_rate": 8.9426774491113e-06,
+      "loss": 0.3782,
+      "num_input_tokens_seen": 113454864,
+      "step": 2055,
+      "train_runtime": 12441.9672,
+      "train_tokens_per_second": 9118.724
+    },
+    {
+      "epoch": 0.42291110654896324,
+      "grad_norm": 0.6475812143434229,
+      "learning_rate": 8.937714294621221e-06,
+      "loss": 0.3477,
+      "num_input_tokens_seen": 113728464,
+      "step": 2060,
+      "train_runtime": 12470.3687,
+      "train_tokens_per_second": 9119.896
+    },
+    {
+      "epoch": 0.42393758981728596,
+      "grad_norm": 0.6161213093099964,
+      "learning_rate": 8.932740902779722e-06,
+      "loss": 0.4016,
+      "num_input_tokens_seen": 113966064,
+      "step": 2065,
+      "train_runtime": 12496.7573,
+      "train_tokens_per_second": 9119.651
+    },
+    {
+      "epoch": 0.4249640730856087,
+      "grad_norm": 0.6260848487018498,
+      "learning_rate": 8.92775728651673e-06,
+      "loss": 0.3685,
+      "num_input_tokens_seen": 114243408,
+      "step": 2070,
+      "train_runtime": 12522.1208,
+      "train_tokens_per_second": 9123.327
+    },
+    {
+      "epoch": 0.4259905563539314,
+      "grad_norm": 0.7169260041833765,
+      "learning_rate": 8.922763458788753e-06,
+      "loss": 0.3798,
+      "num_input_tokens_seen": 114503720,
+      "step": 2075,
+      "train_runtime": 12552.8083,
+      "train_tokens_per_second": 9121.761
+    },
+    {
+      "epoch": 0.4270170396222542,
+      "grad_norm": 0.6272229875853846,
+      "learning_rate": 8.917759432578848e-06,
+      "loss": 0.3715,
+      "num_input_tokens_seen": 114780056,
+      "step": 2080,
+      "train_runtime": 12582.8893,
+      "train_tokens_per_second": 9121.916
+    },
+    {
+      "epoch": 0.4280435228905769,
+      "grad_norm": 0.854691060827857,
+      "learning_rate": 8.912745220896587e-06,
+      "loss": 0.3723,
+      "num_input_tokens_seen": 115023352,
+      "step": 2085,
+      "train_runtime": 12612.7531,
+      "train_tokens_per_second": 9119.607
+    },
+    {
+      "epoch": 0.4290700061588996,
+      "grad_norm": 0.7085698340024562,
+      "learning_rate": 8.907720836778022e-06,
+      "loss": 0.3844,
+      "num_input_tokens_seen": 115252744,
+      "step": 2090,
+      "train_runtime": 12646.1512,
+      "train_tokens_per_second": 9113.662
+    },
+    {
+      "epoch": 0.43009648942722234,
+      "grad_norm": 0.7363420828312812,
+      "learning_rate": 8.902686293285651e-06,
+      "loss": 0.3987,
+      "num_input_tokens_seen": 115525368,
+      "step": 2095,
+      "train_runtime": 12677.3516,
+      "train_tokens_per_second": 9112.737
+    },
+    {
+      "epoch": 0.43112297269554506,
+      "grad_norm": 0.7195386893904244,
+      "learning_rate": 8.897641603508384e-06,
+      "loss": 0.3405,
+      "num_input_tokens_seen": 115788216,
+      "step": 2100,
+      "train_runtime": 12708.0599,
+      "train_tokens_per_second": 9111.4
+    },
+    {
+      "epoch": 0.4321494559638678,
+      "grad_norm": 0.6189931540917261,
+      "learning_rate": 8.892586780561512e-06,
+      "loss": 0.353,
+      "num_input_tokens_seen": 116057456,
+      "step": 2105,
+      "train_runtime": 12734.7952,
+      "train_tokens_per_second": 9113.414
+    },
+    {
+      "epoch": 0.4331759392321905,
+      "grad_norm": 0.8709118729851707,
+      "learning_rate": 8.887521837586668e-06,
+      "loss": 0.3604,
+      "num_input_tokens_seen": 116336008,
+      "step": 2110,
+      "train_runtime": 12759.207,
+      "train_tokens_per_second": 9117.809
+    },
+    {
+      "epoch": 0.4342024225005132,
+      "grad_norm": 0.6184032325854611,
+      "learning_rate": 8.882446787751794e-06,
+      "loss": 0.4022,
+      "num_input_tokens_seen": 116618776,
+      "step": 2115,
+      "train_runtime": 12784.2748,
+      "train_tokens_per_second": 9122.049
+    },
+    {
+      "epoch": 0.435228905768836,
+      "grad_norm": 0.6538064362387122,
+      "learning_rate": 8.877361644251112e-06,
+      "loss": 0.3649,
+      "num_input_tokens_seen": 116892696,
+      "step": 2120,
+      "train_runtime": 12815.4905,
+      "train_tokens_per_second": 9121.203
+    },
+    {
+      "epoch": 0.4362553890371587,
+      "grad_norm": 0.677875689779646,
+      "learning_rate": 8.872266420305084e-06,
+      "loss": 0.3777,
+      "num_input_tokens_seen": 117159048,
+      "step": 2125,
+      "train_runtime": 12844.2824,
+      "train_tokens_per_second": 9121.494
+    },
+    {
+      "epoch": 0.43728187230548143,
+      "grad_norm": 0.7096045494974065,
+      "learning_rate": 8.867161129160378e-06,
+      "loss": 0.3813,
+      "num_input_tokens_seen": 117403016,
+      "step": 2130,
+      "train_runtime": 12864.7685,
+      "train_tokens_per_second": 9125.933
+    },
+    {
+      "epoch": 0.43830835557380415,
+      "grad_norm": 0.7155403416197319,
+      "learning_rate": 8.862045784089836e-06,
+      "loss": 0.3711,
+      "num_input_tokens_seen": 117682400,
+      "step": 2135,
+      "train_runtime": 12896.2334,
+      "train_tokens_per_second": 9125.331
+    },
+    {
+      "epoch": 0.4393348388421269,
+      "grad_norm": 0.5850869501442161,
+      "learning_rate": 8.85692039839244e-06,
+      "loss": 0.359,
+      "num_input_tokens_seen": 117959968,
+      "step": 2140,
+      "train_runtime": 12926.7892,
+      "train_tokens_per_second": 9125.233
+    },
+    {
+      "epoch": 0.4403613221104496,
+      "grad_norm": 0.7132917572169292,
+      "learning_rate": 8.851784985393272e-06,
+      "loss": 0.3455,
+      "num_input_tokens_seen": 118254560,
+      "step": 2145,
+      "train_runtime": 12962.7929,
+      "train_tokens_per_second": 9122.614
+    },
+    {
+      "epoch": 0.4413878053787723,
+      "grad_norm": 0.783810671877599,
+      "learning_rate": 8.846639558443487e-06,
+      "loss": 0.3948,
+      "num_input_tokens_seen": 118509536,
+      "step": 2150,
+      "train_runtime": 12999.7195,
+      "train_tokens_per_second": 9116.315
+    },
+    {
+      "epoch": 0.44241428864709503,
+      "grad_norm": 0.6564780395764842,
+      "learning_rate": 8.84148413092027e-06,
+      "loss": 0.3904,
+      "num_input_tokens_seen": 118790016,
+      "step": 2155,
+      "train_runtime": 13027.3556,
+      "train_tokens_per_second": 9118.506
+    },
+    {
+      "epoch": 0.44344077191541775,
+      "grad_norm": 0.6997192752480172,
+      "learning_rate": 8.836318716226811e-06,
+      "loss": 0.3628,
+      "num_input_tokens_seen": 119070304,
+      "step": 2160,
+      "train_runtime": 13060.7543,
+      "train_tokens_per_second": 9116.648
+    },
+    {
+      "epoch": 0.44446725518374053,
+      "grad_norm": 0.7927124555877658,
+      "learning_rate": 8.831143327792264e-06,
+      "loss": 0.3622,
+      "num_input_tokens_seen": 119308672,
+      "step": 2165,
+      "train_runtime": 13087.4085,
+      "train_tokens_per_second": 9116.295
+    },
+    {
+      "epoch": 0.44549373845206325,
+      "grad_norm": 0.7104177868266475,
+      "learning_rate": 8.825957979071708e-06,
+      "loss": 0.3955,
+      "num_input_tokens_seen": 119561232,
+      "step": 2170,
+      "train_runtime": 13114.426,
+      "train_tokens_per_second": 9116.772
+    },
+    {
+      "epoch": 0.44652022172038597,
+      "grad_norm": 0.6991808174378135,
+      "learning_rate": 8.820762683546122e-06,
+      "loss": 0.4025,
+      "num_input_tokens_seen": 119838176,
+      "step": 2175,
+      "train_runtime": 13145.9653,
+      "train_tokens_per_second": 9115.966
+    },
+    {
+      "epoch": 0.4475467049887087,
+      "grad_norm": 0.6727170682272741,
+      "learning_rate": 8.815557454722346e-06,
+      "loss": 0.357,
+      "num_input_tokens_seen": 120124400,
+      "step": 2180,
+      "train_runtime": 13180.4456,
+      "train_tokens_per_second": 9113.835
+    },
+    {
+      "epoch": 0.4485731882570314,
+      "grad_norm": 0.6413188252575809,
+      "learning_rate": 8.810342306133034e-06,
+      "loss": 0.3646,
+      "num_input_tokens_seen": 120426984,
+      "step": 2185,
+      "train_runtime": 13206.8961,
+      "train_tokens_per_second": 9118.493
+    },
+    {
+      "epoch": 0.4495996715253541,
+      "grad_norm": 0.6792256165110836,
+      "learning_rate": 8.805117251336647e-06,
+      "loss": 0.3707,
+      "num_input_tokens_seen": 120711768,
+      "step": 2190,
+      "train_runtime": 13234.6952,
+      "train_tokens_per_second": 9120.857
+    },
+    {
+      "epoch": 0.45062615479367685,
+      "grad_norm": 0.6459437252734657,
+      "learning_rate": 8.799882303917384e-06,
+      "loss": 0.3687,
+      "num_input_tokens_seen": 120978968,
+      "step": 2195,
+      "train_runtime": 13259.817,
+      "train_tokens_per_second": 9123.728
+    },
+    {
+      "epoch": 0.45165263806199957,
+      "grad_norm": 1.126408798745396,
+      "learning_rate": 8.794637477485174e-06,
+      "loss": 0.38,
+      "num_input_tokens_seen": 121261872,
+      "step": 2200,
+      "train_runtime": 13287.7407,
+      "train_tokens_per_second": 9125.846
+    },
+    {
+      "epoch": 0.45267912133032234,
+      "grad_norm": 0.7548488490433977,
+      "learning_rate": 8.789382785675624e-06,
+      "loss": 0.3654,
+      "num_input_tokens_seen": 121535128,
+      "step": 2205,
+      "train_runtime": 13316.2167,
+      "train_tokens_per_second": 9126.851
+    },
+    {
+      "epoch": 0.45370560459864506,
+      "grad_norm": 0.6555269837899215,
+      "learning_rate": 8.784118242149993e-06,
+      "loss": 0.381,
+      "num_input_tokens_seen": 121816576,
+      "step": 2210,
+      "train_runtime": 13348.1565,
+      "train_tokens_per_second": 9126.097
+    },
+    {
+      "epoch": 0.4547320878669678,
+      "grad_norm": 0.7026141688500164,
+      "learning_rate": 8.778843860595147e-06,
+      "loss": 0.3795,
+      "num_input_tokens_seen": 122091104,
+      "step": 2215,
+      "train_runtime": 13380.2684,
+      "train_tokens_per_second": 9124.713
+    },
+    {
+      "epoch": 0.4557585711352905,
+      "grad_norm": 0.63766855172256,
+      "learning_rate": 8.773559654723535e-06,
+      "loss": 0.3559,
+      "num_input_tokens_seen": 122373528,
+      "step": 2220,
+      "train_runtime": 13413.4704,
+      "train_tokens_per_second": 9123.182
+    },
+    {
+      "epoch": 0.4567850544036132,
+      "grad_norm": 0.738336900988139,
+      "learning_rate": 8.768265638273146e-06,
+      "loss": 0.3629,
+      "num_input_tokens_seen": 122652288,
+      "step": 2225,
+      "train_runtime": 13433.7448,
+      "train_tokens_per_second": 9130.164
+    },
+    {
+      "epoch": 0.45781153767193594,
+      "grad_norm": 0.6273705587311104,
+      "learning_rate": 8.76296182500747e-06,
+      "loss": 0.3809,
+      "num_input_tokens_seen": 122928824,
+      "step": 2230,
+      "train_runtime": 13467.1774,
+      "train_tokens_per_second": 9128.032
+    },
+    {
+      "epoch": 0.45883802094025866,
+      "grad_norm": 0.7462784173042981,
+      "learning_rate": 8.757648228715478e-06,
+      "loss": 0.3571,
+      "num_input_tokens_seen": 123180960,
+      "step": 2235,
+      "train_runtime": 13496.0691,
+      "train_tokens_per_second": 9127.173
+    },
+    {
+      "epoch": 0.4598645042085814,
+      "grad_norm": 0.7752962762150712,
+      "learning_rate": 8.752324863211561e-06,
+      "loss": 0.3712,
+      "num_input_tokens_seen": 123453232,
+      "step": 2240,
+      "train_runtime": 13528.6702,
+      "train_tokens_per_second": 9125.304
+    },
+    {
+      "epoch": 0.4608909874769041,
+      "grad_norm": 0.60703351444238,
+      "learning_rate": 8.74699174233552e-06,
+      "loss": 0.3756,
+      "num_input_tokens_seen": 123726440,
+      "step": 2245,
+      "train_runtime": 13562.1418,
+      "train_tokens_per_second": 9122.928
+    },
+    {
+      "epoch": 0.4619174707452269,
+      "grad_norm": 0.6781754657745135,
+      "learning_rate": 8.741648879952509e-06,
+      "loss": 0.3762,
+      "num_input_tokens_seen": 124005544,
+      "step": 2250,
+      "train_runtime": 13592.7829,
+      "train_tokens_per_second": 9122.896
+    },
+    {
+      "epoch": 0.4629439540135496,
+      "grad_norm": 0.6451034755588178,
+      "learning_rate": 8.736296289953015e-06,
+      "loss": 0.3606,
+      "num_input_tokens_seen": 124305536,
+      "step": 2255,
+      "train_runtime": 13621.9276,
+      "train_tokens_per_second": 9125.4
+    },
+    {
+      "epoch": 0.4639704372818723,
+      "grad_norm": 0.6086256201576795,
+      "learning_rate": 8.730933986252814e-06,
+      "loss": 0.3477,
+      "num_input_tokens_seen": 124618000,
+      "step": 2260,
+      "train_runtime": 13653.2776,
+      "train_tokens_per_second": 9127.332
+    },
+    {
+      "epoch": 0.46499692055019504,
+      "grad_norm": 0.7185903027067392,
+      "learning_rate": 8.725561982792933e-06,
+      "loss": 0.3798,
+      "num_input_tokens_seen": 124896184,
+      "step": 2265,
+      "train_runtime": 13683.6703,
+      "train_tokens_per_second": 9127.389
+    },
+    {
+      "epoch": 0.46602340381851776,
+      "grad_norm": 0.7112058020030725,
+      "learning_rate": 8.720180293539619e-06,
+      "loss": 0.3994,
+      "num_input_tokens_seen": 125187200,
+      "step": 2270,
+      "train_runtime": 13712.0054,
+      "train_tokens_per_second": 9129.751
+    },
+    {
+      "epoch": 0.4670498870868405,
+      "grad_norm": 0.6879254495857431,
+      "learning_rate": 8.7147889324843e-06,
+      "loss": 0.3809,
+      "num_input_tokens_seen": 125467808,
+      "step": 2275,
+      "train_runtime": 13742.1262,
+      "train_tokens_per_second": 9130.16
+    },
+    {
+      "epoch": 0.4680763703551632,
+      "grad_norm": 0.7295066389845437,
+      "learning_rate": 8.709387913643545e-06,
+      "loss": 0.3621,
+      "num_input_tokens_seen": 125710512,
+      "step": 2280,
+      "train_runtime": 13771.4474,
+      "train_tokens_per_second": 9128.344
+    },
+    {
+      "epoch": 0.4691028536234859,
+      "grad_norm": 0.8960586407199946,
+      "learning_rate": 8.703977251059038e-06,
+      "loss": 0.3527,
+      "num_input_tokens_seen": 125997480,
+      "step": 2285,
+      "train_runtime": 13804.6674,
+      "train_tokens_per_second": 9127.165
+    },
+    {
+      "epoch": 0.4701293368918087,
+      "grad_norm": 0.6634813242194921,
+      "learning_rate": 8.698556958797533e-06,
+      "loss": 0.3654,
+      "num_input_tokens_seen": 126274144,
+      "step": 2290,
+      "train_runtime": 13839.143,
+      "train_tokens_per_second": 9124.419
+    },
+    {
+      "epoch": 0.4711558201601314,
+      "grad_norm": 0.6528648025501929,
+      "learning_rate": 8.693127050950814e-06,
+      "loss": 0.3624,
+      "num_input_tokens_seen": 126553464,
+      "step": 2295,
+      "train_runtime": 13866.7936,
+      "train_tokens_per_second": 9126.368
+    },
+    {
+      "epoch": 0.47218230342845413,
+      "grad_norm": 0.6941285173852967,
+      "learning_rate": 8.687687541635673e-06,
+      "loss": 0.4061,
+      "num_input_tokens_seen": 126842600,
+      "step": 2300,
+      "train_runtime": 13902.5172,
+      "train_tokens_per_second": 9123.715
+    },
+    {
+      "epoch": 0.47320878669677685,
+      "grad_norm": 1.5812897755889594,
+      "learning_rate": 8.682238444993857e-06,
+      "loss": 0.3861,
+      "num_input_tokens_seen": 127105336,
+      "step": 2305,
+      "train_runtime": 13935.5785,
+      "train_tokens_per_second": 9120.923
+    },
+    {
+      "epoch": 0.47423526996509957,
+      "grad_norm": 0.7146709760680637,
+      "learning_rate": 8.676779775192042e-06,
+      "loss": 0.3768,
+      "num_input_tokens_seen": 127394688,
+      "step": 2310,
+      "train_runtime": 13968.7515,
+      "train_tokens_per_second": 9119.977
+    },
+    {
+      "epoch": 0.4752617532334223,
+      "grad_norm": 0.6794333747019401,
+      "learning_rate": 8.671311546421787e-06,
+      "loss": 0.3808,
+      "num_input_tokens_seen": 127664672,
+      "step": 2315,
+      "train_runtime": 14000.7523,
+      "train_tokens_per_second": 9118.415
+    },
+    {
+      "epoch": 0.476288236501745,
+      "grad_norm": 0.7395000154645581,
+      "learning_rate": 8.665833772899512e-06,
+      "loss": 0.385,
+      "num_input_tokens_seen": 127931632,
+      "step": 2320,
+      "train_runtime": 14030.0724,
+      "train_tokens_per_second": 9118.387
+    },
+    {
+      "epoch": 0.47731471977006773,
+      "grad_norm": 0.6341452149569766,
+      "learning_rate": 8.660346468866445e-06,
+      "loss": 0.3948,
+      "num_input_tokens_seen": 128224472,
+      "step": 2325,
+      "train_runtime": 14060.7196,
+      "train_tokens_per_second": 9119.339
+    },
+    {
+      "epoch": 0.47834120303839045,
+      "grad_norm": 0.7411003429879205,
+      "learning_rate": 8.654849648588592e-06,
+      "loss": 0.3749,
+      "num_input_tokens_seen": 128512112,
+      "step": 2330,
+      "train_runtime": 14089.7741,
+      "train_tokens_per_second": 9120.949
+    },
+    {
+      "epoch": 0.4793676863067132,
+      "grad_norm": 0.7251523894329502,
+      "learning_rate": 8.6493433263567e-06,
+      "loss": 0.3892,
+      "num_input_tokens_seen": 128810592,
+      "step": 2335,
+      "train_runtime": 14122.8708,
+      "train_tokens_per_second": 9120.709
+    },
+    {
+      "epoch": 0.48039416957503595,
+      "grad_norm": 0.7825552390386923,
+      "learning_rate": 8.643827516486221e-06,
+      "loss": 0.3661,
+      "num_input_tokens_seen": 129066336,
+      "step": 2340,
+      "train_runtime": 14152.0026,
+      "train_tokens_per_second": 9120.005
+    },
+    {
+      "epoch": 0.48142065284335867,
+      "grad_norm": 0.6435310089029639,
+      "learning_rate": 8.638302233317272e-06,
+      "loss": 0.3741,
+      "num_input_tokens_seen": 129345000,
+      "step": 2345,
+      "train_runtime": 14182.1326,
+      "train_tokens_per_second": 9120.279
+    },
+    {
+      "epoch": 0.4824471361116814,
+      "grad_norm": 0.7695655666290389,
+      "learning_rate": 8.632767491214599e-06,
+      "loss": 0.3649,
+      "num_input_tokens_seen": 129595312,
+      "step": 2350,
+      "train_runtime": 14203.3311,
+      "train_tokens_per_second": 9124.29
+    },
+    {
+      "epoch": 0.4834736193800041,
+      "grad_norm": 0.6749147481416143,
+      "learning_rate": 8.627223304567537e-06,
+      "loss": 0.3637,
+      "num_input_tokens_seen": 129876520,
+      "step": 2355,
+      "train_runtime": 14227.6388,
+      "train_tokens_per_second": 9128.466
+    },
+    {
+      "epoch": 0.4845001026483268,
+      "grad_norm": 0.8145842884473318,
+      "learning_rate": 8.621669687789983e-06,
+      "loss": 0.3589,
+      "num_input_tokens_seen": 130154440,
+      "step": 2360,
+      "train_runtime": 14256.8707,
+      "train_tokens_per_second": 9129.243
+    },
+    {
+      "epoch": 0.48552658591664954,
+      "grad_norm": 1.0615231340652196,
+      "learning_rate": 8.616106655320342e-06,
+      "loss": 0.334,
+      "num_input_tokens_seen": 130407048,
+      "step": 2365,
+      "train_runtime": 14280.1066,
+      "train_tokens_per_second": 9132.078
+    },
+    {
+      "epoch": 0.48655306918497226,
+      "grad_norm": 0.7233802352838505,
+      "learning_rate": 8.6105342216215e-06,
+      "loss": 0.3797,
+      "num_input_tokens_seen": 130672792,
+      "step": 2370,
+      "train_runtime": 14312.3987,
+      "train_tokens_per_second": 9130.041
+    },
+    {
+      "epoch": 0.487579552453295,
+      "grad_norm": 0.7399461150429489,
+      "learning_rate": 8.604952401180788e-06,
+      "loss": 0.3768,
+      "num_input_tokens_seen": 130948976,
+      "step": 2375,
+      "train_runtime": 14342.2058,
+      "train_tokens_per_second": 9130.323
+    },
+    {
+      "epoch": 0.48860603572161776,
+      "grad_norm": 0.6735685055643782,
+      "learning_rate": 8.599361208509938e-06,
+      "loss": 0.3913,
+      "num_input_tokens_seen": 131214392,
+      "step": 2380,
+      "train_runtime": 14372.0415,
+      "train_tokens_per_second": 9129.837
+    },
+    {
+      "epoch": 0.4896325189899405,
+      "grad_norm": 0.6699195487919698,
+      "learning_rate": 8.593760658145048e-06,
+      "loss": 0.3703,
+      "num_input_tokens_seen": 131508128,
+      "step": 2385,
+      "train_runtime": 14403.4712,
+      "train_tokens_per_second": 9130.308
+    },
+    {
+      "epoch": 0.4906590022582632,
+      "grad_norm": 0.6669365479116943,
+      "learning_rate": 8.588150764646549e-06,
+      "loss": 0.3944,
+      "num_input_tokens_seen": 131788912,
+      "step": 2390,
+      "train_runtime": 14437.1557,
+      "train_tokens_per_second": 9128.454
+    },
+    {
+      "epoch": 0.4916854855265859,
+      "grad_norm": 0.8309134186746528,
+      "learning_rate": 8.582531542599153e-06,
+      "loss": 0.3636,
+      "num_input_tokens_seen": 132055608,
+      "step": 2395,
+      "train_runtime": 14467.7423,
+      "train_tokens_per_second": 9127.589
+    },
+    {
+      "epoch": 0.49271196879490864,
+      "grad_norm": 0.7140094759145122,
+      "learning_rate": 8.576903006611835e-06,
+      "loss": 0.3891,
+      "num_input_tokens_seen": 132323680,
+      "step": 2400,
+      "train_runtime": 14501.5963,
+      "train_tokens_per_second": 9124.767
+    },
+    {
+      "epoch": 0.49373845206323136,
+      "grad_norm": 0.6336538854166346,
+      "learning_rate": 8.571265171317775e-06,
+      "loss": 0.3554,
+      "num_input_tokens_seen": 132559344,
+      "step": 2405,
+      "train_runtime": 14524.9767,
+      "train_tokens_per_second": 9126.303
+    },
+    {
+      "epoch": 0.4947649353315541,
+      "grad_norm": 0.6805348841169528,
+      "learning_rate": 8.56561805137434e-06,
+      "loss": 0.3886,
+      "num_input_tokens_seen": 132836808,
+      "step": 2410,
+      "train_runtime": 14555.1313,
+      "train_tokens_per_second": 9126.459
+    },
+    {
+      "epoch": 0.4957914185998768,
+      "grad_norm": 0.8330623814054026,
+      "learning_rate": 8.559961661463025e-06,
+      "loss": 0.3772,
+      "num_input_tokens_seen": 133085288,
+      "step": 2415,
+      "train_runtime": 14586.9926,
+      "train_tokens_per_second": 9123.559
+    },
+    {
+      "epoch": 0.4968179018681996,
+      "grad_norm": 0.6450290271811049,
+      "learning_rate": 8.554296016289434e-06,
+      "loss": 0.3822,
+      "num_input_tokens_seen": 133355752,
+      "step": 2420,
+      "train_runtime": 14613.2345,
+      "train_tokens_per_second": 9125.683
+    },
+    {
+      "epoch": 0.4978443851365223,
+      "grad_norm": 0.7200532232474441,
+      "learning_rate": 8.548621130583225e-06,
+      "loss": 0.3701,
+      "num_input_tokens_seen": 133642344,
+      "step": 2425,
+      "train_runtime": 14640.1936,
+      "train_tokens_per_second": 9128.455
+    },
+    {
+      "epoch": 0.498870868404845,
+      "grad_norm": 0.6856265494502186,
+      "learning_rate": 8.542937019098086e-06,
+      "loss": 0.3817,
+      "num_input_tokens_seen": 133924528,
+      "step": 2430,
+      "train_runtime": 14674.1239,
+      "train_tokens_per_second": 9126.577
+    },
+    {
+      "epoch": 0.49989735167316773,
+      "grad_norm": 0.7422341938724425,
+      "learning_rate": 8.537243696611688e-06,
+      "loss": 0.3482,
+      "num_input_tokens_seen": 134200648,
+      "step": 2435,
+      "train_runtime": 14697.566,
+      "train_tokens_per_second": 9130.808
+    },
+    {
+      "epoch": 0.5009238349414905,
+      "grad_norm": 0.7063530633653213,
+      "learning_rate": 8.531541177925649e-06,
+      "loss": 0.3832,
+      "num_input_tokens_seen": 134469528,
+      "step": 2440,
+      "train_runtime": 14723.5463,
+      "train_tokens_per_second": 9132.958
+    },
+    {
+      "epoch": 0.5019503182098132,
+      "grad_norm": 0.6667864011714092,
+      "learning_rate": 8.525829477865494e-06,
+      "loss": 0.3682,
+      "num_input_tokens_seen": 134750304,
+      "step": 2445,
+      "train_runtime": 14755.924,
+      "train_tokens_per_second": 9131.946
+    },
+    {
+      "epoch": 0.502976801478136,
+      "grad_norm": 0.6787243993556761,
+      "learning_rate": 8.520108611280622e-06,
+      "loss": 0.3356,
+      "num_input_tokens_seen": 135027080,
+      "step": 2450,
+      "train_runtime": 14784.1644,
+      "train_tokens_per_second": 9133.224
+    },
+    {
+      "epoch": 0.5040032847464586,
+      "grad_norm": 0.6230783807147213,
+      "learning_rate": 8.514378593044259e-06,
+      "loss": 0.3324,
+      "num_input_tokens_seen": 135307712,
+      "step": 2455,
+      "train_runtime": 14817.3757,
+      "train_tokens_per_second": 9131.692
+    },
+    {
+      "epoch": 0.5050297680147814,
+      "grad_norm": 0.7363909387192605,
+      "learning_rate": 8.508639438053426e-06,
+      "loss": 0.3875,
+      "num_input_tokens_seen": 135592288,
+      "step": 2460,
+      "train_runtime": 14843.7675,
+      "train_tokens_per_second": 9134.628
+    },
+    {
+      "epoch": 0.506056251283104,
+      "grad_norm": 0.5722431937026489,
+      "learning_rate": 8.5028911612289e-06,
+      "loss": 0.372,
+      "num_input_tokens_seen": 135875280,
+      "step": 2465,
+      "train_runtime": 14869.3654,
+      "train_tokens_per_second": 9137.934
+    },
+    {
+      "epoch": 0.5070827345514268,
+      "grad_norm": 0.7034485863865244,
+      "learning_rate": 8.497133777515163e-06,
+      "loss": 0.3828,
+      "num_input_tokens_seen": 136158120,
+      "step": 2470,
+      "train_runtime": 14897.3598,
+      "train_tokens_per_second": 9139.748
+    },
+    {
+      "epoch": 0.5081092178197495,
+      "grad_norm": 0.6806437050985013,
+      "learning_rate": 8.491367301880389e-06,
+      "loss": 0.3631,
+      "num_input_tokens_seen": 136426480,
+      "step": 2475,
+      "train_runtime": 14928.6785,
+      "train_tokens_per_second": 9138.55
+    },
+    {
+      "epoch": 0.5091357010880723,
+      "grad_norm": 0.6821413061618127,
+      "learning_rate": 8.485591749316376e-06,
+      "loss": 0.3849,
+      "num_input_tokens_seen": 136715560,
+      "step": 2480,
+      "train_runtime": 14960.252,
+      "train_tokens_per_second": 9138.587
+    },
+    {
+      "epoch": 0.510162184356395,
+      "grad_norm": 0.7342518774758814,
+      "learning_rate": 8.47980713483853e-06,
+      "loss": 0.3604,
+      "num_input_tokens_seen": 136995336,
+      "step": 2485,
+      "train_runtime": 14984.1076,
+      "train_tokens_per_second": 9142.709
+    },
+    {
+      "epoch": 0.5111886676247177,
+      "grad_norm": 0.6715063847604643,
+      "learning_rate": 8.474013473485808e-06,
+      "loss": 0.3464,
+      "num_input_tokens_seen": 137281832,
+      "step": 2490,
+      "train_runtime": 15013.5459,
+      "train_tokens_per_second": 9143.865
+    },
+    {
+      "epoch": 0.5122151508930405,
+      "grad_norm": 0.649946740448999,
+      "learning_rate": 8.46821078032069e-06,
+      "loss": 0.375,
+      "num_input_tokens_seen": 137567472,
+      "step": 2495,
+      "train_runtime": 15039.0764,
+      "train_tokens_per_second": 9147.335
+    },
+    {
+      "epoch": 0.5132416341613631,
+      "grad_norm": 0.7292667364349488,
+      "learning_rate": 8.462399070429144e-06,
+      "loss": 0.3527,
+      "num_input_tokens_seen": 137865800,
+      "step": 2500,
+      "train_runtime": 15070.4583,
+      "train_tokens_per_second": 9148.083
+    },
+    {
+      "epoch": 0.5142681174296859,
+      "grad_norm": 0.6860378448777334,
+      "learning_rate": 8.456578358920568e-06,
+      "loss": 0.3753,
+      "num_input_tokens_seen": 138148824,
+      "step": 2505,
+      "train_runtime": 15105.4096,
+      "train_tokens_per_second": 9145.652
+    },
+    {
+      "epoch": 0.5152946006980086,
+      "grad_norm": 0.7622412229657831,
+      "learning_rate": 8.450748660927772e-06,
+      "loss": 0.3539,
+      "num_input_tokens_seen": 138414544,
+      "step": 2510,
+      "train_runtime": 15135.8891,
+      "train_tokens_per_second": 9144.791
+    },
+    {
+      "epoch": 0.5163210839663314,
+      "grad_norm": 0.6951562661425762,
+      "learning_rate": 8.444909991606927e-06,
+      "loss": 0.3681,
+      "num_input_tokens_seen": 138698296,
+      "step": 2515,
+      "train_runtime": 15166.0695,
+      "train_tokens_per_second": 9145.303
+    },
+    {
+      "epoch": 0.517347567234654,
+      "grad_norm": 0.8141727165339202,
+      "learning_rate": 8.439062366137523e-06,
+      "loss": 0.3897,
+      "num_input_tokens_seen": 138960712,
+      "step": 2520,
+      "train_runtime": 15199.1564,
+      "train_tokens_per_second": 9142.659
+    },
+    {
+      "epoch": 0.5183740505029768,
+      "grad_norm": 0.7147248874872036,
+      "learning_rate": 8.433205799722342e-06,
+      "loss": 0.3745,
+      "num_input_tokens_seen": 139236888,
+      "step": 2525,
+      "train_runtime": 15230.9767,
+      "train_tokens_per_second": 9141.691
+    },
+    {
+      "epoch": 0.5194005337712996,
+      "grad_norm": 0.8597317367550342,
+      "learning_rate": 8.427340307587403e-06,
+      "loss": 0.3994,
+      "num_input_tokens_seen": 139504976,
+      "step": 2530,
+      "train_runtime": 15263.6673,
+      "train_tokens_per_second": 9139.676
+    },
+    {
+      "epoch": 0.5204270170396222,
+      "grad_norm": 0.6591752243964707,
+      "learning_rate": 8.421465904981939e-06,
+      "loss": 0.3476,
+      "num_input_tokens_seen": 139798336,
+      "step": 2535,
+      "train_runtime": 15295.8572,
+      "train_tokens_per_second": 9139.621
+    },
+    {
+      "epoch": 0.521453500307945,
+      "grad_norm": 0.6402206669546272,
+      "learning_rate": 8.415582607178342e-06,
+      "loss": 0.3665,
+      "num_input_tokens_seen": 140081072,
+      "step": 2540,
+      "train_runtime": 15322.9865,
+      "train_tokens_per_second": 9141.891
+    },
+    {
+      "epoch": 0.5224799835762677,
+      "grad_norm": 0.5959563004511806,
+      "learning_rate": 8.409690429472128e-06,
+      "loss": 0.3726,
+      "num_input_tokens_seen": 140376296,
+      "step": 2545,
+      "train_runtime": 15350.6182,
+      "train_tokens_per_second": 9144.667
+    },
+    {
+      "epoch": 0.5235064668445905,
+      "grad_norm": 0.6162480846360823,
+      "learning_rate": 8.403789387181908e-06,
+      "loss": 0.3594,
+      "num_input_tokens_seen": 140654592,
+      "step": 2550,
+      "train_runtime": 15386.5759,
+      "train_tokens_per_second": 9141.384
+    },
+    {
+      "epoch": 0.5245329501129131,
+      "grad_norm": 0.8796067204620046,
+      "learning_rate": 8.397879495649332e-06,
+      "loss": 0.3702,
+      "num_input_tokens_seen": 140922320,
+      "step": 2555,
+      "train_runtime": 15412.6634,
+      "train_tokens_per_second": 9143.281
+    },
+    {
+      "epoch": 0.5255594333812359,
+      "grad_norm": 0.6235469492599216,
+      "learning_rate": 8.39196077023906e-06,
+      "loss": 0.3738,
+      "num_input_tokens_seen": 141201232,
+      "step": 2560,
+      "train_runtime": 15440.2581,
+      "train_tokens_per_second": 9145.005
+    },
+    {
+      "epoch": 0.5265859166495586,
+      "grad_norm": 0.7140551266161761,
+      "learning_rate": 8.386033226338719e-06,
+      "loss": 0.3525,
+      "num_input_tokens_seen": 141461640,
+      "step": 2565,
+      "train_runtime": 15468.9353,
+      "train_tokens_per_second": 9144.885
+    },
+    {
+      "epoch": 0.5276123999178813,
+      "grad_norm": 0.7461921720829715,
+      "learning_rate": 8.380096879358857e-06,
+      "loss": 0.3884,
+      "num_input_tokens_seen": 141744872,
+      "step": 2570,
+      "train_runtime": 15500.7803,
+      "train_tokens_per_second": 9144.37
+    },
+    {
+      "epoch": 0.5286388831862041,
+      "grad_norm": 0.6262835304089848,
+      "learning_rate": 8.374151744732912e-06,
+      "loss": 0.3963,
+      "num_input_tokens_seen": 142034168,
+      "step": 2575,
+      "train_runtime": 15533.4462,
+      "train_tokens_per_second": 9143.764
+    },
+    {
+      "epoch": 0.5296653664545268,
+      "grad_norm": 0.625752219442821,
+      "learning_rate": 8.368197837917171e-06,
+      "loss": 0.3781,
+      "num_input_tokens_seen": 142325328,
+      "step": 2580,
+      "train_runtime": 15567.4072,
+      "train_tokens_per_second": 9142.52
+    },
+    {
+      "epoch": 0.5306918497228496,
+      "grad_norm": 0.8655987447948922,
+      "learning_rate": 8.362235174390725e-06,
+      "loss": 0.3456,
+      "num_input_tokens_seen": 142596872,
+      "step": 2585,
+      "train_runtime": 15595.0673,
+      "train_tokens_per_second": 9143.716
+    },
+    {
+      "epoch": 0.5317183329911722,
+      "grad_norm": 1.0334848808225787,
+      "learning_rate": 8.356263769655429e-06,
+      "loss": 0.3935,
+      "num_input_tokens_seen": 142891336,
+      "step": 2590,
+      "train_runtime": 15629.014,
+      "train_tokens_per_second": 9142.697
+    },
+    {
+      "epoch": 0.532744816259495,
+      "grad_norm": 0.8200026615697348,
+      "learning_rate": 8.350283639235864e-06,
+      "loss": 0.3731,
+      "num_input_tokens_seen": 143157912,
+      "step": 2595,
+      "train_runtime": 15657.5708,
+      "train_tokens_per_second": 9143.047
+    },
+    {
+      "epoch": 0.5337712995278177,
+      "grad_norm": 0.6562685868247689,
+      "learning_rate": 8.3442947986793e-06,
+      "loss": 0.3651,
+      "num_input_tokens_seen": 143435704,
+      "step": 2600,
+      "train_runtime": 15691.2761,
+      "train_tokens_per_second": 9141.111
+    },
+    {
+      "epoch": 0.5347977827961404,
+      "grad_norm": 0.9122988150085809,
+      "learning_rate": 8.338297263555646e-06,
+      "loss": 0.3708,
+      "num_input_tokens_seen": 143700264,
+      "step": 2605,
+      "train_runtime": 15722.8602,
+      "train_tokens_per_second": 9139.575
+    },
+    {
+      "epoch": 0.5358242660644631,
+      "grad_norm": 0.680359524148968,
+      "learning_rate": 8.332291049457422e-06,
+      "loss": 0.3626,
+      "num_input_tokens_seen": 144009544,
+      "step": 2610,
+      "train_runtime": 15755.9353,
+      "train_tokens_per_second": 9140.019
+    },
+    {
+      "epoch": 0.5368507493327859,
+      "grad_norm": 0.7763427126574969,
+      "learning_rate": 8.326276171999706e-06,
+      "loss": 0.394,
+      "num_input_tokens_seen": 144289064,
+      "step": 2615,
+      "train_runtime": 15787.62,
+      "train_tokens_per_second": 9139.38
+    },
+    {
+      "epoch": 0.5378772326011086,
+      "grad_norm": 0.7488988872702484,
+      "learning_rate": 8.320252646820103e-06,
+      "loss": 0.3789,
+      "num_input_tokens_seen": 144571784,
+      "step": 2620,
+      "train_runtime": 15814.4163,
+      "train_tokens_per_second": 9141.772
+    },
+    {
+      "epoch": 0.5389037158694313,
+      "grad_norm": 0.7207695324431038,
+      "learning_rate": 8.314220489578693e-06,
+      "loss": 0.4146,
+      "num_input_tokens_seen": 144847928,
+      "step": 2625,
+      "train_runtime": 15840.4708,
+      "train_tokens_per_second": 9144.168
+    },
+    {
+      "epoch": 0.5399301991377541,
+      "grad_norm": 0.7613093560399519,
+      "learning_rate": 8.308179715958011e-06,
+      "loss": 0.3658,
+      "num_input_tokens_seen": 145113360,
+      "step": 2630,
+      "train_runtime": 15870.7294,
+      "train_tokens_per_second": 9143.459
+    },
+    {
+      "epoch": 0.5409566824060767,
+      "grad_norm": 0.6986043360892892,
+      "learning_rate": 8.302130341662984e-06,
+      "loss": 0.3921,
+      "num_input_tokens_seen": 145396984,
+      "step": 2635,
+      "train_runtime": 15903.2224,
+      "train_tokens_per_second": 9142.612
+    },
+    {
+      "epoch": 0.5419831656743995,
+      "grad_norm": 0.8161950224164698,
+      "learning_rate": 8.296072382420898e-06,
+      "loss": 0.3454,
+      "num_input_tokens_seen": 145662496,
+      "step": 2640,
+      "train_runtime": 15939.5684,
+      "train_tokens_per_second": 9138.422
+    },
+    {
+      "epoch": 0.5430096489427222,
+      "grad_norm": 0.6272546704509386,
+      "learning_rate": 8.290005853981366e-06,
+      "loss": 0.3666,
+      "num_input_tokens_seen": 145963744,
+      "step": 2645,
+      "train_runtime": 15973.1079,
+      "train_tokens_per_second": 9138.093
+    },
+    {
+      "epoch": 0.544036132211045,
+      "grad_norm": 0.7875319624102359,
+      "learning_rate": 8.283930772116269e-06,
+      "loss": 0.3788,
+      "num_input_tokens_seen": 146238456,
+      "step": 2650,
+      "train_runtime": 16001.9579,
+      "train_tokens_per_second": 9138.785
+    },
+    {
+      "epoch": 0.5450626154793677,
+      "grad_norm": 0.667437463443201,
+      "learning_rate": 8.277847152619736e-06,
+      "loss": 0.3702,
+      "num_input_tokens_seen": 146538000,
+      "step": 2655,
+      "train_runtime": 16030.6962,
+      "train_tokens_per_second": 9141.088
+    },
+    {
+      "epoch": 0.5460890987476904,
+      "grad_norm": 0.8881530513434636,
+      "learning_rate": 8.271755011308084e-06,
+      "loss": 0.389,
+      "num_input_tokens_seen": 146809856,
+      "step": 2660,
+      "train_runtime": 16062.3745,
+      "train_tokens_per_second": 9139.985
+    },
+    {
+      "epoch": 0.5471155820160132,
+      "grad_norm": 0.7055107399766689,
+      "learning_rate": 8.26565436401979e-06,
+      "loss": 0.3613,
+      "num_input_tokens_seen": 147068064,
+      "step": 2665,
+      "train_runtime": 16096.557,
+      "train_tokens_per_second": 9136.616
+    },
+    {
+      "epoch": 0.5481420652843358,
+      "grad_norm": 0.666531202026522,
+      "learning_rate": 8.259545226615445e-06,
+      "loss": 0.3793,
+      "num_input_tokens_seen": 147344104,
+      "step": 2670,
+      "train_runtime": 16126.8777,
+      "train_tokens_per_second": 9136.555
+    },
+    {
+      "epoch": 0.5491685485526586,
+      "grad_norm": 0.6944402324737431,
+      "learning_rate": 8.253427614977707e-06,
+      "loss": 0.343,
+      "num_input_tokens_seen": 147607984,
+      "step": 2675,
+      "train_runtime": 16158.5369,
+      "train_tokens_per_second": 9134.984
+    },
+    {
+      "epoch": 0.5501950318209813,
+      "grad_norm": 0.6473997329926651,
+      "learning_rate": 8.247301545011274e-06,
+      "loss": 0.3781,
+      "num_input_tokens_seen": 147874032,
+      "step": 2680,
+      "train_runtime": 16185.0155,
+      "train_tokens_per_second": 9136.478
+    },
+    {
+      "epoch": 0.5512215150893041,
+      "grad_norm": 0.656004535688685,
+      "learning_rate": 8.241167032642827e-06,
+      "loss": 0.3872,
+      "num_input_tokens_seen": 148148912,
+      "step": 2685,
+      "train_runtime": 16214.6112,
+      "train_tokens_per_second": 9136.754
+    },
+    {
+      "epoch": 0.5522479983576267,
+      "grad_norm": 0.6988452480441923,
+      "learning_rate": 8.235024093821002e-06,
+      "loss": 0.3953,
+      "num_input_tokens_seen": 148414504,
+      "step": 2690,
+      "train_runtime": 16244.2497,
+      "train_tokens_per_second": 9136.433
+    },
+    {
+      "epoch": 0.5532744816259495,
+      "grad_norm": 0.7592652001235926,
+      "learning_rate": 8.228872744516336e-06,
+      "loss": 0.3452,
+      "num_input_tokens_seen": 148668008,
+      "step": 2695,
+      "train_runtime": 16270.0874,
+      "train_tokens_per_second": 9137.505
+    },
+    {
+      "epoch": 0.5543009648942723,
+      "grad_norm": 0.6685032187820517,
+      "learning_rate": 8.222713000721237e-06,
+      "loss": 0.3722,
+      "num_input_tokens_seen": 148951912,
+      "step": 2700,
+      "train_runtime": 16304.7098,
+      "train_tokens_per_second": 9135.514
+    },
+    {
+      "epoch": 0.5553274481625949,
+      "grad_norm": 0.6602428913127454,
+      "learning_rate": 8.216544878449932e-06,
+      "loss": 0.3586,
+      "num_input_tokens_seen": 149217808,
+      "step": 2705,
+      "train_runtime": 16331.9506,
+      "train_tokens_per_second": 9136.558
+    },
+    {
+      "epoch": 0.5563539314309177,
+      "grad_norm": 0.724779114964407,
+      "learning_rate": 8.210368393738437e-06,
+      "loss": 0.3978,
+      "num_input_tokens_seen": 149482264,
+      "step": 2710,
+      "train_runtime": 16360.5902,
+      "train_tokens_per_second": 9136.728
+    },
+    {
+      "epoch": 0.5573804146992404,
+      "grad_norm": 0.6073017502366604,
+      "learning_rate": 8.204183562644504e-06,
+      "loss": 0.3847,
+      "num_input_tokens_seen": 149772480,
+      "step": 2715,
+      "train_runtime": 16391.6556,
+      "train_tokens_per_second": 9137.117
+    },
+    {
+      "epoch": 0.5584068979675632,
+      "grad_norm": 0.6321932383472113,
+      "learning_rate": 8.197990401247582e-06,
+      "loss": 0.3505,
+      "num_input_tokens_seen": 150050176,
+      "step": 2720,
+      "train_runtime": 16416.6884,
+      "train_tokens_per_second": 9140.1
+    },
+    {
+      "epoch": 0.5594333812358858,
+      "grad_norm": 0.7318234218884633,
+      "learning_rate": 8.191788925648784e-06,
+      "loss": 0.3416,
+      "num_input_tokens_seen": 150321896,
+      "step": 2725,
+      "train_runtime": 16441.1607,
+      "train_tokens_per_second": 9143.022
+    },
+    {
+      "epoch": 0.5604598645042086,
+      "grad_norm": 0.6874165728111974,
+      "learning_rate": 8.185579151970836e-06,
+      "loss": 0.3705,
+      "num_input_tokens_seen": 150593224,
+      "step": 2730,
+      "train_runtime": 16471.5052,
+      "train_tokens_per_second": 9142.651
+    },
+    {
+      "epoch": 0.5614863477725313,
+      "grad_norm": 0.6708187242178776,
+      "learning_rate": 8.179361096358033e-06,
+      "loss": 0.3486,
+      "num_input_tokens_seen": 150838760,
+      "step": 2735,
+      "train_runtime": 16506.5337,
+      "train_tokens_per_second": 9138.125
+    },
+    {
+      "epoch": 0.562512831040854,
+      "grad_norm": 0.6503930573779574,
+      "learning_rate": 8.173134774976208e-06,
+      "loss": 0.3479,
+      "num_input_tokens_seen": 151106128,
+      "step": 2740,
+      "train_runtime": 16539.9262,
+      "train_tokens_per_second": 9135.841
+    },
+    {
+      "epoch": 0.5635393143091768,
+      "grad_norm": 0.8823413027706105,
+      "learning_rate": 8.16690020401268e-06,
+      "loss": 0.3683,
+      "num_input_tokens_seen": 151415256,
+      "step": 2745,
+      "train_runtime": 16571.0647,
+      "train_tokens_per_second": 9137.328
+    },
+    {
+      "epoch": 0.5645657975774995,
+      "grad_norm": 0.6649697144376657,
+      "learning_rate": 8.160657399676215e-06,
+      "loss": 0.3767,
+      "num_input_tokens_seen": 151705424,
+      "step": 2750,
+      "train_runtime": 16603.7514,
+      "train_tokens_per_second": 9136.816
+    },
+    {
+      "epoch": 0.5655922808458222,
+      "grad_norm": 0.6828518074942683,
+      "learning_rate": 8.154406378196986e-06,
+      "loss": 0.3518,
+      "num_input_tokens_seen": 151968232,
+      "step": 2755,
+      "train_runtime": 16632.0448,
+      "train_tokens_per_second": 9137.074
+    },
+    {
+      "epoch": 0.5666187641141449,
+      "grad_norm": 0.7384675637606027,
+      "learning_rate": 8.148147155826531e-06,
+      "loss": 0.3771,
+      "num_input_tokens_seen": 152277136,
+      "step": 2760,
+      "train_runtime": 16660.5054,
+      "train_tokens_per_second": 9140.007
+    },
+    {
+      "epoch": 0.5676452473824677,
+      "grad_norm": 0.7554120469222353,
+      "learning_rate": 8.141879748837703e-06,
+      "loss": 0.3716,
+      "num_input_tokens_seen": 152543408,
+      "step": 2765,
+      "train_runtime": 16692.3854,
+      "train_tokens_per_second": 9138.503
+    },
+    {
+      "epoch": 0.5686717306507904,
+      "grad_norm": 0.7253830492883281,
+      "learning_rate": 8.135604173524638e-06,
+      "loss": 0.3594,
+      "num_input_tokens_seen": 152794784,
+      "step": 2770,
+      "train_runtime": 16724.9024,
+      "train_tokens_per_second": 9135.765
+    },
+    {
+      "epoch": 0.5696982139191131,
+      "grad_norm": 0.6730248179132415,
+      "learning_rate": 8.129320446202708e-06,
+      "loss": 0.3552,
+      "num_input_tokens_seen": 153073752,
+      "step": 2775,
+      "train_runtime": 16752.3324,
+      "train_tokens_per_second": 9137.459
+    },
+    {
+      "epoch": 0.5707246971874358,
+      "grad_norm": 15.8603006425389,
+      "learning_rate": 8.123028583208476e-06,
+      "loss": 0.3681,
+      "num_input_tokens_seen": 153376088,
+      "step": 2780,
+      "train_runtime": 16788.9424,
+      "train_tokens_per_second": 9135.542
+    },
+    {
+      "epoch": 0.5717511804557586,
+      "grad_norm": 0.6134495211159131,
+      "learning_rate": 8.11672860089966e-06,
+      "loss": 0.3669,
+      "num_input_tokens_seen": 153678088,
+      "step": 2785,
+      "train_runtime": 16815.7223,
+      "train_tokens_per_second": 9138.953
+    },
+    {
+      "epoch": 0.5727776637240813,
+      "grad_norm": 0.66229969278191,
+      "learning_rate": 8.110420515655082e-06,
+      "loss": 0.3998,
+      "num_input_tokens_seen": 153958112,
+      "step": 2790,
+      "train_runtime": 16849.0637,
+      "train_tokens_per_second": 9137.488
+    },
+    {
+      "epoch": 0.573804146992404,
+      "grad_norm": 0.5938783688760187,
+      "learning_rate": 8.104104343874636e-06,
+      "loss": 0.3629,
+      "num_input_tokens_seen": 154237184,
+      "step": 2795,
+      "train_runtime": 16885.3797,
+      "train_tokens_per_second": 9134.363
+    },
+    {
+      "epoch": 0.5748306302607268,
+      "grad_norm": 0.6584787224220305,
+      "learning_rate": 8.097780101979236e-06,
+      "loss": 0.3554,
+      "num_input_tokens_seen": 154530152,
+      "step": 2800,
+      "train_runtime": 16916.4981,
+      "train_tokens_per_second": 9134.878
+    },
+    {
+      "epoch": 0.5758571135290494,
+      "grad_norm": 0.6406569257316438,
+      "learning_rate": 8.091447806410777e-06,
+      "loss": 0.3558,
+      "num_input_tokens_seen": 154810816,
+      "step": 2805,
+      "train_runtime": 16943.7329,
+      "train_tokens_per_second": 9136.76
+    },
+    {
+      "epoch": 0.5768835967973722,
+      "grad_norm": 0.7830830850642081,
+      "learning_rate": 8.085107473632092e-06,
+      "loss": 0.3856,
+      "num_input_tokens_seen": 155055920,
+      "step": 2810,
+      "train_runtime": 16973.1326,
+      "train_tokens_per_second": 9135.374
+    },
+    {
+      "epoch": 0.5779100800656949,
+      "grad_norm": 0.6122553500489988,
+      "learning_rate": 8.078759120126911e-06,
+      "loss": 0.391,
+      "num_input_tokens_seen": 155345048,
+      "step": 2815,
+      "train_runtime": 17006.233,
+      "train_tokens_per_second": 9134.595
+    },
+    {
+      "epoch": 0.5789365633340177,
+      "grad_norm": 0.7230242330451191,
+      "learning_rate": 8.072402762399815e-06,
+      "loss": 0.3549,
+      "num_input_tokens_seen": 155613152,
+      "step": 2820,
+      "train_runtime": 17033.3195,
+      "train_tokens_per_second": 9135.809
+    },
+    {
+      "epoch": 0.5799630466023403,
+      "grad_norm": 0.6218255230809344,
+      "learning_rate": 8.066038416976194e-06,
+      "loss": 0.3827,
+      "num_input_tokens_seen": 155866832,
+      "step": 2825,
+      "train_runtime": 17065.1232,
+      "train_tokens_per_second": 9133.648
+    },
+    {
+      "epoch": 0.5809895298706631,
+      "grad_norm": 0.5734921163707649,
+      "learning_rate": 8.059666100402209e-06,
+      "loss": 0.3552,
+      "num_input_tokens_seen": 156150944,
+      "step": 2830,
+      "train_runtime": 17095.8699,
+      "train_tokens_per_second": 9133.84
+    },
+    {
+      "epoch": 0.5820160131389859,
+      "grad_norm": 0.6833399774690845,
+      "learning_rate": 8.053285829244734e-06,
+      "loss": 0.3817,
+      "num_input_tokens_seen": 156405608,
+      "step": 2835,
+      "train_runtime": 17127.5556,
+      "train_tokens_per_second": 9131.811
+    },
+    {
+      "epoch": 0.5830424964073085,
+      "grad_norm": 0.6435374065402322,
+      "learning_rate": 8.046897620091336e-06,
+      "loss": 0.3918,
+      "num_input_tokens_seen": 156673952,
+      "step": 2840,
+      "train_runtime": 17156.9346,
+      "train_tokens_per_second": 9131.815
+    },
+    {
+      "epoch": 0.5840689796756313,
+      "grad_norm": 0.8601392674015554,
+      "learning_rate": 8.040501489550212e-06,
+      "loss": 0.3985,
+      "num_input_tokens_seen": 156937592,
+      "step": 2845,
+      "train_runtime": 17186.6108,
+      "train_tokens_per_second": 9131.387
+    },
+    {
+      "epoch": 0.585095462943954,
+      "grad_norm": 0.5922005648448273,
+      "learning_rate": 8.034097454250156e-06,
+      "loss": 0.3696,
+      "num_input_tokens_seen": 157220168,
+      "step": 2850,
+      "train_runtime": 17215.1391,
+      "train_tokens_per_second": 9132.669
+    },
+    {
+      "epoch": 0.5861219462122768,
+      "grad_norm": 0.6272394667051818,
+      "learning_rate": 8.027685530840512e-06,
+      "loss": 0.3656,
+      "num_input_tokens_seen": 157501704,
+      "step": 2855,
+      "train_runtime": 17245.2726,
+      "train_tokens_per_second": 9133.036
+    },
+    {
+      "epoch": 0.5871484294805994,
+      "grad_norm": 0.5762909762688447,
+      "learning_rate": 8.02126573599113e-06,
+      "loss": 0.391,
+      "num_input_tokens_seen": 157755824,
+      "step": 2860,
+      "train_runtime": 17276.839,
+      "train_tokens_per_second": 9131.058
+    },
+    {
+      "epoch": 0.5881749127489222,
+      "grad_norm": 0.8519125024530984,
+      "learning_rate": 8.014838086392326e-06,
+      "loss": 0.3419,
+      "num_input_tokens_seen": 158025424,
+      "step": 2865,
+      "train_runtime": 17307.2896,
+      "train_tokens_per_second": 9130.57
+    },
+    {
+      "epoch": 0.589201396017245,
+      "grad_norm": 0.7110464920444922,
+      "learning_rate": 8.008402598754842e-06,
+      "loss": 0.3772,
+      "num_input_tokens_seen": 158284968,
+      "step": 2870,
+      "train_runtime": 17330.2825,
+      "train_tokens_per_second": 9133.433
+    },
+    {
+      "epoch": 0.5902278792855676,
+      "grad_norm": 0.7212167923265079,
+      "learning_rate": 8.001959289809787e-06,
+      "loss": 0.3521,
+      "num_input_tokens_seen": 158532008,
+      "step": 2875,
+      "train_runtime": 17362.7723,
+      "train_tokens_per_second": 9130.57
+    },
+    {
+      "epoch": 0.5912543625538904,
+      "grad_norm": 0.8030636609383487,
+      "learning_rate": 7.995508176308609e-06,
+      "loss": 0.3464,
+      "num_input_tokens_seen": 158802360,
+      "step": 2880,
+      "train_runtime": 17391.8166,
+      "train_tokens_per_second": 9130.867
+    },
+    {
+      "epoch": 0.5922808458222131,
+      "grad_norm": 0.7666383248384747,
+      "learning_rate": 7.989049275023051e-06,
+      "loss": 0.3681,
+      "num_input_tokens_seen": 159054024,
+      "step": 2885,
+      "train_runtime": 17419.5161,
+      "train_tokens_per_second": 9130.795
+    },
+    {
+      "epoch": 0.5933073290905359,
+      "grad_norm": 0.6987216724048184,
+      "learning_rate": 7.982582602745094e-06,
+      "loss": 0.3801,
+      "num_input_tokens_seen": 159335944,
+      "step": 2890,
+      "train_runtime": 17453.682,
+      "train_tokens_per_second": 9129.073
+    },
+    {
+      "epoch": 0.5943338123588585,
+      "grad_norm": 0.6236648000565571,
+      "learning_rate": 7.976108176286933e-06,
+      "loss": 0.3719,
+      "num_input_tokens_seen": 159620496,
+      "step": 2895,
+      "train_runtime": 17483.1651,
+      "train_tokens_per_second": 9129.954
+    },
+    {
+      "epoch": 0.5953602956271813,
+      "grad_norm": 0.6268001526776982,
+      "learning_rate": 7.96962601248091e-06,
+      "loss": 0.3851,
+      "num_input_tokens_seen": 159889832,
+      "step": 2900,
+      "train_runtime": 17512.7891,
+      "train_tokens_per_second": 9129.89
+    },
+    {
+      "epoch": 0.596386778895504,
+      "grad_norm": 0.6743408669075511,
+      "learning_rate": 7.963136128179495e-06,
+      "loss": 0.3713,
+      "num_input_tokens_seen": 160157016,
+      "step": 2905,
+      "train_runtime": 17545.7781,
+      "train_tokens_per_second": 9127.952
+    },
+    {
+      "epoch": 0.5974132621638267,
+      "grad_norm": 0.6965529554658586,
+      "learning_rate": 7.95663854025522e-06,
+      "loss": 0.3855,
+      "num_input_tokens_seen": 160450024,
+      "step": 2910,
+      "train_runtime": 17567.3936,
+      "train_tokens_per_second": 9133.4
+    },
+    {
+      "epoch": 0.5984397454321495,
+      "grad_norm": 0.689329396534184,
+      "learning_rate": 7.950133265600649e-06,
+      "loss": 0.3843,
+      "num_input_tokens_seen": 160708432,
+      "step": 2915,
+      "train_runtime": 17595.3409,
+      "train_tokens_per_second": 9133.579
+    },
+    {
+      "epoch": 0.5994662287004722,
+      "grad_norm": 0.7200150167049304,
+      "learning_rate": 7.943620321128331e-06,
+      "loss": 0.3619,
+      "num_input_tokens_seen": 160993552,
+      "step": 2920,
+      "train_runtime": 17626.1548,
+      "train_tokens_per_second": 9133.787
+    },
+    {
+      "epoch": 0.600492711968795,
+      "grad_norm": 0.592122616654881,
+      "learning_rate": 7.937099723770756e-06,
+      "loss": 0.3611,
+      "num_input_tokens_seen": 161283160,
+      "step": 2925,
+      "train_runtime": 17656.8788,
+      "train_tokens_per_second": 9134.296
+    },
+    {
+      "epoch": 0.6015191952371176,
+      "grad_norm": 0.6968744312789328,
+      "learning_rate": 7.930571490480307e-06,
+      "loss": 0.3799,
+      "num_input_tokens_seen": 161536096,
+      "step": 2930,
+      "train_runtime": 17690.5018,
+      "train_tokens_per_second": 9131.233
+    },
+    {
+      "epoch": 0.6025456785054404,
+      "grad_norm": 0.694764417485146,
+      "learning_rate": 7.924035638229219e-06,
+      "loss": 0.3478,
+      "num_input_tokens_seen": 161774976,
+      "step": 2935,
+      "train_runtime": 17714.1292,
+      "train_tokens_per_second": 9132.539
+    },
+    {
+      "epoch": 0.603572161773763,
+      "grad_norm": 0.6336732967614932,
+      "learning_rate": 7.91749218400954e-06,
+      "loss": 0.3758,
+      "num_input_tokens_seen": 162075840,
+      "step": 2940,
+      "train_runtime": 17748.3719,
+      "train_tokens_per_second": 9131.871
+    },
+    {
+      "epoch": 0.6045986450420858,
+      "grad_norm": 0.651606482022035,
+      "learning_rate": 7.910941144833076e-06,
+      "loss": 0.3655,
+      "num_input_tokens_seen": 162352800,
+      "step": 2945,
+      "train_runtime": 17779.276,
+      "train_tokens_per_second": 9131.575
+    },
+    {
+      "epoch": 0.6056251283104085,
+      "grad_norm": 0.6421261436810695,
+      "learning_rate": 7.904382537731354e-06,
+      "loss": 0.3738,
+      "num_input_tokens_seen": 162635736,
+      "step": 2950,
+      "train_runtime": 17799.3748,
+      "train_tokens_per_second": 9137.16
+    },
+    {
+      "epoch": 0.6066516115787313,
+      "grad_norm": 0.6509365323484997,
+      "learning_rate": 7.897816379755578e-06,
+      "loss": 0.3284,
+      "num_input_tokens_seen": 162928192,
+      "step": 2955,
+      "train_runtime": 17828.5197,
+      "train_tokens_per_second": 9138.627
+    },
+    {
+      "epoch": 0.607678094847054,
+      "grad_norm": 0.6295283716124837,
+      "learning_rate": 7.891242687976584e-06,
+      "loss": 0.3751,
+      "num_input_tokens_seen": 163223056,
+      "step": 2960,
+      "train_runtime": 17859.7584,
+      "train_tokens_per_second": 9139.153
+    },
+    {
+      "epoch": 0.6087045781153767,
+      "grad_norm": 0.7448733785842406,
+      "learning_rate": 7.884661479484788e-06,
+      "loss": 0.3728,
+      "num_input_tokens_seen": 163492576,
+      "step": 2965,
+      "train_runtime": 17888.7518,
+      "train_tokens_per_second": 9139.407
+    },
+    {
+      "epoch": 0.6097310613836995,
+      "grad_norm": 0.6011945963615855,
+      "learning_rate": 7.878072771390156e-06,
+      "loss": 0.3526,
+      "num_input_tokens_seen": 163775800,
+      "step": 2970,
+      "train_runtime": 17917.8461,
+      "train_tokens_per_second": 9140.373
+    },
+    {
+      "epoch": 0.6107575446520221,
+      "grad_norm": 0.6317656243050457,
+      "learning_rate": 7.871476580822146e-06,
+      "loss": 0.3749,
+      "num_input_tokens_seen": 164078504,
+      "step": 2975,
+      "train_runtime": 17944.5234,
+      "train_tokens_per_second": 9143.654
+    },
+    {
+      "epoch": 0.6117840279203449,
+      "grad_norm": 0.8114094167631916,
+      "learning_rate": 7.864872924929673e-06,
+      "loss": 0.3884,
+      "num_input_tokens_seen": 164364880,
+      "step": 2980,
+      "train_runtime": 17969.7655,
+      "train_tokens_per_second": 9146.746
+    },
+    {
+      "epoch": 0.6128105111886676,
+      "grad_norm": 0.7255470579847625,
+      "learning_rate": 7.85826182088106e-06,
+      "loss": 0.3791,
+      "num_input_tokens_seen": 164636800,
+      "step": 2985,
+      "train_runtime": 17998.8755,
+      "train_tokens_per_second": 9147.06
+    },
+    {
+      "epoch": 0.6138369944569904,
+      "grad_norm": 0.6730773501781524,
+      "learning_rate": 7.851643285863988e-06,
+      "loss": 0.382,
+      "num_input_tokens_seen": 164911056,
+      "step": 2990,
+      "train_runtime": 18030.7858,
+      "train_tokens_per_second": 9146.083
+    },
+    {
+      "epoch": 0.614863477725313,
+      "grad_norm": 1.0755450372773272,
+      "learning_rate": 7.845017337085466e-06,
+      "loss": 0.4141,
+      "num_input_tokens_seen": 165192464,
+      "step": 2995,
+      "train_runtime": 18056.8246,
+      "train_tokens_per_second": 9148.478
+    },
+    {
+      "epoch": 0.6158899609936358,
+      "grad_norm": 0.708415288909431,
+      "learning_rate": 7.83838399177177e-06,
+      "loss": 0.3671,
+      "num_input_tokens_seen": 165465928,
+      "step": 3000,
+      "train_runtime": 18090.2769,
+      "train_tokens_per_second": 9146.677
+    },
+    {
+      "epoch": 0.6169164442619586,
+      "grad_norm": 0.6210271362237241,
+      "learning_rate": 7.831743267168412e-06,
+      "loss": 0.358,
+      "num_input_tokens_seen": 165734056,
+      "step": 3005,
+      "train_runtime": 18114.9895,
+      "train_tokens_per_second": 9149.001
+    },
+    {
+      "epoch": 0.6179429275302812,
+      "grad_norm": 0.6915122272700079,
+      "learning_rate": 7.825095180540086e-06,
+      "loss": 0.3688,
+      "num_input_tokens_seen": 166020464,
+      "step": 3010,
+      "train_runtime": 18144.2025,
+      "train_tokens_per_second": 9150.056
+    },
+    {
+      "epoch": 0.618969410798604,
+      "grad_norm": 0.6045387924212261,
+      "learning_rate": 7.818439749170621e-06,
+      "loss": 0.3857,
+      "num_input_tokens_seen": 166320760,
+      "step": 3015,
+      "train_runtime": 18167.336,
+      "train_tokens_per_second": 9154.934
+    },
+    {
+      "epoch": 0.6199958940669267,
+      "grad_norm": 0.6288084506228738,
+      "learning_rate": 7.811776990362951e-06,
+      "loss": 0.3818,
+      "num_input_tokens_seen": 166583800,
+      "step": 3020,
+      "train_runtime": 18193.0454,
+      "train_tokens_per_second": 9156.455
+    },
+    {
+      "epoch": 0.6210223773352495,
+      "grad_norm": 0.682207768267692,
+      "learning_rate": 7.805106921439053e-06,
+      "loss": 0.3475,
+      "num_input_tokens_seen": 166839064,
+      "step": 3025,
+      "train_runtime": 18215.2162,
+      "train_tokens_per_second": 9159.324
+    },
+    {
+      "epoch": 0.6220488606035721,
+      "grad_norm": 0.6492881584429002,
+      "learning_rate": 7.798429559739911e-06,
+      "loss": 0.3522,
+      "num_input_tokens_seen": 167100432,
+      "step": 3030,
+      "train_runtime": 18248.2428,
+      "train_tokens_per_second": 9157.07
+    },
+    {
+      "epoch": 0.6230753438718949,
+      "grad_norm": 0.720934811776816,
+      "learning_rate": 7.791744922625466e-06,
+      "loss": 0.3813,
+      "num_input_tokens_seen": 167384624,
+      "step": 3035,
+      "train_runtime": 18279.981,
+      "train_tokens_per_second": 9156.718
+    },
+    {
+      "epoch": 0.6241018271402177,
+      "grad_norm": 0.6550585502653522,
+      "learning_rate": 7.78505302747458e-06,
+      "loss": 0.3495,
+      "num_input_tokens_seen": 167642016,
+      "step": 3040,
+      "train_runtime": 18310.8334,
+      "train_tokens_per_second": 9155.346
+    },
+    {
+      "epoch": 0.6251283104085403,
+      "grad_norm": 0.6968172585720633,
+      "learning_rate": 7.778353891684978e-06,
+      "loss": 0.3296,
+      "num_input_tokens_seen": 167939160,
+      "step": 3045,
+      "train_runtime": 18345.7534,
+      "train_tokens_per_second": 9154.116
+    },
+    {
+      "epoch": 0.6261547936768631,
+      "grad_norm": 0.7892292568898572,
+      "learning_rate": 7.771647532673215e-06,
+      "loss": 0.3581,
+      "num_input_tokens_seen": 168222808,
+      "step": 3050,
+      "train_runtime": 18377.4974,
+      "train_tokens_per_second": 9153.739
+    },
+    {
+      "epoch": 0.6271812769451858,
+      "grad_norm": 0.6604736584468172,
+      "learning_rate": 7.764933967874623e-06,
+      "loss": 0.3518,
+      "num_input_tokens_seen": 168509032,
+      "step": 3055,
+      "train_runtime": 18408.7828,
+      "train_tokens_per_second": 9153.73
+    },
+    {
+      "epoch": 0.6282077602135085,
+      "grad_norm": 0.6763875869637858,
+      "learning_rate": 7.758213214743264e-06,
+      "loss": 0.3727,
+      "num_input_tokens_seen": 168765832,
+      "step": 3060,
+      "train_runtime": 18442.1926,
+      "train_tokens_per_second": 9151.072
+    },
+    {
+      "epoch": 0.6292342434818312,
+      "grad_norm": 0.7072667194934091,
+      "learning_rate": 7.751485290751893e-06,
+      "loss": 0.3726,
+      "num_input_tokens_seen": 169043184,
+      "step": 3065,
+      "train_runtime": 18476.8748,
+      "train_tokens_per_second": 9148.906
+    },
+    {
+      "epoch": 0.630260726750154,
+      "grad_norm": 0.7162511883473865,
+      "learning_rate": 7.744750213391907e-06,
+      "loss": 0.3587,
+      "num_input_tokens_seen": 169326592,
+      "step": 3070,
+      "train_runtime": 18507.6952,
+      "train_tokens_per_second": 9148.983
+    },
+    {
+      "epoch": 0.6312872100184767,
+      "grad_norm": 0.6771347327454177,
+      "learning_rate": 7.738008000173301e-06,
+      "loss": 0.3763,
+      "num_input_tokens_seen": 169586184,
+      "step": 3075,
+      "train_runtime": 18538.5724,
+      "train_tokens_per_second": 9147.748
+    },
+    {
+      "epoch": 0.6323136932867994,
+      "grad_norm": 0.7327620307377772,
+      "learning_rate": 7.731258668624619e-06,
+      "loss": 0.3762,
+      "num_input_tokens_seen": 169844928,
+      "step": 3080,
+      "train_runtime": 18564.7027,
+      "train_tokens_per_second": 9148.809
+    },
+    {
+      "epoch": 0.6333401765551222,
+      "grad_norm": 3.193273628394823,
+      "learning_rate": 7.724502236292915e-06,
+      "loss": 0.3978,
+      "num_input_tokens_seen": 170147680,
+      "step": 3085,
+      "train_runtime": 18593.4392,
+      "train_tokens_per_second": 9150.953
+    },
+    {
+      "epoch": 0.6343666598234449,
+      "grad_norm": 0.6513723779549382,
+      "learning_rate": 7.717738720743699e-06,
+      "loss": 0.3623,
+      "num_input_tokens_seen": 170406008,
+      "step": 3090,
+      "train_runtime": 18626.4822,
+      "train_tokens_per_second": 9148.588
+    },
+    {
+      "epoch": 0.6353931430917676,
+      "grad_norm": 0.6844789573327995,
+      "learning_rate": 7.710968139560904e-06,
+      "loss": 0.3646,
+      "num_input_tokens_seen": 170694272,
+      "step": 3095,
+      "train_runtime": 18657.7297,
+      "train_tokens_per_second": 9148.716
+    },
+    {
+      "epoch": 0.6364196263600903,
+      "grad_norm": 0.631107278347099,
+      "learning_rate": 7.704190510346825e-06,
+      "loss": 0.3792,
+      "num_input_tokens_seen": 170978584,
+      "step": 3100,
+      "train_runtime": 18690.4764,
+      "train_tokens_per_second": 9147.899
+    },
+    {
+      "epoch": 0.6374461096284131,
+      "grad_norm": 0.6987933772084053,
+      "learning_rate": 7.697405850722081e-06,
+      "loss": 0.3649,
+      "num_input_tokens_seen": 171260736,
+      "step": 3105,
+      "train_runtime": 18726.04,
+      "train_tokens_per_second": 9145.593
+    },
+    {
+      "epoch": 0.6384725928967357,
+      "grad_norm": 0.6094728052138317,
+      "learning_rate": 7.690614178325576e-06,
+      "loss": 0.3708,
+      "num_input_tokens_seen": 171502056,
+      "step": 3110,
+      "train_runtime": 18751.6618,
+      "train_tokens_per_second": 9145.966
+    },
+    {
+      "epoch": 0.6394990761650585,
+      "grad_norm": 0.7971803841124692,
+      "learning_rate": 7.683815510814437e-06,
+      "loss": 0.3967,
+      "num_input_tokens_seen": 171769864,
+      "step": 3115,
+      "train_runtime": 18773.7204,
+      "train_tokens_per_second": 9149.484
+    },
+    {
+      "epoch": 0.6405255594333812,
+      "grad_norm": 0.685796922454408,
+      "learning_rate": 7.677009865863983e-06,
+      "loss": 0.3816,
+      "num_input_tokens_seen": 172067424,
+      "step": 3120,
+      "train_runtime": 18802.3699,
+      "train_tokens_per_second": 9151.369
+    },
+    {
+      "epoch": 0.641552042701704,
+      "grad_norm": 0.6498535841719868,
+      "learning_rate": 7.67019726116767e-06,
+      "loss": 0.3637,
+      "num_input_tokens_seen": 172316552,
+      "step": 3125,
+      "train_runtime": 18831.8441,
+      "train_tokens_per_second": 9150.275
+    },
+    {
+      "epoch": 0.6425785259700267,
+      "grad_norm": 0.7133342552422645,
+      "learning_rate": 7.66337771443705e-06,
+      "loss": 0.3548,
+      "num_input_tokens_seen": 172584504,
+      "step": 3130,
+      "train_runtime": 18866.4969,
+      "train_tokens_per_second": 9147.671
+    },
+    {
+      "epoch": 0.6436050092383494,
+      "grad_norm": 0.5967868190660772,
+      "learning_rate": 7.656551243401722e-06,
+      "loss": 0.3713,
+      "num_input_tokens_seen": 172871536,
+      "step": 3135,
+      "train_runtime": 18899.9447,
+      "train_tokens_per_second": 9146.669
+    },
+    {
+      "epoch": 0.6446314925066722,
+      "grad_norm": 0.9900122266147519,
+      "learning_rate": 7.649717865809285e-06,
+      "loss": 0.386,
+      "num_input_tokens_seen": 173128768,
+      "step": 3140,
+      "train_runtime": 18925.2472,
+      "train_tokens_per_second": 9148.032
+    },
+    {
+      "epoch": 0.6456579757749948,
+      "grad_norm": 0.5965704148238368,
+      "learning_rate": 7.642877599425301e-06,
+      "loss": 0.3496,
+      "num_input_tokens_seen": 173408800,
+      "step": 3145,
+      "train_runtime": 18958.8441,
+      "train_tokens_per_second": 9146.591
+    },
+    {
+      "epoch": 0.6466844590433176,
+      "grad_norm": 0.6746195125389362,
+      "learning_rate": 7.636030462033233e-06,
+      "loss": 0.3637,
+      "num_input_tokens_seen": 173710480,
+      "step": 3150,
+      "train_runtime": 18981.4828,
+      "train_tokens_per_second": 9151.576
+    },
+    {
+      "epoch": 0.6477109423116403,
+      "grad_norm": 0.6881965197061367,
+      "learning_rate": 7.6291764714344115e-06,
+      "loss": 0.3852,
+      "num_input_tokens_seen": 173996072,
+      "step": 3155,
+      "train_runtime": 19015.4279,
+      "train_tokens_per_second": 9150.258
+    },
+    {
+      "epoch": 0.6487374255799631,
+      "grad_norm": 0.6641675472449187,
+      "learning_rate": 7.622315645447985e-06,
+      "loss": 0.3488,
+      "num_input_tokens_seen": 174265176,
+      "step": 3160,
+      "train_runtime": 19038.0952,
+      "train_tokens_per_second": 9153.499
+    },
+    {
+      "epoch": 0.6497639088482857,
+      "grad_norm": 0.5895017236953844,
+      "learning_rate": 7.61544800191087e-06,
+      "loss": 0.3617,
+      "num_input_tokens_seen": 174556504,
+      "step": 3165,
+      "train_runtime": 19067.6397,
+      "train_tokens_per_second": 9154.594
+    },
+    {
+      "epoch": 0.6507903921166085,
+      "grad_norm": 0.6787711360491077,
+      "learning_rate": 7.608573558677708e-06,
+      "loss": 0.3626,
+      "num_input_tokens_seen": 174837136,
+      "step": 3170,
+      "train_runtime": 19097.9853,
+      "train_tokens_per_second": 9154.742
+    },
+    {
+      "epoch": 0.6518168753849313,
+      "grad_norm": 0.7951796901397448,
+      "learning_rate": 7.601692333620822e-06,
+      "loss": 0.3782,
+      "num_input_tokens_seen": 175127800,
+      "step": 3175,
+      "train_runtime": 19128.2606,
+      "train_tokens_per_second": 9155.448
+    },
+    {
+      "epoch": 0.6528433586532539,
+      "grad_norm": 0.6469605769301698,
+      "learning_rate": 7.594804344630162e-06,
+      "loss": 0.395,
+      "num_input_tokens_seen": 175404464,
+      "step": 3180,
+      "train_runtime": 19159.2767,
+      "train_tokens_per_second": 9155.067
+    },
+    {
+      "epoch": 0.6538698419215767,
+      "grad_norm": 0.6139509159973496,
+      "learning_rate": 7.587909609613264e-06,
+      "loss": 0.3503,
+      "num_input_tokens_seen": 175685352,
+      "step": 3185,
+      "train_runtime": 19192.4979,
+      "train_tokens_per_second": 9153.855
+    },
+    {
+      "epoch": 0.6548963251898994,
+      "grad_norm": 0.7231842952510102,
+      "learning_rate": 7.581008146495208e-06,
+      "loss": 0.3427,
+      "num_input_tokens_seen": 175931240,
+      "step": 3190,
+      "train_runtime": 19222.8858,
+      "train_tokens_per_second": 9152.176
+    },
+    {
+      "epoch": 0.6559228084582222,
+      "grad_norm": 0.6239766537005108,
+      "learning_rate": 7.574099973218554e-06,
+      "loss": 0.3473,
+      "num_input_tokens_seen": 176208760,
+      "step": 3195,
+      "train_runtime": 19257.6766,
+      "train_tokens_per_second": 9150.053
+    },
+    {
+      "epoch": 0.6569492917265448,
+      "grad_norm": 0.6338722544501695,
+      "learning_rate": 7.56718510774332e-06,
+      "loss": 0.3655,
+      "num_input_tokens_seen": 176513256,
+      "step": 3200,
+      "train_runtime": 19285.6775,
+      "train_tokens_per_second": 9152.557
+    },
+    {
+      "epoch": 0.6579757749948676,
+      "grad_norm": 0.6001406208830322,
+      "learning_rate": 7.560263568046915e-06,
+      "loss": 0.3549,
+      "num_input_tokens_seen": 176800672,
+      "step": 3205,
+      "train_runtime": 19316.3096,
+      "train_tokens_per_second": 9152.922
+    },
+    {
+      "epoch": 0.6590022582631903,
+      "grad_norm": 0.617065456903177,
+      "learning_rate": 7.553335372124101e-06,
+      "loss": 0.3724,
+      "num_input_tokens_seen": 177095544,
+      "step": 3210,
+      "train_runtime": 19344.7013,
+      "train_tokens_per_second": 9154.731
+    },
+    {
+      "epoch": 0.660028741531513,
+      "grad_norm": 0.6122373122555842,
+      "learning_rate": 7.546400537986947e-06,
+      "loss": 0.3432,
+      "num_input_tokens_seen": 177364104,
+      "step": 3215,
+      "train_runtime": 19370.3186,
+      "train_tokens_per_second": 9156.489
+    },
+    {
+      "epoch": 0.6610552247998358,
+      "grad_norm": 0.6584911524065958,
+      "learning_rate": 7.539459083664777e-06,
+      "loss": 0.3595,
+      "num_input_tokens_seen": 177633880,
+      "step": 3220,
+      "train_runtime": 19401.3451,
+      "train_tokens_per_second": 9155.751
+    },
+    {
+      "epoch": 0.6620817080681585,
+      "grad_norm": 1.303950818619241,
+      "learning_rate": 7.532511027204131e-06,
+      "loss": 0.3669,
+      "num_input_tokens_seen": 177924776,
+      "step": 3225,
+      "train_runtime": 19433.3778,
+      "train_tokens_per_second": 9155.628
+    },
+    {
+      "epoch": 0.6631081913364812,
+      "grad_norm": 0.7206131797033046,
+      "learning_rate": 7.525556386668708e-06,
+      "loss": 0.37,
+      "num_input_tokens_seen": 178195296,
+      "step": 3230,
+      "train_runtime": 19456.5821,
+      "train_tokens_per_second": 9158.613
+    },
+    {
+      "epoch": 0.6641346746048039,
+      "grad_norm": 0.6170517002062419,
+      "learning_rate": 7.518595180139328e-06,
+      "loss": 0.3633,
+      "num_input_tokens_seen": 178457504,
+      "step": 3235,
+      "train_runtime": 19486.402,
+      "train_tokens_per_second": 9158.053
+    },
+    {
+      "epoch": 0.6651611578731267,
+      "grad_norm": 0.6570040753789808,
+      "learning_rate": 7.511627425713881e-06,
+      "loss": 0.3711,
+      "num_input_tokens_seen": 178725144,
+      "step": 3240,
+      "train_runtime": 19515.7756,
+      "train_tokens_per_second": 9157.983
+    },
+    {
+      "epoch": 0.6661876411414493,
+      "grad_norm": 0.7092592147530916,
+      "learning_rate": 7.504653141507278e-06,
+      "loss": 0.3669,
+      "num_input_tokens_seen": 178989136,
+      "step": 3245,
+      "train_runtime": 19544.9321,
+      "train_tokens_per_second": 9157.828
+    },
+    {
+      "epoch": 0.6672141244097721,
+      "grad_norm": 1.0086601498273946,
+      "learning_rate": 7.4976723456514075e-06,
+      "loss": 0.3537,
+      "num_input_tokens_seen": 179239208,
+      "step": 3250,
+      "train_runtime": 19572.2128,
+      "train_tokens_per_second": 9157.841
+    },
+    {
+      "epoch": 0.6682406076780949,
+      "grad_norm": 0.9465596344968509,
+      "learning_rate": 7.490685056295093e-06,
+      "loss": 0.3509,
+      "num_input_tokens_seen": 179519184,
+      "step": 3255,
+      "train_runtime": 19604.1522,
+      "train_tokens_per_second": 9157.202
+    },
+    {
+      "epoch": 0.6692670909464176,
+      "grad_norm": 0.598565162712866,
+      "learning_rate": 7.483691291604026e-06,
+      "loss": 0.3308,
+      "num_input_tokens_seen": 179811072,
+      "step": 3260,
+      "train_runtime": 19632.4432,
+      "train_tokens_per_second": 9158.874
+    },
+    {
+      "epoch": 0.6702935742147403,
+      "grad_norm": 0.6451505808208982,
+      "learning_rate": 7.476691069760749e-06,
+      "loss": 0.3861,
+      "num_input_tokens_seen": 180087472,
+      "step": 3265,
+      "train_runtime": 19665.2396,
+      "train_tokens_per_second": 9157.655
+    },
+    {
+      "epoch": 0.671320057483063,
+      "grad_norm": 0.8225131262775659,
+      "learning_rate": 7.469684408964579e-06,
+      "loss": 0.3866,
+      "num_input_tokens_seen": 180367360,
+      "step": 3270,
+      "train_runtime": 19694.3178,
+      "train_tokens_per_second": 9158.345
+    },
+    {
+      "epoch": 0.6723465407513858,
+      "grad_norm": 0.7629652410352031,
+      "learning_rate": 7.462671327431579e-06,
+      "loss": 0.3843,
+      "num_input_tokens_seen": 180638440,
+      "step": 3275,
+      "train_runtime": 19723.9094,
+      "train_tokens_per_second": 9158.349
+    },
+    {
+      "epoch": 0.6733730240197084,
+      "grad_norm": 0.6168838669003031,
+      "learning_rate": 7.455651843394508e-06,
+      "loss": 0.3814,
+      "num_input_tokens_seen": 180910264,
+      "step": 3280,
+      "train_runtime": 19749.9632,
+      "train_tokens_per_second": 9160.03
+    },
+    {
+      "epoch": 0.6743995072880312,
+      "grad_norm": 0.666794313629679,
+      "learning_rate": 7.448625975102761e-06,
+      "loss": 0.3643,
+      "num_input_tokens_seen": 181175896,
+      "step": 3285,
+      "train_runtime": 19776.6176,
+      "train_tokens_per_second": 9161.116
+    },
+    {
+      "epoch": 0.6754259905563539,
+      "grad_norm": 0.614107921109957,
+      "learning_rate": 7.441593740822342e-06,
+      "loss": 0.3615,
+      "num_input_tokens_seen": 181455272,
+      "step": 3290,
+      "train_runtime": 19801.1152,
+      "train_tokens_per_second": 9163.892
+    },
+    {
+      "epoch": 0.6764524738246767,
+      "grad_norm": 0.624384289030614,
+      "learning_rate": 7.434555158835794e-06,
+      "loss": 0.366,
+      "num_input_tokens_seen": 181735096,
+      "step": 3295,
+      "train_runtime": 19830.9859,
+      "train_tokens_per_second": 9164.199
+    },
+    {
+      "epoch": 0.6774789570929994,
+      "grad_norm": 0.6588302478436824,
+      "learning_rate": 7.4275102474421716e-06,
+      "loss": 0.3658,
+      "num_input_tokens_seen": 182004248,
+      "step": 3300,
+      "train_runtime": 19861.8346,
+      "train_tokens_per_second": 9163.516
+    },
+    {
+      "epoch": 0.6785054403613221,
+      "grad_norm": 0.9665299393385239,
+      "learning_rate": 7.420459024956982e-06,
+      "loss": 0.3795,
+      "num_input_tokens_seen": 182240000,
+      "step": 3305,
+      "train_runtime": 19886.8837,
+      "train_tokens_per_second": 9163.829
+    },
+    {
+      "epoch": 0.6795319236296449,
+      "grad_norm": 0.6384808719585005,
+      "learning_rate": 7.413401509712138e-06,
+      "loss": 0.3475,
+      "num_input_tokens_seen": 182518880,
+      "step": 3310,
+      "train_runtime": 19920.2265,
+      "train_tokens_per_second": 9162.49
+    },
+    {
+      "epoch": 0.6805584068979675,
+      "grad_norm": 0.6743899573773325,
+      "learning_rate": 7.406337720055919e-06,
+      "loss": 0.3732,
+      "num_input_tokens_seen": 182782792,
+      "step": 3315,
+      "train_runtime": 19947.5142,
+      "train_tokens_per_second": 9163.186
+    },
+    {
+      "epoch": 0.6815848901662903,
+      "grad_norm": 0.5692488320568241,
+      "learning_rate": 7.399267674352906e-06,
+      "loss": 0.3532,
+      "num_input_tokens_seen": 183079968,
+      "step": 3320,
+      "train_runtime": 19976.7342,
+      "train_tokens_per_second": 9164.66
+    },
+    {
+      "epoch": 0.682611373434613,
+      "grad_norm": 0.7446658529265651,
+      "learning_rate": 7.392191390983954e-06,
+      "loss": 0.3784,
+      "num_input_tokens_seen": 183332856,
+      "step": 3325,
+      "train_runtime": 20006.6872,
+      "train_tokens_per_second": 9163.579
+    },
+    {
+      "epoch": 0.6836378567029358,
+      "grad_norm": 0.7291556256899216,
+      "learning_rate": 7.385108888346132e-06,
+      "loss": 0.3688,
+      "num_input_tokens_seen": 183597720,
+      "step": 3330,
+      "train_runtime": 20031.6445,
+      "train_tokens_per_second": 9165.384
+    },
+    {
+      "epoch": 0.6846643399712584,
+      "grad_norm": 0.6337237875337359,
+      "learning_rate": 7.378020184852676e-06,
+      "loss": 0.384,
+      "num_input_tokens_seen": 183886008,
+      "step": 3335,
+      "train_runtime": 20064.847,
+      "train_tokens_per_second": 9164.586
+    },
+    {
+      "epoch": 0.6856908232395812,
+      "grad_norm": 0.5883954484017804,
+      "learning_rate": 7.370925298932944e-06,
+      "loss": 0.3312,
+      "num_input_tokens_seen": 184191800,
+      "step": 3340,
+      "train_runtime": 20096.7061,
+      "train_tokens_per_second": 9165.273
+    },
+    {
+      "epoch": 0.686717306507904,
+      "grad_norm": 0.5953871563383119,
+      "learning_rate": 7.363824249032371e-06,
+      "loss": 0.3266,
+      "num_input_tokens_seen": 184456792,
+      "step": 3345,
+      "train_runtime": 20127.0872,
+      "train_tokens_per_second": 9164.604
+    },
+    {
+      "epoch": 0.6877437897762266,
+      "grad_norm": 0.7053583199459413,
+      "learning_rate": 7.35671705361241e-06,
+      "loss": 0.3768,
+      "num_input_tokens_seen": 184742792,
+      "step": 3350,
+      "train_runtime": 20156.9951,
+      "train_tokens_per_second": 9165.195
+    },
+    {
+      "epoch": 0.6887702730445494,
+      "grad_norm": 0.7037145789654617,
+      "learning_rate": 7.349603731150497e-06,
+      "loss": 0.3694,
+      "num_input_tokens_seen": 185020296,
+      "step": 3355,
+      "train_runtime": 20187.7962,
+      "train_tokens_per_second": 9164.958
+    },
+    {
+      "epoch": 0.6897967563128721,
+      "grad_norm": 0.7056262013340174,
+      "learning_rate": 7.342484300139998e-06,
+      "loss": 0.3613,
+      "num_input_tokens_seen": 185316720,
+      "step": 3360,
+      "train_runtime": 20218.9589,
+      "train_tokens_per_second": 9165.493
+    },
+    {
+      "epoch": 0.6908232395811948,
+      "grad_norm": 0.5962630876573259,
+      "learning_rate": 7.335358779090153e-06,
+      "loss": 0.3748,
+      "num_input_tokens_seen": 185599712,
+      "step": 3365,
+      "train_runtime": 20249.313,
+      "train_tokens_per_second": 9165.729
+    },
+    {
+      "epoch": 0.6918497228495175,
+      "grad_norm": 0.6490829623021795,
+      "learning_rate": 7.328227186526044e-06,
+      "loss": 0.3588,
+      "num_input_tokens_seen": 185864096,
+      "step": 3370,
+      "train_runtime": 20280.1948,
+      "train_tokens_per_second": 9164.808
+    },
+    {
+      "epoch": 0.6928762061178403,
+      "grad_norm": 0.7824255501535479,
+      "learning_rate": 7.321089540988529e-06,
+      "loss": 0.3963,
+      "num_input_tokens_seen": 186126144,
+      "step": 3375,
+      "train_runtime": 20307.1751,
+      "train_tokens_per_second": 9165.536
+    },
+    {
+      "epoch": 0.693902689386163,
+      "grad_norm": 0.659030694837044,
+      "learning_rate": 7.313945861034211e-06,
+      "loss": 0.3609,
+      "num_input_tokens_seen": 186405176,
+      "step": 3380,
+      "train_runtime": 20338.5205,
+      "train_tokens_per_second": 9165.13
+    },
+    {
+      "epoch": 0.6949291726544857,
+      "grad_norm": 0.717289451239136,
+      "learning_rate": 7.3067961652353765e-06,
+      "loss": 0.3485,
+      "num_input_tokens_seen": 186688008,
+      "step": 3385,
+      "train_runtime": 20367.6722,
+      "train_tokens_per_second": 9165.898
+    },
+    {
+      "epoch": 0.6959556559228085,
+      "grad_norm": 0.740802751648692,
+      "learning_rate": 7.299640472179953e-06,
+      "loss": 0.3532,
+      "num_input_tokens_seen": 186948472,
+      "step": 3390,
+      "train_runtime": 20398.3092,
+      "train_tokens_per_second": 9164.9
+    },
+    {
+      "epoch": 0.6969821391911312,
+      "grad_norm": 0.6619848388998758,
+      "learning_rate": 7.292478800471459e-06,
+      "loss": 0.3785,
+      "num_input_tokens_seen": 187191568,
+      "step": 3395,
+      "train_runtime": 20425.3989,
+      "train_tokens_per_second": 9164.647
+    },
+    {
+      "epoch": 0.6980086224594539,
+      "grad_norm": 0.6372482914180075,
+      "learning_rate": 7.285311168728958e-06,
+      "loss": 0.3739,
+      "num_input_tokens_seen": 187496160,
+      "step": 3400,
+      "train_runtime": 20458.0797,
+      "train_tokens_per_second": 9164.895
+    },
+    {
+      "epoch": 0.6990351057277766,
+      "grad_norm": 0.5644759720233054,
+      "learning_rate": 7.278137595587008e-06,
+      "loss": 0.3565,
+      "num_input_tokens_seen": 187776256,
+      "step": 3405,
+      "train_runtime": 20488.2626,
+      "train_tokens_per_second": 9165.065
+    },
+    {
+      "epoch": 0.7000615889960994,
+      "grad_norm": 0.6205089966645964,
+      "learning_rate": 7.270958099695614e-06,
+      "loss": 0.3414,
+      "num_input_tokens_seen": 188033256,
+      "step": 3410,
+      "train_runtime": 20520.7426,
+      "train_tokens_per_second": 9163.082
+    },
+    {
+      "epoch": 0.701088072264422,
+      "grad_norm": 0.7485645360853421,
+      "learning_rate": 7.263772699720179e-06,
+      "loss": 0.3603,
+      "num_input_tokens_seen": 188317200,
+      "step": 3415,
+      "train_runtime": 20556.6944,
+      "train_tokens_per_second": 9160.87
+    },
+    {
+      "epoch": 0.7021145555327448,
+      "grad_norm": 0.649169179906233,
+      "learning_rate": 7.256581414341452e-06,
+      "loss": 0.3653,
+      "num_input_tokens_seen": 188589400,
+      "step": 3420,
+      "train_runtime": 20589.2752,
+      "train_tokens_per_second": 9159.594
+    },
+    {
+      "epoch": 0.7031410388010676,
+      "grad_norm": 0.689699938527501,
+      "learning_rate": 7.24938426225549e-06,
+      "loss": 0.3498,
+      "num_input_tokens_seen": 188864256,
+      "step": 3425,
+      "train_runtime": 20622.3978,
+      "train_tokens_per_second": 9158.21
+    },
+    {
+      "epoch": 0.7041675220693903,
+      "grad_norm": 0.6873074536257411,
+      "learning_rate": 7.242181262173595e-06,
+      "loss": 0.3784,
+      "num_input_tokens_seen": 189132920,
+      "step": 3430,
+      "train_runtime": 20655.496,
+      "train_tokens_per_second": 9156.542
+    },
+    {
+      "epoch": 0.705194005337713,
+      "grad_norm": 0.6478519442445024,
+      "learning_rate": 7.234972432822279e-06,
+      "loss": 0.3518,
+      "num_input_tokens_seen": 189425104,
+      "step": 3435,
+      "train_runtime": 20685.6078,
+      "train_tokens_per_second": 9157.338
+    },
+    {
+      "epoch": 0.7062204886060357,
+      "grad_norm": 0.6260099097097659,
+      "learning_rate": 7.227757792943207e-06,
+      "loss": 0.3468,
+      "num_input_tokens_seen": 189708512,
+      "step": 3440,
+      "train_runtime": 20714.8686,
+      "train_tokens_per_second": 9158.084
+    },
+    {
+      "epoch": 0.7072469718743585,
+      "grad_norm": 0.6358243065027029,
+      "learning_rate": 7.220537361293146e-06,
+      "loss": 0.3956,
+      "num_input_tokens_seen": 189966184,
+      "step": 3445,
+      "train_runtime": 20747.5886,
+      "train_tokens_per_second": 9156.061
+    },
+    {
+      "epoch": 0.7082734551426811,
+      "grad_norm": 0.6328376464152192,
+      "learning_rate": 7.213311156643927e-06,
+      "loss": 0.3607,
+      "num_input_tokens_seen": 190236664,
+      "step": 3450,
+      "train_runtime": 20779.577,
+      "train_tokens_per_second": 9154.982
+    },
+    {
+      "epoch": 0.7092999384110039,
+      "grad_norm": 0.6076641485302251,
+      "learning_rate": 7.206079197782384e-06,
+      "loss": 0.3774,
+      "num_input_tokens_seen": 190498752,
+      "step": 3455,
+      "train_runtime": 20805.6254,
+      "train_tokens_per_second": 9156.118
+    },
+    {
+      "epoch": 0.7103264216793266,
+      "grad_norm": 0.6854786750616896,
+      "learning_rate": 7.198841503510319e-06,
+      "loss": 0.3855,
+      "num_input_tokens_seen": 190785416,
+      "step": 3460,
+      "train_runtime": 20839.0584,
+      "train_tokens_per_second": 9155.184
+    },
+    {
+      "epoch": 0.7113529049476494,
+      "grad_norm": 0.6425190506469415,
+      "learning_rate": 7.191598092644435e-06,
+      "loss": 0.3468,
+      "num_input_tokens_seen": 191051536,
+      "step": 3465,
+      "train_runtime": 20866.1809,
+      "train_tokens_per_second": 9156.038
+    },
+    {
+      "epoch": 0.7123793882159721,
+      "grad_norm": 0.694934818689562,
+      "learning_rate": 7.184348984016305e-06,
+      "loss": 0.3436,
+      "num_input_tokens_seen": 191336904,
+      "step": 3470,
+      "train_runtime": 20899.0784,
+      "train_tokens_per_second": 9155.279
+    },
+    {
+      "epoch": 0.7134058714842948,
+      "grad_norm": 0.6352621844675539,
+      "learning_rate": 7.17709419647231e-06,
+      "loss": 0.3494,
+      "num_input_tokens_seen": 191615944,
+      "step": 3475,
+      "train_runtime": 20930.1909,
+      "train_tokens_per_second": 9155.002
+    },
+    {
+      "epoch": 0.7144323547526176,
+      "grad_norm": 0.6080475944953755,
+      "learning_rate": 7.169833748873598e-06,
+      "loss": 0.3506,
+      "num_input_tokens_seen": 191912664,
+      "step": 3480,
+      "train_runtime": 20959.3212,
+      "train_tokens_per_second": 9156.435
+    },
+    {
+      "epoch": 0.7154588380209402,
+      "grad_norm": 0.761818974468017,
+      "learning_rate": 7.162567660096035e-06,
+      "loss": 0.3831,
+      "num_input_tokens_seen": 192185200,
+      "step": 3485,
+      "train_runtime": 20987.115,
+      "train_tokens_per_second": 9157.295
+    },
+    {
+      "epoch": 0.716485321289263,
+      "grad_norm": 0.6614288624307966,
+      "learning_rate": 7.155295949030143e-06,
+      "loss": 0.3479,
+      "num_input_tokens_seen": 192461864,
+      "step": 3490,
+      "train_runtime": 21020.9708,
+      "train_tokens_per_second": 9155.708
+    },
+    {
+      "epoch": 0.7175118045575857,
+      "grad_norm": 1.101655639413893,
+      "learning_rate": 7.1480186345810735e-06,
+      "loss": 0.3559,
+      "num_input_tokens_seen": 192746592,
+      "step": 3495,
+      "train_runtime": 21050.5474,
+      "train_tokens_per_second": 9156.37
+    },
+    {
+      "epoch": 0.7185382878259085,
+      "grad_norm": 0.790669819059471,
+      "learning_rate": 7.140735735668536e-06,
+      "loss": 0.387,
+      "num_input_tokens_seen": 193004160,
+      "step": 3500,
+      "train_runtime": 21076.7298,
+      "train_tokens_per_second": 9157.216
+    },
+    {
+      "epoch": 0.7195647710942311,
+      "grad_norm": 0.6526173333955143,
+      "learning_rate": 7.133447271226764e-06,
+      "loss": 0.3896,
+      "num_input_tokens_seen": 193303720,
+      "step": 3505,
+      "train_runtime": 21104.9248,
+      "train_tokens_per_second": 9159.176
+    },
+    {
+      "epoch": 0.7205912543625539,
+      "grad_norm": 0.6921798478516266,
+      "learning_rate": 7.126153260204458e-06,
+      "loss": 0.347,
+      "num_input_tokens_seen": 193561584,
+      "step": 3510,
+      "train_runtime": 21130.3065,
+      "train_tokens_per_second": 9160.377
+    },
+    {
+      "epoch": 0.7216177376308767,
+      "grad_norm": 0.7218975312153687,
+      "learning_rate": 7.1188537215647405e-06,
+      "loss": 0.3409,
+      "num_input_tokens_seen": 193816240,
+      "step": 3515,
+      "train_runtime": 21163.294,
+      "train_tokens_per_second": 9158.132
+    },
+    {
+      "epoch": 0.7226442208991993,
+      "grad_norm": 0.75948497029929,
+      "learning_rate": 7.111548674285102e-06,
+      "loss": 0.3554,
+      "num_input_tokens_seen": 194089072,
+      "step": 3520,
+      "train_runtime": 21194.625,
+      "train_tokens_per_second": 9157.467
+    },
+    {
+      "epoch": 0.7236707041675221,
+      "grad_norm": 0.7573749996658645,
+      "learning_rate": 7.1042381373573575e-06,
+      "loss": 0.3777,
+      "num_input_tokens_seen": 194350840,
+      "step": 3525,
+      "train_runtime": 21223.3631,
+      "train_tokens_per_second": 9157.401
+    },
+    {
+      "epoch": 0.7246971874358448,
+      "grad_norm": 0.6319366660799114,
+      "learning_rate": 7.096922129787593e-06,
+      "loss": 0.3323,
+      "num_input_tokens_seen": 194621712,
+      "step": 3530,
+      "train_runtime": 21249.0563,
+      "train_tokens_per_second": 9159.076
+    },
+    {
+      "epoch": 0.7257236707041675,
+      "grad_norm": 0.5881491019787554,
+      "learning_rate": 7.089600670596116e-06,
+      "loss": 0.3616,
+      "num_input_tokens_seen": 194912424,
+      "step": 3535,
+      "train_runtime": 21280.0732,
+      "train_tokens_per_second": 9159.387
+    },
+    {
+      "epoch": 0.7267501539724902,
+      "grad_norm": 0.842193617189023,
+      "learning_rate": 7.082273778817409e-06,
+      "loss": 0.4046,
+      "num_input_tokens_seen": 195163568,
+      "step": 3540,
+      "train_runtime": 21306.4649,
+      "train_tokens_per_second": 9159.829
+    },
+    {
+      "epoch": 0.727776637240813,
+      "grad_norm": 0.6200277190236743,
+      "learning_rate": 7.074941473500075e-06,
+      "loss": 0.3598,
+      "num_input_tokens_seen": 195459040,
+      "step": 3545,
+      "train_runtime": 21337.7362,
+      "train_tokens_per_second": 9160.252
+    },
+    {
+      "epoch": 0.7288031205091356,
+      "grad_norm": 0.6350563279640663,
+      "learning_rate": 7.067603773706799e-06,
+      "loss": 0.3903,
+      "num_input_tokens_seen": 195755232,
+      "step": 3550,
+      "train_runtime": 21372.161,
+      "train_tokens_per_second": 9159.356
+    },
+    {
+      "epoch": 0.7298296037774584,
+      "grad_norm": 0.608443895662852,
+      "learning_rate": 7.060260698514279e-06,
+      "loss": 0.357,
+      "num_input_tokens_seen": 196053520,
+      "step": 3555,
+      "train_runtime": 21406.0788,
+      "train_tokens_per_second": 9158.778
+    },
+    {
+      "epoch": 0.7308560870457812,
+      "grad_norm": 0.5970086201981762,
+      "learning_rate": 7.052912267013199e-06,
+      "loss": 0.3285,
+      "num_input_tokens_seen": 196300800,
+      "step": 3560,
+      "train_runtime": 21441.5865,
+      "train_tokens_per_second": 9155.143
+    },
+    {
+      "epoch": 0.7318825703141039,
+      "grad_norm": 0.8501049310650792,
+      "learning_rate": 7.0455584983081615e-06,
+      "loss": 0.3821,
+      "num_input_tokens_seen": 196555376,
+      "step": 3565,
+      "train_runtime": 21475.3492,
+      "train_tokens_per_second": 9152.604
+    },
+    {
+      "epoch": 0.7329090535824266,
+      "grad_norm": 0.6721645719336174,
+      "learning_rate": 7.0381994115176476e-06,
+      "loss": 0.3497,
+      "num_input_tokens_seen": 196821496,
+      "step": 3570,
+      "train_runtime": 21505.642,
+      "train_tokens_per_second": 9152.087
+    },
+    {
+      "epoch": 0.7339355368507493,
+      "grad_norm": 0.7077016155544539,
+      "learning_rate": 7.030835025773966e-06,
+      "loss": 0.3447,
+      "num_input_tokens_seen": 197097056,
+      "step": 3575,
+      "train_runtime": 21539.5904,
+      "train_tokens_per_second": 9150.455
+    },
+    {
+      "epoch": 0.7349620201190721,
+      "grad_norm": 0.6310311550944414,
+      "learning_rate": 7.023465360223197e-06,
+      "loss": 0.3581,
+      "num_input_tokens_seen": 197343240,
+      "step": 3580,
+      "train_runtime": 21569.3875,
+      "train_tokens_per_second": 9149.228
+    },
+    {
+      "epoch": 0.7359885033873947,
+      "grad_norm": 0.610073663868028,
+      "learning_rate": 7.016090434025151e-06,
+      "loss": 0.3629,
+      "num_input_tokens_seen": 197607496,
+      "step": 3585,
+      "train_runtime": 21598.009,
+      "train_tokens_per_second": 9149.339
+    },
+    {
+      "epoch": 0.7370149866557175,
+      "grad_norm": 0.5885805441478481,
+      "learning_rate": 7.008710266353316e-06,
+      "loss": 0.3536,
+      "num_input_tokens_seen": 197881960,
+      "step": 3590,
+      "train_runtime": 21628.3208,
+      "train_tokens_per_second": 9149.206
+    },
+    {
+      "epoch": 0.7380414699240402,
+      "grad_norm": 0.6610214247194193,
+      "learning_rate": 7.001324876394803e-06,
+      "loss": 0.3547,
+      "num_input_tokens_seen": 198151480,
+      "step": 3595,
+      "train_runtime": 21658.3188,
+      "train_tokens_per_second": 9148.978
+    },
+    {
+      "epoch": 0.739067953192363,
+      "grad_norm": 0.6061407259537787,
+      "learning_rate": 6.993934283350303e-06,
+      "loss": 0.3835,
+      "num_input_tokens_seen": 198423264,
+      "step": 3600,
+      "train_runtime": 21687.1196,
+      "train_tokens_per_second": 9149.36
+    },
+    {
+      "epoch": 0.7400944364606857,
+      "grad_norm": 0.6863116041352865,
+      "learning_rate": 6.986538506434033e-06,
+      "loss": 0.3595,
+      "num_input_tokens_seen": 198708688,
+      "step": 3605,
+      "train_runtime": 21722.0882,
+      "train_tokens_per_second": 9147.771
+    },
+    {
+      "epoch": 0.7411209197290084,
+      "grad_norm": 0.6572492524279813,
+      "learning_rate": 6.979137564873688e-06,
+      "loss": 0.3432,
+      "num_input_tokens_seen": 198975800,
+      "step": 3610,
+      "train_runtime": 21751.4781,
+      "train_tokens_per_second": 9147.691
+    },
+    {
+      "epoch": 0.7421474029973312,
+      "grad_norm": 0.6446032371839616,
+      "learning_rate": 6.9717314779103896e-06,
+      "loss": 0.382,
+      "num_input_tokens_seen": 199275424,
+      "step": 3615,
+      "train_runtime": 21785.9921,
+      "train_tokens_per_second": 9146.952
+    },
+    {
+      "epoch": 0.7431738862656538,
+      "grad_norm": 0.7063090430059813,
+      "learning_rate": 6.964320264798636e-06,
+      "loss": 0.3687,
+      "num_input_tokens_seen": 199556848,
+      "step": 3620,
+      "train_runtime": 21815.332,
+      "train_tokens_per_second": 9147.55
+    },
+    {
+      "epoch": 0.7442003695339766,
+      "grad_norm": 0.7081356120292099,
+      "learning_rate": 6.956903944806254e-06,
+      "loss": 0.3625,
+      "num_input_tokens_seen": 199831672,
+      "step": 3625,
+      "train_runtime": 21849.4444,
+      "train_tokens_per_second": 9145.847
+    },
+    {
+      "epoch": 0.7452268528022993,
+      "grad_norm": 0.7450606721732596,
+      "learning_rate": 6.949482537214346e-06,
+      "loss": 0.3602,
+      "num_input_tokens_seen": 200112160,
+      "step": 3630,
+      "train_runtime": 21882.9523,
+      "train_tokens_per_second": 9144.66
+    },
+    {
+      "epoch": 0.746253336070622,
+      "grad_norm": 0.6489509057076497,
+      "learning_rate": 6.94205606131724e-06,
+      "loss": 0.3532,
+      "num_input_tokens_seen": 200374168,
+      "step": 3635,
+      "train_runtime": 21910.1579,
+      "train_tokens_per_second": 9145.264
+    },
+    {
+      "epoch": 0.7472798193389448,
+      "grad_norm": 0.6383179227405337,
+      "learning_rate": 6.9346245364224455e-06,
+      "loss": 0.3703,
+      "num_input_tokens_seen": 200658192,
+      "step": 3640,
+      "train_runtime": 21936.7305,
+      "train_tokens_per_second": 9147.133
+    },
+    {
+      "epoch": 0.7483063026072675,
+      "grad_norm": 0.663814046034433,
+      "learning_rate": 6.927187981850591e-06,
+      "loss": 0.3616,
+      "num_input_tokens_seen": 200955200,
+      "step": 3645,
+      "train_runtime": 21965.542,
+      "train_tokens_per_second": 9148.657
+    },
+    {
+      "epoch": 0.7493327858755903,
+      "grad_norm": 0.766236046070086,
+      "learning_rate": 6.91974641693539e-06,
+      "loss": 0.3507,
+      "num_input_tokens_seen": 201229744,
+      "step": 3650,
+      "train_runtime": 21995.6005,
+      "train_tokens_per_second": 9148.636
+    },
+    {
+      "epoch": 0.7503592691439129,
+      "grad_norm": 0.6639910569496953,
+      "learning_rate": 6.9122998610235756e-06,
+      "loss": 0.3291,
+      "num_input_tokens_seen": 201506520,
+      "step": 3655,
+      "train_runtime": 22021.1692,
+      "train_tokens_per_second": 9150.582
+    },
+    {
+      "epoch": 0.7513857524122357,
+      "grad_norm": 0.5808126490859832,
+      "learning_rate": 6.904848333474857e-06,
+      "loss": 0.3481,
+      "num_input_tokens_seen": 201780352,
+      "step": 3660,
+      "train_runtime": 22053.9651,
+      "train_tokens_per_second": 9149.391
+    },
+    {
+      "epoch": 0.7524122356805584,
+      "grad_norm": 0.5981956325683252,
+      "learning_rate": 6.897391853661874e-06,
+      "loss": 0.34,
+      "num_input_tokens_seen": 202027984,
+      "step": 3665,
+      "train_runtime": 22086.3575,
+      "train_tokens_per_second": 9147.184
+    },
+    {
+      "epoch": 0.7534387189488811,
+      "grad_norm": 0.6629514709963625,
+      "learning_rate": 6.889930440970134e-06,
+      "loss": 0.3377,
+      "num_input_tokens_seen": 202335104,
+      "step": 3670,
+      "train_runtime": 22116.5032,
+      "train_tokens_per_second": 9148.603
+    },
+    {
+      "epoch": 0.7544652022172038,
+      "grad_norm": 0.5987104811853841,
+      "learning_rate": 6.882464114797974e-06,
+      "loss": 0.3443,
+      "num_input_tokens_seen": 202585256,
+      "step": 3675,
+      "train_runtime": 22148.4428,
+      "train_tokens_per_second": 9146.704
+    },
+    {
+      "epoch": 0.7554916854855266,
+      "grad_norm": 0.8197330245771666,
+      "learning_rate": 6.874992894556506e-06,
+      "loss": 0.3385,
+      "num_input_tokens_seen": 202866224,
+      "step": 3680,
+      "train_runtime": 22175.1402,
+      "train_tokens_per_second": 9148.363
+    },
+    {
+      "epoch": 0.7565181687538494,
+      "grad_norm": 0.7000660461434621,
+      "learning_rate": 6.8675167996695604e-06,
+      "loss": 0.3527,
+      "num_input_tokens_seen": 203150904,
+      "step": 3685,
+      "train_runtime": 22207.0486,
+      "train_tokens_per_second": 9148.037
+    },
+    {
+      "epoch": 0.757544652022172,
+      "grad_norm": 0.6709409106508771,
+      "learning_rate": 6.860035849573645e-06,
+      "loss": 0.3534,
+      "num_input_tokens_seen": 203403088,
+      "step": 3690,
+      "train_runtime": 22237.5957,
+      "train_tokens_per_second": 9146.811
+    },
+    {
+      "epoch": 0.7585711352904948,
+      "grad_norm": 0.6635333387714903,
+      "learning_rate": 6.85255006371789e-06,
+      "loss": 0.3735,
+      "num_input_tokens_seen": 203684888,
+      "step": 3695,
+      "train_runtime": 22267.8465,
+      "train_tokens_per_second": 9147.04
+    },
+    {
+      "epoch": 0.7595976185588175,
+      "grad_norm": 0.6970713171874674,
+      "learning_rate": 6.845059461563996e-06,
+      "loss": 0.3744,
+      "num_input_tokens_seen": 203948456,
+      "step": 3700,
+      "train_runtime": 22293.9925,
+      "train_tokens_per_second": 9148.135
+    },
+    {
+      "epoch": 0.7606241018271402,
+      "grad_norm": 0.6103197488716121,
+      "learning_rate": 6.837564062586187e-06,
+      "loss": 0.3606,
+      "num_input_tokens_seen": 204213872,
+      "step": 3705,
+      "train_runtime": 22326.5012,
+      "train_tokens_per_second": 9146.703
+    },
+    {
+      "epoch": 0.7616505850954629,
+      "grad_norm": 0.7771578590544295,
+      "learning_rate": 6.830063886271158e-06,
+      "loss": 0.3607,
+      "num_input_tokens_seen": 204491584,
+      "step": 3710,
+      "train_runtime": 22353.9974,
+      "train_tokens_per_second": 9147.875
+    },
+    {
+      "epoch": 0.7626770683637857,
+      "grad_norm": 0.6324180610258376,
+      "learning_rate": 6.82255895211802e-06,
+      "loss": 0.3391,
+      "num_input_tokens_seen": 204751544,
+      "step": 3715,
+      "train_runtime": 22384.8362,
+      "train_tokens_per_second": 9146.886
+    },
+    {
+      "epoch": 0.7637035516321083,
+      "grad_norm": 0.5967484992664757,
+      "learning_rate": 6.815049279638259e-06,
+      "loss": 0.3664,
+      "num_input_tokens_seen": 205021200,
+      "step": 3720,
+      "train_runtime": 22414.1025,
+      "train_tokens_per_second": 9146.973
+    },
+    {
+      "epoch": 0.7647300349004311,
+      "grad_norm": 0.6499315396697911,
+      "learning_rate": 6.807534888355676e-06,
+      "loss": 0.3617,
+      "num_input_tokens_seen": 205293568,
+      "step": 3725,
+      "train_runtime": 22442.4599,
+      "train_tokens_per_second": 9147.552
+    },
+    {
+      "epoch": 0.7657565181687539,
+      "grad_norm": 0.5999067421602027,
+      "learning_rate": 6.8000157978063454e-06,
+      "loss": 0.3446,
+      "num_input_tokens_seen": 205582872,
+      "step": 3730,
+      "train_runtime": 22472.1642,
+      "train_tokens_per_second": 9148.334
+    },
+    {
+      "epoch": 0.7667830014370766,
+      "grad_norm": 0.6105708071006253,
+      "learning_rate": 6.792492027538548e-06,
+      "loss": 0.342,
+      "num_input_tokens_seen": 205863152,
+      "step": 3735,
+      "train_runtime": 22507.2494,
+      "train_tokens_per_second": 9146.526
+    },
+    {
+      "epoch": 0.7678094847053993,
+      "grad_norm": 0.596734614103463,
+      "learning_rate": 6.784963597112746e-06,
+      "loss": 0.3628,
+      "num_input_tokens_seen": 206136032,
+      "step": 3740,
+      "train_runtime": 22541.7157,
+      "train_tokens_per_second": 9144.647
+    },
+    {
+      "epoch": 0.768835967973722,
+      "grad_norm": 0.740348282608641,
+      "learning_rate": 6.777430526101503e-06,
+      "loss": 0.3632,
+      "num_input_tokens_seen": 206393840,
+      "step": 3745,
+      "train_runtime": 22572.36,
+      "train_tokens_per_second": 9143.654
+    },
+    {
+      "epoch": 0.7698624512420448,
+      "grad_norm": 0.5977371433103937,
+      "learning_rate": 6.769892834089455e-06,
+      "loss": 0.3484,
+      "num_input_tokens_seen": 206630448,
+      "step": 3750,
+      "train_runtime": 22601.0271,
+      "train_tokens_per_second": 9142.525
+    },
+    {
+      "epoch": 0.7708889345103674,
+      "grad_norm": 0.6826239786774451,
+      "learning_rate": 6.762350540673254e-06,
+      "loss": 0.3598,
+      "num_input_tokens_seen": 206914608,
+      "step": 3755,
+      "train_runtime": 22633.7343,
+      "train_tokens_per_second": 9141.868
+    },
+    {
+      "epoch": 0.7719154177786902,
+      "grad_norm": 0.7011374777022688,
+      "learning_rate": 6.754803665461507e-06,
+      "loss": 0.3745,
+      "num_input_tokens_seen": 207167592,
+      "step": 3760,
+      "train_runtime": 22665.9786,
+      "train_tokens_per_second": 9140.024
+    },
+    {
+      "epoch": 0.7729419010470129,
+      "grad_norm": 0.8199033984229976,
+      "learning_rate": 6.7472522280747385e-06,
+      "loss": 0.3843,
+      "num_input_tokens_seen": 207412256,
+      "step": 3765,
+      "train_runtime": 22695.6018,
+      "train_tokens_per_second": 9138.874
+    },
+    {
+      "epoch": 0.7739683843153357,
+      "grad_norm": 0.5991539352484693,
+      "learning_rate": 6.739696248145331e-06,
+      "loss": 0.3226,
+      "num_input_tokens_seen": 207698912,
+      "step": 3770,
+      "train_runtime": 22732.9357,
+      "train_tokens_per_second": 9136.476
+    },
+    {
+      "epoch": 0.7749948675836584,
+      "grad_norm": 0.6492552817087606,
+      "learning_rate": 6.7321357453174805e-06,
+      "loss": 0.3661,
+      "num_input_tokens_seen": 207941064,
+      "step": 3775,
+      "train_runtime": 22761.9223,
+      "train_tokens_per_second": 9135.479
+    },
+    {
+      "epoch": 0.7760213508519811,
+      "grad_norm": 0.6548908735986675,
+      "learning_rate": 6.724570739247137e-06,
+      "loss": 0.3533,
+      "num_input_tokens_seen": 208225856,
+      "step": 3780,
+      "train_runtime": 22793.4449,
+      "train_tokens_per_second": 9135.339
+    },
+    {
+      "epoch": 0.7770478341203039,
+      "grad_norm": 0.6093496292542123,
+      "learning_rate": 6.717001249601961e-06,
+      "loss": 0.3422,
+      "num_input_tokens_seen": 208513984,
+      "step": 3785,
+      "train_runtime": 22823.9955,
+      "train_tokens_per_second": 9135.735
+    },
+    {
+      "epoch": 0.7780743173886265,
+      "grad_norm": 0.7608504231215985,
+      "learning_rate": 6.70942729606127e-06,
+      "loss": 0.3504,
+      "num_input_tokens_seen": 208759776,
+      "step": 3790,
+      "train_runtime": 22848.9782,
+      "train_tokens_per_second": 9136.504
+    },
+    {
+      "epoch": 0.7791008006569493,
+      "grad_norm": 0.7186401950692669,
+      "learning_rate": 6.701848898315985e-06,
+      "loss": 0.359,
+      "num_input_tokens_seen": 209025456,
+      "step": 3795,
+      "train_runtime": 22878.7809,
+      "train_tokens_per_second": 9136.215
+    },
+    {
+      "epoch": 0.780127283925272,
+      "grad_norm": 0.6031765972729909,
+      "learning_rate": 6.694266076068581e-06,
+      "loss": 0.3576,
+      "num_input_tokens_seen": 209293040,
+      "step": 3800,
+      "train_runtime": 22905.4158,
+      "train_tokens_per_second": 9137.273
+    },
+    {
+      "epoch": 0.7811537671935948,
+      "grad_norm": 0.6264513899905447,
+      "learning_rate": 6.686678849033038e-06,
+      "loss": 0.3633,
+      "num_input_tokens_seen": 209577096,
+      "step": 3805,
+      "train_runtime": 22932.8705,
+      "train_tokens_per_second": 9138.721
+    },
+    {
+      "epoch": 0.7821802504619175,
+      "grad_norm": 0.6598791238172897,
+      "learning_rate": 6.679087236934787e-06,
+      "loss": 0.3582,
+      "num_input_tokens_seen": 209846704,
+      "step": 3810,
+      "train_runtime": 22959.1728,
+      "train_tokens_per_second": 9139.994
+    },
+    {
+      "epoch": 0.7832067337302402,
+      "grad_norm": 0.6825311401598871,
+      "learning_rate": 6.671491259510656e-06,
+      "loss": 0.3913,
+      "num_input_tokens_seen": 210109472,
+      "step": 3815,
+      "train_runtime": 22994.3452,
+      "train_tokens_per_second": 9137.441
+    },
+    {
+      "epoch": 0.784233216998563,
+      "grad_norm": 0.6659510488615709,
+      "learning_rate": 6.663890936508828e-06,
+      "loss": 0.3165,
+      "num_input_tokens_seen": 210378800,
+      "step": 3820,
+      "train_runtime": 23020.7151,
+      "train_tokens_per_second": 9138.674
+    },
+    {
+      "epoch": 0.7852597002668856,
+      "grad_norm": 0.6390766087995401,
+      "learning_rate": 6.656286287688779e-06,
+      "loss": 0.3624,
+      "num_input_tokens_seen": 210641960,
+      "step": 3825,
+      "train_runtime": 23044.9644,
+      "train_tokens_per_second": 9140.477
+    },
+    {
+      "epoch": 0.7862861835352084,
+      "grad_norm": 0.797433208369792,
+      "learning_rate": 6.648677332821234e-06,
+      "loss": 0.371,
+      "num_input_tokens_seen": 210942376,
+      "step": 3830,
+      "train_runtime": 23077.8336,
+      "train_tokens_per_second": 9140.476
+    },
+    {
+      "epoch": 0.7873126668035311,
+      "grad_norm": 0.9858740453260729,
+      "learning_rate": 6.641064091688112e-06,
+      "loss": 0.3753,
+      "num_input_tokens_seen": 211210920,
+      "step": 3835,
+      "train_runtime": 23107.5434,
+      "train_tokens_per_second": 9140.345
+    },
+    {
+      "epoch": 0.7883391500718538,
+      "grad_norm": 0.7542640840297762,
+      "learning_rate": 6.6334465840824725e-06,
+      "loss": 0.3537,
+      "num_input_tokens_seen": 211487880,
+      "step": 3840,
+      "train_runtime": 23139.5375,
+      "train_tokens_per_second": 9139.676
+    },
+    {
+      "epoch": 0.7893656333401765,
+      "grad_norm": 0.7282222776818369,
+      "learning_rate": 6.625824829808475e-06,
+      "loss": 0.3605,
+      "num_input_tokens_seen": 211740576,
+      "step": 3845,
+      "train_runtime": 23167.0326,
+      "train_tokens_per_second": 9139.737
+    },
+    {
+      "epoch": 0.7903921166084993,
+      "grad_norm": 0.7960682833246572,
+      "learning_rate": 6.61819884868131e-06,
+      "loss": 0.3477,
+      "num_input_tokens_seen": 212013088,
+      "step": 3850,
+      "train_runtime": 23198.1949,
+      "train_tokens_per_second": 9139.206
+    },
+    {
+      "epoch": 0.7914185998768221,
+      "grad_norm": 0.6352050288361131,
+      "learning_rate": 6.610568660527164e-06,
+      "loss": 0.3529,
+      "num_input_tokens_seen": 212290080,
+      "step": 3855,
+      "train_runtime": 23230.8454,
+      "train_tokens_per_second": 9138.285
+    },
+    {
+      "epoch": 0.7924450831451447,
+      "grad_norm": 0.5882136838672009,
+      "learning_rate": 6.602934285183158e-06,
+      "loss": 0.3526,
+      "num_input_tokens_seen": 212560096,
+      "step": 3860,
+      "train_runtime": 23262.49,
+      "train_tokens_per_second": 9137.461
+    },
+    {
+      "epoch": 0.7934715664134675,
+      "grad_norm": 0.59408772874702,
+      "learning_rate": 6.5952957424973e-06,
+      "loss": 0.3244,
+      "num_input_tokens_seen": 212841408,
+      "step": 3865,
+      "train_runtime": 23291.5368,
+      "train_tokens_per_second": 9138.144
+    },
+    {
+      "epoch": 0.7944980496817902,
+      "grad_norm": 0.6652788309072202,
+      "learning_rate": 6.587653052328432e-06,
+      "loss": 0.3564,
+      "num_input_tokens_seen": 213105544,
+      "step": 3870,
+      "train_runtime": 23319.7214,
+      "train_tokens_per_second": 9138.426
+    },
+    {
+      "epoch": 0.7955245329501129,
+      "grad_norm": 0.70905894935115,
+      "learning_rate": 6.580006234546178e-06,
+      "loss": 0.3288,
+      "num_input_tokens_seen": 213356416,
+      "step": 3875,
+      "train_runtime": 23345.3034,
+      "train_tokens_per_second": 9139.158
+    },
+    {
+      "epoch": 0.7965510162184356,
+      "grad_norm": 0.6879099342554321,
+      "learning_rate": 6.572355309030894e-06,
+      "loss": 0.3477,
+      "num_input_tokens_seen": 213636376,
+      "step": 3880,
+      "train_runtime": 23382.063,
+      "train_tokens_per_second": 9136.763
+    },
+    {
+      "epoch": 0.7975774994867584,
+      "grad_norm": 0.7391225137739948,
+      "learning_rate": 6.564700295673616e-06,
+      "loss": 0.3272,
+      "num_input_tokens_seen": 213920072,
+      "step": 3885,
+      "train_runtime": 23413.747,
+      "train_tokens_per_second": 9136.516
+    },
+    {
+      "epoch": 0.798603982755081,
+      "grad_norm": 0.9674517472667642,
+      "learning_rate": 6.557041214376006e-06,
+      "loss": 0.3657,
+      "num_input_tokens_seen": 214203952,
+      "step": 3890,
+      "train_runtime": 23444.584,
+      "train_tokens_per_second": 9136.607
+    },
+    {
+      "epoch": 0.7996304660234038,
+      "grad_norm": 1.1081541328520252,
+      "learning_rate": 6.549378085050302e-06,
+      "loss": 0.391,
+      "num_input_tokens_seen": 214483056,
+      "step": 3895,
+      "train_runtime": 23472.93,
+      "train_tokens_per_second": 9137.464
+    },
+    {
+      "epoch": 0.8006569492917266,
+      "grad_norm": 0.692661843811583,
+      "learning_rate": 6.541710927619268e-06,
+      "loss": 0.355,
+      "num_input_tokens_seen": 214741424,
+      "step": 3900,
+      "train_runtime": 23499.5946,
+      "train_tokens_per_second": 9138.091
+    },
+    {
+      "epoch": 0.8016834325600493,
+      "grad_norm": 0.6580866848684346,
+      "learning_rate": 6.534039762016138e-06,
+      "loss": 0.3683,
+      "num_input_tokens_seen": 215039480,
+      "step": 3905,
+      "train_runtime": 23527.7371,
+      "train_tokens_per_second": 9139.828
+    },
+    {
+      "epoch": 0.802709915828372,
+      "grad_norm": 0.638283515894594,
+      "learning_rate": 6.526364608184569e-06,
+      "loss": 0.3612,
+      "num_input_tokens_seen": 215322448,
+      "step": 3910,
+      "train_runtime": 23555.2901,
+      "train_tokens_per_second": 9141.15
+    },
+    {
+      "epoch": 0.8037363990966947,
+      "grad_norm": 0.6923447292707265,
+      "learning_rate": 6.518685486078584e-06,
+      "loss": 0.3403,
+      "num_input_tokens_seen": 215601792,
+      "step": 3915,
+      "train_runtime": 23580.9742,
+      "train_tokens_per_second": 9143.04
+    },
+    {
+      "epoch": 0.8047628823650175,
+      "grad_norm": 0.6621741800095771,
+      "learning_rate": 6.5110024156625254e-06,
+      "loss": 0.3508,
+      "num_input_tokens_seen": 215875040,
+      "step": 3920,
+      "train_runtime": 23614.649,
+      "train_tokens_per_second": 9141.573
+    },
+    {
+      "epoch": 0.8057893656333401,
+      "grad_norm": 1.0307123278436146,
+      "learning_rate": 6.503315416910998e-06,
+      "loss": 0.3902,
+      "num_input_tokens_seen": 216162656,
+      "step": 3925,
+      "train_runtime": 23638.217,
+      "train_tokens_per_second": 9144.626
+    },
+    {
+      "epoch": 0.8068158489016629,
+      "grad_norm": 0.6526984227892959,
+      "learning_rate": 6.495624509808821e-06,
+      "loss": 0.366,
+      "num_input_tokens_seen": 216419936,
+      "step": 3930,
+      "train_runtime": 23670.6891,
+      "train_tokens_per_second": 9142.95
+    },
+    {
+      "epoch": 0.8078423321699856,
+      "grad_norm": 0.8023156813933289,
+      "learning_rate": 6.487929714350977e-06,
+      "loss": 0.3869,
+      "num_input_tokens_seen": 216682184,
+      "step": 3935,
+      "train_runtime": 23697.5863,
+      "train_tokens_per_second": 9143.639
+    },
+    {
+      "epoch": 0.8088688154383084,
+      "grad_norm": 0.7441037257791785,
+      "learning_rate": 6.48023105054255e-06,
+      "loss": 0.3611,
+      "num_input_tokens_seen": 216945712,
+      "step": 3940,
+      "train_runtime": 23729.9477,
+      "train_tokens_per_second": 9142.275
+    },
+    {
+      "epoch": 0.8098952987066311,
+      "grad_norm": 0.605782854968683,
+      "learning_rate": 6.472528538398693e-06,
+      "loss": 0.3651,
+      "num_input_tokens_seen": 217252176,
+      "step": 3945,
+      "train_runtime": 23768.0504,
+      "train_tokens_per_second": 9140.513
+    },
+    {
+      "epoch": 0.8109217819749538,
+      "grad_norm": 0.6502702809787623,
+      "learning_rate": 6.4648221979445504e-06,
+      "loss": 0.3694,
+      "num_input_tokens_seen": 217507008,
+      "step": 3950,
+      "train_runtime": 23797.7069,
+      "train_tokens_per_second": 9139.831
+    },
+    {
+      "epoch": 0.8119482652432766,
+      "grad_norm": 0.8525366210243207,
+      "learning_rate": 6.457112049215228e-06,
+      "loss": 0.3626,
+      "num_input_tokens_seen": 217795272,
+      "step": 3955,
+      "train_runtime": 23827.5156,
+      "train_tokens_per_second": 9140.494
+    },
+    {
+      "epoch": 0.8129747485115992,
+      "grad_norm": 0.6372055132850271,
+      "learning_rate": 6.449398112255735e-06,
+      "loss": 0.3435,
+      "num_input_tokens_seen": 218074632,
+      "step": 3960,
+      "train_runtime": 23858.4689,
+      "train_tokens_per_second": 9140.345
+    },
+    {
+      "epoch": 0.814001231779922,
+      "grad_norm": 0.6604183955775529,
+      "learning_rate": 6.441680407120919e-06,
+      "loss": 0.3645,
+      "num_input_tokens_seen": 218352096,
+      "step": 3965,
+      "train_runtime": 23889.615,
+      "train_tokens_per_second": 9140.042
+    },
+    {
+      "epoch": 0.8150277150482447,
+      "grad_norm": 0.7233675886545727,
+      "learning_rate": 6.433958953875434e-06,
+      "loss": 0.3698,
+      "num_input_tokens_seen": 218637896,
+      "step": 3970,
+      "train_runtime": 23919.9588,
+      "train_tokens_per_second": 9140.396
+    },
+    {
+      "epoch": 0.8160541983165674,
+      "grad_norm": 0.6015608652287643,
+      "learning_rate": 6.426233772593673e-06,
+      "loss": 0.3566,
+      "num_input_tokens_seen": 218927616,
+      "step": 3975,
+      "train_runtime": 23947.2839,
+      "train_tokens_per_second": 9142.065
+    },
+    {
+      "epoch": 0.8170806815848901,
+      "grad_norm": 0.603607371061529,
+      "learning_rate": 6.418504883359723e-06,
+      "loss": 0.3887,
+      "num_input_tokens_seen": 219228256,
+      "step": 3980,
+      "train_runtime": 23983.1524,
+      "train_tokens_per_second": 9140.927
+    },
+    {
+      "epoch": 0.8181071648532129,
+      "grad_norm": 0.6519688700428526,
+      "learning_rate": 6.410772306267311e-06,
+      "loss": 0.3621,
+      "num_input_tokens_seen": 219516392,
+      "step": 3985,
+      "train_runtime": 24011.713,
+      "train_tokens_per_second": 9142.055
+    },
+    {
+      "epoch": 0.8191336481215357,
+      "grad_norm": 0.5957765371282422,
+      "learning_rate": 6.40303606141975e-06,
+      "loss": 0.3616,
+      "num_input_tokens_seen": 219793448,
+      "step": 3990,
+      "train_runtime": 24048.0742,
+      "train_tokens_per_second": 9139.753
+    },
+    {
+      "epoch": 0.8201601313898583,
+      "grad_norm": 0.6245437632927804,
+      "learning_rate": 6.395296168929892e-06,
+      "loss": 0.332,
+      "num_input_tokens_seen": 220077520,
+      "step": 3995,
+      "train_runtime": 24084.5173,
+      "train_tokens_per_second": 9137.718
+    },
+    {
+      "epoch": 0.8211866146581811,
+      "grad_norm": 0.7156311529905475,
+      "learning_rate": 6.38755264892007e-06,
+      "loss": 0.3656,
+      "num_input_tokens_seen": 220365960,
+      "step": 4000,
+      "train_runtime": 24112.8526,
+      "train_tokens_per_second": 9138.942
+    },
+    {
+      "epoch": 0.8222130979265038,
+      "grad_norm": 0.6616317815994622,
+      "learning_rate": 6.379805521522049e-06,
+      "loss": 0.3748,
+      "num_input_tokens_seen": 220635248,
+      "step": 4005,
+      "train_runtime": 24136.1818,
+      "train_tokens_per_second": 9141.266
+    },
+    {
+      "epoch": 0.8232395811948265,
+      "grad_norm": 0.6539076526250981,
+      "learning_rate": 6.37205480687697e-06,
+      "loss": 0.3686,
+      "num_input_tokens_seen": 220898720,
+      "step": 4010,
+      "train_runtime": 24166.3582,
+      "train_tokens_per_second": 9140.753
+    },
+    {
+      "epoch": 0.8242660644631492,
+      "grad_norm": 0.7122282732438637,
+      "learning_rate": 6.364300525135302e-06,
+      "loss": 0.3608,
+      "num_input_tokens_seen": 221162192,
+      "step": 4015,
+      "train_runtime": 24203.5968,
+      "train_tokens_per_second": 9137.575
+    },
+    {
+      "epoch": 0.825292547731472,
+      "grad_norm": 0.5973303152258288,
+      "learning_rate": 6.356542696456787e-06,
+      "loss": 0.3327,
+      "num_input_tokens_seen": 221430888,
+      "step": 4020,
+      "train_runtime": 24233.1021,
+      "train_tokens_per_second": 9137.538
+    },
+    {
+      "epoch": 0.8263190309997948,
+      "grad_norm": 0.6317605069380491,
+      "learning_rate": 6.348781341010394e-06,
+      "loss": 0.3421,
+      "num_input_tokens_seen": 221702136,
+      "step": 4025,
+      "train_runtime": 24265.166,
+      "train_tokens_per_second": 9136.642
+    },
+    {
+      "epoch": 0.8273455142681174,
+      "grad_norm": 0.6034441652021383,
+      "learning_rate": 6.341016478974249e-06,
+      "loss": 0.3865,
+      "num_input_tokens_seen": 221994600,
+      "step": 4030,
+      "train_runtime": 24294.8982,
+      "train_tokens_per_second": 9137.499
+    },
+    {
+      "epoch": 0.8283719975364402,
+      "grad_norm": 0.7588275879517704,
+      "learning_rate": 6.3332481305356065e-06,
+      "loss": 0.3407,
+      "num_input_tokens_seen": 222262744,
+      "step": 4035,
+      "train_runtime": 24331.4594,
+      "train_tokens_per_second": 9134.789
+    },
+    {
+      "epoch": 0.8293984808047629,
+      "grad_norm": 0.6690749142746639,
+      "learning_rate": 6.325476315890779e-06,
+      "loss": 0.3568,
+      "num_input_tokens_seen": 222542944,
+      "step": 4040,
+      "train_runtime": 24361.2066,
+      "train_tokens_per_second": 9135.136
+    },
+    {
+      "epoch": 0.8304249640730856,
+      "grad_norm": 0.6900257848188872,
+      "learning_rate": 6.317701055245091e-06,
+      "loss": 0.3459,
+      "num_input_tokens_seen": 222810936,
+      "step": 4045,
+      "train_runtime": 24395.434,
+      "train_tokens_per_second": 9133.305
+    },
+    {
+      "epoch": 0.8314514473414083,
+      "grad_norm": 0.6534584029952889,
+      "learning_rate": 6.309922368812828e-06,
+      "loss": 0.3593,
+      "num_input_tokens_seen": 223086912,
+      "step": 4050,
+      "train_runtime": 24423.1091,
+      "train_tokens_per_second": 9134.255
+    },
+    {
+      "epoch": 0.8324779306097311,
+      "grad_norm": 0.6894316327169452,
+      "learning_rate": 6.3021402768171794e-06,
+      "loss": 0.3693,
+      "num_input_tokens_seen": 223392680,
+      "step": 4055,
+      "train_runtime": 24452.8868,
+      "train_tokens_per_second": 9135.636
+    },
+    {
+      "epoch": 0.8335044138780537,
+      "grad_norm": 0.6074604433119171,
+      "learning_rate": 6.29435479949019e-06,
+      "loss": 0.3632,
+      "num_input_tokens_seen": 223682200,
+      "step": 4060,
+      "train_runtime": 24484.2267,
+      "train_tokens_per_second": 9135.767
+    },
+    {
+      "epoch": 0.8345308971463765,
+      "grad_norm": 0.6000166451051585,
+      "learning_rate": 6.286565957072704e-06,
+      "loss": 0.3419,
+      "num_input_tokens_seen": 223943144,
+      "step": 4065,
+      "train_runtime": 24515.3765,
+      "train_tokens_per_second": 9134.803
+    },
+    {
+      "epoch": 0.8355573804146993,
+      "grad_norm": 0.642468292904125,
+      "learning_rate": 6.278773769814319e-06,
+      "loss": 0.3703,
+      "num_input_tokens_seen": 224211104,
+      "step": 4070,
+      "train_runtime": 24550.0373,
+      "train_tokens_per_second": 9132.821
+    },
+    {
+      "epoch": 0.836583863683022,
+      "grad_norm": 0.6425460654691968,
+      "learning_rate": 6.270978257973322e-06,
+      "loss": 0.3547,
+      "num_input_tokens_seen": 224492344,
+      "step": 4075,
+      "train_runtime": 24577.3694,
+      "train_tokens_per_second": 9134.108
+    },
+    {
+      "epoch": 0.8376103469513447,
+      "grad_norm": 0.5746677141936519,
+      "learning_rate": 6.263179441816649e-06,
+      "loss": 0.3776,
+      "num_input_tokens_seen": 224763936,
+      "step": 4080,
+      "train_runtime": 24608.4023,
+      "train_tokens_per_second": 9133.626
+    },
+    {
+      "epoch": 0.8386368302196674,
+      "grad_norm": 0.6053922637138237,
+      "learning_rate": 6.2553773416198224e-06,
+      "loss": 0.3227,
+      "num_input_tokens_seen": 225052864,
+      "step": 4085,
+      "train_runtime": 24631.3688,
+      "train_tokens_per_second": 9136.84
+    },
+    {
+      "epoch": 0.8396633134879902,
+      "grad_norm": 0.6521402477922091,
+      "learning_rate": 6.247571977666907e-06,
+      "loss": 0.3462,
+      "num_input_tokens_seen": 225340472,
+      "step": 4090,
+      "train_runtime": 24661.4677,
+      "train_tokens_per_second": 9137.35
+    },
+    {
+      "epoch": 0.8406897967563128,
+      "grad_norm": 0.6830931588297057,
+      "learning_rate": 6.239763370250447e-06,
+      "loss": 0.3682,
+      "num_input_tokens_seen": 225637176,
+      "step": 4095,
+      "train_runtime": 24691.5013,
+      "train_tokens_per_second": 9138.253
+    },
+    {
+      "epoch": 0.8417162800246356,
+      "grad_norm": 0.5865368261548295,
+      "learning_rate": 6.231951539671424e-06,
+      "loss": 0.3557,
+      "num_input_tokens_seen": 225938776,
+      "step": 4100,
+      "train_runtime": 24730.8313,
+      "train_tokens_per_second": 9135.915
+    },
+    {
+      "epoch": 0.8427427632929583,
+      "grad_norm": 0.7374791050620829,
+      "learning_rate": 6.2241365062391965e-06,
+      "loss": 0.366,
+      "num_input_tokens_seen": 226184160,
+      "step": 4105,
+      "train_runtime": 24757.7043,
+      "train_tokens_per_second": 9135.91
+    },
+    {
+      "epoch": 0.843769246561281,
+      "grad_norm": 0.6955282822497169,
+      "learning_rate": 6.2163182902714505e-06,
+      "loss": 0.3732,
+      "num_input_tokens_seen": 226474368,
+      "step": 4110,
+      "train_runtime": 24783.6907,
+      "train_tokens_per_second": 9138.04
+    },
+    {
+      "epoch": 0.8447957298296038,
+      "grad_norm": 0.6920353701852947,
+      "learning_rate": 6.20849691209415e-06,
+      "loss": 0.3479,
+      "num_input_tokens_seen": 226761256,
+      "step": 4115,
+      "train_runtime": 24813.2458,
+      "train_tokens_per_second": 9138.718
+    },
+    {
+      "epoch": 0.8458222130979265,
+      "grad_norm": 0.6817010170332787,
+      "learning_rate": 6.20067239204147e-06,
+      "loss": 0.3834,
+      "num_input_tokens_seen": 227038920,
+      "step": 4120,
+      "train_runtime": 24843.4798,
+      "train_tokens_per_second": 9138.773
+    },
+    {
+      "epoch": 0.8468486963662493,
+      "grad_norm": 0.6839108079047798,
+      "learning_rate": 6.192844750455766e-06,
+      "loss": 0.372,
+      "num_input_tokens_seen": 227330208,
+      "step": 4125,
+      "train_runtime": 24871.4751,
+      "train_tokens_per_second": 9140.198
+    },
+    {
+      "epoch": 0.8478751796345719,
+      "grad_norm": 0.5877321177500816,
+      "learning_rate": 6.1850140076875e-06,
+      "loss": 0.3276,
+      "num_input_tokens_seen": 227618504,
+      "step": 4130,
+      "train_runtime": 24899.6891,
+      "train_tokens_per_second": 9141.42
+    },
+    {
+      "epoch": 0.8489016629028947,
+      "grad_norm": 0.6464043942318515,
+      "learning_rate": 6.1771801840952e-06,
+      "loss": 0.3419,
+      "num_input_tokens_seen": 227899664,
+      "step": 4135,
+      "train_runtime": 24935.3243,
+      "train_tokens_per_second": 9139.631
+    },
+    {
+      "epoch": 0.8499281461712174,
+      "grad_norm": 0.7112348359369695,
+      "learning_rate": 6.169343300045405e-06,
+      "loss": 0.3732,
+      "num_input_tokens_seen": 228156872,
+      "step": 4140,
+      "train_runtime": 24966.7981,
+      "train_tokens_per_second": 9138.411
+    },
+    {
+      "epoch": 0.8509546294395401,
+      "grad_norm": 0.6347078920394771,
+      "learning_rate": 6.161503375912606e-06,
+      "loss": 0.3452,
+      "num_input_tokens_seen": 228431624,
+      "step": 4145,
+      "train_runtime": 24997.5428,
+      "train_tokens_per_second": 9138.163
+    },
+    {
+      "epoch": 0.8519811127078628,
+      "grad_norm": 0.7028410607656278,
+      "learning_rate": 6.153660432079205e-06,
+      "loss": 0.3761,
+      "num_input_tokens_seen": 228700712,
+      "step": 4150,
+      "train_runtime": 25022.5302,
+      "train_tokens_per_second": 9139.792
+    },
+    {
+      "epoch": 0.8530075959761856,
+      "grad_norm": 0.645929033515841,
+      "learning_rate": 6.145814488935448e-06,
+      "loss": 0.3628,
+      "num_input_tokens_seen": 228972288,
+      "step": 4155,
+      "train_runtime": 25058.44,
+      "train_tokens_per_second": 9137.532
+    },
+    {
+      "epoch": 0.8540340792445084,
+      "grad_norm": 0.6538566601317543,
+      "learning_rate": 6.137965566879379e-06,
+      "loss": 0.3947,
+      "num_input_tokens_seen": 229265992,
+      "step": 4160,
+      "train_runtime": 25092.2664,
+      "train_tokens_per_second": 9136.918
+    },
+    {
+      "epoch": 0.855060562512831,
+      "grad_norm": 0.722703974615602,
+      "learning_rate": 6.130113686316792e-06,
+      "loss": 0.3168,
+      "num_input_tokens_seen": 229530032,
+      "step": 4165,
+      "train_runtime": 25117.0505,
+      "train_tokens_per_second": 9138.415
+    },
+    {
+      "epoch": 0.8560870457811538,
+      "grad_norm": 0.8707423036280791,
+      "learning_rate": 6.122258867661165e-06,
+      "loss": 0.3576,
+      "num_input_tokens_seen": 229809528,
+      "step": 4170,
+      "train_runtime": 25147.5739,
+      "train_tokens_per_second": 9138.437
+    },
+    {
+      "epoch": 0.8571135290494765,
+      "grad_norm": 0.6776785917178774,
+      "learning_rate": 6.114401131333624e-06,
+      "loss": 0.3762,
+      "num_input_tokens_seen": 230098472,
+      "step": 4175,
+      "train_runtime": 25181.1216,
+      "train_tokens_per_second": 9137.737
+    },
+    {
+      "epoch": 0.8581400123177992,
+      "grad_norm": 0.670103270189065,
+      "learning_rate": 6.1065404977628715e-06,
+      "loss": 0.3855,
+      "num_input_tokens_seen": 230375152,
+      "step": 4180,
+      "train_runtime": 25203.7535,
+      "train_tokens_per_second": 9140.51
+    },
+    {
+      "epoch": 0.8591664955861219,
+      "grad_norm": 0.6349354254771897,
+      "learning_rate": 6.098676987385147e-06,
+      "loss": 0.352,
+      "num_input_tokens_seen": 230650608,
+      "step": 4185,
+      "train_runtime": 25234.4693,
+      "train_tokens_per_second": 9140.3
+    },
+    {
+      "epoch": 0.8601929788544447,
+      "grad_norm": 0.6464846647455461,
+      "learning_rate": 6.090810620644168e-06,
+      "loss": 0.3421,
+      "num_input_tokens_seen": 230931584,
+      "step": 4190,
+      "train_runtime": 25263.0416,
+      "train_tokens_per_second": 9141.084
+    },
+    {
+      "epoch": 0.8612194621227675,
+      "grad_norm": 0.7080356542532732,
+      "learning_rate": 6.082941417991077e-06,
+      "loss": 0.3204,
+      "num_input_tokens_seen": 231192896,
+      "step": 4195,
+      "train_runtime": 25292.6548,
+      "train_tokens_per_second": 9140.713
+    },
+    {
+      "epoch": 0.8622459453910901,
+      "grad_norm": 0.5847570732937124,
+      "learning_rate": 6.075069399884392e-06,
+      "loss": 0.3438,
+      "num_input_tokens_seen": 231470984,
+      "step": 4200,
+      "train_runtime": 25318.9526,
+      "train_tokens_per_second": 9142.202
+    },
+    {
+      "epoch": 0.8632724286594129,
+      "grad_norm": 1.6949428462024934,
+      "learning_rate": 6.06719458678995e-06,
+      "loss": 0.3527,
+      "num_input_tokens_seen": 231706624,
+      "step": 4205,
+      "train_runtime": 25351.5393,
+      "train_tokens_per_second": 9139.746
+    },
+    {
+      "epoch": 0.8642989119277356,
+      "grad_norm": 0.6296421357648472,
+      "learning_rate": 6.0593169991808535e-06,
+      "loss": 0.3826,
+      "num_input_tokens_seen": 232005912,
+      "step": 4210,
+      "train_runtime": 25379.6171,
+      "train_tokens_per_second": 9141.427
+    },
+    {
+      "epoch": 0.8653253951960583,
+      "grad_norm": 0.6834740235037723,
+      "learning_rate": 6.051436657537415e-06,
+      "loss": 0.3653,
+      "num_input_tokens_seen": 232290616,
+      "step": 4215,
+      "train_runtime": 25415.9282,
+      "train_tokens_per_second": 9139.568
+    },
+    {
+      "epoch": 0.866351878464381,
+      "grad_norm": 0.7552438463521511,
+      "learning_rate": 6.0435535823471145e-06,
+      "loss": 0.3399,
+      "num_input_tokens_seen": 232562712,
+      "step": 4220,
+      "train_runtime": 25444.9252,
+      "train_tokens_per_second": 9139.847
+    },
+    {
+      "epoch": 0.8673783617327038,
+      "grad_norm": 0.5958195136489561,
+      "learning_rate": 6.035667794104532e-06,
+      "loss": 0.3243,
+      "num_input_tokens_seen": 232819392,
+      "step": 4225,
+      "train_runtime": 25473.1033,
+      "train_tokens_per_second": 9139.813
+    },
+    {
+      "epoch": 0.8684048450010264,
+      "grad_norm": 0.5322868993097685,
+      "learning_rate": 6.027779313311306e-06,
+      "loss": 0.3359,
+      "num_input_tokens_seen": 233066672,
+      "step": 4230,
+      "train_runtime": 25498.2431,
+      "train_tokens_per_second": 9140.499
+    },
+    {
+      "epoch": 0.8694313282693492,
+      "grad_norm": 0.8818943848380144,
+      "learning_rate": 6.01988816047607e-06,
+      "loss": 0.3615,
+      "num_input_tokens_seen": 233354360,
+      "step": 4235,
+      "train_runtime": 25528.5706,
+      "train_tokens_per_second": 9140.91
+    },
+    {
+      "epoch": 0.870457811537672,
+      "grad_norm": 0.6221875892989196,
+      "learning_rate": 6.0119943561144075e-06,
+      "loss": 0.3598,
+      "num_input_tokens_seen": 233639680,
+      "step": 4240,
+      "train_runtime": 25561.4791,
+      "train_tokens_per_second": 9140.304
+    },
+    {
+      "epoch": 0.8714842948059947,
+      "grad_norm": 0.6871116540647696,
+      "learning_rate": 6.004097920748798e-06,
+      "loss": 0.3367,
+      "num_input_tokens_seen": 233904392,
+      "step": 4245,
+      "train_runtime": 25594.496,
+      "train_tokens_per_second": 9138.855
+    },
+    {
+      "epoch": 0.8725107780743174,
+      "grad_norm": 0.688699540455089,
+      "learning_rate": 5.996198874908554e-06,
+      "loss": 0.3104,
+      "num_input_tokens_seen": 234182016,
+      "step": 4250,
+      "train_runtime": 25626.4563,
+      "train_tokens_per_second": 9138.291
+    },
+    {
+      "epoch": 0.8735372613426401,
+      "grad_norm": 0.6428182280012115,
+      "learning_rate": 5.988297239129783e-06,
+      "loss": 0.3649,
+      "num_input_tokens_seen": 234427728,
+      "step": 4255,
+      "train_runtime": 25653.7528,
+      "train_tokens_per_second": 9138.146
+    },
+    {
+      "epoch": 0.8745637446109629,
+      "grad_norm": 0.6752652848445431,
+      "learning_rate": 5.980393033955319e-06,
+      "loss": 0.3508,
+      "num_input_tokens_seen": 234674736,
+      "step": 4260,
+      "train_runtime": 25680.4044,
+      "train_tokens_per_second": 9138.28
+    },
+    {
+      "epoch": 0.8755902278792855,
+      "grad_norm": 0.6125776686775365,
+      "learning_rate": 5.97248627993468e-06,
+      "loss": 0.3664,
+      "num_input_tokens_seen": 234954576,
+      "step": 4265,
+      "train_runtime": 25706.315,
+      "train_tokens_per_second": 9139.956
+    },
+    {
+      "epoch": 0.8766167111476083,
+      "grad_norm": 0.6878479859402575,
+      "learning_rate": 5.96457699762401e-06,
+      "loss": 0.3562,
+      "num_input_tokens_seen": 235245616,
+      "step": 4270,
+      "train_runtime": 25735.299,
+      "train_tokens_per_second": 9140.971
+    },
+    {
+      "epoch": 0.877643194415931,
+      "grad_norm": 0.6560473633430887,
+      "learning_rate": 5.956665207586023e-06,
+      "loss": 0.3662,
+      "num_input_tokens_seen": 235531600,
+      "step": 4275,
+      "train_runtime": 25768.0285,
+      "train_tokens_per_second": 9140.459
+    },
+    {
+      "epoch": 0.8786696776842537,
+      "grad_norm": 0.6272916129754538,
+      "learning_rate": 5.948750930389958e-06,
+      "loss": 0.3277,
+      "num_input_tokens_seen": 235799912,
+      "step": 4280,
+      "train_runtime": 25804.277,
+      "train_tokens_per_second": 9138.017
+    },
+    {
+      "epoch": 0.8796961609525765,
+      "grad_norm": 0.6918708176261733,
+      "learning_rate": 5.940834186611515e-06,
+      "loss": 0.3503,
+      "num_input_tokens_seen": 236092840,
+      "step": 4285,
+      "train_runtime": 25833.5635,
+      "train_tokens_per_second": 9138.996
+    },
+    {
+      "epoch": 0.8807226442208992,
+      "grad_norm": 0.6381579192385177,
+      "learning_rate": 5.932914996832809e-06,
+      "loss": 0.3668,
+      "num_input_tokens_seen": 236385968,
+      "step": 4290,
+      "train_runtime": 25856.4621,
+      "train_tokens_per_second": 9142.239
+    },
+    {
+      "epoch": 0.881749127489222,
+      "grad_norm": 0.6281531604303734,
+      "learning_rate": 5.9249933816423166e-06,
+      "loss": 0.3486,
+      "num_input_tokens_seen": 236675368,
+      "step": 4295,
+      "train_runtime": 25890.746,
+      "train_tokens_per_second": 9141.311
+    },
+    {
+      "epoch": 0.8827756107575446,
+      "grad_norm": 0.6303283496781344,
+      "learning_rate": 5.917069361634815e-06,
+      "loss": 0.3103,
+      "num_input_tokens_seen": 236929368,
+      "step": 4300,
+      "train_runtime": 25923.6077,
+      "train_tokens_per_second": 9139.521
+    },
+    {
+      "epoch": 0.8838020940258674,
+      "grad_norm": 0.6638431396538005,
+      "learning_rate": 5.909142957411337e-06,
+      "loss": 0.3519,
+      "num_input_tokens_seen": 237202640,
+      "step": 4305,
+      "train_runtime": 25950.9537,
+      "train_tokens_per_second": 9140.421
+    },
+    {
+      "epoch": 0.8848285772941901,
+      "grad_norm": 0.637817983388844,
+      "learning_rate": 5.901214189579112e-06,
+      "loss": 0.3813,
+      "num_input_tokens_seen": 237474408,
+      "step": 4310,
+      "train_runtime": 25978.1641,
+      "train_tokens_per_second": 9141.308
+    },
+    {
+      "epoch": 0.8858550605625128,
+      "grad_norm": 0.6816448229875455,
+      "learning_rate": 5.893283078751516e-06,
+      "loss": 0.3415,
+      "num_input_tokens_seen": 237752096,
+      "step": 4315,
+      "train_runtime": 26007.3747,
+      "train_tokens_per_second": 9141.718
+    },
+    {
+      "epoch": 0.8868815438308355,
+      "grad_norm": 0.5891189196062315,
+      "learning_rate": 5.885349645548018e-06,
+      "loss": 0.3499,
+      "num_input_tokens_seen": 238009976,
+      "step": 4320,
+      "train_runtime": 26036.9018,
+      "train_tokens_per_second": 9141.256
+    },
+    {
+      "epoch": 0.8879080270991583,
+      "grad_norm": 0.7104028266895371,
+      "learning_rate": 5.877413910594119e-06,
+      "loss": 0.366,
+      "num_input_tokens_seen": 238274760,
+      "step": 4325,
+      "train_runtime": 26064.6076,
+      "train_tokens_per_second": 9141.698
+    },
+    {
+      "epoch": 0.8889345103674811,
+      "grad_norm": 0.6079223325591576,
+      "learning_rate": 5.8694758945213084e-06,
+      "loss": 0.3539,
+      "num_input_tokens_seen": 238544064,
+      "step": 4330,
+      "train_runtime": 26098.5058,
+      "train_tokens_per_second": 9140.143
+    },
+    {
+      "epoch": 0.8899609936358037,
+      "grad_norm": 0.7169098496003778,
+      "learning_rate": 5.8615356179670055e-06,
+      "loss": 0.3577,
+      "num_input_tokens_seen": 238789408,
+      "step": 4335,
+      "train_runtime": 26127.0805,
+      "train_tokens_per_second": 9139.537
+    },
+    {
+      "epoch": 0.8909874769041265,
+      "grad_norm": 0.6056386473690265,
+      "learning_rate": 5.8535931015745065e-06,
+      "loss": 0.3393,
+      "num_input_tokens_seen": 239089272,
+      "step": 4340,
+      "train_runtime": 26158.6816,
+      "train_tokens_per_second": 9139.959
+    },
+    {
+      "epoch": 0.8920139601724492,
+      "grad_norm": 0.5854695225983222,
+      "learning_rate": 5.8456483659929305e-06,
+      "loss": 0.3757,
+      "num_input_tokens_seen": 239378008,
+      "step": 4345,
+      "train_runtime": 26181.0064,
+      "train_tokens_per_second": 9143.194
+    },
+    {
+      "epoch": 0.8930404434407719,
+      "grad_norm": 0.7433168148997688,
+      "learning_rate": 5.837701431877166e-06,
+      "loss": 0.3553,
+      "num_input_tokens_seen": 239617848,
+      "step": 4350,
+      "train_runtime": 26213.2912,
+      "train_tokens_per_second": 9141.082
+    },
+    {
+      "epoch": 0.8940669267090946,
+      "grad_norm": 0.6774457394451351,
+      "learning_rate": 5.8297523198878184e-06,
+      "loss": 0.3451,
+      "num_input_tokens_seen": 239892088,
+      "step": 4355,
+      "train_runtime": 26240.6294,
+      "train_tokens_per_second": 9142.01
+    },
+    {
+      "epoch": 0.8950934099774174,
+      "grad_norm": 0.5923755051361288,
+      "learning_rate": 5.821801050691153e-06,
+      "loss": 0.3561,
+      "num_input_tokens_seen": 240189008,
+      "step": 4360,
+      "train_runtime": 26274.1078,
+      "train_tokens_per_second": 9141.662
+    },
+    {
+      "epoch": 0.89611989324574,
+      "grad_norm": 0.6486437353213257,
+      "learning_rate": 5.813847644959044e-06,
+      "loss": 0.3503,
+      "num_input_tokens_seen": 240466432,
+      "step": 4365,
+      "train_runtime": 26305.4616,
+      "train_tokens_per_second": 9141.312
+    },
+    {
+      "epoch": 0.8971463765140628,
+      "grad_norm": 0.6808661940072406,
+      "learning_rate": 5.805892123368923e-06,
+      "loss": 0.3623,
+      "num_input_tokens_seen": 240739368,
+      "step": 4370,
+      "train_runtime": 26333.6954,
+      "train_tokens_per_second": 9141.876
+    },
+    {
+      "epoch": 0.8981728597823856,
+      "grad_norm": 0.7072059045725927,
+      "learning_rate": 5.797934506603717e-06,
+      "loss": 0.3253,
+      "num_input_tokens_seen": 241024608,
+      "step": 4375,
+      "train_runtime": 26368.5202,
+      "train_tokens_per_second": 9140.619
+    },
+    {
+      "epoch": 0.8991993430507083,
+      "grad_norm": 0.607263945050118,
+      "learning_rate": 5.789974815351806e-06,
+      "loss": 0.334,
+      "num_input_tokens_seen": 241298584,
+      "step": 4380,
+      "train_runtime": 26399.5314,
+      "train_tokens_per_second": 9140.26
+    },
+    {
+      "epoch": 0.900225826319031,
+      "grad_norm": 0.7553276862433587,
+      "learning_rate": 5.78201307030696e-06,
+      "loss": 0.3743,
+      "num_input_tokens_seen": 241560272,
+      "step": 4385,
+      "train_runtime": 26424.1114,
+      "train_tokens_per_second": 9141.661
+    },
+    {
+      "epoch": 0.9012523095873537,
+      "grad_norm": 0.6998122967369986,
+      "learning_rate": 5.77404929216829e-06,
+      "loss": 0.3408,
+      "num_input_tokens_seen": 241829760,
+      "step": 4390,
+      "train_runtime": 26457.3739,
+      "train_tokens_per_second": 9140.354
+    },
+    {
+      "epoch": 0.9022787928556765,
+      "grad_norm": 0.6760524731303175,
+      "learning_rate": 5.766083501640191e-06,
+      "loss": 0.3628,
+      "num_input_tokens_seen": 242105472,
+      "step": 4395,
+      "train_runtime": 26486.1021,
+      "train_tokens_per_second": 9140.849
+    },
+    {
+      "epoch": 0.9033052761239991,
+      "grad_norm": 0.5493921111487259,
+      "learning_rate": 5.75811571943229e-06,
+      "loss": 0.3148,
+      "num_input_tokens_seen": 242397152,
+      "step": 4400,
+      "train_runtime": 26513.4764,
+      "train_tokens_per_second": 9142.413
+    },
+    {
+      "epoch": 0.9043317593923219,
+      "grad_norm": 0.6032414256933115,
+      "learning_rate": 5.750145966259393e-06,
+      "loss": 0.3208,
+      "num_input_tokens_seen": 242676416,
+      "step": 4405,
+      "train_runtime": 26539.5748,
+      "train_tokens_per_second": 9143.945
+    },
+    {
+      "epoch": 0.9053582426606447,
+      "grad_norm": 0.5998341954173685,
+      "learning_rate": 5.742174262841431e-06,
+      "loss": 0.3402,
+      "num_input_tokens_seen": 242961400,
+      "step": 4410,
+      "train_runtime": 26572.4281,
+      "train_tokens_per_second": 9143.365
+    },
+    {
+      "epoch": 0.9063847259289673,
+      "grad_norm": 0.6586905556511449,
+      "learning_rate": 5.734200629903403e-06,
+      "loss": 0.3541,
+      "num_input_tokens_seen": 243244336,
+      "step": 4415,
+      "train_runtime": 26600.0081,
+      "train_tokens_per_second": 9144.521
+    },
+    {
+      "epoch": 0.9074112091972901,
+      "grad_norm": 0.5832123559643287,
+      "learning_rate": 5.7262250881753255e-06,
+      "loss": 0.3553,
+      "num_input_tokens_seen": 243526664,
+      "step": 4420,
+      "train_runtime": 26632.5732,
+      "train_tokens_per_second": 9143.94
+    },
+    {
+      "epoch": 0.9084376924656128,
+      "grad_norm": 0.6339099332991818,
+      "learning_rate": 5.718247658392178e-06,
+      "loss": 0.3604,
+      "num_input_tokens_seen": 243815632,
+      "step": 4425,
+      "train_runtime": 26665.589,
+      "train_tokens_per_second": 9143.456
+    },
+    {
+      "epoch": 0.9094641757339356,
+      "grad_norm": 0.6694618903866862,
+      "learning_rate": 5.710268361293848e-06,
+      "loss": 0.3548,
+      "num_input_tokens_seen": 244115704,
+      "step": 4430,
+      "train_runtime": 26698.7128,
+      "train_tokens_per_second": 9143.351
+    },
+    {
+      "epoch": 0.9104906590022582,
+      "grad_norm": 0.5498681477394135,
+      "learning_rate": 5.70228721762508e-06,
+      "loss": 0.3352,
+      "num_input_tokens_seen": 244384448,
+      "step": 4435,
+      "train_runtime": 26730.9239,
+      "train_tokens_per_second": 9142.387
+    },
+    {
+      "epoch": 0.911517142270581,
+      "grad_norm": 0.5960205349819031,
+      "learning_rate": 5.694304248135413e-06,
+      "loss": 0.3158,
+      "num_input_tokens_seen": 244637632,
+      "step": 4440,
+      "train_runtime": 26759.932,
+      "train_tokens_per_second": 9141.938
+    },
+    {
+      "epoch": 0.9125436255389037,
+      "grad_norm": 0.6714462047165061,
+      "learning_rate": 5.686319473579141e-06,
+      "loss": 0.3426,
+      "num_input_tokens_seen": 244915776,
+      "step": 4445,
+      "train_runtime": 26784.9803,
+      "train_tokens_per_second": 9143.773
+    },
+    {
+      "epoch": 0.9135701088072264,
+      "grad_norm": 0.6702218827989639,
+      "learning_rate": 5.678332914715242e-06,
+      "loss": 0.3641,
+      "num_input_tokens_seen": 245180312,
+      "step": 4450,
+      "train_runtime": 26819.5654,
+      "train_tokens_per_second": 9141.845
+    },
+    {
+      "epoch": 0.9145965920755492,
+      "grad_norm": 0.704829081303585,
+      "learning_rate": 5.670344592307342e-06,
+      "loss": 0.3465,
+      "num_input_tokens_seen": 245445592,
+      "step": 4455,
+      "train_runtime": 26840.3622,
+      "train_tokens_per_second": 9144.645
+    },
+    {
+      "epoch": 0.9156230753438719,
+      "grad_norm": 0.7330209243845758,
+      "learning_rate": 5.662354527123646e-06,
+      "loss": 0.3685,
+      "num_input_tokens_seen": 245714200,
+      "step": 4460,
+      "train_runtime": 26876.365,
+      "train_tokens_per_second": 9142.39
+    },
+    {
+      "epoch": 0.9166495586121947,
+      "grad_norm": 0.6180692274756142,
+      "learning_rate": 5.6543627399368905e-06,
+      "loss": 0.3912,
+      "num_input_tokens_seen": 246006256,
+      "step": 4465,
+      "train_runtime": 26912.6827,
+      "train_tokens_per_second": 9140.904
+    },
+    {
+      "epoch": 0.9176760418805173,
+      "grad_norm": 0.6738667586033908,
+      "learning_rate": 5.646369251524292e-06,
+      "loss": 0.3612,
+      "num_input_tokens_seen": 246302064,
+      "step": 4470,
+      "train_runtime": 26941.5334,
+      "train_tokens_per_second": 9142.095
+    },
+    {
+      "epoch": 0.9187025251488401,
+      "grad_norm": 0.8219431471346768,
+      "learning_rate": 5.6383740826674856e-06,
+      "loss": 0.376,
+      "num_input_tokens_seen": 246580688,
+      "step": 4475,
+      "train_runtime": 26969.7097,
+      "train_tokens_per_second": 9142.875
+    },
+    {
+      "epoch": 0.9197290084171628,
+      "grad_norm": 0.7125002201740719,
+      "learning_rate": 5.630377254152478e-06,
+      "loss": 0.341,
+      "num_input_tokens_seen": 246846144,
+      "step": 4480,
+      "train_runtime": 27002.1329,
+      "train_tokens_per_second": 9141.728
+    },
+    {
+      "epoch": 0.9207554916854855,
+      "grad_norm": 0.5841837868871039,
+      "learning_rate": 5.622378786769591e-06,
+      "loss": 0.3759,
+      "num_input_tokens_seen": 247123392,
+      "step": 4485,
+      "train_runtime": 27029.6631,
+      "train_tokens_per_second": 9142.674
+    },
+    {
+      "epoch": 0.9217819749538082,
+      "grad_norm": 0.7568793871492225,
+      "learning_rate": 5.614378701313405e-06,
+      "loss": 0.3566,
+      "num_input_tokens_seen": 247381704,
+      "step": 4490,
+      "train_runtime": 27055.3697,
+      "train_tokens_per_second": 9143.534
+    },
+    {
+      "epoch": 0.922808458222131,
+      "grad_norm": 0.6528957266466389,
+      "learning_rate": 5.606377018582708e-06,
+      "loss": 0.3729,
+      "num_input_tokens_seen": 247636400,
+      "step": 4495,
+      "train_runtime": 27081.8176,
+      "train_tokens_per_second": 9144.01
+    },
+    {
+      "epoch": 0.9238349414904538,
+      "grad_norm": 0.5994226610223004,
+      "learning_rate": 5.598373759380445e-06,
+      "loss": 0.3361,
+      "num_input_tokens_seen": 247908904,
+      "step": 4500,
+      "train_runtime": 27107.7845,
+      "train_tokens_per_second": 9145.303
+    },
+    {
+      "epoch": 0.9248614247587764,
+      "grad_norm": 0.6312765626387813,
+      "learning_rate": 5.590368944513649e-06,
+      "loss": 0.3552,
+      "num_input_tokens_seen": 248194040,
+      "step": 4505,
+      "train_runtime": 27133.916,
+      "train_tokens_per_second": 9147.004
+    },
+    {
+      "epoch": 0.9258879080270992,
+      "grad_norm": 0.670887689134295,
+      "learning_rate": 5.582362594793409e-06,
+      "loss": 0.3548,
+      "num_input_tokens_seen": 248484184,
+      "step": 4510,
+      "train_runtime": 27170.7279,
+      "train_tokens_per_second": 9145.29
+    },
+    {
+      "epoch": 0.9269143912954219,
+      "grad_norm": 0.6799082070877618,
+      "learning_rate": 5.5743547310347975e-06,
+      "loss": 0.3224,
+      "num_input_tokens_seen": 248752656,
+      "step": 4515,
+      "train_runtime": 27206.6626,
+      "train_tokens_per_second": 9143.079
+    },
+    {
+      "epoch": 0.9279408745637446,
+      "grad_norm": 0.6450705509056616,
+      "learning_rate": 5.566345374056825e-06,
+      "loss": 0.3429,
+      "num_input_tokens_seen": 249026640,
+      "step": 4520,
+      "train_runtime": 27241.655,
+      "train_tokens_per_second": 9141.392
+    },
+    {
+      "epoch": 0.9289673578320673,
+      "grad_norm": 0.6293397539647454,
+      "learning_rate": 5.5583345446823856e-06,
+      "loss": 0.3472,
+      "num_input_tokens_seen": 249304304,
+      "step": 4525,
+      "train_runtime": 27267.581,
+      "train_tokens_per_second": 9142.883
+    },
+    {
+      "epoch": 0.9299938411003901,
+      "grad_norm": 0.6011826064652812,
+      "learning_rate": 5.550322263738197e-06,
+      "loss": 0.3424,
+      "num_input_tokens_seen": 249571072,
+      "step": 4530,
+      "train_runtime": 27298.7787,
+      "train_tokens_per_second": 9142.207
+    },
+    {
+      "epoch": 0.9310203243687127,
+      "grad_norm": 0.5608100025181015,
+      "learning_rate": 5.5423085520547556e-06,
+      "loss": 0.3599,
+      "num_input_tokens_seen": 249867832,
+      "step": 4535,
+      "train_runtime": 27328.6704,
+      "train_tokens_per_second": 9143.066
+    },
+    {
+      "epoch": 0.9320468076370355,
+      "grad_norm": 0.6653034631083543,
+      "learning_rate": 5.534293430466276e-06,
+      "loss": 0.3472,
+      "num_input_tokens_seen": 250144088,
+      "step": 4540,
+      "train_runtime": 27354.3145,
+      "train_tokens_per_second": 9144.594
+    },
+    {
+      "epoch": 0.9330732909053583,
+      "grad_norm": 0.5749865710486631,
+      "learning_rate": 5.526276919810637e-06,
+      "loss": 0.347,
+      "num_input_tokens_seen": 250410368,
+      "step": 4545,
+      "train_runtime": 27382.388,
+      "train_tokens_per_second": 9144.943
+    },
+    {
+      "epoch": 0.934099774173681,
+      "grad_norm": 0.6960635874424217,
+      "learning_rate": 5.5182590409293305e-06,
+      "loss": 0.3237,
+      "num_input_tokens_seen": 250675720,
+      "step": 4550,
+      "train_runtime": 27412.9033,
+      "train_tokens_per_second": 9144.443
+    },
+    {
+      "epoch": 0.9351262574420037,
+      "grad_norm": 0.690572766462507,
+      "learning_rate": 5.510239814667401e-06,
+      "loss": 0.3989,
+      "num_input_tokens_seen": 250968512,
+      "step": 4555,
+      "train_runtime": 27444.8264,
+      "train_tokens_per_second": 9144.474
+    },
+    {
+      "epoch": 0.9361527407103264,
+      "grad_norm": 0.5901269309034463,
+      "learning_rate": 5.5022192618734035e-06,
+      "loss": 0.3469,
+      "num_input_tokens_seen": 251245648,
+      "step": 4560,
+      "train_runtime": 27474.4834,
+      "train_tokens_per_second": 9144.691
+    },
+    {
+      "epoch": 0.9371792239786492,
+      "grad_norm": 0.7108565107479332,
+      "learning_rate": 5.494197403399337e-06,
+      "loss": 0.371,
+      "num_input_tokens_seen": 251518792,
+      "step": 4565,
+      "train_runtime": 27499.7033,
+      "train_tokens_per_second": 9146.237
+    },
+    {
+      "epoch": 0.9382057072469718,
+      "grad_norm": 0.5976328230115633,
+      "learning_rate": 5.486174260100595e-06,
+      "loss": 0.3592,
+      "num_input_tokens_seen": 251801792,
+      "step": 4570,
+      "train_runtime": 27530.0761,
+      "train_tokens_per_second": 9146.426
+    },
+    {
+      "epoch": 0.9392321905152946,
+      "grad_norm": 0.6746439869824401,
+      "learning_rate": 5.478149852835912e-06,
+      "loss": 0.3851,
+      "num_input_tokens_seen": 252097144,
+      "step": 4575,
+      "train_runtime": 27560.0558,
+      "train_tokens_per_second": 9147.193
+    },
+    {
+      "epoch": 0.9402586737836174,
+      "grad_norm": 0.6690757804821494,
+      "learning_rate": 5.4701242024673105e-06,
+      "loss": 0.3601,
+      "num_input_tokens_seen": 252349240,
+      "step": 4580,
+      "train_runtime": 27590.3955,
+      "train_tokens_per_second": 9146.271
+    },
+    {
+      "epoch": 0.94128515705194,
+      "grad_norm": 0.7039559711896304,
+      "learning_rate": 5.462097329860039e-06,
+      "loss": 0.3515,
+      "num_input_tokens_seen": 252595248,
+      "step": 4585,
+      "train_runtime": 27617.9932,
+      "train_tokens_per_second": 9146.039
+    },
+    {
+      "epoch": 0.9423116403202628,
+      "grad_norm": 0.7363193147070622,
+      "learning_rate": 5.454069255882531e-06,
+      "loss": 0.3489,
+      "num_input_tokens_seen": 252849784,
+      "step": 4590,
+      "train_runtime": 27651.7093,
+      "train_tokens_per_second": 9144.092
+    },
+    {
+      "epoch": 0.9433381235885855,
+      "grad_norm": 0.653813660740808,
+      "learning_rate": 5.446040001406341e-06,
+      "loss": 0.3646,
+      "num_input_tokens_seen": 253121736,
+      "step": 4595,
+      "train_runtime": 27682.2657,
+      "train_tokens_per_second": 9143.823
+    },
+    {
+      "epoch": 0.9443646068569083,
+      "grad_norm": 0.5862307905693872,
+      "learning_rate": 5.438009587306089e-06,
+      "loss": 0.3266,
+      "num_input_tokens_seen": 253392832,
+      "step": 4600,
+      "train_runtime": 27719.2621,
+      "train_tokens_per_second": 9141.399
+    },
+    {
+      "epoch": 0.9453910901252309,
+      "grad_norm": 0.8501636021454011,
+      "learning_rate": 5.429978034459412e-06,
+      "loss": 0.3384,
+      "num_input_tokens_seen": 253666240,
+      "step": 4605,
+      "train_runtime": 27746.9922,
+      "train_tokens_per_second": 9142.117
+    },
+    {
+      "epoch": 0.9464175733935537,
+      "grad_norm": 0.6124016485899665,
+      "learning_rate": 5.42194536374691e-06,
+      "loss": 0.3647,
+      "num_input_tokens_seen": 253942008,
+      "step": 4610,
+      "train_runtime": 27777.9292,
+      "train_tokens_per_second": 9141.862
+    },
+    {
+      "epoch": 0.9474440566618764,
+      "grad_norm": 0.6795589665762509,
+      "learning_rate": 5.4139115960520875e-06,
+      "loss": 0.3715,
+      "num_input_tokens_seen": 254208448,
+      "step": 4615,
+      "train_runtime": 27813.4472,
+      "train_tokens_per_second": 9139.768
+    },
+    {
+      "epoch": 0.9484705399301991,
+      "grad_norm": 0.6997352282336915,
+      "learning_rate": 5.405876752261297e-06,
+      "loss": 0.3558,
+      "num_input_tokens_seen": 254476904,
+      "step": 4620,
+      "train_runtime": 27844.2282,
+      "train_tokens_per_second": 9139.305
+    },
+    {
+      "epoch": 0.9494970231985219,
+      "grad_norm": 0.7592219309784003,
+      "learning_rate": 5.397840853263696e-06,
+      "loss": 0.3557,
+      "num_input_tokens_seen": 254747256,
+      "step": 4625,
+      "train_runtime": 27875.1004,
+      "train_tokens_per_second": 9138.882
+    },
+    {
+      "epoch": 0.9505235064668446,
+      "grad_norm": 0.63162315595167,
+      "learning_rate": 5.389803919951181e-06,
+      "loss": 0.3437,
+      "num_input_tokens_seen": 255026208,
+      "step": 4630,
+      "train_runtime": 27906.8257,
+      "train_tokens_per_second": 9138.489
+    },
+    {
+      "epoch": 0.9515499897351674,
+      "grad_norm": 0.6430646412783846,
+      "learning_rate": 5.3817659732183354e-06,
+      "loss": 0.3442,
+      "num_input_tokens_seen": 255296472,
+      "step": 4635,
+      "train_runtime": 27943.208,
+      "train_tokens_per_second": 9136.262
+    },
+    {
+      "epoch": 0.95257647300349,
+      "grad_norm": 0.6319027309667108,
+      "learning_rate": 5.3737270339623845e-06,
+      "loss": 0.355,
+      "num_input_tokens_seen": 255567624,
+      "step": 4640,
+      "train_runtime": 27981.006,
+      "train_tokens_per_second": 9133.611
+    },
+    {
+      "epoch": 0.9536029562718128,
+      "grad_norm": 0.58379476064289,
+      "learning_rate": 5.365687123083126e-06,
+      "loss": 0.3262,
+      "num_input_tokens_seen": 255860512,
+      "step": 4645,
+      "train_runtime": 28013.227,
+      "train_tokens_per_second": 9133.561
+    },
+    {
+      "epoch": 0.9546294395401355,
+      "grad_norm": 0.6156394074170514,
+      "learning_rate": 5.357646261482889e-06,
+      "loss": 0.3318,
+      "num_input_tokens_seen": 256163984,
+      "step": 4650,
+      "train_runtime": 28040.0171,
+      "train_tokens_per_second": 9135.657
+    },
+    {
+      "epoch": 0.9556559228084582,
+      "grad_norm": 0.6544714788300031,
+      "learning_rate": 5.3496044700664725e-06,
+      "loss": 0.3552,
+      "num_input_tokens_seen": 256456480,
+      "step": 4655,
+      "train_runtime": 28064.3714,
+      "train_tokens_per_second": 9138.152
+    },
+    {
+      "epoch": 0.9566824060767809,
+      "grad_norm": 0.5966957587356085,
+      "learning_rate": 5.341561769741092e-06,
+      "loss": 0.3293,
+      "num_input_tokens_seen": 256758544,
+      "step": 4660,
+      "train_runtime": 28095.741,
+      "train_tokens_per_second": 9138.7
+    },
+    {
+      "epoch": 0.9577088893451037,
+      "grad_norm": 0.6555285239916481,
+      "learning_rate": 5.333518181416328e-06,
+      "loss": 0.3731,
+      "num_input_tokens_seen": 257048088,
+      "step": 4665,
+      "train_runtime": 28125.9116,
+      "train_tokens_per_second": 9139.191
+    },
+    {
+      "epoch": 0.9587353726134265,
+      "grad_norm": 0.6539718852687597,
+      "learning_rate": 5.325473726004068e-06,
+      "loss": 0.3121,
+      "num_input_tokens_seen": 257349816,
+      "step": 4670,
+      "train_runtime": 28160.1123,
+      "train_tokens_per_second": 9138.806
+    },
+    {
+      "epoch": 0.9597618558817491,
+      "grad_norm": 0.6259771275659021,
+      "learning_rate": 5.317428424418452e-06,
+      "loss": 0.336,
+      "num_input_tokens_seen": 257654696,
+      "step": 4675,
+      "train_runtime": 28194.1459,
+      "train_tokens_per_second": 9138.588
+    },
+    {
+      "epoch": 0.9607883391500719,
+      "grad_norm": 0.6887658148127057,
+      "learning_rate": 5.309382297575826e-06,
+      "loss": 0.3551,
+      "num_input_tokens_seen": 257956856,
+      "step": 4680,
+      "train_runtime": 28224.2961,
+      "train_tokens_per_second": 9139.532
+    },
+    {
+      "epoch": 0.9618148224183946,
+      "grad_norm": 0.9098321181941857,
+      "learning_rate": 5.301335366394675e-06,
+      "loss": 0.344,
+      "num_input_tokens_seen": 258224736,
+      "step": 4685,
+      "train_runtime": 28253.6581,
+      "train_tokens_per_second": 9139.515
+    },
+    {
+      "epoch": 0.9628413056867173,
+      "grad_norm": 0.631900715002507,
+      "learning_rate": 5.29328765179558e-06,
+      "loss": 0.3467,
+      "num_input_tokens_seen": 258503064,
+      "step": 4690,
+      "train_runtime": 28287.2673,
+      "train_tokens_per_second": 9138.495
+    },
+    {
+      "epoch": 0.96386778895504,
+      "grad_norm": 0.5963775399177595,
+      "learning_rate": 5.285239174701156e-06,
+      "loss": 0.3553,
+      "num_input_tokens_seen": 258792920,
+      "step": 4695,
+      "train_runtime": 28316.2714,
+      "train_tokens_per_second": 9139.371
+    },
+    {
+      "epoch": 0.9648942722233628,
+      "grad_norm": 0.6801319140970866,
+      "learning_rate": 5.277189956035998e-06,
+      "loss": 0.3512,
+      "num_input_tokens_seen": 259062192,
+      "step": 4700,
+      "train_runtime": 28347.0213,
+      "train_tokens_per_second": 9138.956
+    },
+    {
+      "epoch": 0.9659207554916854,
+      "grad_norm": 0.6129886165033186,
+      "learning_rate": 5.269140016726636e-06,
+      "loss": 0.3548,
+      "num_input_tokens_seen": 259367432,
+      "step": 4705,
+      "train_runtime": 28379.977,
+      "train_tokens_per_second": 9139.099
+    },
+    {
+      "epoch": 0.9669472387600082,
+      "grad_norm": 0.6700661103209735,
+      "learning_rate": 5.261089377701466e-06,
+      "loss": 0.3634,
+      "num_input_tokens_seen": 259657648,
+      "step": 4710,
+      "train_runtime": 28409.823,
+      "train_tokens_per_second": 9139.714
+    },
+    {
+      "epoch": 0.967973722028331,
+      "grad_norm": 0.658532242961507,
+      "learning_rate": 5.253038059890711e-06,
+      "loss": 0.3573,
+      "num_input_tokens_seen": 259945584,
+      "step": 4715,
+      "train_runtime": 28438.4302,
+      "train_tokens_per_second": 9140.645
+    },
+    {
+      "epoch": 0.9690002052966536,
+      "grad_norm": 0.6577604416482349,
+      "learning_rate": 5.244986084226349e-06,
+      "loss": 0.3663,
+      "num_input_tokens_seen": 260210296,
+      "step": 4720,
+      "train_runtime": 28470.0122,
+      "train_tokens_per_second": 9139.803
+    },
+    {
+      "epoch": 0.9700266885649764,
+      "grad_norm": 0.6542357522455757,
+      "learning_rate": 5.236933471642076e-06,
+      "loss": 0.3736,
+      "num_input_tokens_seen": 260486616,
+      "step": 4725,
+      "train_runtime": 28498.5867,
+      "train_tokens_per_second": 9140.335
+    },
+    {
+      "epoch": 0.9710531718332991,
+      "grad_norm": 0.6096835626591642,
+      "learning_rate": 5.228880243073244e-06,
+      "loss": 0.3707,
+      "num_input_tokens_seen": 260773600,
+      "step": 4730,
+      "train_runtime": 28529.8424,
+      "train_tokens_per_second": 9140.38
+    },
+    {
+      "epoch": 0.9720796551016219,
+      "grad_norm": 0.6542709443443375,
+      "learning_rate": 5.2208264194568e-06,
+      "loss": 0.4129,
+      "num_input_tokens_seen": 261058816,
+      "step": 4735,
+      "train_runtime": 28562.9613,
+      "train_tokens_per_second": 9139.767
+    },
+    {
+      "epoch": 0.9731061383699445,
+      "grad_norm": 0.7184797180369246,
+      "learning_rate": 5.212772021731244e-06,
+      "loss": 0.3353,
+      "num_input_tokens_seen": 261333176,
+      "step": 4740,
+      "train_runtime": 28595.7198,
+      "train_tokens_per_second": 9138.891
+    },
+    {
+      "epoch": 0.9741326216382673,
+      "grad_norm": 0.6308642157098003,
+      "learning_rate": 5.204717070836566e-06,
+      "loss": 0.3419,
+      "num_input_tokens_seen": 261625240,
+      "step": 4745,
+      "train_runtime": 28630.5946,
+      "train_tokens_per_second": 9137.96
+    },
+    {
+      "epoch": 0.97515910490659,
+      "grad_norm": 0.6128689791696088,
+      "learning_rate": 5.196661587714195e-06,
+      "loss": 0.341,
+      "num_input_tokens_seen": 261906080,
+      "step": 4750,
+      "train_runtime": 28664.1849,
+      "train_tokens_per_second": 9137.05
+    },
+    {
+      "epoch": 0.9761855881749127,
+      "grad_norm": 0.5921475887283982,
+      "learning_rate": 5.188605593306943e-06,
+      "loss": 0.3403,
+      "num_input_tokens_seen": 262142992,
+      "step": 4755,
+      "train_runtime": 28689.5016,
+      "train_tokens_per_second": 9137.245
+    },
+    {
+      "epoch": 0.9772120714432355,
+      "grad_norm": 0.6281460295864397,
+      "learning_rate": 5.180549108558951e-06,
+      "loss": 0.3371,
+      "num_input_tokens_seen": 262395520,
+      "step": 4760,
+      "train_runtime": 28716.7398,
+      "train_tokens_per_second": 9137.371
+    },
+    {
+      "epoch": 0.9782385547115582,
+      "grad_norm": 0.6229790000163968,
+      "learning_rate": 5.172492154415635e-06,
+      "loss": 0.3418,
+      "num_input_tokens_seen": 262664448,
+      "step": 4765,
+      "train_runtime": 28751.0561,
+      "train_tokens_per_second": 9135.819
+    },
+    {
+      "epoch": 0.979265037979881,
+      "grad_norm": 0.7658132948535467,
+      "learning_rate": 5.164434751823635e-06,
+      "loss": 0.3544,
+      "num_input_tokens_seen": 262919560,
+      "step": 4770,
+      "train_runtime": 28776.4772,
+      "train_tokens_per_second": 9136.614
+    },
+    {
+      "epoch": 0.9802915212482036,
+      "grad_norm": 0.6061661536297872,
+      "learning_rate": 5.1563769217307505e-06,
+      "loss": 0.3375,
+      "num_input_tokens_seen": 263207176,
+      "step": 4775,
+      "train_runtime": 28804.6623,
+      "train_tokens_per_second": 9137.659
+    },
+    {
+      "epoch": 0.9813180045165264,
+      "grad_norm": 0.6332163589318958,
+      "learning_rate": 5.148318685085895e-06,
+      "loss": 0.3259,
+      "num_input_tokens_seen": 263491280,
+      "step": 4780,
+      "train_runtime": 28832.73,
+      "train_tokens_per_second": 9138.617
+    },
+    {
+      "epoch": 0.9823444877848491,
+      "grad_norm": 0.5962198611771214,
+      "learning_rate": 5.1402600628390445e-06,
+      "loss": 0.3447,
+      "num_input_tokens_seen": 263768944,
+      "step": 4785,
+      "train_runtime": 28859.8697,
+      "train_tokens_per_second": 9139.644
+    },
+    {
+      "epoch": 0.9833709710531718,
+      "grad_norm": 0.6587733920166751,
+      "learning_rate": 5.132201075941166e-06,
+      "loss": 0.345,
+      "num_input_tokens_seen": 264044864,
+      "step": 4790,
+      "train_runtime": 28889.5234,
+      "train_tokens_per_second": 9139.814
+    },
+    {
+      "epoch": 0.9843974543214946,
+      "grad_norm": 0.6706957844172302,
+      "learning_rate": 5.124141745344186e-06,
+      "loss": 0.3159,
+      "num_input_tokens_seen": 264314368,
+      "step": 4795,
+      "train_runtime": 28920.3303,
+      "train_tokens_per_second": 9139.397
+    },
+    {
+      "epoch": 0.9854239375898173,
+      "grad_norm": 0.6459559332585905,
+      "learning_rate": 5.1160820920009155e-06,
+      "loss": 0.357,
+      "num_input_tokens_seen": 264597400,
+      "step": 4800,
+      "train_runtime": 28955.279,
+      "train_tokens_per_second": 9138.14
+    },
+    {
+      "epoch": 0.98645042085814,
+      "grad_norm": 0.6364063740247085,
+      "learning_rate": 5.108022136865011e-06,
+      "loss": 0.325,
+      "num_input_tokens_seen": 264858440,
+      "step": 4805,
+      "train_runtime": 28983.7822,
+      "train_tokens_per_second": 9138.16
+    },
+    {
+      "epoch": 0.9874769041264627,
+      "grad_norm": 0.5832234436830201,
+      "learning_rate": 5.099961900890914e-06,
+      "loss": 0.326,
+      "num_input_tokens_seen": 265139872,
+      "step": 4810,
+      "train_runtime": 29017.7069,
+      "train_tokens_per_second": 9137.175
+    },
+    {
+      "epoch": 0.9885033873947855,
+      "grad_norm": 0.7226218201977719,
+      "learning_rate": 5.0919014050337875e-06,
+      "loss": 0.3482,
+      "num_input_tokens_seen": 265419712,
+      "step": 4815,
+      "train_runtime": 29051.9599,
+      "train_tokens_per_second": 9136.035
+    },
+    {
+      "epoch": 0.9895298706631082,
+      "grad_norm": 0.6606494042057718,
+      "learning_rate": 5.0838406702494816e-06,
+      "loss": 0.3362,
+      "num_input_tokens_seen": 265667312,
+      "step": 4820,
+      "train_runtime": 29078.1841,
+      "train_tokens_per_second": 9136.31
+    },
+    {
+      "epoch": 0.9905563539314309,
+      "grad_norm": 0.6394386320703596,
+      "learning_rate": 5.075779717494458e-06,
+      "loss": 0.3324,
+      "num_input_tokens_seen": 265974904,
+      "step": 4825,
+      "train_runtime": 29107.7181,
+      "train_tokens_per_second": 9137.608
+    },
+    {
+      "epoch": 0.9915828371997536,
+      "grad_norm": 0.7054619780361043,
+      "learning_rate": 5.067718567725751e-06,
+      "loss": 0.3534,
+      "num_input_tokens_seen": 266256544,
+      "step": 4830,
+      "train_runtime": 29135.2891,
+      "train_tokens_per_second": 9138.627
+    },
+    {
+      "epoch": 0.9926093204680764,
+      "grad_norm": 0.7543425966019021,
+      "learning_rate": 5.0596572419009045e-06,
+      "loss": 0.3413,
+      "num_input_tokens_seen": 266533584,
+      "step": 4835,
+      "train_runtime": 29165.1991,
+      "train_tokens_per_second": 9138.754
+    },
+    {
+      "epoch": 0.9936358037363991,
+      "grad_norm": 0.7059354413957727,
+      "learning_rate": 5.051595760977922e-06,
+      "loss": 0.3261,
+      "num_input_tokens_seen": 266815096,
+      "step": 4840,
+      "train_runtime": 29194.9326,
+      "train_tokens_per_second": 9139.089
+    },
+    {
+      "epoch": 0.9946622870047218,
+      "grad_norm": 0.6976943833410832,
+      "learning_rate": 5.043534145915209e-06,
+      "loss": 0.3723,
+      "num_input_tokens_seen": 267086184,
+      "step": 4845,
+      "train_runtime": 29227.945,
+      "train_tokens_per_second": 9138.042
+    },
+    {
+      "epoch": 0.9956887702730446,
+      "grad_norm": 0.6081911612720114,
+      "learning_rate": 5.035472417671517e-06,
+      "loss": 0.3442,
+      "num_input_tokens_seen": 267385416,
+      "step": 4850,
+      "train_runtime": 29261.1735,
+      "train_tokens_per_second": 9137.891
+    },
+    {
+      "epoch": 0.9967152535413673,
+      "grad_norm": 0.680915416372239,
+      "learning_rate": 5.027410597205898e-06,
+      "loss": 0.3443,
+      "num_input_tokens_seen": 267664352,
+      "step": 4855,
+      "train_runtime": 29287.4166,
+      "train_tokens_per_second": 9139.227
+    },
+    {
+      "epoch": 0.99774173680969,
+      "grad_norm": 0.6256701448462965,
+      "learning_rate": 5.019348705477639e-06,
+      "loss": 0.3368,
+      "num_input_tokens_seen": 267958320,
+      "step": 4860,
+      "train_runtime": 29315.807,
+      "train_tokens_per_second": 9140.404
+    },
+    {
+      "epoch": 0.9987682200780127,
+      "grad_norm": 0.7539487935094128,
+      "learning_rate": 5.011286763446214e-06,
+      "loss": 0.3714,
+      "num_input_tokens_seen": 268236264,
+      "step": 4865,
+      "train_runtime": 29351.9701,
+      "train_tokens_per_second": 9138.612
+    },
+    {
+      "epoch": 0.9997947033463355,
+      "grad_norm": 0.7253875840136152,
+      "learning_rate": 5.003224792071225e-06,
+      "loss": 0.3553,
+      "num_input_tokens_seen": 268482200,
+      "step": 4870,
+      "train_runtime": 29380.1747,
+      "train_tokens_per_second": 9138.21
+    },
+    {
+      "epoch": 1.0008211866146581,
+      "grad_norm": 0.5909443791362023,
+      "learning_rate": 4.995162812312357e-06,
+      "loss": 0.2939,
+      "num_input_tokens_seen": 268767416,
+      "step": 4875,
+      "train_runtime": 29513.4969,
+      "train_tokens_per_second": 9106.593
+    },
+    {
+      "epoch": 1.001847669882981,
+      "grad_norm": 0.7354600932911579,
+      "learning_rate": 4.987100845129309e-06,
+      "loss": 0.3002,
+      "num_input_tokens_seen": 269038384,
+      "step": 4880,
+      "train_runtime": 29544.4168,
+      "train_tokens_per_second": 9106.234
+    },
+    {
+      "epoch": 1.0028741531513037,
+      "grad_norm": 0.7245050328990359,
+      "learning_rate": 4.979038911481752e-06,
+      "loss": 0.3041,
+      "num_input_tokens_seen": 269296304,
+      "step": 4885,
+      "train_runtime": 29575.1328,
+      "train_tokens_per_second": 9105.498
+    },
+    {
+      "epoch": 1.0039006364196263,
+      "grad_norm": 0.7251085936838166,
+      "learning_rate": 4.970977032329268e-06,
+      "loss": 0.2939,
+      "num_input_tokens_seen": 269583520,
+      "step": 4890,
+      "train_runtime": 29604.9957,
+      "train_tokens_per_second": 9106.014
+    },
+    {
+      "epoch": 1.004927119687949,
+      "grad_norm": 2.265405789508219,
+      "learning_rate": 4.962915228631298e-06,
+      "loss": 0.2877,
+      "num_input_tokens_seen": 269858072,
+      "step": 4895,
+      "train_runtime": 29632.8917,
+      "train_tokens_per_second": 9106.707
+    },
+    {
+      "epoch": 1.005953602956272,
+      "grad_norm": 0.7787000651112569,
+      "learning_rate": 4.954853521347087e-06,
+      "loss": 0.2903,
+      "num_input_tokens_seen": 270140016,
+      "step": 4900,
+      "train_runtime": 29664.1259,
+      "train_tokens_per_second": 9106.623
+    },
+    {
+      "epoch": 1.0069800862245946,
+      "grad_norm": 0.5918361785462564,
+      "learning_rate": 4.946791931435629e-06,
+      "loss": 0.2894,
+      "num_input_tokens_seen": 270415248,
+      "step": 4905,
+      "train_runtime": 29686.2028,
+      "train_tokens_per_second": 9109.122
+    },
+    {
+      "epoch": 1.0080065694929172,
+      "grad_norm": 0.6121529391959222,
+      "learning_rate": 4.938730479855614e-06,
+      "loss": 0.2949,
+      "num_input_tokens_seen": 270680752,
+      "step": 4910,
+      "train_runtime": 29716.3333,
+      "train_tokens_per_second": 9108.821
+    },
+    {
+      "epoch": 1.00903305276124,
+      "grad_norm": 0.7163920374350438,
+      "learning_rate": 4.930669187565371e-06,
+      "loss": 0.2985,
+      "num_input_tokens_seen": 270958712,
+      "step": 4915,
+      "train_runtime": 29746.502,
+      "train_tokens_per_second": 9108.927
+    },
+    {
+      "epoch": 1.0100595360295628,
+      "grad_norm": 0.6257528879239934,
+      "learning_rate": 4.922608075522811e-06,
+      "loss": 0.2837,
+      "num_input_tokens_seen": 271240016,
+      "step": 4920,
+      "train_runtime": 29774.9329,
+      "train_tokens_per_second": 9109.677
+    },
+    {
+      "epoch": 1.0110860192978854,
+      "grad_norm": 0.6141616517692382,
+      "learning_rate": 4.9145471646853866e-06,
+      "loss": 0.3214,
+      "num_input_tokens_seen": 271530072,
+      "step": 4925,
+      "train_runtime": 29801.4038,
+      "train_tokens_per_second": 9111.318
+    },
+    {
+      "epoch": 1.012112502566208,
+      "grad_norm": 0.6411964355086108,
+      "learning_rate": 4.906486476010018e-06,
+      "loss": 0.2893,
+      "num_input_tokens_seen": 271803640,
+      "step": 4930,
+      "train_runtime": 29830.1349,
+      "train_tokens_per_second": 9111.713
+    },
+    {
+      "epoch": 1.013138985834531,
+      "grad_norm": 0.7064261972718641,
+      "learning_rate": 4.89842603045305e-06,
+      "loss": 0.2953,
+      "num_input_tokens_seen": 272093248,
+      "step": 4935,
+      "train_runtime": 29865.7913,
+      "train_tokens_per_second": 9110.532
+    },
+    {
+      "epoch": 1.0141654691028537,
+      "grad_norm": 0.954876511784353,
+      "learning_rate": 4.8903658489702e-06,
+      "loss": 0.2854,
+      "num_input_tokens_seen": 272356224,
+      "step": 4940,
+      "train_runtime": 29898.2995,
+      "train_tokens_per_second": 9109.422
+    },
+    {
+      "epoch": 1.0151919523711763,
+      "grad_norm": 0.6592827552500885,
+      "learning_rate": 4.882305952516492e-06,
+      "loss": 0.3015,
+      "num_input_tokens_seen": 272647552,
+      "step": 4945,
+      "train_runtime": 29929.9832,
+      "train_tokens_per_second": 9109.512
+    },
+    {
+      "epoch": 1.016218435639499,
+      "grad_norm": 0.6255795334853693,
+      "learning_rate": 4.874246362046216e-06,
+      "loss": 0.2902,
+      "num_input_tokens_seen": 272932280,
+      "step": 4950,
+      "train_runtime": 29966.1892,
+      "train_tokens_per_second": 9108.008
+    },
+    {
+      "epoch": 1.0172449189078219,
+      "grad_norm": 0.5871263733386882,
+      "learning_rate": 4.866187098512856e-06,
+      "loss": 0.301,
+      "num_input_tokens_seen": 273232976,
+      "step": 4955,
+      "train_runtime": 29997.7086,
+      "train_tokens_per_second": 9108.462
+    },
+    {
+      "epoch": 1.0182714021761445,
+      "grad_norm": 0.7313601853281902,
+      "learning_rate": 4.858128182869059e-06,
+      "loss": 0.3071,
+      "num_input_tokens_seen": 273498848,
+      "step": 4960,
+      "train_runtime": 30020.7056,
+      "train_tokens_per_second": 9110.34
+    },
+    {
+      "epoch": 1.0192978854444672,
+      "grad_norm": 0.6701453842750955,
+      "learning_rate": 4.8500696360665615e-06,
+      "loss": 0.2901,
+      "num_input_tokens_seen": 273778288,
+      "step": 4965,
+      "train_runtime": 30053.9559,
+      "train_tokens_per_second": 9109.559
+    },
+    {
+      "epoch": 1.02032436871279,
+      "grad_norm": 0.6811683263875807,
+      "learning_rate": 4.8420114790561365e-06,
+      "loss": 0.286,
+      "num_input_tokens_seen": 274053216,
+      "step": 4970,
+      "train_runtime": 30079.4284,
+      "train_tokens_per_second": 9110.985
+    },
+    {
+      "epoch": 1.0213508519811128,
+      "grad_norm": 0.6124095911043747,
+      "learning_rate": 4.833953732787549e-06,
+      "loss": 0.2777,
+      "num_input_tokens_seen": 274310296,
+      "step": 4975,
+      "train_runtime": 30108.4753,
+      "train_tokens_per_second": 9110.734
+    },
+    {
+      "epoch": 1.0223773352494354,
+      "grad_norm": 0.5374487335475947,
+      "learning_rate": 4.8258964182094985e-06,
+      "loss": 0.3064,
+      "num_input_tokens_seen": 274576824,
+      "step": 4980,
+      "train_runtime": 30136.6731,
+      "train_tokens_per_second": 9111.053
+    },
+    {
+      "epoch": 1.023403818517758,
+      "grad_norm": 0.6283678286321782,
+      "learning_rate": 4.817839556269556e-06,
+      "loss": 0.2915,
+      "num_input_tokens_seen": 274860544,
+      "step": 4985,
+      "train_runtime": 30170.4351,
+      "train_tokens_per_second": 9110.261
+    },
+    {
+      "epoch": 1.024430301786081,
+      "grad_norm": 0.6196645935467849,
+      "learning_rate": 4.809783167914118e-06,
+      "loss": 0.2905,
+      "num_input_tokens_seen": 275132432,
+      "step": 4990,
+      "train_runtime": 30198.8219,
+      "train_tokens_per_second": 9110.701
+    },
+    {
+      "epoch": 1.0254567850544036,
+      "grad_norm": 0.7187907280728117,
+      "learning_rate": 4.801727274088353e-06,
+      "loss": 0.315,
+      "num_input_tokens_seen": 275385376,
+      "step": 4995,
+      "train_runtime": 30229.7613,
+      "train_tokens_per_second": 9109.744
+    },
+    {
+      "epoch": 1.0264832683227263,
+      "grad_norm": 0.6384038715001752,
+      "learning_rate": 4.793671895736138e-06,
+      "loss": 0.2908,
+      "num_input_tokens_seen": 275680872,
+      "step": 5000,
+      "train_runtime": 30262.7892,
+      "train_tokens_per_second": 9109.566
+    },
+    {
+      "epoch": 1.027509751591049,
+      "grad_norm": 0.6628959712600392,
+      "learning_rate": 4.785617053800017e-06,
+      "loss": 0.2964,
+      "num_input_tokens_seen": 275955352,
+      "step": 5005,
+      "train_runtime": 30294.5545,
+      "train_tokens_per_second": 9109.074
+    },
+    {
+      "epoch": 1.0285362348593718,
+      "grad_norm": 0.6693141978521356,
+      "learning_rate": 4.777562769221129e-06,
+      "loss": 0.2886,
+      "num_input_tokens_seen": 276257936,
+      "step": 5010,
+      "train_runtime": 30324.4061,
+      "train_tokens_per_second": 9110.086
+    },
+    {
+      "epoch": 1.0295627181276945,
+      "grad_norm": 0.6854000300203442,
+      "learning_rate": 4.769509062939177e-06,
+      "loss": 0.3296,
+      "num_input_tokens_seen": 276523656,
+      "step": 5015,
+      "train_runtime": 30356.572,
+      "train_tokens_per_second": 9109.186
+    },
+    {
+      "epoch": 1.0305892013960172,
+      "grad_norm": 0.8344170818554825,
+      "learning_rate": 4.761455955892351e-06,
+      "loss": 0.2914,
+      "num_input_tokens_seen": 276804672,
+      "step": 5020,
+      "train_runtime": 30379.242,
+      "train_tokens_per_second": 9111.639
+    },
+    {
+      "epoch": 1.03161568466434,
+      "grad_norm": 0.7596874998759221,
+      "learning_rate": 4.753403469017284e-06,
+      "loss": 0.3088,
+      "num_input_tokens_seen": 277083176,
+      "step": 5025,
+      "train_runtime": 30417.3035,
+      "train_tokens_per_second": 9109.393
+    },
+    {
+      "epoch": 1.0326421679326627,
+      "grad_norm": 0.6499271098507563,
+      "learning_rate": 4.745351623249003e-06,
+      "loss": 0.281,
+      "num_input_tokens_seen": 277355344,
+      "step": 5030,
+      "train_runtime": 30448.1524,
+      "train_tokens_per_second": 9109.103
+    },
+    {
+      "epoch": 1.0336686512009854,
+      "grad_norm": 0.7536064420318498,
+      "learning_rate": 4.737300439520862e-06,
+      "loss": 0.3055,
+      "num_input_tokens_seen": 277620656,
+      "step": 5035,
+      "train_runtime": 30483.0468,
+      "train_tokens_per_second": 9107.379
+    },
+    {
+      "epoch": 1.034695134469308,
+      "grad_norm": 0.5436688675866136,
+      "learning_rate": 4.729249938764496e-06,
+      "loss": 0.2832,
+      "num_input_tokens_seen": 277898096,
+      "step": 5040,
+      "train_runtime": 30515.4031,
+      "train_tokens_per_second": 9106.814
+    },
+    {
+      "epoch": 1.035721617737631,
+      "grad_norm": 0.5646444098021194,
+      "learning_rate": 4.7212001419097585e-06,
+      "loss": 0.3048,
+      "num_input_tokens_seen": 278186496,
+      "step": 5045,
+      "train_runtime": 30542.4146,
+      "train_tokens_per_second": 9108.202
+    },
+    {
+      "epoch": 1.0367481010059536,
+      "grad_norm": 0.8398189701691778,
+      "learning_rate": 4.713151069884687e-06,
+      "loss": 0.278,
+      "num_input_tokens_seen": 278451720,
+      "step": 5050,
+      "train_runtime": 30579.2434,
+      "train_tokens_per_second": 9105.906
+    },
+    {
+      "epoch": 1.0377745842742763,
+      "grad_norm": 0.655273633748474,
+      "learning_rate": 4.705102743615425e-06,
+      "loss": 0.2709,
+      "num_input_tokens_seen": 278733104,
+      "step": 5055,
+      "train_runtime": 30611.9675,
+      "train_tokens_per_second": 9105.364
+    },
+    {
+      "epoch": 1.0388010675425992,
+      "grad_norm": 0.6325778278164816,
+      "learning_rate": 4.697055184026173e-06,
+      "loss": 0.3041,
+      "num_input_tokens_seen": 278994888,
+      "step": 5060,
+      "train_runtime": 30641.9958,
+      "train_tokens_per_second": 9104.984
+    },
+    {
+      "epoch": 1.0398275508109218,
+      "grad_norm": 0.6591300073160415,
+      "learning_rate": 4.689008412039146e-06,
+      "loss": 0.3213,
+      "num_input_tokens_seen": 279268680,
+      "step": 5065,
+      "train_runtime": 30674.8181,
+      "train_tokens_per_second": 9104.167
+    },
+    {
+      "epoch": 1.0408540340792445,
+      "grad_norm": 0.6019441697307025,
+      "learning_rate": 4.680962448574512e-06,
+      "loss": 0.2934,
+      "num_input_tokens_seen": 279564976,
+      "step": 5070,
+      "train_runtime": 30707.3594,
+      "train_tokens_per_second": 9104.169
+    },
+    {
+      "epoch": 1.0418805173475671,
+      "grad_norm": 0.7924149678931984,
+      "learning_rate": 4.67291731455033e-06,
+      "loss": 0.2802,
+      "num_input_tokens_seen": 279836056,
+      "step": 5075,
+      "train_runtime": 30735.1972,
+      "train_tokens_per_second": 9104.742
+    },
+    {
+      "epoch": 1.04290700061589,
+      "grad_norm": 0.7019177489082794,
+      "learning_rate": 4.664873030882507e-06,
+      "loss": 0.2959,
+      "num_input_tokens_seen": 280091616,
+      "step": 5080,
+      "train_runtime": 30767.6877,
+      "train_tokens_per_second": 9103.434
+    },
+    {
+      "epoch": 1.0439334838842127,
+      "grad_norm": 0.7232782469433794,
+      "learning_rate": 4.656829618484741e-06,
+      "loss": 0.2938,
+      "num_input_tokens_seen": 280369128,
+      "step": 5085,
+      "train_runtime": 30798.9758,
+      "train_tokens_per_second": 9103.196
+    },
+    {
+      "epoch": 1.0449599671525354,
+      "grad_norm": 0.650281143085423,
+      "learning_rate": 4.648787098268463e-06,
+      "loss": 0.2668,
+      "num_input_tokens_seen": 280648224,
+      "step": 5090,
+      "train_runtime": 30826.3314,
+      "train_tokens_per_second": 9104.172
+    },
+    {
+      "epoch": 1.045986450420858,
+      "grad_norm": 0.657607606685648,
+      "learning_rate": 4.640745491142782e-06,
+      "loss": 0.2834,
+      "num_input_tokens_seen": 280931784,
+      "step": 5095,
+      "train_runtime": 30854.7612,
+      "train_tokens_per_second": 9104.974
+    },
+    {
+      "epoch": 1.047012933689181,
+      "grad_norm": 0.6352921296142084,
+      "learning_rate": 4.6327048180144335e-06,
+      "loss": 0.2971,
+      "num_input_tokens_seen": 281192768,
+      "step": 5100,
+      "train_runtime": 30882.4212,
+      "train_tokens_per_second": 9105.27
+    },
+    {
+      "epoch": 1.0480394169575036,
+      "grad_norm": 0.7754567271872,
+      "learning_rate": 4.624665099787732e-06,
+      "loss": 0.3003,
+      "num_input_tokens_seen": 281462840,
+      "step": 5105,
+      "train_runtime": 30908.6401,
+      "train_tokens_per_second": 9106.284
+    },
+    {
+      "epoch": 1.0490659002258262,
+      "grad_norm": 0.7440816370271657,
+      "learning_rate": 4.6166263573645e-06,
+      "loss": 0.2885,
+      "num_input_tokens_seen": 281710520,
+      "step": 5110,
+      "train_runtime": 30937.1842,
+      "train_tokens_per_second": 9105.888
+    },
+    {
+      "epoch": 1.0500923834941491,
+      "grad_norm": 0.600204316105717,
+      "learning_rate": 4.608588611644026e-06,
+      "loss": 0.2871,
+      "num_input_tokens_seen": 281981736,
+      "step": 5115,
+      "train_runtime": 30972.7275,
+      "train_tokens_per_second": 9104.195
+    },
+    {
+      "epoch": 1.0511188667624718,
+      "grad_norm": 0.7568911929643095,
+      "learning_rate": 4.600551883523012e-06,
+      "loss": 0.2752,
+      "num_input_tokens_seen": 282255392,
+      "step": 5120,
+      "train_runtime": 31004.3325,
+      "train_tokens_per_second": 9103.74
+    },
+    {
+      "epoch": 1.0521453500307945,
+      "grad_norm": 0.7011806260130155,
+      "learning_rate": 4.592516193895509e-06,
+      "loss": 0.3039,
+      "num_input_tokens_seen": 282515856,
+      "step": 5125,
+      "train_runtime": 31038.0654,
+      "train_tokens_per_second": 9102.238
+    },
+    {
+      "epoch": 1.0531718332991171,
+      "grad_norm": 0.7643471886361182,
+      "learning_rate": 4.5844815636528685e-06,
+      "loss": 0.3125,
+      "num_input_tokens_seen": 282750568,
+      "step": 5130,
+      "train_runtime": 31061.4391,
+      "train_tokens_per_second": 9102.945
+    },
+    {
+      "epoch": 1.05419831656744,
+      "grad_norm": 0.728494573262625,
+      "learning_rate": 4.576448013683691e-06,
+      "loss": 0.2948,
+      "num_input_tokens_seen": 283023600,
+      "step": 5135,
+      "train_runtime": 31091.311,
+      "train_tokens_per_second": 9102.981
+    },
+    {
+      "epoch": 1.0552247998357627,
+      "grad_norm": 0.6285958978417208,
+      "learning_rate": 4.568415564873766e-06,
+      "loss": 0.3001,
+      "num_input_tokens_seen": 283305672,
+      "step": 5140,
+      "train_runtime": 31124.5236,
+      "train_tokens_per_second": 9102.33
+    },
+    {
+      "epoch": 1.0562512831040853,
+      "grad_norm": 0.7807178834569509,
+      "learning_rate": 4.560384238106022e-06,
+      "loss": 0.3073,
+      "num_input_tokens_seen": 283572688,
+      "step": 5145,
+      "train_runtime": 31150.3096,
+      "train_tokens_per_second": 9103.367
+    },
+    {
+      "epoch": 1.0572777663724082,
+      "grad_norm": 0.5830052361683197,
+      "learning_rate": 4.5523540542604645e-06,
+      "loss": 0.3021,
+      "num_input_tokens_seen": 283845784,
+      "step": 5150,
+      "train_runtime": 31178.6732,
+      "train_tokens_per_second": 9103.844
+    },
+    {
+      "epoch": 1.0583042496407309,
+      "grad_norm": 0.6740037683142238,
+      "learning_rate": 4.544325034214139e-06,
+      "loss": 0.3106,
+      "num_input_tokens_seen": 284087456,
+      "step": 5155,
+      "train_runtime": 31212.1466,
+      "train_tokens_per_second": 9101.824
+    },
+    {
+      "epoch": 1.0593307329090536,
+      "grad_norm": 0.63572674451117,
+      "learning_rate": 4.536297198841054e-06,
+      "loss": 0.2889,
+      "num_input_tokens_seen": 284379840,
+      "step": 5160,
+      "train_runtime": 31238.0427,
+      "train_tokens_per_second": 9103.638
+    },
+    {
+      "epoch": 1.0603572161773762,
+      "grad_norm": 0.7284120908891873,
+      "learning_rate": 4.5282705690121445e-06,
+      "loss": 0.3161,
+      "num_input_tokens_seen": 284660608,
+      "step": 5165,
+      "train_runtime": 31271.8233,
+      "train_tokens_per_second": 9102.783
+    },
+    {
+      "epoch": 1.061383699445699,
+      "grad_norm": 0.6808213981849749,
+      "learning_rate": 4.5202451655952064e-06,
+      "loss": 0.2971,
+      "num_input_tokens_seen": 284929616,
+      "step": 5170,
+      "train_runtime": 31302.3881,
+      "train_tokens_per_second": 9102.488
+    },
+    {
+      "epoch": 1.0624101827140218,
+      "grad_norm": 0.6392472592182883,
+      "learning_rate": 4.512221009454855e-06,
+      "loss": 0.31,
+      "num_input_tokens_seen": 285208464,
+      "step": 5175,
+      "train_runtime": 31330.6393,
+      "train_tokens_per_second": 9103.18
+    },
+    {
+      "epoch": 1.0634366659823444,
+      "grad_norm": 0.5840843633038963,
+      "learning_rate": 4.504198121452456e-06,
+      "loss": 0.2871,
+      "num_input_tokens_seen": 285470768,
+      "step": 5180,
+      "train_runtime": 31365.9174,
+      "train_tokens_per_second": 9101.305
+    },
+    {
+      "epoch": 1.0644631492506673,
+      "grad_norm": 0.7094226385399854,
+      "learning_rate": 4.49617652244608e-06,
+      "loss": 0.3143,
+      "num_input_tokens_seen": 285749152,
+      "step": 5185,
+      "train_runtime": 31394.7162,
+      "train_tokens_per_second": 9101.823
+    },
+    {
+      "epoch": 1.06548963251899,
+      "grad_norm": 0.6587299406318894,
+      "learning_rate": 4.488156233290445e-06,
+      "loss": 0.3272,
+      "num_input_tokens_seen": 285983168,
+      "step": 5190,
+      "train_runtime": 31422.9584,
+      "train_tokens_per_second": 9101.09
+    },
+    {
+      "epoch": 1.0665161157873126,
+      "grad_norm": 0.5866467475781378,
+      "learning_rate": 4.4801372748368686e-06,
+      "loss": 0.2707,
+      "num_input_tokens_seen": 286261464,
+      "step": 5195,
+      "train_runtime": 31454.9417,
+      "train_tokens_per_second": 9100.683
+    },
+    {
+      "epoch": 1.0675425990556353,
+      "grad_norm": 0.6779024159498442,
+      "learning_rate": 4.472119667933204e-06,
+      "loss": 0.3205,
+      "num_input_tokens_seen": 286535680,
+      "step": 5200,
+      "train_runtime": 31488.4713,
+      "train_tokens_per_second": 9099.701
+    },
+    {
+      "epoch": 1.0685690823239582,
+      "grad_norm": 0.7567502517519482,
+      "learning_rate": 4.464103433423791e-06,
+      "loss": 0.2762,
+      "num_input_tokens_seen": 286795120,
+      "step": 5205,
+      "train_runtime": 31519.7054,
+      "train_tokens_per_second": 9098.915
+    },
+    {
+      "epoch": 1.0695955655922809,
+      "grad_norm": 0.7877620821276659,
+      "learning_rate": 4.456088592149405e-06,
+      "loss": 0.2915,
+      "num_input_tokens_seen": 287043168,
+      "step": 5210,
+      "train_runtime": 31546.8157,
+      "train_tokens_per_second": 9098.959
+    },
+    {
+      "epoch": 1.0706220488606035,
+      "grad_norm": 0.6397347610976213,
+      "learning_rate": 4.448075164947197e-06,
+      "loss": 0.3004,
+      "num_input_tokens_seen": 287315376,
+      "step": 5215,
+      "train_runtime": 31575.9113,
+      "train_tokens_per_second": 9099.195
+    },
+    {
+      "epoch": 1.0716485321289264,
+      "grad_norm": 0.6952510660066078,
+      "learning_rate": 4.44006317265064e-06,
+      "loss": 0.3049,
+      "num_input_tokens_seen": 287614616,
+      "step": 5220,
+      "train_runtime": 31605.5319,
+      "train_tokens_per_second": 9100.135
+    },
+    {
+      "epoch": 1.072675015397249,
+      "grad_norm": 0.7663603140862428,
+      "learning_rate": 4.4320526360894755e-06,
+      "loss": 0.3068,
+      "num_input_tokens_seen": 287861488,
+      "step": 5225,
+      "train_runtime": 31633.7547,
+      "train_tokens_per_second": 9099.82
+    },
+    {
+      "epoch": 1.0737014986655717,
+      "grad_norm": 0.7855571389411248,
+      "learning_rate": 4.424043576089668e-06,
+      "loss": 0.2833,
+      "num_input_tokens_seen": 288157168,
+      "step": 5230,
+      "train_runtime": 31664.6907,
+      "train_tokens_per_second": 9100.268
+    },
+    {
+      "epoch": 1.0747279819338944,
+      "grad_norm": 0.593668628869913,
+      "learning_rate": 4.416036013473336e-06,
+      "loss": 0.2777,
+      "num_input_tokens_seen": 288428816,
+      "step": 5235,
+      "train_runtime": 31701.8065,
+      "train_tokens_per_second": 9098.182
+    },
+    {
+      "epoch": 1.0757544652022173,
+      "grad_norm": 0.6310575702653457,
+      "learning_rate": 4.408029969058703e-06,
+      "loss": 0.2842,
+      "num_input_tokens_seen": 288714624,
+      "step": 5240,
+      "train_runtime": 31733.6423,
+      "train_tokens_per_second": 9098.061
+    },
+    {
+      "epoch": 1.07678094847054,
+      "grad_norm": 0.5979499980313704,
+      "learning_rate": 4.400025463660058e-06,
+      "loss": 0.3174,
+      "num_input_tokens_seen": 289018088,
+      "step": 5245,
+      "train_runtime": 31765.8118,
+      "train_tokens_per_second": 9098.401
+    },
+    {
+      "epoch": 1.0778074317388626,
+      "grad_norm": 0.595670347673524,
+      "learning_rate": 4.392022518087675e-06,
+      "loss": 0.2817,
+      "num_input_tokens_seen": 289304360,
+      "step": 5250,
+      "train_runtime": 31797.7041,
+      "train_tokens_per_second": 9098.278
+    },
+    {
+      "epoch": 1.0788339150071853,
+      "grad_norm": 0.6090396299709708,
+      "learning_rate": 4.3840211531477774e-06,
+      "loss": 0.3004,
+      "num_input_tokens_seen": 289572512,
+      "step": 5255,
+      "train_runtime": 31831.4522,
+      "train_tokens_per_second": 9097.056
+    },
+    {
+      "epoch": 1.0798603982755082,
+      "grad_norm": 0.6934463983256236,
+      "learning_rate": 4.376021389642481e-06,
+      "loss": 0.3046,
+      "num_input_tokens_seen": 289863704,
+      "step": 5260,
+      "train_runtime": 31857.975,
+      "train_tokens_per_second": 9098.623
+    },
+    {
+      "epoch": 1.0808868815438308,
+      "grad_norm": 0.9120769218971705,
+      "learning_rate": 4.368023248369739e-06,
+      "loss": 0.3073,
+      "num_input_tokens_seen": 290119056,
+      "step": 5265,
+      "train_runtime": 31883.8177,
+      "train_tokens_per_second": 9099.257
+    },
+    {
+      "epoch": 1.0819133648121535,
+      "grad_norm": 0.6952330711310782,
+      "learning_rate": 4.360026750123284e-06,
+      "loss": 0.2927,
+      "num_input_tokens_seen": 290388848,
+      "step": 5270,
+      "train_runtime": 31915.5131,
+      "train_tokens_per_second": 9098.674
+    },
+    {
+      "epoch": 1.0829398480804764,
+      "grad_norm": 0.7283962456693076,
+      "learning_rate": 4.352031915692576e-06,
+      "loss": 0.2963,
+      "num_input_tokens_seen": 290678256,
+      "step": 5275,
+      "train_runtime": 31943.5958,
+      "train_tokens_per_second": 9099.735
+    },
+    {
+      "epoch": 1.083966331348799,
+      "grad_norm": 0.7419732268099207,
+      "learning_rate": 4.344038765862757e-06,
+      "loss": 0.2698,
+      "num_input_tokens_seen": 290945904,
+      "step": 5280,
+      "train_runtime": 31972.741,
+      "train_tokens_per_second": 9099.811
+    },
+    {
+      "epoch": 1.0849928146171217,
+      "grad_norm": 0.688618493407117,
+      "learning_rate": 4.336047321414579e-06,
+      "loss": 0.2922,
+      "num_input_tokens_seen": 291213640,
+      "step": 5285,
+      "train_runtime": 31999.9077,
+      "train_tokens_per_second": 9100.452
+    },
+    {
+      "epoch": 1.0860192978854444,
+      "grad_norm": 0.6204340505929726,
+      "learning_rate": 4.328057603124368e-06,
+      "loss": 0.2879,
+      "num_input_tokens_seen": 291493136,
+      "step": 5290,
+      "train_runtime": 32025.9704,
+      "train_tokens_per_second": 9101.774
+    },
+    {
+      "epoch": 1.0870457811537673,
+      "grad_norm": 0.5647923541678943,
+      "learning_rate": 4.320069631763957e-06,
+      "loss": 0.3047,
+      "num_input_tokens_seen": 291757704,
+      "step": 5295,
+      "train_runtime": 32053.5557,
+      "train_tokens_per_second": 9102.195
+    },
+    {
+      "epoch": 1.08807226442209,
+      "grad_norm": 0.7087419870454613,
+      "learning_rate": 4.312083428100642e-06,
+      "loss": 0.2984,
+      "num_input_tokens_seen": 292045576,
+      "step": 5300,
+      "train_runtime": 32091.8286,
+      "train_tokens_per_second": 9100.31
+    },
+    {
+      "epoch": 1.0890987476904126,
+      "grad_norm": 0.6081310866238191,
+      "learning_rate": 4.304099012897121e-06,
+      "loss": 0.3112,
+      "num_input_tokens_seen": 292299376,
+      "step": 5305,
+      "train_runtime": 32116.3856,
+      "train_tokens_per_second": 9101.254
+    },
+    {
+      "epoch": 1.0901252309587353,
+      "grad_norm": 0.7348479157536204,
+      "learning_rate": 4.296116406911443e-06,
+      "loss": 0.3022,
+      "num_input_tokens_seen": 292582288,
+      "step": 5310,
+      "train_runtime": 32142.3065,
+      "train_tokens_per_second": 9102.716
+    },
+    {
+      "epoch": 1.0911517142270581,
+      "grad_norm": 0.6019237352527048,
+      "learning_rate": 4.288135630896949e-06,
+      "loss": 0.2897,
+      "num_input_tokens_seen": 292878144,
+      "step": 5315,
+      "train_runtime": 32173.1087,
+      "train_tokens_per_second": 9103.197
+    },
+    {
+      "epoch": 1.0921781974953808,
+      "grad_norm": 1.4318760262644639,
+      "learning_rate": 4.280156705602232e-06,
+      "loss": 0.3205,
+      "num_input_tokens_seen": 293139672,
+      "step": 5320,
+      "train_runtime": 32202.6409,
+      "train_tokens_per_second": 9102.97
+    },
+    {
+      "epoch": 1.0932046807637035,
+      "grad_norm": 0.5720383648306809,
+      "learning_rate": 4.2721796517710654e-06,
+      "loss": 0.291,
+      "num_input_tokens_seen": 293420872,
+      "step": 5325,
+      "train_runtime": 32229.6375,
+      "train_tokens_per_second": 9104.07
+    },
+    {
+      "epoch": 1.0942311640320264,
+      "grad_norm": 0.6536027582877287,
+      "learning_rate": 4.2642044901423555e-06,
+      "loss": 0.2921,
+      "num_input_tokens_seen": 293690624,
+      "step": 5330,
+      "train_runtime": 32254.3278,
+      "train_tokens_per_second": 9105.464
+    },
+    {
+      "epoch": 1.095257647300349,
+      "grad_norm": 0.9744839622199749,
+      "learning_rate": 4.256231241450101e-06,
+      "loss": 0.2856,
+      "num_input_tokens_seen": 293947088,
+      "step": 5335,
+      "train_runtime": 32281.4736,
+      "train_tokens_per_second": 9105.752
+    },
+    {
+      "epoch": 1.0962841305686717,
+      "grad_norm": 0.627279156252746,
+      "learning_rate": 4.248259926423313e-06,
+      "loss": 0.2833,
+      "num_input_tokens_seen": 294221360,
+      "step": 5340,
+      "train_runtime": 32304.6715,
+      "train_tokens_per_second": 9107.703
+    },
+    {
+      "epoch": 1.0973106138369944,
+      "grad_norm": 0.6847924952662712,
+      "learning_rate": 4.240290565785985e-06,
+      "loss": 0.2897,
+      "num_input_tokens_seen": 294497360,
+      "step": 5345,
+      "train_runtime": 32331.0797,
+      "train_tokens_per_second": 9108.801
+    },
+    {
+      "epoch": 1.0983370971053172,
+      "grad_norm": 0.7691128161217603,
+      "learning_rate": 4.232323180257024e-06,
+      "loss": 0.2711,
+      "num_input_tokens_seen": 294777088,
+      "step": 5350,
+      "train_runtime": 32358.7823,
+      "train_tokens_per_second": 9109.647
+    },
+    {
+      "epoch": 1.09936358037364,
+      "grad_norm": 0.619707914558093,
+      "learning_rate": 4.2243577905502065e-06,
+      "loss": 0.3118,
+      "num_input_tokens_seen": 295082344,
+      "step": 5355,
+      "train_runtime": 32390.7877,
+      "train_tokens_per_second": 9110.07
+    },
+    {
+      "epoch": 1.1003900636419626,
+      "grad_norm": 0.6703508690975893,
+      "learning_rate": 4.216394417374118e-06,
+      "loss": 0.3016,
+      "num_input_tokens_seen": 295356816,
+      "step": 5360,
+      "train_runtime": 32420.614,
+      "train_tokens_per_second": 9110.155
+    },
+    {
+      "epoch": 1.1014165469102855,
+      "grad_norm": 0.6870865575554828,
+      "learning_rate": 4.2084330814321e-06,
+      "loss": 0.2873,
+      "num_input_tokens_seen": 295638992,
+      "step": 5365,
+      "train_runtime": 32450.6899,
+      "train_tokens_per_second": 9110.407
+    },
+    {
+      "epoch": 1.1024430301786081,
+      "grad_norm": 0.6445487419965991,
+      "learning_rate": 4.2004738034222e-06,
+      "loss": 0.2767,
+      "num_input_tokens_seen": 295921880,
+      "step": 5370,
+      "train_runtime": 32479.9316,
+      "train_tokens_per_second": 9110.915
+    },
+    {
+      "epoch": 1.1034695134469308,
+      "grad_norm": 0.6501613452465083,
+      "learning_rate": 4.192516604037116e-06,
+      "loss": 0.2901,
+      "num_input_tokens_seen": 296209400,
+      "step": 5375,
+      "train_runtime": 32513.0421,
+      "train_tokens_per_second": 9110.479
+    },
+    {
+      "epoch": 1.1044959967152534,
+      "grad_norm": 0.6749975229319636,
+      "learning_rate": 4.184561503964138e-06,
+      "loss": 0.3029,
+      "num_input_tokens_seen": 296500888,
+      "step": 5380,
+      "train_runtime": 32540.1412,
+      "train_tokens_per_second": 9111.85
+    },
+    {
+      "epoch": 1.1055224799835763,
+      "grad_norm": 0.5789829802638805,
+      "learning_rate": 4.1766085238851e-06,
+      "loss": 0.2865,
+      "num_input_tokens_seen": 296785600,
+      "step": 5385,
+      "train_runtime": 32575.7025,
+      "train_tokens_per_second": 9110.643
+    },
+    {
+      "epoch": 1.106548963251899,
+      "grad_norm": 0.6213250441114163,
+      "learning_rate": 4.168657684476326e-06,
+      "loss": 0.2862,
+      "num_input_tokens_seen": 297072760,
+      "step": 5390,
+      "train_runtime": 32606.9008,
+      "train_tokens_per_second": 9110.733
+    },
+    {
+      "epoch": 1.1075754465202217,
+      "grad_norm": 0.632922258810323,
+      "learning_rate": 4.1607090064085745e-06,
+      "loss": 0.2654,
+      "num_input_tokens_seen": 297351912,
+      "step": 5395,
+      "train_runtime": 32639.7819,
+      "train_tokens_per_second": 9110.107
+    },
+    {
+      "epoch": 1.1086019297885445,
+      "grad_norm": 0.6843660916984037,
+      "learning_rate": 4.152762510346981e-06,
+      "loss": 0.2934,
+      "num_input_tokens_seen": 297616752,
+      "step": 5400,
+      "train_runtime": 32669.9752,
+      "train_tokens_per_second": 9109.794
+    },
+    {
+      "epoch": 1.1096284130568672,
+      "grad_norm": 0.5857821627200764,
+      "learning_rate": 4.1448182169510136e-06,
+      "loss": 0.2955,
+      "num_input_tokens_seen": 297877672,
+      "step": 5405,
+      "train_runtime": 32699.2875,
+      "train_tokens_per_second": 9109.607
+    },
+    {
+      "epoch": 1.1106548963251899,
+      "grad_norm": 0.8900042016375527,
+      "learning_rate": 4.13687614687441e-06,
+      "loss": 0.3181,
+      "num_input_tokens_seen": 298128472,
+      "step": 5410,
+      "train_runtime": 32730.0914,
+      "train_tokens_per_second": 9108.697
+    },
+    {
+      "epoch": 1.1116813795935125,
+      "grad_norm": 0.6738870724361132,
+      "learning_rate": 4.128936320765131e-06,
+      "loss": 0.2944,
+      "num_input_tokens_seen": 298408672,
+      "step": 5415,
+      "train_runtime": 32757.6372,
+      "train_tokens_per_second": 9109.591
+    },
+    {
+      "epoch": 1.1127078628618354,
+      "grad_norm": 0.5497469878670879,
+      "learning_rate": 4.120998759265296e-06,
+      "loss": 0.2804,
+      "num_input_tokens_seen": 298675904,
+      "step": 5420,
+      "train_runtime": 32789.0012,
+      "train_tokens_per_second": 9109.027
+    },
+    {
+      "epoch": 1.113734346130158,
+      "grad_norm": 0.6402149460505865,
+      "learning_rate": 4.113063483011149e-06,
+      "loss": 0.3023,
+      "num_input_tokens_seen": 298977904,
+      "step": 5425,
+      "train_runtime": 32817.5361,
+      "train_tokens_per_second": 9110.309
+    },
+    {
+      "epoch": 1.1147608293984808,
+      "grad_norm": 0.6893994700663192,
+      "learning_rate": 4.105130512632983e-06,
+      "loss": 0.2733,
+      "num_input_tokens_seen": 299232712,
+      "step": 5430,
+      "train_runtime": 32846.2766,
+      "train_tokens_per_second": 9110.095
+    },
+    {
+      "epoch": 1.1157873126668036,
+      "grad_norm": 0.7007763967235385,
+      "learning_rate": 4.0971998687551e-06,
+      "loss": 0.3028,
+      "num_input_tokens_seen": 299514600,
+      "step": 5435,
+      "train_runtime": 32876.8328,
+      "train_tokens_per_second": 9110.202
+    },
+    {
+      "epoch": 1.1168137959351263,
+      "grad_norm": 0.7137239872944446,
+      "learning_rate": 4.0892715719957485e-06,
+      "loss": 0.2852,
+      "num_input_tokens_seen": 299775960,
+      "step": 5440,
+      "train_runtime": 32909.6397,
+      "train_tokens_per_second": 9109.062
+    },
+    {
+      "epoch": 1.117840279203449,
+      "grad_norm": 0.6884517461268562,
+      "learning_rate": 4.081345642967086e-06,
+      "loss": 0.3168,
+      "num_input_tokens_seen": 300066048,
+      "step": 5445,
+      "train_runtime": 32941.9654,
+      "train_tokens_per_second": 9108.93
+    },
+    {
+      "epoch": 1.1188667624717716,
+      "grad_norm": 0.6780755598313941,
+      "learning_rate": 4.0734221022751025e-06,
+      "loss": 0.2873,
+      "num_input_tokens_seen": 300353184,
+      "step": 5450,
+      "train_runtime": 32975.2791,
+      "train_tokens_per_second": 9108.435
+    },
+    {
+      "epoch": 1.1198932457400945,
+      "grad_norm": 0.6272991888006443,
+      "learning_rate": 4.065500970519585e-06,
+      "loss": 0.3107,
+      "num_input_tokens_seen": 300651664,
+      "step": 5455,
+      "train_runtime": 33004.5808,
+      "train_tokens_per_second": 9109.392
+    },
+    {
+      "epoch": 1.1209197290084172,
+      "grad_norm": 0.6084631825838913,
+      "learning_rate": 4.057582268294057e-06,
+      "loss": 0.2907,
+      "num_input_tokens_seen": 300911784,
+      "step": 5460,
+      "train_runtime": 33028.9113,
+      "train_tokens_per_second": 9110.557
+    },
+    {
+      "epoch": 1.1219462122767399,
+      "grad_norm": 0.6958199347167279,
+      "learning_rate": 4.049666016185726e-06,
+      "loss": 0.3032,
+      "num_input_tokens_seen": 301178000,
+      "step": 5465,
+      "train_runtime": 33059.2738,
+      "train_tokens_per_second": 9110.242
+    },
+    {
+      "epoch": 1.1229726955450625,
+      "grad_norm": 0.6371367350451463,
+      "learning_rate": 4.041752234775427e-06,
+      "loss": 0.3097,
+      "num_input_tokens_seen": 301435736,
+      "step": 5470,
+      "train_runtime": 33087.0078,
+      "train_tokens_per_second": 9110.396
+    },
+    {
+      "epoch": 1.1239991788133854,
+      "grad_norm": 0.6238944626029613,
+      "learning_rate": 4.033840944637571e-06,
+      "loss": 0.2901,
+      "num_input_tokens_seen": 301726464,
+      "step": 5475,
+      "train_runtime": 33114.3959,
+      "train_tokens_per_second": 9111.64
+    },
+    {
+      "epoch": 1.125025662081708,
+      "grad_norm": 0.7035697646537921,
+      "learning_rate": 4.025932166340099e-06,
+      "loss": 0.3128,
+      "num_input_tokens_seen": 302029920,
+      "step": 5480,
+      "train_runtime": 33143.1686,
+      "train_tokens_per_second": 9112.886
+    },
+    {
+      "epoch": 1.1260521453500307,
+      "grad_norm": 0.610598874741099,
+      "learning_rate": 4.018025920444415e-06,
+      "loss": 0.27,
+      "num_input_tokens_seen": 302309984,
+      "step": 5485,
+      "train_runtime": 33172.2627,
+      "train_tokens_per_second": 9113.336
+    },
+    {
+      "epoch": 1.1270786286183536,
+      "grad_norm": 0.6668126286532043,
+      "learning_rate": 4.01012222750534e-06,
+      "loss": 0.3054,
+      "num_input_tokens_seen": 302602576,
+      "step": 5490,
+      "train_runtime": 33203.912,
+      "train_tokens_per_second": 9113.462
+    },
+    {
+      "epoch": 1.1281051118866763,
+      "grad_norm": 0.6666951971710134,
+      "learning_rate": 4.002221108071063e-06,
+      "loss": 0.3106,
+      "num_input_tokens_seen": 302886408,
+      "step": 5495,
+      "train_runtime": 33232.675,
+      "train_tokens_per_second": 9114.115
+    },
+    {
+      "epoch": 1.129131595154999,
+      "grad_norm": 0.947568721001854,
+      "learning_rate": 3.9943225826830745e-06,
+      "loss": 0.2995,
+      "num_input_tokens_seen": 303171992,
+      "step": 5500,
+      "train_runtime": 33257.8829,
+      "train_tokens_per_second": 9115.793
+    },
+    {
+      "epoch": 1.1301580784233216,
+      "grad_norm": 0.614941866200498,
+      "learning_rate": 3.986426671876128e-06,
+      "loss": 0.3024,
+      "num_input_tokens_seen": 303467944,
+      "step": 5505,
+      "train_runtime": 33289.9666,
+      "train_tokens_per_second": 9115.898
+    },
+    {
+      "epoch": 1.1311845616916445,
+      "grad_norm": 0.675762411991752,
+      "learning_rate": 3.978533396178171e-06,
+      "loss": 0.3083,
+      "num_input_tokens_seen": 303744352,
+      "step": 5510,
+      "train_runtime": 33315.6303,
+      "train_tokens_per_second": 9117.173
+    },
+    {
+      "epoch": 1.1322110449599672,
+      "grad_norm": 0.6707733074403013,
+      "learning_rate": 3.970642776110314e-06,
+      "loss": 0.3276,
+      "num_input_tokens_seen": 304030224,
+      "step": 5515,
+      "train_runtime": 33346.6317,
+      "train_tokens_per_second": 9117.269
+    },
+    {
+      "epoch": 1.1332375282282898,
+      "grad_norm": 0.7037082022330918,
+      "learning_rate": 3.962754832186748e-06,
+      "loss": 0.2978,
+      "num_input_tokens_seen": 304279144,
+      "step": 5520,
+      "train_runtime": 33376.5008,
+      "train_tokens_per_second": 9116.568
+    },
+    {
+      "epoch": 1.1342640114966125,
+      "grad_norm": 0.6075423642478847,
+      "learning_rate": 3.954869584914717e-06,
+      "loss": 0.2905,
+      "num_input_tokens_seen": 304567752,
+      "step": 5525,
+      "train_runtime": 33404.4253,
+      "train_tokens_per_second": 9117.587
+    },
+    {
+      "epoch": 1.1352904947649354,
+      "grad_norm": 0.6457468041989161,
+      "learning_rate": 3.946987054794447e-06,
+      "loss": 0.2784,
+      "num_input_tokens_seen": 304861424,
+      "step": 5530,
+      "train_runtime": 33433.4615,
+      "train_tokens_per_second": 9118.452
+    },
+    {
+      "epoch": 1.136316978033258,
+      "grad_norm": 0.7351552713967859,
+      "learning_rate": 3.939107262319108e-06,
+      "loss": 0.3037,
+      "num_input_tokens_seen": 305111432,
+      "step": 5535,
+      "train_runtime": 33451.6576,
+      "train_tokens_per_second": 9120.966
+    },
+    {
+      "epoch": 1.1373434613015807,
+      "grad_norm": 0.7513293731141045,
+      "learning_rate": 3.931230227974746e-06,
+      "loss": 0.2864,
+      "num_input_tokens_seen": 305396352,
+      "step": 5540,
+      "train_runtime": 33479.8467,
+      "train_tokens_per_second": 9121.797
+    },
+    {
+      "epoch": 1.1383699445699036,
+      "grad_norm": 0.6295263818912261,
+      "learning_rate": 3.923355972240238e-06,
+      "loss": 0.2786,
+      "num_input_tokens_seen": 305661792,
+      "step": 5545,
+      "train_runtime": 33504.9972,
+      "train_tokens_per_second": 9122.872
+    },
+    {
+      "epoch": 1.1393964278382263,
+      "grad_norm": 0.673630271875065,
+      "learning_rate": 3.915484515587239e-06,
+      "loss": 0.3035,
+      "num_input_tokens_seen": 305917064,
+      "step": 5550,
+      "train_runtime": 33530.502,
+      "train_tokens_per_second": 9123.546
+    },
+    {
+      "epoch": 1.140422911106549,
+      "grad_norm": 0.6533812813047974,
+      "learning_rate": 3.907615878480127e-06,
+      "loss": 0.2954,
+      "num_input_tokens_seen": 306194112,
+      "step": 5555,
+      "train_runtime": 33559.5006,
+      "train_tokens_per_second": 9123.917
+    },
+    {
+      "epoch": 1.1414493943748716,
+      "grad_norm": 0.6805579736616899,
+      "learning_rate": 3.8997500813759446e-06,
+      "loss": 0.2914,
+      "num_input_tokens_seen": 306445872,
+      "step": 5560,
+      "train_runtime": 33589.6929,
+      "train_tokens_per_second": 9123.211
+    },
+    {
+      "epoch": 1.1424758776431945,
+      "grad_norm": 0.6852550277878736,
+      "learning_rate": 3.891887144724354e-06,
+      "loss": 0.2755,
+      "num_input_tokens_seen": 306705400,
+      "step": 5565,
+      "train_runtime": 33613.8951,
+      "train_tokens_per_second": 9124.364
+    },
+    {
+      "epoch": 1.1435023609115171,
+      "grad_norm": 0.624978286912364,
+      "learning_rate": 3.884027088967585e-06,
+      "loss": 0.2875,
+      "num_input_tokens_seen": 306990536,
+      "step": 5570,
+      "train_runtime": 33646.8244,
+      "train_tokens_per_second": 9123.908
+    },
+    {
+      "epoch": 1.1445288441798398,
+      "grad_norm": 0.6633970710707415,
+      "learning_rate": 3.876169934540372e-06,
+      "loss": 0.2687,
+      "num_input_tokens_seen": 307260968,
+      "step": 5575,
+      "train_runtime": 33672.9503,
+      "train_tokens_per_second": 9124.86
+    },
+    {
+      "epoch": 1.1455553274481627,
+      "grad_norm": 0.6887745577215199,
+      "learning_rate": 3.868315701869907e-06,
+      "loss": 0.3015,
+      "num_input_tokens_seen": 307550096,
+      "step": 5580,
+      "train_runtime": 33705.4157,
+      "train_tokens_per_second": 9124.649
+    },
+    {
+      "epoch": 1.1465818107164854,
+      "grad_norm": 0.6815201885284509,
+      "learning_rate": 3.860464411375791e-06,
+      "loss": 0.2995,
+      "num_input_tokens_seen": 307823200,
+      "step": 5585,
+      "train_runtime": 33739.793,
+      "train_tokens_per_second": 9123.447
+    },
+    {
+      "epoch": 1.147608293984808,
+      "grad_norm": 0.6531342176821333,
+      "learning_rate": 3.852616083469968e-06,
+      "loss": 0.2914,
+      "num_input_tokens_seen": 308124024,
+      "step": 5590,
+      "train_runtime": 33774.4304,
+      "train_tokens_per_second": 9122.997
+    },
+    {
+      "epoch": 1.1486347772531307,
+      "grad_norm": 0.6882573898675105,
+      "learning_rate": 3.844770738556688e-06,
+      "loss": 0.3141,
+      "num_input_tokens_seen": 308406264,
+      "step": 5595,
+      "train_runtime": 33802.8467,
+      "train_tokens_per_second": 9123.677
+    },
+    {
+      "epoch": 1.1496612605214536,
+      "grad_norm": 0.7170970734561046,
+      "learning_rate": 3.8369283970324355e-06,
+      "loss": 0.287,
+      "num_input_tokens_seen": 308681976,
+      "step": 5600,
+      "train_runtime": 33829.3661,
+      "train_tokens_per_second": 9124.675
+    },
+    {
+      "epoch": 1.1506877437897762,
+      "grad_norm": 0.6271764598902886,
+      "learning_rate": 3.8290890792859e-06,
+      "loss": 0.2857,
+      "num_input_tokens_seen": 308966576,
+      "step": 5605,
+      "train_runtime": 33856.7867,
+      "train_tokens_per_second": 9125.691
+    },
+    {
+      "epoch": 1.151714227058099,
+      "grad_norm": 0.7123549931745463,
+      "learning_rate": 3.821252805697898e-06,
+      "loss": 0.2776,
+      "num_input_tokens_seen": 309243056,
+      "step": 5610,
+      "train_runtime": 33889.8136,
+      "train_tokens_per_second": 9124.956
+    },
+    {
+      "epoch": 1.1527407103264218,
+      "grad_norm": 0.6674340316011919,
+      "learning_rate": 3.8134195966413383e-06,
+      "loss": 0.2882,
+      "num_input_tokens_seen": 309490672,
+      "step": 5615,
+      "train_runtime": 33918.1756,
+      "train_tokens_per_second": 9124.626
+    },
+    {
+      "epoch": 1.1537671935947444,
+      "grad_norm": 0.6867549917612106,
+      "learning_rate": 3.805589472481159e-06,
+      "loss": 0.2754,
+      "num_input_tokens_seen": 309780192,
+      "step": 5620,
+      "train_runtime": 33949.5612,
+      "train_tokens_per_second": 9124.719
+    },
+    {
+      "epoch": 1.154793676863067,
+      "grad_norm": 0.692003064648111,
+      "learning_rate": 3.7977624535742815e-06,
+      "loss": 0.2942,
+      "num_input_tokens_seen": 310086008,
+      "step": 5625,
+      "train_runtime": 33979.4899,
+      "train_tokens_per_second": 9125.682
+    },
+    {
+      "epoch": 1.1558201601313898,
+      "grad_norm": 0.7506895170392874,
+      "learning_rate": 3.7899385602695505e-06,
+      "loss": 0.3041,
+      "num_input_tokens_seen": 310370792,
+      "step": 5630,
+      "train_runtime": 34009.5104,
+      "train_tokens_per_second": 9126.0
+    },
+    {
+      "epoch": 1.1568466433997127,
+      "grad_norm": 0.734162089565356,
+      "learning_rate": 3.7821178129076864e-06,
+      "loss": 0.2915,
+      "num_input_tokens_seen": 310647024,
+      "step": 5635,
+      "train_runtime": 34034.4823,
+      "train_tokens_per_second": 9127.42
+    },
+    {
+      "epoch": 1.1578731266680353,
+      "grad_norm": 0.6967465483941289,
+      "learning_rate": 3.7743002318212324e-06,
+      "loss": 0.2984,
+      "num_input_tokens_seen": 310900816,
+      "step": 5640,
+      "train_runtime": 34062.3993,
+      "train_tokens_per_second": 9127.39
+    },
+    {
+      "epoch": 1.158899609936358,
+      "grad_norm": 0.7084554107781367,
+      "learning_rate": 3.766485837334498e-06,
+      "loss": 0.3119,
+      "num_input_tokens_seen": 311163240,
+      "step": 5645,
+      "train_runtime": 34100.4606,
+      "train_tokens_per_second": 9124.898
+    },
+    {
+      "epoch": 1.1599260932046809,
+      "grad_norm": 0.7172121016727513,
+      "learning_rate": 3.758674649763506e-06,
+      "loss": 0.288,
+      "num_input_tokens_seen": 311459152,
+      "step": 5650,
+      "train_runtime": 34133.7207,
+      "train_tokens_per_second": 9124.676
+    },
+    {
+      "epoch": 1.1609525764730035,
+      "grad_norm": 0.692382827685263,
+      "learning_rate": 3.750866689415945e-06,
+      "loss": 0.2986,
+      "num_input_tokens_seen": 311725120,
+      "step": 5655,
+      "train_runtime": 34164.2304,
+      "train_tokens_per_second": 9124.313
+    },
+    {
+      "epoch": 1.1619790597413262,
+      "grad_norm": 0.7551182865034742,
+      "learning_rate": 3.7430619765911137e-06,
+      "loss": 0.2943,
+      "num_input_tokens_seen": 312007952,
+      "step": 5660,
+      "train_runtime": 34199.4657,
+      "train_tokens_per_second": 9123.182
+    },
+    {
+      "epoch": 1.1630055430096489,
+      "grad_norm": 0.6218805510949,
+      "learning_rate": 3.7352605315798658e-06,
+      "loss": 0.2839,
+      "num_input_tokens_seen": 312298008,
+      "step": 5665,
+      "train_runtime": 34236.3779,
+      "train_tokens_per_second": 9121.818
+    },
+    {
+      "epoch": 1.1640320262779718,
+      "grad_norm": 0.7802022750136441,
+      "learning_rate": 3.7274623746645587e-06,
+      "loss": 0.3151,
+      "num_input_tokens_seen": 312554440,
+      "step": 5670,
+      "train_runtime": 34270.7072,
+      "train_tokens_per_second": 9120.163
+    },
+    {
+      "epoch": 1.1650585095462944,
+      "grad_norm": 0.7045100048923442,
+      "learning_rate": 3.7196675261190063e-06,
+      "loss": 0.2744,
+      "num_input_tokens_seen": 312829504,
+      "step": 5675,
+      "train_runtime": 34303.158,
+      "train_tokens_per_second": 9119.554
+    },
+    {
+      "epoch": 1.166084992814617,
+      "grad_norm": 0.737538655099362,
+      "learning_rate": 3.7118760062084146e-06,
+      "loss": 0.2807,
+      "num_input_tokens_seen": 313106496,
+      "step": 5680,
+      "train_runtime": 34327.3753,
+      "train_tokens_per_second": 9121.19
+    },
+    {
+      "epoch": 1.16711147608294,
+      "grad_norm": 0.6869016646450301,
+      "learning_rate": 3.7040878351893418e-06,
+      "loss": 0.3131,
+      "num_input_tokens_seen": 313384160,
+      "step": 5685,
+      "train_runtime": 34360.6227,
+      "train_tokens_per_second": 9120.445
+    },
+    {
+      "epoch": 1.1681379593512626,
+      "grad_norm": 0.7896660699201127,
+      "learning_rate": 3.6963030333096316e-06,
+      "loss": 0.3041,
+      "num_input_tokens_seen": 313660784,
+      "step": 5690,
+      "train_runtime": 34384.7913,
+      "train_tokens_per_second": 9122.079
+    },
+    {
+      "epoch": 1.1691644426195853,
+      "grad_norm": 0.634030684865466,
+      "learning_rate": 3.6885216208083807e-06,
+      "loss": 0.2812,
+      "num_input_tokens_seen": 313943216,
+      "step": 5695,
+      "train_runtime": 34418.1446,
+      "train_tokens_per_second": 9121.445
+    },
+    {
+      "epoch": 1.170190925887908,
+      "grad_norm": 0.6352796299986511,
+      "learning_rate": 3.680743617915861e-06,
+      "loss": 0.2998,
+      "num_input_tokens_seen": 314235576,
+      "step": 5700,
+      "train_runtime": 34447.852,
+      "train_tokens_per_second": 9122.066
+    },
+    {
+      "epoch": 1.1712174091562308,
+      "grad_norm": 0.6575226772412485,
+      "learning_rate": 3.6729690448534884e-06,
+      "loss": 0.3125,
+      "num_input_tokens_seen": 314523768,
+      "step": 5705,
+      "train_runtime": 34473.8601,
+      "train_tokens_per_second": 9123.544
+    },
+    {
+      "epoch": 1.1722438924245535,
+      "grad_norm": 0.6717498724885491,
+      "learning_rate": 3.665197921833759e-06,
+      "loss": 0.3114,
+      "num_input_tokens_seen": 314792136,
+      "step": 5710,
+      "train_runtime": 34504.4488,
+      "train_tokens_per_second": 9123.233
+    },
+    {
+      "epoch": 1.1732703756928762,
+      "grad_norm": 0.7517955301504996,
+      "learning_rate": 3.657430269060201e-06,
+      "loss": 0.3112,
+      "num_input_tokens_seen": 315061944,
+      "step": 5715,
+      "train_runtime": 34532.4836,
+      "train_tokens_per_second": 9123.64
+    },
+    {
+      "epoch": 1.1742968589611988,
+      "grad_norm": 0.7018283563727894,
+      "learning_rate": 3.649666106727318e-06,
+      "loss": 0.3086,
+      "num_input_tokens_seen": 315353312,
+      "step": 5720,
+      "train_runtime": 34569.5493,
+      "train_tokens_per_second": 9122.286
+    },
+    {
+      "epoch": 1.1753233422295217,
+      "grad_norm": 0.6748393142942318,
+      "learning_rate": 3.6419054550205403e-06,
+      "loss": 0.2924,
+      "num_input_tokens_seen": 315617384,
+      "step": 5725,
+      "train_runtime": 34598.0014,
+      "train_tokens_per_second": 9122.417
+    },
+    {
+      "epoch": 1.1763498254978444,
+      "grad_norm": 0.6207824818175754,
+      "learning_rate": 3.6341483341161734e-06,
+      "loss": 0.277,
+      "num_input_tokens_seen": 315911840,
+      "step": 5730,
+      "train_runtime": 34628.2251,
+      "train_tokens_per_second": 9122.958
+    },
+    {
+      "epoch": 1.177376308766167,
+      "grad_norm": 0.6607854815637404,
+      "learning_rate": 3.6263947641813415e-06,
+      "loss": 0.273,
+      "num_input_tokens_seen": 316175496,
+      "step": 5735,
+      "train_runtime": 34664.0689,
+      "train_tokens_per_second": 9121.131
+    },
+    {
+      "epoch": 1.1784027920344897,
+      "grad_norm": 0.8569459544339443,
+      "learning_rate": 3.6186447653739333e-06,
+      "loss": 0.3101,
+      "num_input_tokens_seen": 316439808,
+      "step": 5740,
+      "train_runtime": 34696.9908,
+      "train_tokens_per_second": 9120.094
+    },
+    {
+      "epoch": 1.1794292753028126,
+      "grad_norm": 0.6213092025340214,
+      "learning_rate": 3.6108983578425584e-06,
+      "loss": 0.3058,
+      "num_input_tokens_seen": 316717328,
+      "step": 5745,
+      "train_runtime": 34730.6601,
+      "train_tokens_per_second": 9119.243
+    },
+    {
+      "epoch": 1.1804557585711353,
+      "grad_norm": 0.5939559358966061,
+      "learning_rate": 3.603155561726489e-06,
+      "loss": 0.2658,
+      "num_input_tokens_seen": 317004640,
+      "step": 5750,
+      "train_runtime": 34760.2749,
+      "train_tokens_per_second": 9119.739
+    },
+    {
+      "epoch": 1.181482241839458,
+      "grad_norm": 0.6789285746784032,
+      "learning_rate": 3.5954163971556076e-06,
+      "loss": 0.299,
+      "num_input_tokens_seen": 317262688,
+      "step": 5755,
+      "train_runtime": 34792.9995,
+      "train_tokens_per_second": 9118.578
+    },
+    {
+      "epoch": 1.1825087251077808,
+      "grad_norm": 0.8068879156640408,
+      "learning_rate": 3.587680884250353e-06,
+      "loss": 0.2896,
+      "num_input_tokens_seen": 317511344,
+      "step": 5760,
+      "train_runtime": 34819.9713,
+      "train_tokens_per_second": 9118.656
+    },
+    {
+      "epoch": 1.1835352083761035,
+      "grad_norm": 0.6664054733248244,
+      "learning_rate": 3.5799490431216737e-06,
+      "loss": 0.3173,
+      "num_input_tokens_seen": 317777656,
+      "step": 5765,
+      "train_runtime": 34850.8733,
+      "train_tokens_per_second": 9118.212
+    },
+    {
+      "epoch": 1.1845616916444262,
+      "grad_norm": 0.7208080129743141,
+      "learning_rate": 3.572220893870971e-06,
+      "loss": 0.2948,
+      "num_input_tokens_seen": 318053224,
+      "step": 5770,
+      "train_runtime": 34879.2285,
+      "train_tokens_per_second": 9118.7
+    },
+    {
+      "epoch": 1.1855881749127488,
+      "grad_norm": 0.6783951560086422,
+      "learning_rate": 3.5644964565900484e-06,
+      "loss": 0.3009,
+      "num_input_tokens_seen": 318336512,
+      "step": 5775,
+      "train_runtime": 34908.8969,
+      "train_tokens_per_second": 9119.065
+    },
+    {
+      "epoch": 1.1866146581810717,
+      "grad_norm": 0.8161309433456831,
+      "learning_rate": 3.556775751361055e-06,
+      "loss": 0.3004,
+      "num_input_tokens_seen": 318583224,
+      "step": 5780,
+      "train_runtime": 34938.6783,
+      "train_tokens_per_second": 9118.354
+    },
+    {
+      "epoch": 1.1876411414493944,
+      "grad_norm": 0.6670408010631228,
+      "learning_rate": 3.5490587982564463e-06,
+      "loss": 0.2888,
+      "num_input_tokens_seen": 318864656,
+      "step": 5785,
+      "train_runtime": 34972.6402,
+      "train_tokens_per_second": 9117.546
+    },
+    {
+      "epoch": 1.188667624717717,
+      "grad_norm": 0.6241302948547998,
+      "learning_rate": 3.5413456173389128e-06,
+      "loss": 0.3131,
+      "num_input_tokens_seen": 319144224,
+      "step": 5790,
+      "train_runtime": 34999.2632,
+      "train_tokens_per_second": 9118.598
+    },
+    {
+      "epoch": 1.18969410798604,
+      "grad_norm": 0.695809667993355,
+      "learning_rate": 3.533636228661342e-06,
+      "loss": 0.296,
+      "num_input_tokens_seen": 319410272,
+      "step": 5795,
+      "train_runtime": 35022.911,
+      "train_tokens_per_second": 9120.038
+    },
+    {
+      "epoch": 1.1907205912543626,
+      "grad_norm": 0.7799040127604728,
+      "learning_rate": 3.525930652266767e-06,
+      "loss": 0.2982,
+      "num_input_tokens_seen": 319687696,
+      "step": 5800,
+      "train_runtime": 35057.3725,
+      "train_tokens_per_second": 9118.986
+    },
+    {
+      "epoch": 1.1917470745226852,
+      "grad_norm": 0.6512409152671484,
+      "learning_rate": 3.5182289081883012e-06,
+      "loss": 0.28,
+      "num_input_tokens_seen": 319957392,
+      "step": 5805,
+      "train_runtime": 35082.7971,
+      "train_tokens_per_second": 9120.065
+    },
+    {
+      "epoch": 1.192773557791008,
+      "grad_norm": 0.6589726595157118,
+      "learning_rate": 3.510531016449102e-06,
+      "loss": 0.2978,
+      "num_input_tokens_seen": 320216392,
+      "step": 5810,
+      "train_runtime": 35121.8001,
+      "train_tokens_per_second": 9117.312
+    },
+    {
+      "epoch": 1.1938000410593308,
+      "grad_norm": 0.7194892287627295,
+      "learning_rate": 3.5028369970623056e-06,
+      "loss": 0.2932,
+      "num_input_tokens_seen": 320498544,
+      "step": 5815,
+      "train_runtime": 35155.6485,
+      "train_tokens_per_second": 9116.559
+    },
+    {
+      "epoch": 1.1948265243276535,
+      "grad_norm": 0.686250135381765,
+      "learning_rate": 3.4951468700309855e-06,
+      "loss": 0.2818,
+      "num_input_tokens_seen": 320758664,
+      "step": 5820,
+      "train_runtime": 35182.3944,
+      "train_tokens_per_second": 9117.022
+    },
+    {
+      "epoch": 1.1958530075959761,
+      "grad_norm": 0.6964332191801913,
+      "learning_rate": 3.4874606553480934e-06,
+      "loss": 0.3102,
+      "num_input_tokens_seen": 321015648,
+      "step": 5825,
+      "train_runtime": 35215.2941,
+      "train_tokens_per_second": 9115.802
+    },
+    {
+      "epoch": 1.196879490864299,
+      "grad_norm": 0.6701121057801793,
+      "learning_rate": 3.4797783729964085e-06,
+      "loss": 0.2945,
+      "num_input_tokens_seen": 321275968,
+      "step": 5830,
+      "train_runtime": 35235.9688,
+      "train_tokens_per_second": 9117.841
+    },
+    {
+      "epoch": 1.1979059741326217,
+      "grad_norm": 0.8980114578024757,
+      "learning_rate": 3.472100042948488e-06,
+      "loss": 0.3167,
+      "num_input_tokens_seen": 321550696,
+      "step": 5835,
+      "train_runtime": 35265.2894,
+      "train_tokens_per_second": 9118.051
+    },
+    {
+      "epoch": 1.1989324574009443,
+      "grad_norm": 0.6882583339616317,
+      "learning_rate": 3.4644256851666157e-06,
+      "loss": 0.3062,
+      "num_input_tokens_seen": 321824224,
+      "step": 5840,
+      "train_runtime": 35298.7466,
+      "train_tokens_per_second": 9117.157
+    },
+    {
+      "epoch": 1.199958940669267,
+      "grad_norm": 0.6657288459921779,
+      "learning_rate": 3.456755319602746e-06,
+      "loss": 0.276,
+      "num_input_tokens_seen": 322083664,
+      "step": 5845,
+      "train_runtime": 35330.0479,
+      "train_tokens_per_second": 9116.423
+    },
+    {
+      "epoch": 1.20098542393759,
+      "grad_norm": 0.6642933019706218,
+      "learning_rate": 3.4490889661984515e-06,
+      "loss": 0.3022,
+      "num_input_tokens_seen": 322375200,
+      "step": 5850,
+      "train_runtime": 35362.6708,
+      "train_tokens_per_second": 9116.257
+    },
+    {
+      "epoch": 1.2020119072059126,
+      "grad_norm": 0.7128055533366179,
+      "learning_rate": 3.441426644884882e-06,
+      "loss": 0.2942,
+      "num_input_tokens_seen": 322634928,
+      "step": 5855,
+      "train_runtime": 35393.2194,
+      "train_tokens_per_second": 9115.727
+    },
+    {
+      "epoch": 1.2030383904742352,
+      "grad_norm": 0.7511192445996998,
+      "learning_rate": 3.4337683755826966e-06,
+      "loss": 0.3006,
+      "num_input_tokens_seen": 322912800,
+      "step": 5860,
+      "train_runtime": 35418.2526,
+      "train_tokens_per_second": 9117.13
+    },
+    {
+      "epoch": 1.204064873742558,
+      "grad_norm": 0.7528735100479516,
+      "learning_rate": 3.4261141782020246e-06,
+      "loss": 0.2961,
+      "num_input_tokens_seen": 323161240,
+      "step": 5865,
+      "train_runtime": 35448.3309,
+      "train_tokens_per_second": 9116.402
+    },
+    {
+      "epoch": 1.2050913570108808,
+      "grad_norm": 0.7865504350748835,
+      "learning_rate": 3.4184640726424034e-06,
+      "loss": 0.2944,
+      "num_input_tokens_seen": 323435584,
+      "step": 5870,
+      "train_runtime": 35478.8494,
+      "train_tokens_per_second": 9116.293
+    },
+    {
+      "epoch": 1.2061178402792034,
+      "grad_norm": 0.804513780577209,
+      "learning_rate": 3.4108180787927425e-06,
+      "loss": 0.279,
+      "num_input_tokens_seen": 323693768,
+      "step": 5875,
+      "train_runtime": 35508.1971,
+      "train_tokens_per_second": 9116.029
+    },
+    {
+      "epoch": 1.207144323547526,
+      "grad_norm": 0.6507330796834635,
+      "learning_rate": 3.403176216531251e-06,
+      "loss": 0.276,
+      "num_input_tokens_seen": 323995112,
+      "step": 5880,
+      "train_runtime": 35540.9431,
+      "train_tokens_per_second": 9116.109
+    },
+    {
+      "epoch": 1.208170806815849,
+      "grad_norm": 0.6590281331720018,
+      "learning_rate": 3.395538505725401e-06,
+      "loss": 0.2962,
+      "num_input_tokens_seen": 324253888,
+      "step": 5885,
+      "train_runtime": 35565.9879,
+      "train_tokens_per_second": 9116.966
+    },
+    {
+      "epoch": 1.2091972900841716,
+      "grad_norm": 0.5991626478332235,
+      "learning_rate": 3.3879049662318746e-06,
+      "loss": 0.2944,
+      "num_input_tokens_seen": 324548104,
+      "step": 5890,
+      "train_runtime": 35587.6001,
+      "train_tokens_per_second": 9119.696
+    },
+    {
+      "epoch": 1.2102237733524943,
+      "grad_norm": 0.7810528146448231,
+      "learning_rate": 3.3802756178965048e-06,
+      "loss": 0.3023,
+      "num_input_tokens_seen": 324800960,
+      "step": 5895,
+      "train_runtime": 35619.0048,
+      "train_tokens_per_second": 9118.754
+    },
+    {
+      "epoch": 1.2112502566208172,
+      "grad_norm": 0.7376705912402627,
+      "learning_rate": 3.37265048055423e-06,
+      "loss": 0.2854,
+      "num_input_tokens_seen": 325051000,
+      "step": 5900,
+      "train_runtime": 35645.1472,
+      "train_tokens_per_second": 9119.081
+    },
+    {
+      "epoch": 1.2122767398891399,
+      "grad_norm": 0.5817763450942739,
+      "learning_rate": 3.3650295740290396e-06,
+      "loss": 0.281,
+      "num_input_tokens_seen": 325328352,
+      "step": 5905,
+      "train_runtime": 35668.5665,
+      "train_tokens_per_second": 9120.87
+    },
+    {
+      "epoch": 1.2133032231574625,
+      "grad_norm": 0.610715899252456,
+      "learning_rate": 3.357412918133926e-06,
+      "loss": 0.2894,
+      "num_input_tokens_seen": 325598080,
+      "step": 5910,
+      "train_runtime": 35699.2194,
+      "train_tokens_per_second": 9120.594
+    },
+    {
+      "epoch": 1.2143297064257852,
+      "grad_norm": 0.6130023992803658,
+      "learning_rate": 3.3498005326708305e-06,
+      "loss": 0.2945,
+      "num_input_tokens_seen": 325853264,
+      "step": 5915,
+      "train_runtime": 35727.3219,
+      "train_tokens_per_second": 9120.562
+    },
+    {
+      "epoch": 1.215356189694108,
+      "grad_norm": 0.5682061987530126,
+      "learning_rate": 3.342192437430587e-06,
+      "loss": 0.2897,
+      "num_input_tokens_seen": 326157488,
+      "step": 5920,
+      "train_runtime": 35757.5165,
+      "train_tokens_per_second": 9121.369
+    },
+    {
+      "epoch": 1.2163826729624307,
+      "grad_norm": 0.6358837672820936,
+      "learning_rate": 3.334588652192885e-06,
+      "loss": 0.2859,
+      "num_input_tokens_seen": 326448176,
+      "step": 5925,
+      "train_runtime": 35790.5687,
+      "train_tokens_per_second": 9121.067
+    },
+    {
+      "epoch": 1.2174091562307534,
+      "grad_norm": 0.6284725634756756,
+      "learning_rate": 3.3269891967261993e-06,
+      "loss": 0.2952,
+      "num_input_tokens_seen": 326711184,
+      "step": 5930,
+      "train_runtime": 35820.6581,
+      "train_tokens_per_second": 9120.748
+    },
+    {
+      "epoch": 1.218435639499076,
+      "grad_norm": 0.6611570316458776,
+      "learning_rate": 3.3193940907877532e-06,
+      "loss": 0.2889,
+      "num_input_tokens_seen": 326998496,
+      "step": 5935,
+      "train_runtime": 35844.6521,
+      "train_tokens_per_second": 9122.658
+    },
+    {
+      "epoch": 1.219462122767399,
+      "grad_norm": 0.5956082984509181,
+      "learning_rate": 3.3118033541234596e-06,
+      "loss": 0.2813,
+      "num_input_tokens_seen": 327286016,
+      "step": 5940,
+      "train_runtime": 35871.0013,
+      "train_tokens_per_second": 9123.972
+    },
+    {
+      "epoch": 1.2204886060357216,
+      "grad_norm": 0.666697046734497,
+      "learning_rate": 3.3042170064678757e-06,
+      "loss": 0.3128,
+      "num_input_tokens_seen": 327565680,
+      "step": 5945,
+      "train_runtime": 35902.0814,
+      "train_tokens_per_second": 9123.863
+    },
+    {
+      "epoch": 1.2215150893040443,
+      "grad_norm": 0.6218234324169812,
+      "learning_rate": 3.296635067544145e-06,
+      "loss": 0.2936,
+      "num_input_tokens_seen": 327854360,
+      "step": 5950,
+      "train_runtime": 35930.5006,
+      "train_tokens_per_second": 9124.681
+    },
+    {
+      "epoch": 1.222541572572367,
+      "grad_norm": 0.7080519924524092,
+      "learning_rate": 3.2890575570639506e-06,
+      "loss": 0.2745,
+      "num_input_tokens_seen": 328139872,
+      "step": 5955,
+      "train_runtime": 35961.0448,
+      "train_tokens_per_second": 9124.87
+    },
+    {
+      "epoch": 1.2235680558406898,
+      "grad_norm": 0.7052798351434344,
+      "learning_rate": 3.281484494727457e-06,
+      "loss": 0.2612,
+      "num_input_tokens_seen": 328405224,
+      "step": 5960,
+      "train_runtime": 35983.7211,
+      "train_tokens_per_second": 9126.494
+    },
+    {
+      "epoch": 1.2245945391090125,
+      "grad_norm": 0.6414777814740729,
+      "learning_rate": 3.273915900223277e-06,
+      "loss": 0.293,
+      "num_input_tokens_seen": 328687872,
+      "step": 5965,
+      "train_runtime": 36007.4475,
+      "train_tokens_per_second": 9128.33
+    },
+    {
+      "epoch": 1.2256210223773352,
+      "grad_norm": 0.7661704720112442,
+      "learning_rate": 3.266351793228393e-06,
+      "loss": 0.2856,
+      "num_input_tokens_seen": 328942096,
+      "step": 5970,
+      "train_runtime": 36034.1337,
+      "train_tokens_per_second": 9128.625
+    },
+    {
+      "epoch": 1.226647505645658,
+      "grad_norm": 0.624970709918718,
+      "learning_rate": 3.258792193408129e-06,
+      "loss": 0.2997,
+      "num_input_tokens_seen": 329197104,
+      "step": 5975,
+      "train_runtime": 36061.296,
+      "train_tokens_per_second": 9128.821
+    },
+    {
+      "epoch": 1.2276739889139807,
+      "grad_norm": 0.6960393142185567,
+      "learning_rate": 3.2512371204160912e-06,
+      "loss": 0.3042,
+      "num_input_tokens_seen": 329471744,
+      "step": 5980,
+      "train_runtime": 36090.4,
+      "train_tokens_per_second": 9129.069
+    },
+    {
+      "epoch": 1.2287004721823034,
+      "grad_norm": 0.6968385576835884,
+      "learning_rate": 3.243686593894115e-06,
+      "loss": 0.3012,
+      "num_input_tokens_seen": 329753360,
+      "step": 5985,
+      "train_runtime": 36124.442,
+      "train_tokens_per_second": 9128.262
+    },
+    {
+      "epoch": 1.229726955450626,
+      "grad_norm": 0.658043821091868,
+      "learning_rate": 3.2361406334722155e-06,
+      "loss": 0.2836,
+      "num_input_tokens_seen": 330037016,
+      "step": 5990,
+      "train_runtime": 36150.4292,
+      "train_tokens_per_second": 9129.546
+    },
+    {
+      "epoch": 1.230753438718949,
+      "grad_norm": 0.6568002049475922,
+      "learning_rate": 3.228599258768537e-06,
+      "loss": 0.2979,
+      "num_input_tokens_seen": 330316320,
+      "step": 5995,
+      "train_runtime": 36183.6711,
+      "train_tokens_per_second": 9128.878
+    },
+    {
+      "epoch": 1.2317799219872716,
+      "grad_norm": 0.651552794070971,
+      "learning_rate": 3.2210624893893037e-06,
+      "loss": 0.28,
+      "num_input_tokens_seen": 330600640,
+      "step": 6000,
+      "train_runtime": 36212.0386,
+      "train_tokens_per_second": 9129.578
+    },
+    {
+      "epoch": 1.2328064052555943,
+      "grad_norm": 0.6206049790690967,
+      "learning_rate": 3.2135303449287646e-06,
+      "loss": 0.2903,
+      "num_input_tokens_seen": 330881760,
+      "step": 6005,
+      "train_runtime": 36242.7474,
+      "train_tokens_per_second": 9129.599
+    },
+    {
+      "epoch": 1.2338328885239171,
+      "grad_norm": 0.5890319411007776,
+      "learning_rate": 3.2060028449691425e-06,
+      "loss": 0.3084,
+      "num_input_tokens_seen": 331161480,
+      "step": 6010,
+      "train_runtime": 36273.4189,
+      "train_tokens_per_second": 9129.591
+    },
+    {
+      "epoch": 1.2348593717922398,
+      "grad_norm": 0.705117075077152,
+      "learning_rate": 3.1984800090805932e-06,
+      "loss": 0.3046,
+      "num_input_tokens_seen": 331432368,
+      "step": 6015,
+      "train_runtime": 36301.298,
+      "train_tokens_per_second": 9130.042
+    },
+    {
+      "epoch": 1.2358858550605625,
+      "grad_norm": 0.6048996367706566,
+      "learning_rate": 3.190961856821138e-06,
+      "loss": 0.2779,
+      "num_input_tokens_seen": 331720768,
+      "step": 6020,
+      "train_runtime": 36336.0561,
+      "train_tokens_per_second": 9129.245
+    },
+    {
+      "epoch": 1.2369123383288851,
+      "grad_norm": 0.7014095287320355,
+      "learning_rate": 3.1834484077366267e-06,
+      "loss": 0.2967,
+      "num_input_tokens_seen": 332023168,
+      "step": 6025,
+      "train_runtime": 36369.397,
+      "train_tokens_per_second": 9129.191
+    },
+    {
+      "epoch": 1.237938821597208,
+      "grad_norm": 0.691711749975623,
+      "learning_rate": 3.1759396813606787e-06,
+      "loss": 0.2837,
+      "num_input_tokens_seen": 332280944,
+      "step": 6030,
+      "train_runtime": 36401.8312,
+      "train_tokens_per_second": 9128.138
+    },
+    {
+      "epoch": 1.2389653048655307,
+      "grad_norm": 0.657770924107305,
+      "learning_rate": 3.1684356972146392e-06,
+      "loss": 0.2708,
+      "num_input_tokens_seen": 332565968,
+      "step": 6035,
+      "train_runtime": 36436.2218,
+      "train_tokens_per_second": 9127.345
+    },
+    {
+      "epoch": 1.2399917881338534,
+      "grad_norm": 0.7000680731956157,
+      "learning_rate": 3.160936474807521e-06,
+      "loss": 0.2932,
+      "num_input_tokens_seen": 332828296,
+      "step": 6040,
+      "train_runtime": 36460.3682,
+      "train_tokens_per_second": 9128.495
+    },
+    {
+      "epoch": 1.2410182714021762,
+      "grad_norm": 0.6516543250082267,
+      "learning_rate": 3.153442033635957e-06,
+      "loss": 0.2871,
+      "num_input_tokens_seen": 333121880,
+      "step": 6045,
+      "train_runtime": 36486.6982,
+      "train_tokens_per_second": 9129.954
+    },
+    {
+      "epoch": 1.242044754670499,
+      "grad_norm": 0.664671768877068,
+      "learning_rate": 3.1459523931841535e-06,
+      "loss": 0.2758,
+      "num_input_tokens_seen": 333404496,
+      "step": 6050,
+      "train_runtime": 36509.1915,
+      "train_tokens_per_second": 9132.07
+    },
+    {
+      "epoch": 1.2430712379388216,
+      "grad_norm": 0.7599548780175434,
+      "learning_rate": 3.138467572923833e-06,
+      "loss": 0.2781,
+      "num_input_tokens_seen": 333699408,
+      "step": 6055,
+      "train_runtime": 36540.0584,
+      "train_tokens_per_second": 9132.427
+    },
+    {
+      "epoch": 1.2440977212071442,
+      "grad_norm": 0.6450512150934127,
+      "learning_rate": 3.1309875923141853e-06,
+      "loss": 0.3068,
+      "num_input_tokens_seen": 333988872,
+      "step": 6060,
+      "train_runtime": 36564.126,
+      "train_tokens_per_second": 9134.332
+    },
+    {
+      "epoch": 1.2451242044754671,
+      "grad_norm": 0.6172872081230096,
+      "learning_rate": 3.1235124708018183e-06,
+      "loss": 0.2756,
+      "num_input_tokens_seen": 334267136,
+      "step": 6065,
+      "train_runtime": 36598.8996,
+      "train_tokens_per_second": 9133.256
+    },
+    {
+      "epoch": 1.2461506877437898,
+      "grad_norm": 0.6359738319402124,
+      "learning_rate": 3.116042227820711e-06,
+      "loss": 0.2908,
+      "num_input_tokens_seen": 334568032,
+      "step": 6070,
+      "train_runtime": 36631.6889,
+      "train_tokens_per_second": 9133.295
+    },
+    {
+      "epoch": 1.2471771710121125,
+      "grad_norm": 0.6402124543372725,
+      "learning_rate": 3.108576882792155e-06,
+      "loss": 0.2989,
+      "num_input_tokens_seen": 334855688,
+      "step": 6075,
+      "train_runtime": 36660.1152,
+      "train_tokens_per_second": 9134.06
+    },
+    {
+      "epoch": 1.2482036542804353,
+      "grad_norm": 0.8757584474612019,
+      "learning_rate": 3.101116455124709e-06,
+      "loss": 0.3122,
+      "num_input_tokens_seen": 335127888,
+      "step": 6080,
+      "train_runtime": 36695.0869,
+      "train_tokens_per_second": 9132.773
+    },
+    {
+      "epoch": 1.249230137548758,
+      "grad_norm": 0.5839295466242271,
+      "learning_rate": 3.093660964214148e-06,
+      "loss": 0.2946,
+      "num_input_tokens_seen": 335407752,
+      "step": 6085,
+      "train_runtime": 36727.7752,
+      "train_tokens_per_second": 9132.264
+    },
+    {
+      "epoch": 1.2502566208170807,
+      "grad_norm": 0.6613520013374334,
+      "learning_rate": 3.086210429443413e-06,
+      "loss": 0.2874,
+      "num_input_tokens_seen": 335660360,
+      "step": 6090,
+      "train_runtime": 36753.7388,
+      "train_tokens_per_second": 9132.686
+    },
+    {
+      "epoch": 1.2512831040854033,
+      "grad_norm": 0.6195188486662263,
+      "learning_rate": 3.07876487018256e-06,
+      "loss": 0.3027,
+      "num_input_tokens_seen": 335950888,
+      "step": 6095,
+      "train_runtime": 36782.5841,
+      "train_tokens_per_second": 9133.423
+    },
+    {
+      "epoch": 1.2523095873537262,
+      "grad_norm": 0.6580663339632391,
+      "learning_rate": 3.0713243057887034e-06,
+      "loss": 0.2847,
+      "num_input_tokens_seen": 336208768,
+      "step": 6100,
+      "train_runtime": 36808.5937,
+      "train_tokens_per_second": 9133.975
+    },
+    {
+      "epoch": 1.2533360706220489,
+      "grad_norm": 0.6754315363102434,
+      "learning_rate": 3.063888755605984e-06,
+      "loss": 0.2968,
+      "num_input_tokens_seen": 336480824,
+      "step": 6105,
+      "train_runtime": 36842.4268,
+      "train_tokens_per_second": 9132.971
+    },
+    {
+      "epoch": 1.2543625538903715,
+      "grad_norm": 0.6461740138972248,
+      "learning_rate": 3.056458238965495e-06,
+      "loss": 0.2849,
+      "num_input_tokens_seen": 336780656,
+      "step": 6110,
+      "train_runtime": 36872.9872,
+      "train_tokens_per_second": 9133.533
+    },
+    {
+      "epoch": 1.2553890371586944,
+      "grad_norm": 0.6369194721870857,
+      "learning_rate": 3.04903277518525e-06,
+      "loss": 0.2966,
+      "num_input_tokens_seen": 337065672,
+      "step": 6115,
+      "train_runtime": 36899.117,
+      "train_tokens_per_second": 9134.79
+    },
+    {
+      "epoch": 1.256415520427017,
+      "grad_norm": 0.6143888292213631,
+      "learning_rate": 3.041612383570122e-06,
+      "loss": 0.2966,
+      "num_input_tokens_seen": 337348736,
+      "step": 6120,
+      "train_runtime": 36933.8161,
+      "train_tokens_per_second": 9133.872
+    },
+    {
+      "epoch": 1.2574420036953398,
+      "grad_norm": 0.6317553060565108,
+      "learning_rate": 3.034197083411802e-06,
+      "loss": 0.3152,
+      "num_input_tokens_seen": 337643368,
+      "step": 6125,
+      "train_runtime": 36971.6274,
+      "train_tokens_per_second": 9132.499
+    },
+    {
+      "epoch": 1.2584684869636624,
+      "grad_norm": 0.7136112336443893,
+      "learning_rate": 3.0267868939887396e-06,
+      "loss": 0.3038,
+      "num_input_tokens_seen": 337939752,
+      "step": 6130,
+      "train_runtime": 37001.5173,
+      "train_tokens_per_second": 9133.132
+    },
+    {
+      "epoch": 1.259494970231985,
+      "grad_norm": 0.7097596134161139,
+      "learning_rate": 3.0193818345660986e-06,
+      "loss": 0.2695,
+      "num_input_tokens_seen": 338227504,
+      "step": 6135,
+      "train_runtime": 37030.5568,
+      "train_tokens_per_second": 9133.741
+    },
+    {
+      "epoch": 1.260521453500308,
+      "grad_norm": 0.6625193943633193,
+      "learning_rate": 3.011981924395708e-06,
+      "loss": 0.3098,
+      "num_input_tokens_seen": 338510016,
+      "step": 6140,
+      "train_runtime": 37054.6044,
+      "train_tokens_per_second": 9135.437
+    },
+    {
+      "epoch": 1.2615479367686306,
+      "grad_norm": 0.6880201170725786,
+      "learning_rate": 3.0045871827160083e-06,
+      "loss": 0.3105,
+      "num_input_tokens_seen": 338767560,
+      "step": 6145,
+      "train_runtime": 37087.3169,
+      "train_tokens_per_second": 9134.324
+    },
+    {
+      "epoch": 1.2625744200369535,
+      "grad_norm": 0.6576527029765754,
+      "learning_rate": 2.997197628752001e-06,
+      "loss": 0.2803,
+      "num_input_tokens_seen": 339030840,
+      "step": 6150,
+      "train_runtime": 37122.007,
+      "train_tokens_per_second": 9132.88
+    },
+    {
+      "epoch": 1.2636009033052762,
+      "grad_norm": 0.7283223469933805,
+      "learning_rate": 2.9898132817152003e-06,
+      "loss": 0.306,
+      "num_input_tokens_seen": 339311264,
+      "step": 6155,
+      "train_runtime": 37152.0942,
+      "train_tokens_per_second": 9133.032
+    },
+    {
+      "epoch": 1.2646273865735989,
+      "grad_norm": 0.6582169784644492,
+      "learning_rate": 2.982434160803589e-06,
+      "loss": 0.2966,
+      "num_input_tokens_seen": 339573416,
+      "step": 6160,
+      "train_runtime": 37186.2718,
+      "train_tokens_per_second": 9131.688
+    },
+    {
+      "epoch": 1.2656538698419215,
+      "grad_norm": 0.616101914689766,
+      "learning_rate": 2.9750602852015574e-06,
+      "loss": 0.2983,
+      "num_input_tokens_seen": 339859328,
+      "step": 6165,
+      "train_runtime": 37214.2873,
+      "train_tokens_per_second": 9132.496
+    },
+    {
+      "epoch": 1.2666803531102442,
+      "grad_norm": 0.6618711729021667,
+      "learning_rate": 2.96769167407986e-06,
+      "loss": 0.3087,
+      "num_input_tokens_seen": 340123712,
+      "step": 6170,
+      "train_runtime": 37238.4903,
+      "train_tokens_per_second": 9133.66
+    },
+    {
+      "epoch": 1.267706836378567,
+      "grad_norm": 0.6837020297961852,
+      "learning_rate": 2.9603283465955637e-06,
+      "loss": 0.3077,
+      "num_input_tokens_seen": 340390544,
+      "step": 6175,
+      "train_runtime": 37264.9969,
+      "train_tokens_per_second": 9134.324
+    },
+    {
+      "epoch": 1.2687333196468897,
+      "grad_norm": 0.7365151254409835,
+      "learning_rate": 2.952970321892003e-06,
+      "loss": 0.3133,
+      "num_input_tokens_seen": 340650016,
+      "step": 6180,
+      "train_runtime": 37289.5306,
+      "train_tokens_per_second": 9135.272
+    },
+    {
+      "epoch": 1.2697598029152124,
+      "grad_norm": 0.6243815137140124,
+      "learning_rate": 2.9456176190987233e-06,
+      "loss": 0.2916,
+      "num_input_tokens_seen": 340933024,
+      "step": 6185,
+      "train_runtime": 37321.0333,
+      "train_tokens_per_second": 9135.144
+    },
+    {
+      "epoch": 1.2707862861835353,
+      "grad_norm": 0.6831681303846019,
+      "learning_rate": 2.93827025733143e-06,
+      "loss": 0.2753,
+      "num_input_tokens_seen": 341194424,
+      "step": 6190,
+      "train_runtime": 37351.0597,
+      "train_tokens_per_second": 9134.799
+    },
+    {
+      "epoch": 1.271812769451858,
+      "grad_norm": 1.1023008745862108,
+      "learning_rate": 2.9309282556919525e-06,
+      "loss": 0.2877,
+      "num_input_tokens_seen": 341471584,
+      "step": 6195,
+      "train_runtime": 37387.5945,
+      "train_tokens_per_second": 9133.286
+    },
+    {
+      "epoch": 1.2728392527201806,
+      "grad_norm": 0.7141394362302543,
+      "learning_rate": 2.9235916332681756e-06,
+      "loss": 0.2953,
+      "num_input_tokens_seen": 341738696,
+      "step": 6200,
+      "train_runtime": 37413.0966,
+      "train_tokens_per_second": 9134.2
+    },
+    {
+      "epoch": 1.2738657359885033,
+      "grad_norm": 0.737084039538406,
+      "learning_rate": 2.9162604091340046e-06,
+      "loss": 0.2931,
+      "num_input_tokens_seen": 342011592,
+      "step": 6205,
+      "train_runtime": 37446.6264,
+      "train_tokens_per_second": 9133.309
+    },
+    {
+      "epoch": 1.2748922192568262,
+      "grad_norm": 0.75091654031835,
+      "learning_rate": 2.9089346023493043e-06,
+      "loss": 0.2962,
+      "num_input_tokens_seen": 342291400,
+      "step": 6210,
+      "train_runtime": 37474.6919,
+      "train_tokens_per_second": 9133.935
+    },
+    {
+      "epoch": 1.2759187025251488,
+      "grad_norm": 0.6714089036773458,
+      "learning_rate": 2.901614231959864e-06,
+      "loss": 0.286,
+      "num_input_tokens_seen": 342583784,
+      "step": 6215,
+      "train_runtime": 37505.8513,
+      "train_tokens_per_second": 9134.142
+    },
+    {
+      "epoch": 1.2769451857934715,
+      "grad_norm": 0.6707452104936544,
+      "learning_rate": 2.894299316997333e-06,
+      "loss": 0.2947,
+      "num_input_tokens_seen": 342848152,
+      "step": 6220,
+      "train_runtime": 37533.7677,
+      "train_tokens_per_second": 9134.392
+    },
+    {
+      "epoch": 1.2779716690617944,
+      "grad_norm": 0.6486163358657222,
+      "learning_rate": 2.8869898764791794e-06,
+      "loss": 0.2876,
+      "num_input_tokens_seen": 343119352,
+      "step": 6225,
+      "train_runtime": 37561.9457,
+      "train_tokens_per_second": 9134.76
+    },
+    {
+      "epoch": 1.278998152330117,
+      "grad_norm": 0.6450565396702729,
+      "learning_rate": 2.879685929408638e-06,
+      "loss": 0.3032,
+      "num_input_tokens_seen": 343412216,
+      "step": 6230,
+      "train_runtime": 37592.3508,
+      "train_tokens_per_second": 9135.162
+    },
+    {
+      "epoch": 1.2800246355984397,
+      "grad_norm": 0.6668900850259314,
+      "learning_rate": 2.872387494774663e-06,
+      "loss": 0.2996,
+      "num_input_tokens_seen": 343695352,
+      "step": 6235,
+      "train_runtime": 37619.3154,
+      "train_tokens_per_second": 9136.141
+    },
+    {
+      "epoch": 1.2810511188667624,
+      "grad_norm": 0.664860203867104,
+      "learning_rate": 2.8650945915518744e-06,
+      "loss": 0.2962,
+      "num_input_tokens_seen": 343973736,
+      "step": 6240,
+      "train_runtime": 37654.2903,
+      "train_tokens_per_second": 9135.048
+    },
+    {
+      "epoch": 1.2820776021350853,
+      "grad_norm": 0.6763147119368383,
+      "learning_rate": 2.8578072387005142e-06,
+      "loss": 0.2824,
+      "num_input_tokens_seen": 344235624,
+      "step": 6245,
+      "train_runtime": 37688.061,
+      "train_tokens_per_second": 9133.811
+    },
+    {
+      "epoch": 1.283104085403408,
+      "grad_norm": 0.9606494501203379,
+      "learning_rate": 2.8505254551663956e-06,
+      "loss": 0.2927,
+      "num_input_tokens_seen": 344503320,
+      "step": 6250,
+      "train_runtime": 37723.8689,
+      "train_tokens_per_second": 9132.237
+    },
+    {
+      "epoch": 1.2841305686717306,
+      "grad_norm": 0.8884583731954401,
+      "learning_rate": 2.8432492598808514e-06,
+      "loss": 0.3246,
+      "num_input_tokens_seen": 344779040,
+      "step": 6255,
+      "train_runtime": 37754.4837,
+      "train_tokens_per_second": 9132.135
+    },
+    {
+      "epoch": 1.2851570519400535,
+      "grad_norm": 0.6884173583538721,
+      "learning_rate": 2.8359786717606795e-06,
+      "loss": 0.3196,
+      "num_input_tokens_seen": 345046400,
+      "step": 6260,
+      "train_runtime": 37789.5307,
+      "train_tokens_per_second": 9130.741
+    },
+    {
+      "epoch": 1.2861835352083761,
+      "grad_norm": 0.6080520770877227,
+      "learning_rate": 2.8287137097081117e-06,
+      "loss": 0.2857,
+      "num_input_tokens_seen": 345301224,
+      "step": 6265,
+      "train_runtime": 37813.5069,
+      "train_tokens_per_second": 9131.69
+    },
+    {
+      "epoch": 1.2872100184766988,
+      "grad_norm": 0.6361333340050421,
+      "learning_rate": 2.821454392610746e-06,
+      "loss": 0.2866,
+      "num_input_tokens_seen": 345587488,
+      "step": 6270,
+      "train_runtime": 37840.4127,
+      "train_tokens_per_second": 9132.762
+    },
+    {
+      "epoch": 1.2882365017450215,
+      "grad_norm": 0.6598850366012137,
+      "learning_rate": 2.8142007393415045e-06,
+      "loss": 0.2919,
+      "num_input_tokens_seen": 345862712,
+      "step": 6275,
+      "train_runtime": 37869.9008,
+      "train_tokens_per_second": 9132.918
+    },
+    {
+      "epoch": 1.2892629850133444,
+      "grad_norm": 0.5800161948197182,
+      "learning_rate": 2.8069527687585864e-06,
+      "loss": 0.3051,
+      "num_input_tokens_seen": 346164904,
+      "step": 6280,
+      "train_runtime": 37901.9805,
+      "train_tokens_per_second": 9133.161
+    },
+    {
+      "epoch": 1.290289468281667,
+      "grad_norm": 0.7311138471169327,
+      "learning_rate": 2.799710499705415e-06,
+      "loss": 0.3068,
+      "num_input_tokens_seen": 346438864,
+      "step": 6285,
+      "train_runtime": 37929.5272,
+      "train_tokens_per_second": 9133.751
+    },
+    {
+      "epoch": 1.2913159515499897,
+      "grad_norm": 0.8719437142093381,
+      "learning_rate": 2.7924739510105943e-06,
+      "loss": 0.306,
+      "num_input_tokens_seen": 346694912,
+      "step": 6290,
+      "train_runtime": 37954.8038,
+      "train_tokens_per_second": 9134.415
+    },
+    {
+      "epoch": 1.2923424348183126,
+      "grad_norm": 0.5782945178831149,
+      "learning_rate": 2.7852431414878513e-06,
+      "loss": 0.2812,
+      "num_input_tokens_seen": 346965520,
+      "step": 6295,
+      "train_runtime": 37990.5208,
+      "train_tokens_per_second": 9132.95
+    },
+    {
+      "epoch": 1.2933689180866352,
+      "grad_norm": 0.7873161872973119,
+      "learning_rate": 2.7780180899359945e-06,
+      "loss": 0.3007,
+      "num_input_tokens_seen": 347248840,
+      "step": 6300,
+      "train_runtime": 38020.3218,
+      "train_tokens_per_second": 9133.243
+    },
+    {
+      "epoch": 1.294395401354958,
+      "grad_norm": 0.676273489479129,
+      "learning_rate": 2.7707988151388688e-06,
+      "loss": 0.2873,
+      "num_input_tokens_seen": 347490400,
+      "step": 6305,
+      "train_runtime": 38048.7449,
+      "train_tokens_per_second": 9132.769
+    },
+    {
+      "epoch": 1.2954218846232806,
+      "grad_norm": 0.7852160594577124,
+      "learning_rate": 2.7635853358652877e-06,
+      "loss": 0.2886,
+      "num_input_tokens_seen": 347757448,
+      "step": 6310,
+      "train_runtime": 38080.6495,
+      "train_tokens_per_second": 9132.13
+    },
+    {
+      "epoch": 1.2964483678916034,
+      "grad_norm": 0.7464026268231113,
+      "learning_rate": 2.756377670869006e-06,
+      "loss": 0.3177,
+      "num_input_tokens_seen": 348033600,
+      "step": 6315,
+      "train_runtime": 38111.6624,
+      "train_tokens_per_second": 9131.945
+    },
+    {
+      "epoch": 1.2974748511599261,
+      "grad_norm": 0.7312397621984632,
+      "learning_rate": 2.7491758388886634e-06,
+      "loss": 0.3022,
+      "num_input_tokens_seen": 348302032,
+      "step": 6320,
+      "train_runtime": 38141.4865,
+      "train_tokens_per_second": 9131.842
+    },
+    {
+      "epoch": 1.2985013344282488,
+      "grad_norm": 0.8648631388766304,
+      "learning_rate": 2.741979858647731e-06,
+      "loss": 0.2918,
+      "num_input_tokens_seen": 348581080,
+      "step": 6325,
+      "train_runtime": 38172.6172,
+      "train_tokens_per_second": 9131.705
+    },
+    {
+      "epoch": 1.2995278176965717,
+      "grad_norm": 0.608649125693906,
+      "learning_rate": 2.734789748854468e-06,
+      "loss": 0.2632,
+      "num_input_tokens_seen": 348837920,
+      "step": 6330,
+      "train_runtime": 38206.9155,
+      "train_tokens_per_second": 9130.23
+    },
+    {
+      "epoch": 1.3005543009648943,
+      "grad_norm": 0.6763220547141338,
+      "learning_rate": 2.727605528201871e-06,
+      "loss": 0.3009,
+      "num_input_tokens_seen": 349089640,
+      "step": 6335,
+      "train_runtime": 38235.4149,
+      "train_tokens_per_second": 9130.008
+    },
+    {
+      "epoch": 1.301580784233217,
+      "grad_norm": 0.6372660553084611,
+      "learning_rate": 2.7204272153676254e-06,
+      "loss": 0.2856,
+      "num_input_tokens_seen": 349379552,
+      "step": 6340,
+      "train_runtime": 38269.1965,
+      "train_tokens_per_second": 9129.524
+    },
+    {
+      "epoch": 1.3026072675015397,
+      "grad_norm": 0.7999433213521334,
+      "learning_rate": 2.7132548290140593e-06,
+      "loss": 0.3132,
+      "num_input_tokens_seen": 349686744,
+      "step": 6345,
+      "train_runtime": 38300.0675,
+      "train_tokens_per_second": 9130.186
+    },
+    {
+      "epoch": 1.3036337507698623,
+      "grad_norm": 0.7101317696425322,
+      "learning_rate": 2.7060883877880885e-06,
+      "loss": 0.3185,
+      "num_input_tokens_seen": 349980952,
+      "step": 6350,
+      "train_runtime": 38329.7827,
+      "train_tokens_per_second": 9130.784
+    },
+    {
+      "epoch": 1.3046602340381852,
+      "grad_norm": 0.6304316885099528,
+      "learning_rate": 2.6989279103211825e-06,
+      "loss": 0.2727,
+      "num_input_tokens_seen": 350282240,
+      "step": 6355,
+      "train_runtime": 38363.7423,
+      "train_tokens_per_second": 9130.555
+    },
+    {
+      "epoch": 1.3056867173065079,
+      "grad_norm": 0.770524519105958,
+      "learning_rate": 2.691773415229293e-06,
+      "loss": 0.2861,
+      "num_input_tokens_seen": 350552232,
+      "step": 6360,
+      "train_runtime": 38394.6946,
+      "train_tokens_per_second": 9130.226
+    },
+    {
+      "epoch": 1.3067132005748308,
+      "grad_norm": 0.703319270587308,
+      "learning_rate": 2.684624921112827e-06,
+      "loss": 0.3014,
+      "num_input_tokens_seen": 350807024,
+      "step": 6365,
+      "train_runtime": 38420.2197,
+      "train_tokens_per_second": 9130.792
+    },
+    {
+      "epoch": 1.3077396838431534,
+      "grad_norm": 0.706780889155962,
+      "learning_rate": 2.6774824465565865e-06,
+      "loss": 0.2946,
+      "num_input_tokens_seen": 351095744,
+      "step": 6370,
+      "train_runtime": 38444.41,
+      "train_tokens_per_second": 9132.556
+    },
+    {
+      "epoch": 1.308766167111476,
+      "grad_norm": 0.6268966042971468,
+      "learning_rate": 2.6703460101297285e-06,
+      "loss": 0.2811,
+      "num_input_tokens_seen": 351373600,
+      "step": 6375,
+      "train_runtime": 38480.0873,
+      "train_tokens_per_second": 9131.31
+    },
+    {
+      "epoch": 1.3097926503797988,
+      "grad_norm": 0.655217119186077,
+      "learning_rate": 2.663215630385707e-06,
+      "loss": 0.2799,
+      "num_input_tokens_seen": 351654216,
+      "step": 6380,
+      "train_runtime": 38513.2064,
+      "train_tokens_per_second": 9130.744
+    },
+    {
+      "epoch": 1.3108191336481214,
+      "grad_norm": 0.7217018065692212,
+      "learning_rate": 2.656091325862234e-06,
+      "loss": 0.3118,
+      "num_input_tokens_seen": 351939064,
+      "step": 6385,
+      "train_runtime": 38540.7342,
+      "train_tokens_per_second": 9131.613
+    },
+    {
+      "epoch": 1.3118456169164443,
+      "grad_norm": 0.6737226430934566,
+      "learning_rate": 2.648973115081217e-06,
+      "loss": 0.2845,
+      "num_input_tokens_seen": 352205336,
+      "step": 6390,
+      "train_runtime": 38569.6221,
+      "train_tokens_per_second": 9131.677
+    },
+    {
+      "epoch": 1.312872100184767,
+      "grad_norm": 0.8558856033411144,
+      "learning_rate": 2.6418610165487356e-06,
+      "loss": 0.2928,
+      "num_input_tokens_seen": 352459776,
+      "step": 6395,
+      "train_runtime": 38602.6314,
+      "train_tokens_per_second": 9130.46
+    },
+    {
+      "epoch": 1.3138985834530899,
+      "grad_norm": 0.747178207147192,
+      "learning_rate": 2.634755048754969e-06,
+      "loss": 0.3106,
+      "num_input_tokens_seen": 352732760,
+      "step": 6400,
+      "train_runtime": 38636.8283,
+      "train_tokens_per_second": 9129.444
+    },
+    {
+      "epoch": 1.3149250667214125,
+      "grad_norm": 0.6606908030798695,
+      "learning_rate": 2.6276552301741566e-06,
+      "loss": 0.2918,
+      "num_input_tokens_seen": 352991448,
+      "step": 6405,
+      "train_runtime": 38671.5453,
+      "train_tokens_per_second": 9127.937
+    },
+    {
+      "epoch": 1.3159515499897352,
+      "grad_norm": 0.6676350646059017,
+      "learning_rate": 2.6205615792645623e-06,
+      "loss": 0.2902,
+      "num_input_tokens_seen": 353277544,
+      "step": 6410,
+      "train_runtime": 38706.2537,
+      "train_tokens_per_second": 9127.144
+    },
+    {
+      "epoch": 1.3169780332580578,
+      "grad_norm": 0.6630177844874262,
+      "learning_rate": 2.6134741144683994e-06,
+      "loss": 0.3122,
+      "num_input_tokens_seen": 353584032,
+      "step": 6415,
+      "train_runtime": 38734.7998,
+      "train_tokens_per_second": 9128.33
+    },
+    {
+      "epoch": 1.3180045165263805,
+      "grad_norm": 0.6381691064395613,
+      "learning_rate": 2.606392854211809e-06,
+      "loss": 0.2644,
+      "num_input_tokens_seen": 353832504,
+      "step": 6420,
+      "train_runtime": 38761.9446,
+      "train_tokens_per_second": 9128.348
+    },
+    {
+      "epoch": 1.3190309997947034,
+      "grad_norm": 0.6805144390322483,
+      "learning_rate": 2.599317816904796e-06,
+      "loss": 0.2683,
+      "num_input_tokens_seen": 354102656,
+      "step": 6425,
+      "train_runtime": 38791.6936,
+      "train_tokens_per_second": 9128.311
+    },
+    {
+      "epoch": 1.320057483063026,
+      "grad_norm": 0.7010570949969497,
+      "learning_rate": 2.5922490209411937e-06,
+      "loss": 0.3024,
+      "num_input_tokens_seen": 354387400,
+      "step": 6430,
+      "train_runtime": 38828.4479,
+      "train_tokens_per_second": 9127.004
+    },
+    {
+      "epoch": 1.3210839663313487,
+      "grad_norm": 0.8156144706524634,
+      "learning_rate": 2.585186484698605e-06,
+      "loss": 0.2842,
+      "num_input_tokens_seen": 354656952,
+      "step": 6435,
+      "train_runtime": 38855.2317,
+      "train_tokens_per_second": 9127.65
+    },
+    {
+      "epoch": 1.3221104495996716,
+      "grad_norm": 0.744884320928106,
+      "learning_rate": 2.578130226538352e-06,
+      "loss": 0.2766,
+      "num_input_tokens_seen": 354915176,
+      "step": 6440,
+      "train_runtime": 38887.6144,
+      "train_tokens_per_second": 9126.689
+    },
+    {
+      "epoch": 1.3231369328679943,
+      "grad_norm": 0.6035498586165806,
+      "learning_rate": 2.571080264805447e-06,
+      "loss": 0.2938,
+      "num_input_tokens_seen": 355193688,
+      "step": 6445,
+      "train_runtime": 38916.379,
+      "train_tokens_per_second": 9127.1
+    },
+    {
+      "epoch": 1.324163416136317,
+      "grad_norm": 0.7471590732177349,
+      "learning_rate": 2.564036617828526e-06,
+      "loss": 0.2835,
+      "num_input_tokens_seen": 355479496,
+      "step": 6450,
+      "train_runtime": 38946.9712,
+      "train_tokens_per_second": 9127.269
+    },
+    {
+      "epoch": 1.3251898994046396,
+      "grad_norm": 0.6994427934741206,
+      "learning_rate": 2.556999303919807e-06,
+      "loss": 0.2789,
+      "num_input_tokens_seen": 355771056,
+      "step": 6455,
+      "train_runtime": 38975.3478,
+      "train_tokens_per_second": 9128.105
+    },
+    {
+      "epoch": 1.3262163826729625,
+      "grad_norm": 0.7514609211802262,
+      "learning_rate": 2.5499683413750463e-06,
+      "loss": 0.3136,
+      "num_input_tokens_seen": 356029136,
+      "step": 6460,
+      "train_runtime": 39006.2752,
+      "train_tokens_per_second": 9127.484
+    },
+    {
+      "epoch": 1.3272428659412852,
+      "grad_norm": 0.7029432204970839,
+      "learning_rate": 2.542943748473485e-06,
+      "loss": 0.2785,
+      "num_input_tokens_seen": 356317992,
+      "step": 6465,
+      "train_runtime": 39041.4175,
+      "train_tokens_per_second": 9126.666
+    },
+    {
+      "epoch": 1.3282693492096078,
+      "grad_norm": 0.6969765011701352,
+      "learning_rate": 2.5359255434778085e-06,
+      "loss": 0.2917,
+      "num_input_tokens_seen": 356603128,
+      "step": 6470,
+      "train_runtime": 39078.5236,
+      "train_tokens_per_second": 9125.297
+    },
+    {
+      "epoch": 1.3292958324779307,
+      "grad_norm": 0.6104523334992173,
+      "learning_rate": 2.5289137446340867e-06,
+      "loss": 0.3238,
+      "num_input_tokens_seen": 356862912,
+      "step": 6475,
+      "train_runtime": 39107.2772,
+      "train_tokens_per_second": 9125.23
+    },
+    {
+      "epoch": 1.3303223157462534,
+      "grad_norm": 0.6510591955933783,
+      "learning_rate": 2.521908370171745e-06,
+      "loss": 0.2926,
+      "num_input_tokens_seen": 357132536,
+      "step": 6480,
+      "train_runtime": 39130.8897,
+      "train_tokens_per_second": 9126.614
+    },
+    {
+      "epoch": 1.331348799014576,
+      "grad_norm": 0.6248236428625131,
+      "learning_rate": 2.514909438303502e-06,
+      "loss": 0.2866,
+      "num_input_tokens_seen": 357384064,
+      "step": 6485,
+      "train_runtime": 39165.4921,
+      "train_tokens_per_second": 9124.973
+    },
+    {
+      "epoch": 1.3323752822828987,
+      "grad_norm": 0.8946914267270204,
+      "learning_rate": 2.507916967225322e-06,
+      "loss": 0.311,
+      "num_input_tokens_seen": 357658008,
+      "step": 6490,
+      "train_runtime": 39193.3627,
+      "train_tokens_per_second": 9125.474
+    },
+    {
+      "epoch": 1.3334017655512216,
+      "grad_norm": 0.6490691943786241,
+      "learning_rate": 2.5009309751163767e-06,
+      "loss": 0.2828,
+      "num_input_tokens_seen": 357917416,
+      "step": 6495,
+      "train_runtime": 39223.7323,
+      "train_tokens_per_second": 9125.022
+    },
+    {
+      "epoch": 1.3344282488195442,
+      "grad_norm": 0.7016711424982762,
+      "learning_rate": 2.4939514801389968e-06,
+      "loss": 0.3012,
+      "num_input_tokens_seen": 358192760,
+      "step": 6500,
+      "train_runtime": 39255.9085,
+      "train_tokens_per_second": 9124.557
+    },
+    {
+      "epoch": 1.335454732087867,
+      "grad_norm": 0.6952752451302698,
+      "learning_rate": 2.4869785004386165e-06,
+      "loss": 0.2527,
+      "num_input_tokens_seen": 358456496,
+      "step": 6505,
+      "train_runtime": 39279.0764,
+      "train_tokens_per_second": 9125.889
+    },
+    {
+      "epoch": 1.3364812153561898,
+      "grad_norm": 0.6009608859341691,
+      "learning_rate": 2.4800120541437333e-06,
+      "loss": 0.2852,
+      "num_input_tokens_seen": 358720880,
+      "step": 6510,
+      "train_runtime": 39305.7276,
+      "train_tokens_per_second": 9126.428
+    },
+    {
+      "epoch": 1.3375076986245125,
+      "grad_norm": 0.7030371626088499,
+      "learning_rate": 2.4730521593658597e-06,
+      "loss": 0.3008,
+      "num_input_tokens_seen": 358969672,
+      "step": 6515,
+      "train_runtime": 39331.7494,
+      "train_tokens_per_second": 9126.715
+    },
+    {
+      "epoch": 1.3385341818928351,
+      "grad_norm": 0.6098851908676337,
+      "learning_rate": 2.466098834199474e-06,
+      "loss": 0.292,
+      "num_input_tokens_seen": 359252936,
+      "step": 6520,
+      "train_runtime": 39367.5439,
+      "train_tokens_per_second": 9125.612
+    },
+    {
+      "epoch": 1.3395606651611578,
+      "grad_norm": 0.6682171031173756,
+      "learning_rate": 2.4591520967219757e-06,
+      "loss": 0.2678,
+      "num_input_tokens_seen": 359524344,
+      "step": 6525,
+      "train_runtime": 39402.0628,
+      "train_tokens_per_second": 9124.506
+    },
+    {
+      "epoch": 1.3405871484294807,
+      "grad_norm": 0.7279131845903446,
+      "learning_rate": 2.452211964993635e-06,
+      "loss": 0.2801,
+      "num_input_tokens_seen": 359817480,
+      "step": 6530,
+      "train_runtime": 39430.0925,
+      "train_tokens_per_second": 9125.454
+    },
+    {
+      "epoch": 1.3416136316978033,
+      "grad_norm": 0.7794427837725226,
+      "learning_rate": 2.4452784570575544e-06,
+      "loss": 0.304,
+      "num_input_tokens_seen": 360077096,
+      "step": 6535,
+      "train_runtime": 39454.1447,
+      "train_tokens_per_second": 9126.471
+    },
+    {
+      "epoch": 1.342640114966126,
+      "grad_norm": 0.7069231868882841,
+      "learning_rate": 2.438351590939612e-06,
+      "loss": 0.2917,
+      "num_input_tokens_seen": 360354240,
+      "step": 6540,
+      "train_runtime": 39481.6908,
+      "train_tokens_per_second": 9127.123
+    },
+    {
+      "epoch": 1.343666598234449,
+      "grad_norm": 0.6591397022408749,
+      "learning_rate": 2.4314313846484148e-06,
+      "loss": 0.3026,
+      "num_input_tokens_seen": 360649280,
+      "step": 6545,
+      "train_runtime": 39520.0881,
+      "train_tokens_per_second": 9125.721
+    },
+    {
+      "epoch": 1.3446930815027716,
+      "grad_norm": 0.7062766228514463,
+      "learning_rate": 2.424517856175258e-06,
+      "loss": 0.2816,
+      "num_input_tokens_seen": 360922792,
+      "step": 6550,
+      "train_runtime": 39552.0635,
+      "train_tokens_per_second": 9125.258
+    },
+    {
+      "epoch": 1.3457195647710942,
+      "grad_norm": 0.6954545242205027,
+      "learning_rate": 2.4176110234940813e-06,
+      "loss": 0.2991,
+      "num_input_tokens_seen": 361193320,
+      "step": 6555,
+      "train_runtime": 39580.5836,
+      "train_tokens_per_second": 9125.518
+    },
+    {
+      "epoch": 1.3467460480394169,
+      "grad_norm": 0.6677208918096188,
+      "learning_rate": 2.410710904561409e-06,
+      "loss": 0.2941,
+      "num_input_tokens_seen": 361446400,
+      "step": 6560,
+      "train_runtime": 39610.5616,
+      "train_tokens_per_second": 9125.001
+    },
+    {
+      "epoch": 1.3477725313077396,
+      "grad_norm": 0.753053437235032,
+      "learning_rate": 2.4038175173163144e-06,
+      "loss": 0.2891,
+      "num_input_tokens_seen": 361727520,
+      "step": 6565,
+      "train_runtime": 39640.8326,
+      "train_tokens_per_second": 9125.124
+    },
+    {
+      "epoch": 1.3487990145760624,
+      "grad_norm": 0.7114612892364812,
+      "learning_rate": 2.396930879680368e-06,
+      "loss": 0.3088,
+      "num_input_tokens_seen": 361998392,
+      "step": 6570,
+      "train_runtime": 39663.7461,
+      "train_tokens_per_second": 9126.682
+    },
+    {
+      "epoch": 1.349825497844385,
+      "grad_norm": 0.7407180544592107,
+      "learning_rate": 2.390051009557594e-06,
+      "loss": 0.2889,
+      "num_input_tokens_seen": 362277432,
+      "step": 6575,
+      "train_runtime": 39691.0475,
+      "train_tokens_per_second": 9127.434
+    },
+    {
+      "epoch": 1.350851981112708,
+      "grad_norm": 0.6258191148613007,
+      "learning_rate": 2.3831779248344218e-06,
+      "loss": 0.2966,
+      "num_input_tokens_seen": 362550600,
+      "step": 6580,
+      "train_runtime": 39723.9009,
+      "train_tokens_per_second": 9126.762
+    },
+    {
+      "epoch": 1.3518784643810307,
+      "grad_norm": 0.6308899635712916,
+      "learning_rate": 2.376311643379637e-06,
+      "loss": 0.3046,
+      "num_input_tokens_seen": 362825128,
+      "step": 6585,
+      "train_runtime": 39758.5158,
+      "train_tokens_per_second": 9125.721
+    },
+    {
+      "epoch": 1.3529049476493533,
+      "grad_norm": 0.6208485017952117,
+      "learning_rate": 2.3694521830443483e-06,
+      "loss": 0.2789,
+      "num_input_tokens_seen": 363105120,
+      "step": 6590,
+      "train_runtime": 39788.6022,
+      "train_tokens_per_second": 9125.858
+    },
+    {
+      "epoch": 1.353931430917676,
+      "grad_norm": 0.627621806019884,
+      "learning_rate": 2.3625995616619156e-06,
+      "loss": 0.2924,
+      "num_input_tokens_seen": 363408480,
+      "step": 6595,
+      "train_runtime": 39817.34,
+      "train_tokens_per_second": 9126.89
+    },
+    {
+      "epoch": 1.3549579141859986,
+      "grad_norm": 0.6694570398965487,
+      "learning_rate": 2.355753797047929e-06,
+      "loss": 0.3168,
+      "num_input_tokens_seen": 363698744,
+      "step": 6600,
+      "train_runtime": 39849.3725,
+      "train_tokens_per_second": 9126.837
+    },
+    {
+      "epoch": 1.3559843974543215,
+      "grad_norm": 0.7842134345396427,
+      "learning_rate": 2.348914907000147e-06,
+      "loss": 0.2719,
+      "num_input_tokens_seen": 363981528,
+      "step": 6605,
+      "train_runtime": 39885.5153,
+      "train_tokens_per_second": 9125.657
+    },
+    {
+      "epoch": 1.3570108807226442,
+      "grad_norm": 0.583509311512718,
+      "learning_rate": 2.3420829092984632e-06,
+      "loss": 0.2885,
+      "num_input_tokens_seen": 364268648,
+      "step": 6610,
+      "train_runtime": 39916.877,
+      "train_tokens_per_second": 9125.68
+    },
+    {
+      "epoch": 1.358037363990967,
+      "grad_norm": 0.6389475742504179,
+      "learning_rate": 2.335257821704846e-06,
+      "loss": 0.2911,
+      "num_input_tokens_seen": 364561984,
+      "step": 6615,
+      "train_runtime": 39951.5514,
+      "train_tokens_per_second": 9125.102
+    },
+    {
+      "epoch": 1.3590638472592897,
+      "grad_norm": 0.5893594244370135,
+      "learning_rate": 2.3284396619632948e-06,
+      "loss": 0.2916,
+      "num_input_tokens_seen": 364812520,
+      "step": 6620,
+      "train_runtime": 39978.5809,
+      "train_tokens_per_second": 9125.199
+    },
+    {
+      "epoch": 1.3600903305276124,
+      "grad_norm": 0.683772712287262,
+      "learning_rate": 2.321628447799808e-06,
+      "loss": 0.3082,
+      "num_input_tokens_seen": 365087768,
+      "step": 6625,
+      "train_runtime": 40009.5642,
+      "train_tokens_per_second": 9125.012
+    },
+    {
+      "epoch": 1.361116813795935,
+      "grad_norm": 0.7391966750708234,
+      "learning_rate": 2.31482419692232e-06,
+      "loss": 0.3001,
+      "num_input_tokens_seen": 365365248,
+      "step": 6630,
+      "train_runtime": 40032.0215,
+      "train_tokens_per_second": 9126.825
+    },
+    {
+      "epoch": 1.3621432970642577,
+      "grad_norm": 0.6921703317774391,
+      "learning_rate": 2.308026927020664e-06,
+      "loss": 0.2753,
+      "num_input_tokens_seen": 365651568,
+      "step": 6635,
+      "train_runtime": 40063.3693,
+      "train_tokens_per_second": 9126.83
+    },
+    {
+      "epoch": 1.3631697803325806,
+      "grad_norm": 0.7332766376120624,
+      "learning_rate": 2.3012366557665226e-06,
+      "loss": 0.2943,
+      "num_input_tokens_seen": 365927112,
+      "step": 6640,
+      "train_runtime": 40093.7777,
+      "train_tokens_per_second": 9126.781
+    },
+    {
+      "epoch": 1.3641962636009033,
+      "grad_norm": 0.6400124900700764,
+      "learning_rate": 2.2944534008133856e-06,
+      "loss": 0.3058,
+      "num_input_tokens_seen": 366211224,
+      "step": 6645,
+      "train_runtime": 40126.7493,
+      "train_tokens_per_second": 9126.362
+    },
+    {
+      "epoch": 1.365222746869226,
+      "grad_norm": 0.8685646214617104,
+      "learning_rate": 2.2876771797964986e-06,
+      "loss": 0.3138,
+      "num_input_tokens_seen": 366482216,
+      "step": 6650,
+      "train_runtime": 40153.352,
+      "train_tokens_per_second": 9127.064
+    },
+    {
+      "epoch": 1.3662492301375488,
+      "grad_norm": 0.7111093799696572,
+      "learning_rate": 2.2809080103328205e-06,
+      "loss": 0.2972,
+      "num_input_tokens_seen": 366739480,
+      "step": 6655,
+      "train_runtime": 40183.5212,
+      "train_tokens_per_second": 9126.614
+    },
+    {
+      "epoch": 1.3672757134058715,
+      "grad_norm": 0.6998228214886973,
+      "learning_rate": 2.274145910020982e-06,
+      "loss": 0.2836,
+      "num_input_tokens_seen": 366998792,
+      "step": 6660,
+      "train_runtime": 40208.9563,
+      "train_tokens_per_second": 9127.29
+    },
+    {
+      "epoch": 1.3683021966741942,
+      "grad_norm": 0.7387565585240519,
+      "learning_rate": 2.267390896441232e-06,
+      "loss": 0.2818,
+      "num_input_tokens_seen": 367272712,
+      "step": 6665,
+      "train_runtime": 40247.3908,
+      "train_tokens_per_second": 9125.379
+    },
+    {
+      "epoch": 1.3693286799425168,
+      "grad_norm": 0.6479923903024966,
+      "learning_rate": 2.2606429871553935e-06,
+      "loss": 0.2976,
+      "num_input_tokens_seen": 367564192,
+      "step": 6670,
+      "train_runtime": 40275.8585,
+      "train_tokens_per_second": 9126.167
+    },
+    {
+      "epoch": 1.3703551632108397,
+      "grad_norm": 0.883196910643702,
+      "learning_rate": 2.253902199706819e-06,
+      "loss": 0.2784,
+      "num_input_tokens_seen": 367816112,
+      "step": 6675,
+      "train_runtime": 40303.7131,
+      "train_tokens_per_second": 9126.11
+    },
+    {
+      "epoch": 1.3713816464791624,
+      "grad_norm": 0.8450896251835531,
+      "learning_rate": 2.247168551620351e-06,
+      "loss": 0.2959,
+      "num_input_tokens_seen": 368087520,
+      "step": 6680,
+      "train_runtime": 40324.9074,
+      "train_tokens_per_second": 9128.044
+    },
+    {
+      "epoch": 1.372408129747485,
+      "grad_norm": 0.6878835028196867,
+      "learning_rate": 2.2404420604022677e-06,
+      "loss": 0.2704,
+      "num_input_tokens_seen": 368369736,
+      "step": 6685,
+      "train_runtime": 40351.8191,
+      "train_tokens_per_second": 9128.95
+    },
+    {
+      "epoch": 1.373434613015808,
+      "grad_norm": 0.6429145244480807,
+      "learning_rate": 2.2337227435402392e-06,
+      "loss": 0.2757,
+      "num_input_tokens_seen": 368668952,
+      "step": 6690,
+      "train_runtime": 40383.7466,
+      "train_tokens_per_second": 9129.142
+    },
+    {
+      "epoch": 1.3744610962841306,
+      "grad_norm": 0.648716919047243,
+      "learning_rate": 2.227010618503286e-06,
+      "loss": 0.255,
+      "num_input_tokens_seen": 368944240,
+      "step": 6695,
+      "train_runtime": 40409.2114,
+      "train_tokens_per_second": 9130.201
+    },
+    {
+      "epoch": 1.3754875795524533,
+      "grad_norm": 0.617159799516413,
+      "learning_rate": 2.2203057027417293e-06,
+      "loss": 0.27,
+      "num_input_tokens_seen": 369225768,
+      "step": 6700,
+      "train_runtime": 40436.3395,
+      "train_tokens_per_second": 9131.038
+    },
+    {
+      "epoch": 1.376514062820776,
+      "grad_norm": 0.66930022033639,
+      "learning_rate": 2.213608013687149e-06,
+      "loss": 0.2865,
+      "num_input_tokens_seen": 369509088,
+      "step": 6705,
+      "train_runtime": 40467.3099,
+      "train_tokens_per_second": 9131.051
+    },
+    {
+      "epoch": 1.3775405460890988,
+      "grad_norm": 0.8447787980265163,
+      "learning_rate": 2.206917568752335e-06,
+      "loss": 0.2902,
+      "num_input_tokens_seen": 369780840,
+      "step": 6710,
+      "train_runtime": 40499.2945,
+      "train_tokens_per_second": 9130.55
+    },
+    {
+      "epoch": 1.3785670293574215,
+      "grad_norm": 0.6510497273411973,
+      "learning_rate": 2.200234385331247e-06,
+      "loss": 0.28,
+      "num_input_tokens_seen": 370050976,
+      "step": 6715,
+      "train_runtime": 40534.8618,
+      "train_tokens_per_second": 9129.203
+    },
+    {
+      "epoch": 1.3795935126257441,
+      "grad_norm": 0.7966538254535465,
+      "learning_rate": 2.1935584807989653e-06,
+      "loss": 0.2777,
+      "num_input_tokens_seen": 370343960,
+      "step": 6720,
+      "train_runtime": 40565.1595,
+      "train_tokens_per_second": 9129.607
+    },
+    {
+      "epoch": 1.380619995894067,
+      "grad_norm": 0.5982636011164046,
+      "learning_rate": 2.1868898725116417e-06,
+      "loss": 0.3238,
+      "num_input_tokens_seen": 370613152,
+      "step": 6725,
+      "train_runtime": 40591.5072,
+      "train_tokens_per_second": 9130.313
+    },
+    {
+      "epoch": 1.3816464791623897,
+      "grad_norm": 0.670901782316349,
+      "learning_rate": 2.1802285778064615e-06,
+      "loss": 0.2876,
+      "num_input_tokens_seen": 370895272,
+      "step": 6730,
+      "train_runtime": 40624.9366,
+      "train_tokens_per_second": 9129.744
+    },
+    {
+      "epoch": 1.3826729624307124,
+      "grad_norm": 0.6510155152721939,
+      "learning_rate": 2.1735746140016016e-06,
+      "loss": 0.2699,
+      "num_input_tokens_seen": 371177368,
+      "step": 6735,
+      "train_runtime": 40657.5245,
+      "train_tokens_per_second": 9129.365
+    },
+    {
+      "epoch": 1.383699445699035,
+      "grad_norm": 0.7899285312375668,
+      "learning_rate": 2.1669279983961735e-06,
+      "loss": 0.2945,
+      "num_input_tokens_seen": 371447216,
+      "step": 6740,
+      "train_runtime": 40689.5565,
+      "train_tokens_per_second": 9128.81
+    },
+    {
+      "epoch": 1.384725928967358,
+      "grad_norm": 0.6379939742388341,
+      "learning_rate": 2.1602887482701863e-06,
+      "loss": 0.2773,
+      "num_input_tokens_seen": 371702880,
+      "step": 6745,
+      "train_runtime": 40718.5704,
+      "train_tokens_per_second": 9128.584
+    },
+    {
+      "epoch": 1.3857524122356806,
+      "grad_norm": 0.6184078883807373,
+      "learning_rate": 2.1536568808845e-06,
+      "loss": 0.3064,
+      "num_input_tokens_seen": 371992664,
+      "step": 6750,
+      "train_runtime": 40750.4483,
+      "train_tokens_per_second": 9128.554
+    },
+    {
+      "epoch": 1.3867788955040032,
+      "grad_norm": 0.7371570266825161,
+      "learning_rate": 2.147032413480782e-06,
+      "loss": 0.2955,
+      "num_input_tokens_seen": 372274376,
+      "step": 6755,
+      "train_runtime": 40781.2579,
+      "train_tokens_per_second": 9128.565
+    },
+    {
+      "epoch": 1.3878053787723261,
+      "grad_norm": 0.6623796366660992,
+      "learning_rate": 2.140415363281459e-06,
+      "loss": 0.2917,
+      "num_input_tokens_seen": 372556320,
+      "step": 6760,
+      "train_runtime": 40815.2134,
+      "train_tokens_per_second": 9127.879
+    },
+    {
+      "epoch": 1.3888318620406488,
+      "grad_norm": 0.7380111722463538,
+      "learning_rate": 2.1338057474896753e-06,
+      "loss": 0.2897,
+      "num_input_tokens_seen": 372850248,
+      "step": 6765,
+      "train_runtime": 40847.9676,
+      "train_tokens_per_second": 9127.755
+    },
+    {
+      "epoch": 1.3898583453089715,
+      "grad_norm": 0.681393479929912,
+      "learning_rate": 2.1272035832892517e-06,
+      "loss": 0.3057,
+      "num_input_tokens_seen": 373101312,
+      "step": 6770,
+      "train_runtime": 40878.3217,
+      "train_tokens_per_second": 9127.119
+    },
+    {
+      "epoch": 1.3908848285772941,
+      "grad_norm": 0.6738137427688541,
+      "learning_rate": 2.1206088878446268e-06,
+      "loss": 0.2907,
+      "num_input_tokens_seen": 373382320,
+      "step": 6775,
+      "train_runtime": 40905.3357,
+      "train_tokens_per_second": 9127.961
+    },
+    {
+      "epoch": 1.3919113118456168,
+      "grad_norm": 0.7547756989621721,
+      "learning_rate": 2.114021678300827e-06,
+      "loss": 0.279,
+      "num_input_tokens_seen": 373657304,
+      "step": 6780,
+      "train_runtime": 40934.7027,
+      "train_tokens_per_second": 9128.13
+    },
+    {
+      "epoch": 1.3929377951139397,
+      "grad_norm": 0.7355862999966134,
+      "learning_rate": 2.1074419717834193e-06,
+      "loss": 0.2969,
+      "num_input_tokens_seen": 373925584,
+      "step": 6785,
+      "train_runtime": 40963.6131,
+      "train_tokens_per_second": 9128.237
+    },
+    {
+      "epoch": 1.3939642783822623,
+      "grad_norm": 0.668205980085425,
+      "learning_rate": 2.1008697853984605e-06,
+      "loss": 0.3149,
+      "num_input_tokens_seen": 374227256,
+      "step": 6790,
+      "train_runtime": 40994.7422,
+      "train_tokens_per_second": 9128.665
+    },
+    {
+      "epoch": 1.3949907616505852,
+      "grad_norm": 0.6524406584030676,
+      "learning_rate": 2.0943051362324568e-06,
+      "loss": 0.2728,
+      "num_input_tokens_seen": 374507000,
+      "step": 6795,
+      "train_runtime": 41026.9877,
+      "train_tokens_per_second": 9128.308
+    },
+    {
+      "epoch": 1.3960172449189079,
+      "grad_norm": 0.6076986589788961,
+      "learning_rate": 2.087748041352319e-06,
+      "loss": 0.2631,
+      "num_input_tokens_seen": 374786656,
+      "step": 6800,
+      "train_runtime": 41057.7182,
+      "train_tokens_per_second": 9128.287
+    },
+    {
+      "epoch": 1.3970437281872305,
+      "grad_norm": 0.8183725354300722,
+      "learning_rate": 2.08119851780532e-06,
+      "loss": 0.2849,
+      "num_input_tokens_seen": 375049600,
+      "step": 6805,
+      "train_runtime": 41085.1189,
+      "train_tokens_per_second": 9128.6
+    },
+    {
+      "epoch": 1.3980702114555532,
+      "grad_norm": 0.7490575633327407,
+      "learning_rate": 2.074656582619047e-06,
+      "loss": 0.2934,
+      "num_input_tokens_seen": 375306440,
+      "step": 6810,
+      "train_runtime": 41110.5865,
+      "train_tokens_per_second": 9129.192
+    },
+    {
+      "epoch": 1.3990966947238759,
+      "grad_norm": 0.6842856576925328,
+      "learning_rate": 2.068122252801357e-06,
+      "loss": 0.2979,
+      "num_input_tokens_seen": 375601912,
+      "step": 6815,
+      "train_runtime": 41142.7381,
+      "train_tokens_per_second": 9129.24
+    },
+    {
+      "epoch": 1.4001231779921988,
+      "grad_norm": 0.6824316572129177,
+      "learning_rate": 2.0615955453403384e-06,
+      "loss": 0.3028,
+      "num_input_tokens_seen": 375867912,
+      "step": 6820,
+      "train_runtime": 41172.3577,
+      "train_tokens_per_second": 9129.133
+    },
+    {
+      "epoch": 1.4011496612605214,
+      "grad_norm": 0.6745741857959566,
+      "learning_rate": 2.0550764772042596e-06,
+      "loss": 0.2981,
+      "num_input_tokens_seen": 376153752,
+      "step": 6825,
+      "train_runtime": 41207.4313,
+      "train_tokens_per_second": 9128.299
+    },
+    {
+      "epoch": 1.4021761445288443,
+      "grad_norm": 0.7270357623418761,
+      "learning_rate": 2.048565065341531e-06,
+      "loss": 0.2826,
+      "num_input_tokens_seen": 376417496,
+      "step": 6830,
+      "train_runtime": 41236.4184,
+      "train_tokens_per_second": 9128.278
+    },
+    {
+      "epoch": 1.403202627797167,
+      "grad_norm": 0.7658918040883771,
+      "learning_rate": 2.042061326680653e-06,
+      "loss": 0.3071,
+      "num_input_tokens_seen": 376683456,
+      "step": 6835,
+      "train_runtime": 41267.8455,
+      "train_tokens_per_second": 9127.771
+    },
+    {
+      "epoch": 1.4042291110654896,
+      "grad_norm": 0.6171861804205186,
+      "learning_rate": 2.0355652781301844e-06,
+      "loss": 0.2796,
+      "num_input_tokens_seen": 376962312,
+      "step": 6840,
+      "train_runtime": 41299.5216,
+      "train_tokens_per_second": 9127.522
+    },
+    {
+      "epoch": 1.4052555943338123,
+      "grad_norm": 0.653544282163319,
+      "learning_rate": 2.0290769365786857e-06,
+      "loss": 0.2731,
+      "num_input_tokens_seen": 377234376,
+      "step": 6845,
+      "train_runtime": 41331.2053,
+      "train_tokens_per_second": 9127.108
+    },
+    {
+      "epoch": 1.406282077602135,
+      "grad_norm": 0.804341318518001,
+      "learning_rate": 2.0225963188946835e-06,
+      "loss": 0.3025,
+      "num_input_tokens_seen": 377499312,
+      "step": 6850,
+      "train_runtime": 41362.6159,
+      "train_tokens_per_second": 9126.582
+    },
+    {
+      "epoch": 1.4073085608704579,
+      "grad_norm": 0.6865398544932734,
+      "learning_rate": 2.0161234419266175e-06,
+      "loss": 0.2837,
+      "num_input_tokens_seen": 377765488,
+      "step": 6855,
+      "train_runtime": 41393.9018,
+      "train_tokens_per_second": 9126.115
+    },
+    {
+      "epoch": 1.4083350441387805,
+      "grad_norm": 0.7365574134676242,
+      "learning_rate": 2.009658322502812e-06,
+      "loss": 0.3004,
+      "num_input_tokens_seen": 378011672,
+      "step": 6860,
+      "train_runtime": 41421.7626,
+      "train_tokens_per_second": 9125.92
+    },
+    {
+      "epoch": 1.4093615274071032,
+      "grad_norm": 0.6597804634429908,
+      "learning_rate": 2.0032009774314186e-06,
+      "loss": 0.2973,
+      "num_input_tokens_seen": 378315056,
+      "step": 6865,
+      "train_runtime": 41455.4214,
+      "train_tokens_per_second": 9125.828
+    },
+    {
+      "epoch": 1.410388010675426,
+      "grad_norm": 0.68790059136237,
+      "learning_rate": 1.996751423500376e-06,
+      "loss": 0.3062,
+      "num_input_tokens_seen": 378592600,
+      "step": 6870,
+      "train_runtime": 41486.3165,
+      "train_tokens_per_second": 9125.722
+    },
+    {
+      "epoch": 1.4114144939437487,
+      "grad_norm": 0.6565992270335824,
+      "learning_rate": 1.9903096774773684e-06,
+      "loss": 0.2795,
+      "num_input_tokens_seen": 378870600,
+      "step": 6875,
+      "train_runtime": 41519.7737,
+      "train_tokens_per_second": 9125.064
+    },
+    {
+      "epoch": 1.4124409772120714,
+      "grad_norm": 0.7128074326273595,
+      "learning_rate": 1.9838757561097817e-06,
+      "loss": 0.2888,
+      "num_input_tokens_seen": 379154608,
+      "step": 6880,
+      "train_runtime": 41546.787,
+      "train_tokens_per_second": 9125.967
+    },
+    {
+      "epoch": 1.413467460480394,
+      "grad_norm": 0.6175162956182079,
+      "learning_rate": 1.9774496761246575e-06,
+      "loss": 0.2493,
+      "num_input_tokens_seen": 379436696,
+      "step": 6885,
+      "train_runtime": 41574.7237,
+      "train_tokens_per_second": 9126.62
+    },
+    {
+      "epoch": 1.414493943748717,
+      "grad_norm": 0.8718106829733486,
+      "learning_rate": 1.971031454228651e-06,
+      "loss": 0.2872,
+      "num_input_tokens_seen": 379696768,
+      "step": 6890,
+      "train_runtime": 41605.0856,
+      "train_tokens_per_second": 9126.21
+    },
+    {
+      "epoch": 1.4155204270170396,
+      "grad_norm": 0.5895075330055656,
+      "learning_rate": 1.9646211071079914e-06,
+      "loss": 0.2865,
+      "num_input_tokens_seen": 379994536,
+      "step": 6895,
+      "train_runtime": 41635.6948,
+      "train_tokens_per_second": 9126.653
+    },
+    {
+      "epoch": 1.4165469102853623,
+      "grad_norm": 0.7110282648392519,
+      "learning_rate": 1.9582186514284337e-06,
+      "loss": 0.2885,
+      "num_input_tokens_seen": 380261720,
+      "step": 6900,
+      "train_runtime": 41666.4854,
+      "train_tokens_per_second": 9126.321
+    },
+    {
+      "epoch": 1.4175733935536852,
+      "grad_norm": 0.763650376702062,
+      "learning_rate": 1.9518241038352083e-06,
+      "loss": 0.2975,
+      "num_input_tokens_seen": 380519480,
+      "step": 6905,
+      "train_runtime": 41699.6646,
+      "train_tokens_per_second": 9125.241
+    },
+    {
+      "epoch": 1.4185998768220078,
+      "grad_norm": 0.7499495043953227,
+      "learning_rate": 1.945437480952999e-06,
+      "loss": 0.2711,
+      "num_input_tokens_seen": 380774896,
+      "step": 6910,
+      "train_runtime": 41731.1306,
+      "train_tokens_per_second": 9124.481
+    },
+    {
+      "epoch": 1.4196263600903305,
+      "grad_norm": 0.6377441691660825,
+      "learning_rate": 1.939058799385879e-06,
+      "loss": 0.2878,
+      "num_input_tokens_seen": 381064672,
+      "step": 6915,
+      "train_runtime": 41766.3025,
+      "train_tokens_per_second": 9123.735
+    },
+    {
+      "epoch": 1.4206528433586532,
+      "grad_norm": 0.6464389023688649,
+      "learning_rate": 1.9326880757172774e-06,
+      "loss": 0.2977,
+      "num_input_tokens_seen": 381327552,
+      "step": 6920,
+      "train_runtime": 41800.9901,
+      "train_tokens_per_second": 9122.453
+    },
+    {
+      "epoch": 1.421679326626976,
+      "grad_norm": 0.6539424795543127,
+      "learning_rate": 1.926325326509933e-06,
+      "loss": 0.2756,
+      "num_input_tokens_seen": 381622560,
+      "step": 6925,
+      "train_runtime": 41824.8117,
+      "train_tokens_per_second": 9124.31
+    },
+    {
+      "epoch": 1.4227058098952987,
+      "grad_norm": 0.6238085916832637,
+      "learning_rate": 1.919970568305854e-06,
+      "loss": 0.2941,
+      "num_input_tokens_seen": 381912192,
+      "step": 6930,
+      "train_runtime": 41849.4207,
+      "train_tokens_per_second": 9125.866
+    },
+    {
+      "epoch": 1.4237322931636214,
+      "grad_norm": 0.7050825295296314,
+      "learning_rate": 1.913623817626272e-06,
+      "loss": 0.2898,
+      "num_input_tokens_seen": 382167936,
+      "step": 6935,
+      "train_runtime": 41879.3656,
+      "train_tokens_per_second": 9125.447
+    },
+    {
+      "epoch": 1.4247587764319443,
+      "grad_norm": 0.712894104339634,
+      "learning_rate": 1.907285090971603e-06,
+      "loss": 0.298,
+      "num_input_tokens_seen": 382455312,
+      "step": 6940,
+      "train_runtime": 41913.7926,
+      "train_tokens_per_second": 9124.808
+    },
+    {
+      "epoch": 1.425785259700267,
+      "grad_norm": 0.6487106594986023,
+      "learning_rate": 1.9009544048213963e-06,
+      "loss": 0.2925,
+      "num_input_tokens_seen": 382722352,
+      "step": 6945,
+      "train_runtime": 41947.486,
+      "train_tokens_per_second": 9123.845
+    },
+    {
+      "epoch": 1.4268117429685896,
+      "grad_norm": 0.6983437231796298,
+      "learning_rate": 1.894631775634308e-06,
+      "loss": 0.2916,
+      "num_input_tokens_seen": 383013064,
+      "step": 6950,
+      "train_runtime": 41978.5526,
+      "train_tokens_per_second": 9124.018
+    },
+    {
+      "epoch": 1.4278382262369123,
+      "grad_norm": 0.6789066316634869,
+      "learning_rate": 1.8883172198480355e-06,
+      "loss": 0.2927,
+      "num_input_tokens_seen": 383322872,
+      "step": 6955,
+      "train_runtime": 42006.8717,
+      "train_tokens_per_second": 9125.242
+    },
+    {
+      "epoch": 1.4288647095052351,
+      "grad_norm": 0.7237707494711455,
+      "learning_rate": 1.8820107538792914e-06,
+      "loss": 0.2874,
+      "num_input_tokens_seen": 383612664,
+      "step": 6960,
+      "train_runtime": 42040.9384,
+      "train_tokens_per_second": 9124.741
+    },
+    {
+      "epoch": 1.4298911927735578,
+      "grad_norm": 0.7259676705598558,
+      "learning_rate": 1.8757123941237604e-06,
+      "loss": 0.2772,
+      "num_input_tokens_seen": 383887768,
+      "step": 6965,
+      "train_runtime": 42072.0016,
+      "train_tokens_per_second": 9124.543
+    },
+    {
+      "epoch": 1.4309176760418805,
+      "grad_norm": 0.9465766492015032,
+      "learning_rate": 1.8694221569560484e-06,
+      "loss": 0.2887,
+      "num_input_tokens_seen": 384132800,
+      "step": 6970,
+      "train_runtime": 42098.6409,
+      "train_tokens_per_second": 9124.589
+    },
+    {
+      "epoch": 1.4319441593102034,
+      "grad_norm": 0.6516784804414664,
+      "learning_rate": 1.8631400587296433e-06,
+      "loss": 0.2982,
+      "num_input_tokens_seen": 384407568,
+      "step": 6975,
+      "train_runtime": 42125.6218,
+      "train_tokens_per_second": 9125.268
+    },
+    {
+      "epoch": 1.432970642578526,
+      "grad_norm": 0.6177974658282939,
+      "learning_rate": 1.8568661157768753e-06,
+      "loss": 0.2901,
+      "num_input_tokens_seen": 384687912,
+      "step": 6980,
+      "train_runtime": 42154.0194,
+      "train_tokens_per_second": 9125.771
+    },
+    {
+      "epoch": 1.4339971258468487,
+      "grad_norm": 0.6623848117619114,
+      "learning_rate": 1.8506003444088716e-06,
+      "loss": 0.2958,
+      "num_input_tokens_seen": 384964608,
+      "step": 6985,
+      "train_runtime": 42184.2026,
+      "train_tokens_per_second": 9125.8
+    },
+    {
+      "epoch": 1.4350236091151714,
+      "grad_norm": 0.6418716741977396,
+      "learning_rate": 1.844342760915514e-06,
+      "loss": 0.2829,
+      "num_input_tokens_seen": 385252328,
+      "step": 6990,
+      "train_runtime": 42218.7065,
+      "train_tokens_per_second": 9125.157
+    },
+    {
+      "epoch": 1.436050092383494,
+      "grad_norm": 0.5533198771782328,
+      "learning_rate": 1.838093381565398e-06,
+      "loss": 0.2893,
+      "num_input_tokens_seen": 385524600,
+      "step": 6995,
+      "train_runtime": 42246.6146,
+      "train_tokens_per_second": 9125.574
+    },
+    {
+      "epoch": 1.437076575651817,
+      "grad_norm": 0.7152385600853659,
+      "learning_rate": 1.8318522226057934e-06,
+      "loss": 0.2841,
+      "num_input_tokens_seen": 385801816,
+      "step": 7000,
+      "train_runtime": 42276.7944,
+      "train_tokens_per_second": 9125.617
+    },
+    {
+      "epoch": 1.4381030589201396,
+      "grad_norm": 0.6287723550813584,
+      "learning_rate": 1.82561930026259e-06,
+      "loss": 0.2844,
+      "num_input_tokens_seen": 386088008,
+      "step": 7005,
+      "train_runtime": 42304.0719,
+      "train_tokens_per_second": 9126.498
+    },
+    {
+      "epoch": 1.4391295421884625,
+      "grad_norm": 0.6960072082831291,
+      "learning_rate": 1.819394630740271e-06,
+      "loss": 0.2895,
+      "num_input_tokens_seen": 386344328,
+      "step": 7010,
+      "train_runtime": 42332.8643,
+      "train_tokens_per_second": 9126.345
+    },
+    {
+      "epoch": 1.4401560254567851,
+      "grad_norm": 0.6232550868424318,
+      "learning_rate": 1.8131782302218602e-06,
+      "loss": 0.2812,
+      "num_input_tokens_seen": 386639784,
+      "step": 7015,
+      "train_runtime": 42363.1982,
+      "train_tokens_per_second": 9126.785
+    },
+    {
+      "epoch": 1.4411825087251078,
+      "grad_norm": 0.736512624934245,
+      "learning_rate": 1.8069701148688873e-06,
+      "loss": 0.2906,
+      "num_input_tokens_seen": 386905424,
+      "step": 7020,
+      "train_runtime": 42396.1192,
+      "train_tokens_per_second": 9125.963
+    },
+    {
+      "epoch": 1.4422089919934304,
+      "grad_norm": 0.7206429423438141,
+      "learning_rate": 1.8007703008213384e-06,
+      "loss": 0.3093,
+      "num_input_tokens_seen": 387189656,
+      "step": 7025,
+      "train_runtime": 42432.0366,
+      "train_tokens_per_second": 9124.937
+    },
+    {
+      "epoch": 1.443235475261753,
+      "grad_norm": 0.7491250416988712,
+      "learning_rate": 1.794578804197618e-06,
+      "loss": 0.275,
+      "num_input_tokens_seen": 387468632,
+      "step": 7030,
+      "train_runtime": 42459.6588,
+      "train_tokens_per_second": 9125.571
+    },
+    {
+      "epoch": 1.444261958530076,
+      "grad_norm": 0.6211507907904463,
+      "learning_rate": 1.788395641094509e-06,
+      "loss": 0.287,
+      "num_input_tokens_seen": 387764136,
+      "step": 7035,
+      "train_runtime": 42489.5456,
+      "train_tokens_per_second": 9126.107
+    },
+    {
+      "epoch": 1.4452884417983987,
+      "grad_norm": 0.6624110976449131,
+      "learning_rate": 1.7822208275871267e-06,
+      "loss": 0.2843,
+      "num_input_tokens_seen": 388038728,
+      "step": 7040,
+      "train_runtime": 42513.2114,
+      "train_tokens_per_second": 9127.486
+    },
+    {
+      "epoch": 1.4463149250667215,
+      "grad_norm": 0.7119278503748192,
+      "learning_rate": 1.7760543797288787e-06,
+      "loss": 0.2859,
+      "num_input_tokens_seen": 388299592,
+      "step": 7045,
+      "train_runtime": 42543.1725,
+      "train_tokens_per_second": 9127.189
+    },
+    {
+      "epoch": 1.4473414083350442,
+      "grad_norm": 0.750225664754766,
+      "learning_rate": 1.7698963135514253e-06,
+      "loss": 0.3068,
+      "num_input_tokens_seen": 388574024,
+      "step": 7050,
+      "train_runtime": 42574.925,
+      "train_tokens_per_second": 9126.828
+    },
+    {
+      "epoch": 1.4483678916033669,
+      "grad_norm": 0.6278718476140441,
+      "learning_rate": 1.7637466450646346e-06,
+      "loss": 0.2916,
+      "num_input_tokens_seen": 388839712,
+      "step": 7055,
+      "train_runtime": 42607.6306,
+      "train_tokens_per_second": 9126.058
+    },
+    {
+      "epoch": 1.4493943748716895,
+      "grad_norm": 0.846237568004612,
+      "learning_rate": 1.7576053902565421e-06,
+      "loss": 0.316,
+      "num_input_tokens_seen": 389125168,
+      "step": 7060,
+      "train_runtime": 42631.7927,
+      "train_tokens_per_second": 9127.582
+    },
+    {
+      "epoch": 1.4504208581400122,
+      "grad_norm": 0.6955679609039622,
+      "learning_rate": 1.7514725650933102e-06,
+      "loss": 0.2776,
+      "num_input_tokens_seen": 389399568,
+      "step": 7065,
+      "train_runtime": 42667.1335,
+      "train_tokens_per_second": 9126.453
+    },
+    {
+      "epoch": 1.451447341408335,
+      "grad_norm": 0.6867199416842031,
+      "learning_rate": 1.7453481855191834e-06,
+      "loss": 0.2998,
+      "num_input_tokens_seen": 389658368,
+      "step": 7070,
+      "train_runtime": 42699.0139,
+      "train_tokens_per_second": 9125.699
+    },
+    {
+      "epoch": 1.4524738246766578,
+      "grad_norm": 0.7629285576998989,
+      "learning_rate": 1.7392322674564538e-06,
+      "loss": 0.2619,
+      "num_input_tokens_seen": 389916088,
+      "step": 7075,
+      "train_runtime": 42732.6068,
+      "train_tokens_per_second": 9124.557
+    },
+    {
+      "epoch": 1.4535003079449804,
+      "grad_norm": 0.6834227336019226,
+      "learning_rate": 1.7331248268054123e-06,
+      "loss": 0.2826,
+      "num_input_tokens_seen": 390211984,
+      "step": 7080,
+      "train_runtime": 42765.9581,
+      "train_tokens_per_second": 9124.36
+    },
+    {
+      "epoch": 1.4545267912133033,
+      "grad_norm": 0.6997262464374445,
+      "learning_rate": 1.727025879444305e-06,
+      "loss": 0.2954,
+      "num_input_tokens_seen": 390495152,
+      "step": 7085,
+      "train_runtime": 42801.3171,
+      "train_tokens_per_second": 9123.438
+    },
+    {
+      "epoch": 1.455553274481626,
+      "grad_norm": 0.7964009689984521,
+      "learning_rate": 1.7209354412293077e-06,
+      "loss": 0.3349,
+      "num_input_tokens_seen": 390787104,
+      "step": 7090,
+      "train_runtime": 42832.0337,
+      "train_tokens_per_second": 9123.711
+    },
+    {
+      "epoch": 1.4565797577499486,
+      "grad_norm": 0.666367865107754,
+      "learning_rate": 1.7148535279944656e-06,
+      "loss": 0.2676,
+      "num_input_tokens_seen": 391073248,
+      "step": 7095,
+      "train_runtime": 42866.0952,
+      "train_tokens_per_second": 9123.137
+    },
+    {
+      "epoch": 1.4576062410182713,
+      "grad_norm": 0.7843391903631476,
+      "learning_rate": 1.7087801555516636e-06,
+      "loss": 0.279,
+      "num_input_tokens_seen": 391322552,
+      "step": 7100,
+      "train_runtime": 42897.7992,
+      "train_tokens_per_second": 9122.206
+    },
+    {
+      "epoch": 1.4586327242865942,
+      "grad_norm": 0.6054339892350564,
+      "learning_rate": 1.702715339690581e-06,
+      "loss": 0.2941,
+      "num_input_tokens_seen": 391608648,
+      "step": 7105,
+      "train_runtime": 42929.977,
+      "train_tokens_per_second": 9122.033
+    },
+    {
+      "epoch": 1.4596592075549168,
+      "grad_norm": 0.7046724700602465,
+      "learning_rate": 1.6966590961786529e-06,
+      "loss": 0.2884,
+      "num_input_tokens_seen": 391903792,
+      "step": 7110,
+      "train_runtime": 42960.0136,
+      "train_tokens_per_second": 9122.525
+    },
+    {
+      "epoch": 1.4606856908232395,
+      "grad_norm": 0.6117337336096451,
+      "learning_rate": 1.6906114407610258e-06,
+      "loss": 0.2887,
+      "num_input_tokens_seen": 392163336,
+      "step": 7115,
+      "train_runtime": 42990.294,
+      "train_tokens_per_second": 9122.137
+    },
+    {
+      "epoch": 1.4617121740915624,
+      "grad_norm": 0.5920641847827747,
+      "learning_rate": 1.6845723891605198e-06,
+      "loss": 0.2878,
+      "num_input_tokens_seen": 392461248,
+      "step": 7120,
+      "train_runtime": 43021.194,
+      "train_tokens_per_second": 9122.509
+    },
+    {
+      "epoch": 1.462738657359885,
+      "grad_norm": 0.6944447779745019,
+      "learning_rate": 1.6785419570775886e-06,
+      "loss": 0.2893,
+      "num_input_tokens_seen": 392745376,
+      "step": 7125,
+      "train_runtime": 43050.2291,
+      "train_tokens_per_second": 9122.957
+    },
+    {
+      "epoch": 1.4637651406282077,
+      "grad_norm": 0.6345188813586844,
+      "learning_rate": 1.672520160190275e-06,
+      "loss": 0.3008,
+      "num_input_tokens_seen": 393002736,
+      "step": 7130,
+      "train_runtime": 43081.7444,
+      "train_tokens_per_second": 9122.257
+    },
+    {
+      "epoch": 1.4647916238965304,
+      "grad_norm": 0.6862337572632851,
+      "learning_rate": 1.666507014154169e-06,
+      "loss": 0.2974,
+      "num_input_tokens_seen": 393276808,
+      "step": 7135,
+      "train_runtime": 43111.9234,
+      "train_tokens_per_second": 9122.228
+    },
+    {
+      "epoch": 1.4658181071648533,
+      "grad_norm": 0.6570874639472464,
+      "learning_rate": 1.6605025346023723e-06,
+      "loss": 0.2945,
+      "num_input_tokens_seen": 393564864,
+      "step": 7140,
+      "train_runtime": 43138.8781,
+      "train_tokens_per_second": 9123.206
+    },
+    {
+      "epoch": 1.466844590433176,
+      "grad_norm": 0.6246454635261469,
+      "learning_rate": 1.6545067371454588e-06,
+      "loss": 0.2929,
+      "num_input_tokens_seen": 393847816,
+      "step": 7145,
+      "train_runtime": 43163.3189,
+      "train_tokens_per_second": 9124.595
+    },
+    {
+      "epoch": 1.4678710737014986,
+      "grad_norm": 0.6823661782851369,
+      "learning_rate": 1.6485196373714252e-06,
+      "loss": 0.2763,
+      "num_input_tokens_seen": 394122288,
+      "step": 7150,
+      "train_runtime": 43188.6767,
+      "train_tokens_per_second": 9125.593
+    },
+    {
+      "epoch": 1.4688975569698215,
+      "grad_norm": 0.6780519303439085,
+      "learning_rate": 1.642541250845659e-06,
+      "loss": 0.2914,
+      "num_input_tokens_seen": 394415264,
+      "step": 7155,
+      "train_runtime": 43221.2025,
+      "train_tokens_per_second": 9125.504
+    },
+    {
+      "epoch": 1.4699240402381442,
+      "grad_norm": 0.6912620543142503,
+      "learning_rate": 1.6365715931108922e-06,
+      "loss": 0.2974,
+      "num_input_tokens_seen": 394662904,
+      "step": 7160,
+      "train_runtime": 43253.9417,
+      "train_tokens_per_second": 9124.322
+    },
+    {
+      "epoch": 1.4709505235064668,
+      "grad_norm": 0.5727463840608166,
+      "learning_rate": 1.6306106796871663e-06,
+      "loss": 0.2748,
+      "num_input_tokens_seen": 394919584,
+      "step": 7165,
+      "train_runtime": 43278.8304,
+      "train_tokens_per_second": 9125.006
+    },
+    {
+      "epoch": 1.4719770067747895,
+      "grad_norm": 0.73374746662879,
+      "learning_rate": 1.6246585260717873e-06,
+      "loss": 0.307,
+      "num_input_tokens_seen": 395200520,
+      "step": 7170,
+      "train_runtime": 43312.049,
+      "train_tokens_per_second": 9124.494
+    },
+    {
+      "epoch": 1.4730034900431124,
+      "grad_norm": 0.7448280753219572,
+      "learning_rate": 1.618715147739286e-06,
+      "loss": 0.2906,
+      "num_input_tokens_seen": 395458464,
+      "step": 7175,
+      "train_runtime": 43338.2977,
+      "train_tokens_per_second": 9124.919
+    },
+    {
+      "epoch": 1.474029973311435,
+      "grad_norm": 0.7130390107607212,
+      "learning_rate": 1.612780560141386e-06,
+      "loss": 0.2866,
+      "num_input_tokens_seen": 395760784,
+      "step": 7180,
+      "train_runtime": 43365.4823,
+      "train_tokens_per_second": 9126.17
+    },
+    {
+      "epoch": 1.4750564565797577,
+      "grad_norm": 0.6321551203947324,
+      "learning_rate": 1.606854778706946e-06,
+      "loss": 0.2788,
+      "num_input_tokens_seen": 396037984,
+      "step": 7185,
+      "train_runtime": 43399.9369,
+      "train_tokens_per_second": 9125.312
+    },
+    {
+      "epoch": 1.4760829398480806,
+      "grad_norm": 0.6798411421627804,
+      "learning_rate": 1.6009378188419384e-06,
+      "loss": 0.2891,
+      "num_input_tokens_seen": 396318312,
+      "step": 7190,
+      "train_runtime": 43423.6898,
+      "train_tokens_per_second": 9126.777
+    },
+    {
+      "epoch": 1.4771094231164033,
+      "grad_norm": 0.5607421519508156,
+      "learning_rate": 1.5950296959293955e-06,
+      "loss": 0.2858,
+      "num_input_tokens_seen": 396593808,
+      "step": 7195,
+      "train_runtime": 43454.8313,
+      "train_tokens_per_second": 9126.576
+    },
+    {
+      "epoch": 1.478135906384726,
+      "grad_norm": 0.7231225322146951,
+      "learning_rate": 1.5891304253293822e-06,
+      "loss": 0.2892,
+      "num_input_tokens_seen": 396859272,
+      "step": 7200,
+      "train_runtime": 43485.6921,
+      "train_tokens_per_second": 9126.203
+    },
+    {
+      "epoch": 1.4791623896530486,
+      "grad_norm": 0.6718079582542518,
+      "learning_rate": 1.5832400223789423e-06,
+      "loss": 0.2983,
+      "num_input_tokens_seen": 397149640,
+      "step": 7205,
+      "train_runtime": 43505.8413,
+      "train_tokens_per_second": 9128.651
+    },
+    {
+      "epoch": 1.4801888729213712,
+      "grad_norm": 0.6591665493912727,
+      "learning_rate": 1.577358502392069e-06,
+      "loss": 0.2917,
+      "num_input_tokens_seen": 397408072,
+      "step": 7210,
+      "train_runtime": 43534.8156,
+      "train_tokens_per_second": 9128.512
+    },
+    {
+      "epoch": 1.4812153561896941,
+      "grad_norm": 0.6333739190304549,
+      "learning_rate": 1.5714858806596584e-06,
+      "loss": 0.2724,
+      "num_input_tokens_seen": 397700176,
+      "step": 7215,
+      "train_runtime": 43568.2307,
+      "train_tokens_per_second": 9128.215
+    },
+    {
+      "epoch": 1.4822418394580168,
+      "grad_norm": 0.6047930458601233,
+      "learning_rate": 1.5656221724494758e-06,
+      "loss": 0.2848,
+      "num_input_tokens_seen": 397987176,
+      "step": 7220,
+      "train_runtime": 43600.937,
+      "train_tokens_per_second": 9127.95
+    },
+    {
+      "epoch": 1.4832683227263397,
+      "grad_norm": 0.6662038772630142,
+      "learning_rate": 1.5597673930061114e-06,
+      "loss": 0.3162,
+      "num_input_tokens_seen": 398268200,
+      "step": 7225,
+      "train_runtime": 43631.4859,
+      "train_tokens_per_second": 9128.0
+    },
+    {
+      "epoch": 1.4842948059946623,
+      "grad_norm": 0.5743637166870917,
+      "learning_rate": 1.553921557550942e-06,
+      "loss": 0.275,
+      "num_input_tokens_seen": 398538568,
+      "step": 7230,
+      "train_runtime": 43667.8407,
+      "train_tokens_per_second": 9126.592
+    },
+    {
+      "epoch": 1.485321289262985,
+      "grad_norm": 0.6434602988246784,
+      "learning_rate": 1.5480846812820927e-06,
+      "loss": 0.2891,
+      "num_input_tokens_seen": 398810016,
+      "step": 7235,
+      "train_runtime": 43693.9535,
+      "train_tokens_per_second": 9127.35
+    },
+    {
+      "epoch": 1.4863477725313077,
+      "grad_norm": 0.6537123211850155,
+      "learning_rate": 1.5422567793743948e-06,
+      "loss": 0.2859,
+      "num_input_tokens_seen": 399088408,
+      "step": 7240,
+      "train_runtime": 43719.2769,
+      "train_tokens_per_second": 9128.431
+    },
+    {
+      "epoch": 1.4873742557996303,
+      "grad_norm": 0.7054079702080568,
+      "learning_rate": 1.536437866979348e-06,
+      "loss": 0.2787,
+      "num_input_tokens_seen": 399381776,
+      "step": 7245,
+      "train_runtime": 43750.052,
+      "train_tokens_per_second": 9128.715
+    },
+    {
+      "epoch": 1.4884007390679532,
+      "grad_norm": 0.6297125036984785,
+      "learning_rate": 1.5306279592250839e-06,
+      "loss": 0.3044,
+      "num_input_tokens_seen": 399667968,
+      "step": 7250,
+      "train_runtime": 43781.4268,
+      "train_tokens_per_second": 9128.71
+    },
+    {
+      "epoch": 1.489427222336276,
+      "grad_norm": 0.6831476639313856,
+      "learning_rate": 1.5248270712163204e-06,
+      "loss": 0.2832,
+      "num_input_tokens_seen": 399946872,
+      "step": 7255,
+      "train_runtime": 43806.6412,
+      "train_tokens_per_second": 9129.823
+    },
+    {
+      "epoch": 1.4904537056045988,
+      "grad_norm": 0.6794650010106923,
+      "learning_rate": 1.5190352180343282e-06,
+      "loss": 0.2941,
+      "num_input_tokens_seen": 400231856,
+      "step": 7260,
+      "train_runtime": 43840.2688,
+      "train_tokens_per_second": 9129.32
+    },
+    {
+      "epoch": 1.4914801888729214,
+      "grad_norm": 0.771651003143777,
+      "learning_rate": 1.5132524147368826e-06,
+      "loss": 0.2897,
+      "num_input_tokens_seen": 400515888,
+      "step": 7265,
+      "train_runtime": 43872.7089,
+      "train_tokens_per_second": 9129.044
+    },
+    {
+      "epoch": 1.492506672141244,
+      "grad_norm": 0.7255959098030506,
+      "learning_rate": 1.5074786763582404e-06,
+      "loss": 0.2945,
+      "num_input_tokens_seen": 400779336,
+      "step": 7270,
+      "train_runtime": 43904.9606,
+      "train_tokens_per_second": 9128.338
+    },
+    {
+      "epoch": 1.4935331554095668,
+      "grad_norm": 0.6874675589927857,
+      "learning_rate": 1.5017140179090856e-06,
+      "loss": 0.2731,
+      "num_input_tokens_seen": 401068032,
+      "step": 7275,
+      "train_runtime": 43940.0958,
+      "train_tokens_per_second": 9127.609
+    },
+    {
+      "epoch": 1.4945596386778894,
+      "grad_norm": 0.6729392021367923,
+      "learning_rate": 1.4959584543764966e-06,
+      "loss": 0.3089,
+      "num_input_tokens_seen": 401327176,
+      "step": 7280,
+      "train_runtime": 43969.1364,
+      "train_tokens_per_second": 9127.475
+    },
+    {
+      "epoch": 1.4955861219462123,
+      "grad_norm": 0.678042651654815,
+      "learning_rate": 1.4902120007239062e-06,
+      "loss": 0.2666,
+      "num_input_tokens_seen": 401612608,
+      "step": 7285,
+      "train_runtime": 44003.1057,
+      "train_tokens_per_second": 9126.915
+    },
+    {
+      "epoch": 1.496612605214535,
+      "grad_norm": 0.6323983534914425,
+      "learning_rate": 1.4844746718910658e-06,
+      "loss": 0.2806,
+      "num_input_tokens_seen": 401901864,
+      "step": 7290,
+      "train_runtime": 44031.2469,
+      "train_tokens_per_second": 9127.651
+    },
+    {
+      "epoch": 1.4976390884828576,
+      "grad_norm": 0.7475544571414471,
+      "learning_rate": 1.4787464827940002e-06,
+      "loss": 0.2915,
+      "num_input_tokens_seen": 402179152,
+      "step": 7295,
+      "train_runtime": 44063.634,
+      "train_tokens_per_second": 9127.235
+    },
+    {
+      "epoch": 1.4986655717511805,
+      "grad_norm": 0.6942954474701847,
+      "learning_rate": 1.473027448324974e-06,
+      "loss": 0.3036,
+      "num_input_tokens_seen": 402469664,
+      "step": 7300,
+      "train_runtime": 44091.5447,
+      "train_tokens_per_second": 9128.046
+    },
+    {
+      "epoch": 1.4996920550195032,
+      "grad_norm": 0.6106990600743277,
+      "learning_rate": 1.4673175833524545e-06,
+      "loss": 0.2547,
+      "num_input_tokens_seen": 402746608,
+      "step": 7305,
+      "train_runtime": 44124.5109,
+      "train_tokens_per_second": 9127.503
+    },
+    {
+      "epoch": 1.5007185382878259,
+      "grad_norm": 0.697691256172081,
+      "learning_rate": 1.4616169027210675e-06,
+      "loss": 0.2999,
+      "num_input_tokens_seen": 403013384,
+      "step": 7310,
+      "train_runtime": 44155.6693,
+      "train_tokens_per_second": 9127.104
+    },
+    {
+      "epoch": 1.5017450215561485,
+      "grad_norm": 0.6596563332788774,
+      "learning_rate": 1.4559254212515579e-06,
+      "loss": 0.2858,
+      "num_input_tokens_seen": 403299216,
+      "step": 7315,
+      "train_runtime": 44184.9087,
+      "train_tokens_per_second": 9127.533
+    },
+    {
+      "epoch": 1.5027715048244714,
+      "grad_norm": 0.6865222457391658,
+      "learning_rate": 1.4502431537407568e-06,
+      "loss": 0.2567,
+      "num_input_tokens_seen": 403564960,
+      "step": 7320,
+      "train_runtime": 44211.7376,
+      "train_tokens_per_second": 9128.005
+    },
+    {
+      "epoch": 1.503797988092794,
+      "grad_norm": 0.6565878291950061,
+      "learning_rate": 1.444570114961545e-06,
+      "loss": 0.2779,
+      "num_input_tokens_seen": 403862512,
+      "step": 7325,
+      "train_runtime": 44243.2792,
+      "train_tokens_per_second": 9128.223
+    },
+    {
+      "epoch": 1.504824471361117,
+      "grad_norm": 0.6954331121590422,
+      "learning_rate": 1.4389063196628044e-06,
+      "loss": 0.303,
+      "num_input_tokens_seen": 404143712,
+      "step": 7330,
+      "train_runtime": 44272.0445,
+      "train_tokens_per_second": 9128.644
+    },
+    {
+      "epoch": 1.5058509546294396,
+      "grad_norm": 0.6675807645797636,
+      "learning_rate": 1.4332517825693888e-06,
+      "loss": 0.289,
+      "num_input_tokens_seen": 404433152,
+      "step": 7335,
+      "train_runtime": 44302.662,
+      "train_tokens_per_second": 9128.868
+    },
+    {
+      "epoch": 1.5068774378977623,
+      "grad_norm": 0.5951945349280648,
+      "learning_rate": 1.4276065183820803e-06,
+      "loss": 0.2834,
+      "num_input_tokens_seen": 404732368,
+      "step": 7340,
+      "train_runtime": 44327.9031,
+      "train_tokens_per_second": 9130.42
+    },
+    {
+      "epoch": 1.507903921166085,
+      "grad_norm": 0.6181049428156958,
+      "learning_rate": 1.4219705417775543e-06,
+      "loss": 0.319,
+      "num_input_tokens_seen": 405009632,
+      "step": 7345,
+      "train_runtime": 44354.4809,
+      "train_tokens_per_second": 9131.2
+    },
+    {
+      "epoch": 1.5089304044344076,
+      "grad_norm": 0.7238960733182497,
+      "learning_rate": 1.4163438674083403e-06,
+      "loss": 0.2988,
+      "num_input_tokens_seen": 405306744,
+      "step": 7350,
+      "train_runtime": 44381.1478,
+      "train_tokens_per_second": 9132.408
+    },
+    {
+      "epoch": 1.5099568877027303,
+      "grad_norm": 0.6515705227503776,
+      "learning_rate": 1.41072650990278e-06,
+      "loss": 0.2832,
+      "num_input_tokens_seen": 405577880,
+      "step": 7355,
+      "train_runtime": 44409.9761,
+      "train_tokens_per_second": 9132.585
+    },
+    {
+      "epoch": 1.5109833709710532,
+      "grad_norm": 0.7591232702649267,
+      "learning_rate": 1.4051184838650029e-06,
+      "loss": 0.2891,
+      "num_input_tokens_seen": 405874448,
+      "step": 7360,
+      "train_runtime": 44438.3101,
+      "train_tokens_per_second": 9133.436
+    },
+    {
+      "epoch": 1.512009854239376,
+      "grad_norm": 0.6533341258371985,
+      "learning_rate": 1.3995198038748652e-06,
+      "loss": 0.2615,
+      "num_input_tokens_seen": 406126496,
+      "step": 7365,
+      "train_runtime": 44468.8036,
+      "train_tokens_per_second": 9132.841
+    },
+    {
+      "epoch": 1.5130363375076987,
+      "grad_norm": 0.6501854468772995,
+      "learning_rate": 1.3939304844879348e-06,
+      "loss": 0.2848,
+      "num_input_tokens_seen": 406398656,
+      "step": 7370,
+      "train_runtime": 44497.2048,
+      "train_tokens_per_second": 9133.128
+    },
+    {
+      "epoch": 1.5140628207760214,
+      "grad_norm": 0.760469507163229,
+      "learning_rate": 1.388350540235437e-06,
+      "loss": 0.3093,
+      "num_input_tokens_seen": 406662352,
+      "step": 7375,
+      "train_runtime": 44521.8495,
+      "train_tokens_per_second": 9133.995
+    },
+    {
+      "epoch": 1.515089304044344,
+      "grad_norm": 0.8248382013374049,
+      "learning_rate": 1.3827799856242307e-06,
+      "loss": 0.2924,
+      "num_input_tokens_seen": 406948168,
+      "step": 7380,
+      "train_runtime": 44545.0959,
+      "train_tokens_per_second": 9135.645
+    },
+    {
+      "epoch": 1.5161157873126667,
+      "grad_norm": 0.6845080096783555,
+      "learning_rate": 1.3772188351367577e-06,
+      "loss": 0.2985,
+      "num_input_tokens_seen": 407226608,
+      "step": 7385,
+      "train_runtime": 44578.4724,
+      "train_tokens_per_second": 9135.051
+    },
+    {
+      "epoch": 1.5171422705809894,
+      "grad_norm": 0.5672081856534902,
+      "learning_rate": 1.3716671032310131e-06,
+      "loss": 0.2884,
+      "num_input_tokens_seen": 407498328,
+      "step": 7390,
+      "train_runtime": 44608.0149,
+      "train_tokens_per_second": 9135.092
+    },
+    {
+      "epoch": 1.5181687538493123,
+      "grad_norm": 0.745685495894887,
+      "learning_rate": 1.3661248043405046e-06,
+      "loss": 0.3186,
+      "num_input_tokens_seen": 407795408,
+      "step": 7395,
+      "train_runtime": 44637.9764,
+      "train_tokens_per_second": 9135.616
+    },
+    {
+      "epoch": 1.519195237117635,
+      "grad_norm": 0.6262001016162689,
+      "learning_rate": 1.3605919528742157e-06,
+      "loss": 0.2875,
+      "num_input_tokens_seen": 408079512,
+      "step": 7400,
+      "train_runtime": 44672.8818,
+      "train_tokens_per_second": 9134.837
+    },
+    {
+      "epoch": 1.5202217203859578,
+      "grad_norm": 0.747466073904737,
+      "learning_rate": 1.3550685632165683e-06,
+      "loss": 0.273,
+      "num_input_tokens_seen": 408356432,
+      "step": 7405,
+      "train_runtime": 44698.5822,
+      "train_tokens_per_second": 9135.78
+    },
+    {
+      "epoch": 1.5212482036542805,
+      "grad_norm": 0.6657009971692032,
+      "learning_rate": 1.3495546497273865e-06,
+      "loss": 0.3035,
+      "num_input_tokens_seen": 408622648,
+      "step": 7410,
+      "train_runtime": 44734.3365,
+      "train_tokens_per_second": 9134.43
+    },
+    {
+      "epoch": 1.5222746869226031,
+      "grad_norm": 0.6213125381256387,
+      "learning_rate": 1.3440502267418565e-06,
+      "loss": 0.2718,
+      "num_input_tokens_seen": 408912608,
+      "step": 7415,
+      "train_runtime": 44764.2801,
+      "train_tokens_per_second": 9134.797
+    },
+    {
+      "epoch": 1.5233011701909258,
+      "grad_norm": 0.6623676004944248,
+      "learning_rate": 1.3385553085704916e-06,
+      "loss": 0.2956,
+      "num_input_tokens_seen": 409165112,
+      "step": 7420,
+      "train_runtime": 44796.8565,
+      "train_tokens_per_second": 9133.791
+    },
+    {
+      "epoch": 1.5243276534592485,
+      "grad_norm": 0.607351460093954,
+      "learning_rate": 1.3330699094990924e-06,
+      "loss": 0.2903,
+      "num_input_tokens_seen": 409421216,
+      "step": 7425,
+      "train_runtime": 44824.6328,
+      "train_tokens_per_second": 9133.844
+    },
+    {
+      "epoch": 1.5253541367275714,
+      "grad_norm": 0.6301133322090412,
+      "learning_rate": 1.3275940437887176e-06,
+      "loss": 0.2849,
+      "num_input_tokens_seen": 409687832,
+      "step": 7430,
+      "train_runtime": 44848.4822,
+      "train_tokens_per_second": 9134.932
+    },
+    {
+      "epoch": 1.526380619995894,
+      "grad_norm": 0.6762168845395247,
+      "learning_rate": 1.3221277256756332e-06,
+      "loss": 0.2948,
+      "num_input_tokens_seen": 409975792,
+      "step": 7435,
+      "train_runtime": 44876.0282,
+      "train_tokens_per_second": 9135.741
+    },
+    {
+      "epoch": 1.527407103264217,
+      "grad_norm": 0.7428530038680458,
+      "learning_rate": 1.3166709693712898e-06,
+      "loss": 0.2737,
+      "num_input_tokens_seen": 410246536,
+      "step": 7440,
+      "train_runtime": 44909.2989,
+      "train_tokens_per_second": 9135.002
+    },
+    {
+      "epoch": 1.5284335865325396,
+      "grad_norm": 0.6787283408607883,
+      "learning_rate": 1.3112237890622703e-06,
+      "loss": 0.2962,
+      "num_input_tokens_seen": 410515976,
+      "step": 7445,
+      "train_runtime": 44935.6838,
+      "train_tokens_per_second": 9135.634
+    },
+    {
+      "epoch": 1.5294600698008622,
+      "grad_norm": 0.6357393102906406,
+      "learning_rate": 1.3057861989102722e-06,
+      "loss": 0.2873,
+      "num_input_tokens_seen": 410824120,
+      "step": 7450,
+      "train_runtime": 44967.6203,
+      "train_tokens_per_second": 9135.999
+    },
+    {
+      "epoch": 1.530486553069185,
+      "grad_norm": 0.7300732398337838,
+      "learning_rate": 1.3003582130520536e-06,
+      "loss": 0.2863,
+      "num_input_tokens_seen": 411098272,
+      "step": 7455,
+      "train_runtime": 45000.2262,
+      "train_tokens_per_second": 9135.471
+    },
+    {
+      "epoch": 1.5315130363375076,
+      "grad_norm": 1.0018082331061282,
+      "learning_rate": 1.294939845599405e-06,
+      "loss": 0.3011,
+      "num_input_tokens_seen": 411383728,
+      "step": 7460,
+      "train_runtime": 45028.8928,
+      "train_tokens_per_second": 9135.995
+    },
+    {
+      "epoch": 1.5325395196058305,
+      "grad_norm": 0.7169308036428433,
+      "learning_rate": 1.2895311106391106e-06,
+      "loss": 0.3058,
+      "num_input_tokens_seen": 411665864,
+      "step": 7465,
+      "train_runtime": 45063.1486,
+      "train_tokens_per_second": 9135.311
+    },
+    {
+      "epoch": 1.5335660028741531,
+      "grad_norm": 0.6071560096467135,
+      "learning_rate": 1.2841320222329119e-06,
+      "loss": 0.2694,
+      "num_input_tokens_seen": 411966728,
+      "step": 7470,
+      "train_runtime": 45093.1274,
+      "train_tokens_per_second": 9135.909
+    },
+    {
+      "epoch": 1.534592486142476,
+      "grad_norm": 0.7963109557127629,
+      "learning_rate": 1.2787425944174708e-06,
+      "loss": 0.3266,
+      "num_input_tokens_seen": 412258352,
+      "step": 7475,
+      "train_runtime": 45127.0651,
+      "train_tokens_per_second": 9135.501
+    },
+    {
+      "epoch": 1.5356189694107987,
+      "grad_norm": 0.8718228220982424,
+      "learning_rate": 1.273362841204333e-06,
+      "loss": 0.2841,
+      "num_input_tokens_seen": 412484880,
+      "step": 7480,
+      "train_runtime": 45157.4761,
+      "train_tokens_per_second": 9134.365
+    },
+    {
+      "epoch": 1.5366454526791213,
+      "grad_norm": 0.7593790307620355,
+      "learning_rate": 1.2679927765798949e-06,
+      "loss": 0.2674,
+      "num_input_tokens_seen": 412707992,
+      "step": 7485,
+      "train_runtime": 45183.8192,
+      "train_tokens_per_second": 9133.978
+    },
+    {
+      "epoch": 1.537671935947444,
+      "grad_norm": 0.6399264889344344,
+      "learning_rate": 1.2626324145053626e-06,
+      "loss": 0.2711,
+      "num_input_tokens_seen": 412966704,
+      "step": 7490,
+      "train_runtime": 45215.628,
+      "train_tokens_per_second": 9133.274
+    },
+    {
+      "epoch": 1.5386984192157667,
+      "grad_norm": 0.8056098675754333,
+      "learning_rate": 1.2572817689167132e-06,
+      "loss": 0.282,
+      "num_input_tokens_seen": 413239856,
+      "step": 7495,
+      "train_runtime": 45246.4836,
+      "train_tokens_per_second": 9133.082
+    },
+    {
+      "epoch": 1.5397249024840896,
+      "grad_norm": 0.7463330712574192,
+      "learning_rate": 1.2519408537246664e-06,
+      "loss": 0.3112,
+      "num_input_tokens_seen": 413506760,
+      "step": 7500,
+      "train_runtime": 45277.5406,
+      "train_tokens_per_second": 9132.712
+    },
+    {
+      "epoch": 1.5407513857524122,
+      "grad_norm": 0.5671827605283214,
+      "learning_rate": 1.246609682814648e-06,
+      "loss": 0.2696,
+      "num_input_tokens_seen": 413794072,
+      "step": 7505,
+      "train_runtime": 45305.5227,
+      "train_tokens_per_second": 9133.413
+    },
+    {
+      "epoch": 1.541777869020735,
+      "grad_norm": 0.610253810639789,
+      "learning_rate": 1.2412882700467442e-06,
+      "loss": 0.3026,
+      "num_input_tokens_seen": 414070704,
+      "step": 7510,
+      "train_runtime": 45340.6895,
+      "train_tokens_per_second": 9132.431
+    },
+    {
+      "epoch": 1.5428043522890578,
+      "grad_norm": 0.6110731981285108,
+      "learning_rate": 1.2359766292556764e-06,
+      "loss": 0.2707,
+      "num_input_tokens_seen": 414329688,
+      "step": 7515,
+      "train_runtime": 45370.78,
+      "train_tokens_per_second": 9132.082
+    },
+    {
+      "epoch": 1.5438308355573804,
+      "grad_norm": 0.6294527543350031,
+      "learning_rate": 1.2306747742507564e-06,
+      "loss": 0.2918,
+      "num_input_tokens_seen": 414599624,
+      "step": 7520,
+      "train_runtime": 45400.5095,
+      "train_tokens_per_second": 9132.048
+    },
+    {
+      "epoch": 1.544857318825703,
+      "grad_norm": 0.6404775180985235,
+      "learning_rate": 1.2253827188158585e-06,
+      "loss": 0.2913,
+      "num_input_tokens_seen": 414881456,
+      "step": 7525,
+      "train_runtime": 45431.3573,
+      "train_tokens_per_second": 9132.051
+    },
+    {
+      "epoch": 1.5458838020940258,
+      "grad_norm": 0.9663690933647143,
+      "learning_rate": 1.2201004767093782e-06,
+      "loss": 0.2746,
+      "num_input_tokens_seen": 415140992,
+      "step": 7530,
+      "train_runtime": 45456.5814,
+      "train_tokens_per_second": 9132.693
+    },
+    {
+      "epoch": 1.5469102853623486,
+      "grad_norm": 0.6835120514933375,
+      "learning_rate": 1.214828061664196e-06,
+      "loss": 0.2804,
+      "num_input_tokens_seen": 415396064,
+      "step": 7535,
+      "train_runtime": 45493.7218,
+      "train_tokens_per_second": 9130.844
+    },
+    {
+      "epoch": 1.5479367686306713,
+      "grad_norm": 0.8871917935207436,
+      "learning_rate": 1.2095654873876517e-06,
+      "loss": 0.2596,
+      "num_input_tokens_seen": 415664960,
+      "step": 7540,
+      "train_runtime": 45525.6993,
+      "train_tokens_per_second": 9130.337
+    },
+    {
+      "epoch": 1.5489632518989942,
+      "grad_norm": 0.7396694687859762,
+      "learning_rate": 1.2043127675614891e-06,
+      "loss": 0.2943,
+      "num_input_tokens_seen": 415948552,
+      "step": 7545,
+      "train_runtime": 45563.8744,
+      "train_tokens_per_second": 9128.911
+    },
+    {
+      "epoch": 1.5499897351673169,
+      "grad_norm": 0.5860810806276339,
+      "learning_rate": 1.1990699158418401e-06,
+      "loss": 0.2705,
+      "num_input_tokens_seen": 416205144,
+      "step": 7550,
+      "train_runtime": 45593.3677,
+      "train_tokens_per_second": 9128.634
+    },
+    {
+      "epoch": 1.5510162184356395,
+      "grad_norm": 0.6461918272252107,
+      "learning_rate": 1.1938369458591815e-06,
+      "loss": 0.2794,
+      "num_input_tokens_seen": 416486104,
+      "step": 7555,
+      "train_runtime": 45622.2351,
+      "train_tokens_per_second": 9129.016
+    },
+    {
+      "epoch": 1.5520427017039622,
+      "grad_norm": 0.8803184034014958,
+      "learning_rate": 1.188613871218296e-06,
+      "loss": 0.2858,
+      "num_input_tokens_seen": 416773456,
+      "step": 7560,
+      "train_runtime": 45648.9663,
+      "train_tokens_per_second": 9129.965
+    },
+    {
+      "epoch": 1.5530691849722849,
+      "grad_norm": 0.6865619406390884,
+      "learning_rate": 1.1834007054982416e-06,
+      "loss": 0.2658,
+      "num_input_tokens_seen": 417019176,
+      "step": 7565,
+      "train_runtime": 45676.1165,
+      "train_tokens_per_second": 9129.918
+    },
+    {
+      "epoch": 1.5540956682406075,
+      "grad_norm": 0.7609924611155514,
+      "learning_rate": 1.1781974622523145e-06,
+      "loss": 0.2973,
+      "num_input_tokens_seen": 417299640,
+      "step": 7570,
+      "train_runtime": 45703.1487,
+      "train_tokens_per_second": 9130.654
+    },
+    {
+      "epoch": 1.5551221515089304,
+      "grad_norm": 0.5974500263506823,
+      "learning_rate": 1.1730041550080152e-06,
+      "loss": 0.2952,
+      "num_input_tokens_seen": 417585264,
+      "step": 7575,
+      "train_runtime": 45731.6387,
+      "train_tokens_per_second": 9131.211
+    },
+    {
+      "epoch": 1.5561486347772533,
+      "grad_norm": 0.7050421143561159,
+      "learning_rate": 1.1678207972670113e-06,
+      "loss": 0.2647,
+      "num_input_tokens_seen": 417855568,
+      "step": 7580,
+      "train_runtime": 45763.301,
+      "train_tokens_per_second": 9130.8
+    },
+    {
+      "epoch": 1.557175118045576,
+      "grad_norm": 0.7268841316283392,
+      "learning_rate": 1.1626474025051043e-06,
+      "loss": 0.2943,
+      "num_input_tokens_seen": 418127152,
+      "step": 7585,
+      "train_runtime": 45790.9428,
+      "train_tokens_per_second": 9131.22
+    },
+    {
+      "epoch": 1.5582016013138986,
+      "grad_norm": 0.6827967339271552,
+      "learning_rate": 1.157483984172193e-06,
+      "loss": 0.3175,
+      "num_input_tokens_seen": 418401488,
+      "step": 7590,
+      "train_runtime": 45821.4933,
+      "train_tokens_per_second": 9131.119
+    },
+    {
+      "epoch": 1.5592280845822213,
+      "grad_norm": 0.652945857130692,
+      "learning_rate": 1.1523305556922397e-06,
+      "loss": 0.2888,
+      "num_input_tokens_seen": 418696744,
+      "step": 7595,
+      "train_runtime": 45851.0789,
+      "train_tokens_per_second": 9131.666
+    },
+    {
+      "epoch": 1.560254567850544,
+      "grad_norm": 0.7216848662554934,
+      "learning_rate": 1.1471871304632354e-06,
+      "loss": 0.2726,
+      "num_input_tokens_seen": 418978280,
+      "step": 7600,
+      "train_runtime": 45878.1077,
+      "train_tokens_per_second": 9132.423
+    },
+    {
+      "epoch": 1.5612810511188666,
+      "grad_norm": 0.635787028208562,
+      "learning_rate": 1.1420537218571615e-06,
+      "loss": 0.2788,
+      "num_input_tokens_seen": 419267176,
+      "step": 7605,
+      "train_runtime": 45911.7697,
+      "train_tokens_per_second": 9132.019
+    },
+    {
+      "epoch": 1.5623075343871895,
+      "grad_norm": 0.6281856112193759,
+      "learning_rate": 1.136930343219964e-06,
+      "loss": 0.3103,
+      "num_input_tokens_seen": 419530832,
+      "step": 7610,
+      "train_runtime": 45940.7278,
+      "train_tokens_per_second": 9132.002
+    },
+    {
+      "epoch": 1.5633340176555122,
+      "grad_norm": 0.7696293915949742,
+      "learning_rate": 1.1318170078715068e-06,
+      "loss": 0.2952,
+      "num_input_tokens_seen": 419811048,
+      "step": 7615,
+      "train_runtime": 45977.237,
+      "train_tokens_per_second": 9130.846
+    },
+    {
+      "epoch": 1.564360500923835,
+      "grad_norm": 0.7498776893409057,
+      "learning_rate": 1.126713729105548e-06,
+      "loss": 0.2714,
+      "num_input_tokens_seen": 420087888,
+      "step": 7620,
+      "train_runtime": 46007.3059,
+      "train_tokens_per_second": 9130.895
+    },
+    {
+      "epoch": 1.5653869841921577,
+      "grad_norm": 0.6014785643871569,
+      "learning_rate": 1.1216205201896923e-06,
+      "loss": 0.2793,
+      "num_input_tokens_seen": 420371392,
+      "step": 7625,
+      "train_runtime": 46031.9309,
+      "train_tokens_per_second": 9132.169
+    },
+    {
+      "epoch": 1.5664134674604804,
+      "grad_norm": 0.7835795721382438,
+      "learning_rate": 1.1165373943653746e-06,
+      "loss": 0.3008,
+      "num_input_tokens_seen": 420647688,
+      "step": 7630,
+      "train_runtime": 46057.2504,
+      "train_tokens_per_second": 9133.148
+    },
+    {
+      "epoch": 1.567439950728803,
+      "grad_norm": 0.6562423595338617,
+      "learning_rate": 1.11146436484781e-06,
+      "loss": 0.28,
+      "num_input_tokens_seen": 420909536,
+      "step": 7635,
+      "train_runtime": 46086.4271,
+      "train_tokens_per_second": 9133.048
+    },
+    {
+      "epoch": 1.5684664339971257,
+      "grad_norm": 0.6885232508918427,
+      "learning_rate": 1.1064014448259658e-06,
+      "loss": 0.2808,
+      "num_input_tokens_seen": 421178536,
+      "step": 7640,
+      "train_runtime": 46120.6551,
+      "train_tokens_per_second": 9132.102
+    },
+    {
+      "epoch": 1.5694929172654486,
+      "grad_norm": 0.7463895353826966,
+      "learning_rate": 1.1013486474625263e-06,
+      "loss": 0.2823,
+      "num_input_tokens_seen": 421434176,
+      "step": 7645,
+      "train_runtime": 46152.6435,
+      "train_tokens_per_second": 9131.312
+    },
+    {
+      "epoch": 1.5705194005337713,
+      "grad_norm": 0.7142014267513574,
+      "learning_rate": 1.0963059858938597e-06,
+      "loss": 0.2809,
+      "num_input_tokens_seen": 421708360,
+      "step": 7650,
+      "train_runtime": 46183.5797,
+      "train_tokens_per_second": 9131.132
+    },
+    {
+      "epoch": 1.5715458838020941,
+      "grad_norm": 0.6696257580210784,
+      "learning_rate": 1.0912734732299818e-06,
+      "loss": 0.2726,
+      "num_input_tokens_seen": 421969016,
+      "step": 7655,
+      "train_runtime": 46208.6964,
+      "train_tokens_per_second": 9131.81
+    },
+    {
+      "epoch": 1.5725723670704168,
+      "grad_norm": 0.6629706500315824,
+      "learning_rate": 1.0862511225545224e-06,
+      "loss": 0.2555,
+      "num_input_tokens_seen": 422247864,
+      "step": 7660,
+      "train_runtime": 46242.382,
+      "train_tokens_per_second": 9131.188
+    },
+    {
+      "epoch": 1.5735988503387395,
+      "grad_norm": 0.7234100780782806,
+      "learning_rate": 1.0812389469246965e-06,
+      "loss": 0.3007,
+      "num_input_tokens_seen": 422526328,
+      "step": 7665,
+      "train_runtime": 46274.993,
+      "train_tokens_per_second": 9130.77
+    },
+    {
+      "epoch": 1.5746253336070621,
+      "grad_norm": 0.8571272779539639,
+      "learning_rate": 1.0762369593712608e-06,
+      "loss": 0.301,
+      "num_input_tokens_seen": 422825536,
+      "step": 7670,
+      "train_runtime": 46312.3781,
+      "train_tokens_per_second": 9129.86
+    },
+    {
+      "epoch": 1.5756518168753848,
+      "grad_norm": 0.6746183243822663,
+      "learning_rate": 1.0712451728984836e-06,
+      "loss": 0.3189,
+      "num_input_tokens_seen": 423113704,
+      "step": 7675,
+      "train_runtime": 46335.1516,
+      "train_tokens_per_second": 9131.592
+    },
+    {
+      "epoch": 1.5766783001437077,
+      "grad_norm": 0.635023138746647,
+      "learning_rate": 1.0662636004841187e-06,
+      "loss": 0.2722,
+      "num_input_tokens_seen": 423400208,
+      "step": 7680,
+      "train_runtime": 46364.6469,
+      "train_tokens_per_second": 9131.962
+    },
+    {
+      "epoch": 1.5777047834120304,
+      "grad_norm": 0.653589041007687,
+      "learning_rate": 1.061292255079361e-06,
+      "loss": 0.296,
+      "num_input_tokens_seen": 423698072,
+      "step": 7685,
+      "train_runtime": 46400.4465,
+      "train_tokens_per_second": 9131.336
+    },
+    {
+      "epoch": 1.5787312666803532,
+      "grad_norm": 0.5988304289522718,
+      "learning_rate": 1.0563311496088157e-06,
+      "loss": 0.2934,
+      "num_input_tokens_seen": 423985424,
+      "step": 7690,
+      "train_runtime": 46427.1325,
+      "train_tokens_per_second": 9132.277
+    },
+    {
+      "epoch": 1.579757749948676,
+      "grad_norm": 0.5992531912457367,
+      "learning_rate": 1.05138029697047e-06,
+      "loss": 0.2765,
+      "num_input_tokens_seen": 424283024,
+      "step": 7695,
+      "train_runtime": 46460.0556,
+      "train_tokens_per_second": 9132.211
+    },
+    {
+      "epoch": 1.5807842332169986,
+      "grad_norm": 0.7127941579259701,
+      "learning_rate": 1.0464397100356521e-06,
+      "loss": 0.3176,
+      "num_input_tokens_seen": 424534640,
+      "step": 7700,
+      "train_runtime": 46489.4674,
+      "train_tokens_per_second": 9131.846
+    },
+    {
+      "epoch": 1.5818107164853212,
+      "grad_norm": 0.6413871963574115,
+      "learning_rate": 1.041509401649003e-06,
+      "loss": 0.2943,
+      "num_input_tokens_seen": 424834360,
+      "step": 7705,
+      "train_runtime": 46520.2796,
+      "train_tokens_per_second": 9132.24
+    },
+    {
+      "epoch": 1.582837199753644,
+      "grad_norm": 0.67055467624613,
+      "learning_rate": 1.0365893846284408e-06,
+      "loss": 0.2889,
+      "num_input_tokens_seen": 425085488,
+      "step": 7710,
+      "train_runtime": 46548.7716,
+      "train_tokens_per_second": 9132.045
+    },
+    {
+      "epoch": 1.5838636830219668,
+      "grad_norm": 0.7085206116169199,
+      "learning_rate": 1.0316796717651262e-06,
+      "loss": 0.2933,
+      "num_input_tokens_seen": 425361200,
+      "step": 7715,
+      "train_runtime": 46573.7322,
+      "train_tokens_per_second": 9133.071
+    },
+    {
+      "epoch": 1.5848901662902894,
+      "grad_norm": 0.701998763359103,
+      "learning_rate": 1.0267802758234373e-06,
+      "loss": 0.3068,
+      "num_input_tokens_seen": 425612808,
+      "step": 7720,
+      "train_runtime": 46606.5511,
+      "train_tokens_per_second": 9132.038
+    },
+    {
+      "epoch": 1.5859166495586123,
+      "grad_norm": 0.6195614490320093,
+      "learning_rate": 1.0218912095409217e-06,
+      "loss": 0.2711,
+      "num_input_tokens_seen": 425899664,
+      "step": 7725,
+      "train_runtime": 46636.4559,
+      "train_tokens_per_second": 9132.333
+    },
+    {
+      "epoch": 1.586943132826935,
+      "grad_norm": 0.6820866237121077,
+      "learning_rate": 1.0170124856282749e-06,
+      "loss": 0.2752,
+      "num_input_tokens_seen": 426181048,
+      "step": 7730,
+      "train_runtime": 46668.5211,
+      "train_tokens_per_second": 9132.088
+    },
+    {
+      "epoch": 1.5879696160952577,
+      "grad_norm": 0.7155183668129014,
+      "learning_rate": 1.0121441167693068e-06,
+      "loss": 0.3029,
+      "num_input_tokens_seen": 426442784,
+      "step": 7735,
+      "train_runtime": 46697.9369,
+      "train_tokens_per_second": 9131.941
+    },
+    {
+      "epoch": 1.5889960993635803,
+      "grad_norm": 0.667351430531221,
+      "learning_rate": 1.0072861156209035e-06,
+      "loss": 0.3108,
+      "num_input_tokens_seen": 426739120,
+      "step": 7740,
+      "train_runtime": 46728.4194,
+      "train_tokens_per_second": 9132.325
+    },
+    {
+      "epoch": 1.590022582631903,
+      "grad_norm": 0.7092732205185082,
+      "learning_rate": 1.0024384948129973e-06,
+      "loss": 0.2805,
+      "num_input_tokens_seen": 427018264,
+      "step": 7745,
+      "train_runtime": 46761.2219,
+      "train_tokens_per_second": 9131.888
+    },
+    {
+      "epoch": 1.5910490659002259,
+      "grad_norm": 0.6396225133892764,
+      "learning_rate": 9.97601266948533e-07,
+      "loss": 0.2856,
+      "num_input_tokens_seen": 427303784,
+      "step": 7750,
+      "train_runtime": 46796.5243,
+      "train_tokens_per_second": 9131.101
+    },
+    {
+      "epoch": 1.5920755491685485,
+      "grad_norm": 0.6833711259852852,
+      "learning_rate": 9.92774444603436e-07,
+      "loss": 0.2999,
+      "num_input_tokens_seen": 427582808,
+      "step": 7755,
+      "train_runtime": 46826.1284,
+      "train_tokens_per_second": 9131.287
+    },
+    {
+      "epoch": 1.5931020324368714,
+      "grad_norm": 0.6642999537387282,
+      "learning_rate": 9.879580403265799e-07,
+      "loss": 0.2975,
+      "num_input_tokens_seen": 427862504,
+      "step": 7760,
+      "train_runtime": 46855.1736,
+      "train_tokens_per_second": 9131.596
+    },
+    {
+      "epoch": 1.594128515705194,
+      "grad_norm": 0.6291855825027791,
+      "learning_rate": 9.83152066639751e-07,
+      "loss": 0.2954,
+      "num_input_tokens_seen": 428129648,
+      "step": 7765,
+      "train_runtime": 46890.3515,
+      "train_tokens_per_second": 9130.442
+    },
+    {
+      "epoch": 1.5951549989735168,
+      "grad_norm": 0.659975213918398,
+      "learning_rate": 9.78356536037623e-07,
+      "loss": 0.2879,
+      "num_input_tokens_seen": 428433760,
+      "step": 7770,
+      "train_runtime": 46922.9289,
+      "train_tokens_per_second": 9130.584
+    },
+    {
+      "epoch": 1.5961814822418394,
+      "grad_norm": 0.6276112722267811,
+      "learning_rate": 9.735714609877127e-07,
+      "loss": 0.2941,
+      "num_input_tokens_seen": 428683880,
+      "step": 7775,
+      "train_runtime": 46951.3912,
+      "train_tokens_per_second": 9130.377
+    },
+    {
+      "epoch": 1.597207965510162,
+      "grad_norm": 0.7685645679655322,
+      "learning_rate": 9.687968539303588e-07,
+      "loss": 0.2901,
+      "num_input_tokens_seen": 428954296,
+      "step": 7780,
+      "train_runtime": 46981.0415,
+      "train_tokens_per_second": 9130.37
+    },
+    {
+      "epoch": 1.5982344487784848,
+      "grad_norm": 0.6777392881013219,
+      "learning_rate": 9.640327272786826e-07,
+      "loss": 0.2946,
+      "num_input_tokens_seen": 429237280,
+      "step": 7785,
+      "train_runtime": 47015.7453,
+      "train_tokens_per_second": 9129.65
+    },
+    {
+      "epoch": 1.5992609320468076,
+      "grad_norm": 0.6556337778748671,
+      "learning_rate": 9.592790934185615e-07,
+      "loss": 0.2807,
+      "num_input_tokens_seen": 429530112,
+      "step": 7790,
+      "train_runtime": 47049.0803,
+      "train_tokens_per_second": 9129.405
+    },
+    {
+      "epoch": 1.6002874153151305,
+      "grad_norm": 0.7413916248292691,
+      "learning_rate": 9.54535964708591e-07,
+      "loss": 0.2742,
+      "num_input_tokens_seen": 429809464,
+      "step": 7795,
+      "train_runtime": 47079.4428,
+      "train_tokens_per_second": 9129.451
+    },
+    {
+      "epoch": 1.6013138985834532,
+      "grad_norm": 0.6246280573020946,
+      "learning_rate": 9.498033534800555e-07,
+      "loss": 0.2903,
+      "num_input_tokens_seen": 430107872,
+      "step": 7800,
+      "train_runtime": 47103.9795,
+      "train_tokens_per_second": 9131.03
+    },
+    {
+      "epoch": 1.6023403818517759,
+      "grad_norm": 0.6845471433015657,
+      "learning_rate": 9.450812720368957e-07,
+      "loss": 0.2911,
+      "num_input_tokens_seen": 430389416,
+      "step": 7805,
+      "train_runtime": 47136.5258,
+      "train_tokens_per_second": 9130.699
+    },
+    {
+      "epoch": 1.6033668651200985,
+      "grad_norm": 0.6263880259404115,
+      "learning_rate": 9.403697326556776e-07,
+      "loss": 0.2874,
+      "num_input_tokens_seen": 430644256,
+      "step": 7810,
+      "train_runtime": 47165.7763,
+      "train_tokens_per_second": 9130.439
+    },
+    {
+      "epoch": 1.6043933483884212,
+      "grad_norm": 0.7054172504898565,
+      "learning_rate": 9.356687475855586e-07,
+      "loss": 0.3001,
+      "num_input_tokens_seen": 430885776,
+      "step": 7815,
+      "train_runtime": 47194.1629,
+      "train_tokens_per_second": 9130.065
+    },
+    {
+      "epoch": 1.6054198316567438,
+      "grad_norm": 0.6372109271664709,
+      "learning_rate": 9.309783290482572e-07,
+      "loss": 0.289,
+      "num_input_tokens_seen": 431171512,
+      "step": 7820,
+      "train_runtime": 47225.0954,
+      "train_tokens_per_second": 9130.135
+    },
+    {
+      "epoch": 1.6064463149250667,
+      "grad_norm": 0.7105081361259638,
+      "learning_rate": 9.262984892380217e-07,
+      "loss": 0.2653,
+      "num_input_tokens_seen": 431462264,
+      "step": 7825,
+      "train_runtime": 47251.037,
+      "train_tokens_per_second": 9131.276
+    },
+    {
+      "epoch": 1.6074727981933894,
+      "grad_norm": 0.6906681900048771,
+      "learning_rate": 9.216292403215965e-07,
+      "loss": 0.281,
+      "num_input_tokens_seen": 431745752,
+      "step": 7830,
+      "train_runtime": 47278.5351,
+      "train_tokens_per_second": 9131.961
+    },
+    {
+      "epoch": 1.6084992814617123,
+      "grad_norm": 0.7910618935887346,
+      "learning_rate": 9.169705944381912e-07,
+      "loss": 0.2606,
+      "num_input_tokens_seen": 432013224,
+      "step": 7835,
+      "train_runtime": 47304.1879,
+      "train_tokens_per_second": 9132.663
+    },
+    {
+      "epoch": 1.609525764730035,
+      "grad_norm": 0.6349912117271084,
+      "learning_rate": 9.123225636994498e-07,
+      "loss": 0.2778,
+      "num_input_tokens_seen": 432316936,
+      "step": 7840,
+      "train_runtime": 47335.5949,
+      "train_tokens_per_second": 9133.02
+    },
+    {
+      "epoch": 1.6105522479983576,
+      "grad_norm": 0.7013367973864304,
+      "learning_rate": 9.07685160189422e-07,
+      "loss": 0.2777,
+      "num_input_tokens_seen": 432603712,
+      "step": 7845,
+      "train_runtime": 47369.828,
+      "train_tokens_per_second": 9132.474
+    },
+    {
+      "epoch": 1.6115787312666803,
+      "grad_norm": 0.6357920966645663,
+      "learning_rate": 9.030583959645245e-07,
+      "loss": 0.2877,
+      "num_input_tokens_seen": 432898328,
+      "step": 7850,
+      "train_runtime": 47397.2859,
+      "train_tokens_per_second": 9133.399
+    },
+    {
+      "epoch": 1.612605214535003,
+      "grad_norm": 0.6747847841923503,
+      "learning_rate": 8.984422830535123e-07,
+      "loss": 0.2988,
+      "num_input_tokens_seen": 433143976,
+      "step": 7855,
+      "train_runtime": 47427.1507,
+      "train_tokens_per_second": 9132.827
+    },
+    {
+      "epoch": 1.6136316978033258,
+      "grad_norm": 0.6649496316328591,
+      "learning_rate": 8.938368334574549e-07,
+      "loss": 0.2689,
+      "num_input_tokens_seen": 433438488,
+      "step": 7860,
+      "train_runtime": 47457.2268,
+      "train_tokens_per_second": 9133.245
+    },
+    {
+      "epoch": 1.6146581810716485,
+      "grad_norm": 0.6584989534422363,
+      "learning_rate": 8.892420591496947e-07,
+      "loss": 0.2938,
+      "num_input_tokens_seen": 433703800,
+      "step": 7865,
+      "train_runtime": 47492.9131,
+      "train_tokens_per_second": 9131.969
+    },
+    {
+      "epoch": 1.6156846643399714,
+      "grad_norm": 0.7494438051549067,
+      "learning_rate": 8.846579720758219e-07,
+      "loss": 0.3274,
+      "num_input_tokens_seen": 433984720,
+      "step": 7870,
+      "train_runtime": 47522.9875,
+      "train_tokens_per_second": 9132.101
+    },
+    {
+      "epoch": 1.616711147608294,
+      "grad_norm": 0.7217493784526718,
+      "learning_rate": 8.800845841536404e-07,
+      "loss": 0.2757,
+      "num_input_tokens_seen": 434262432,
+      "step": 7875,
+      "train_runtime": 47551.7288,
+      "train_tokens_per_second": 9132.422
+    },
+    {
+      "epoch": 1.6177376308766167,
+      "grad_norm": 0.743002018056527,
+      "learning_rate": 8.755219072731397e-07,
+      "loss": 0.2814,
+      "num_input_tokens_seen": 434506824,
+      "step": 7880,
+      "train_runtime": 47575.3374,
+      "train_tokens_per_second": 9133.027
+    },
+    {
+      "epoch": 1.6187641141449394,
+      "grad_norm": 0.7710289761954622,
+      "learning_rate": 8.709699532964616e-07,
+      "loss": 0.3216,
+      "num_input_tokens_seen": 434794272,
+      "step": 7885,
+      "train_runtime": 47607.0023,
+      "train_tokens_per_second": 9132.99
+    },
+    {
+      "epoch": 1.619790597413262,
+      "grad_norm": 0.767427867118848,
+      "learning_rate": 8.664287340578697e-07,
+      "loss": 0.2809,
+      "num_input_tokens_seen": 435028440,
+      "step": 7890,
+      "train_runtime": 47632.4701,
+      "train_tokens_per_second": 9133.023
+    },
+    {
+      "epoch": 1.620817080681585,
+      "grad_norm": 0.6727969091123991,
+      "learning_rate": 8.618982613637222e-07,
+      "loss": 0.2752,
+      "num_input_tokens_seen": 435315456,
+      "step": 7895,
+      "train_runtime": 47662.5161,
+      "train_tokens_per_second": 9133.287
+    },
+    {
+      "epoch": 1.6218435639499076,
+      "grad_norm": 0.6387244602751139,
+      "learning_rate": 8.573785469924367e-07,
+      "loss": 0.3075,
+      "num_input_tokens_seen": 435605128,
+      "step": 7900,
+      "train_runtime": 47690.8837,
+      "train_tokens_per_second": 9133.929
+    },
+    {
+      "epoch": 1.6228700472182305,
+      "grad_norm": 0.6385097555377032,
+      "learning_rate": 8.528696026944577e-07,
+      "loss": 0.2603,
+      "num_input_tokens_seen": 435886552,
+      "step": 7905,
+      "train_runtime": 47724.7966,
+      "train_tokens_per_second": 9133.335
+    },
+    {
+      "epoch": 1.6238965304865531,
+      "grad_norm": 0.746422237940599,
+      "learning_rate": 8.483714401922322e-07,
+      "loss": 0.2629,
+      "num_input_tokens_seen": 436165320,
+      "step": 7910,
+      "train_runtime": 47756.4585,
+      "train_tokens_per_second": 9133.117
+    },
+    {
+      "epoch": 1.6249230137548758,
+      "grad_norm": 0.6518407334785854,
+      "learning_rate": 8.438840711801794e-07,
+      "loss": 0.2916,
+      "num_input_tokens_seen": 436432248,
+      "step": 7915,
+      "train_runtime": 47789.0285,
+      "train_tokens_per_second": 9132.478
+    },
+    {
+      "epoch": 1.6259494970231985,
+      "grad_norm": 0.6765652330372787,
+      "learning_rate": 8.394075073246521e-07,
+      "loss": 0.299,
+      "num_input_tokens_seen": 436706488,
+      "step": 7920,
+      "train_runtime": 47821.4726,
+      "train_tokens_per_second": 9132.017
+    },
+    {
+      "epoch": 1.6269759802915211,
+      "grad_norm": 0.6499901084983544,
+      "learning_rate": 8.349417602639154e-07,
+      "loss": 0.2886,
+      "num_input_tokens_seen": 436985328,
+      "step": 7925,
+      "train_runtime": 47857.5988,
+      "train_tokens_per_second": 9130.95
+    },
+    {
+      "epoch": 1.628002463559844,
+      "grad_norm": 0.6856278515193532,
+      "learning_rate": 8.304868416081102e-07,
+      "loss": 0.2951,
+      "num_input_tokens_seen": 437243424,
+      "step": 7930,
+      "train_runtime": 47888.2433,
+      "train_tokens_per_second": 9130.496
+    },
+    {
+      "epoch": 1.6290289468281667,
+      "grad_norm": 0.6904519790327983,
+      "learning_rate": 8.26042762939227e-07,
+      "loss": 0.2902,
+      "num_input_tokens_seen": 437518192,
+      "step": 7935,
+      "train_runtime": 47913.289,
+      "train_tokens_per_second": 9131.458
+    },
+    {
+      "epoch": 1.6300554300964896,
+      "grad_norm": 0.746560674064675,
+      "learning_rate": 8.216095358110742e-07,
+      "loss": 0.2931,
+      "num_input_tokens_seen": 437803256,
+      "step": 7940,
+      "train_runtime": 47949.3687,
+      "train_tokens_per_second": 9130.532
+    },
+    {
+      "epoch": 1.6310819133648122,
+      "grad_norm": 0.6277597301075255,
+      "learning_rate": 8.171871717492463e-07,
+      "loss": 0.2616,
+      "num_input_tokens_seen": 438085048,
+      "step": 7945,
+      "train_runtime": 47981.8835,
+      "train_tokens_per_second": 9130.218
+    },
+    {
+      "epoch": 1.632108396633135,
+      "grad_norm": 0.8435459152615113,
+      "learning_rate": 8.127756822511001e-07,
+      "loss": 0.2894,
+      "num_input_tokens_seen": 438323144,
+      "step": 7950,
+      "train_runtime": 48012.5099,
+      "train_tokens_per_second": 9129.353
+    },
+    {
+      "epoch": 1.6331348799014576,
+      "grad_norm": 0.6523098768756739,
+      "learning_rate": 8.083750787857153e-07,
+      "loss": 0.2926,
+      "num_input_tokens_seen": 438599088,
+      "step": 7955,
+      "train_runtime": 48041.3463,
+      "train_tokens_per_second": 9129.617
+    },
+    {
+      "epoch": 1.6341613631697802,
+      "grad_norm": 0.7325812892976818,
+      "learning_rate": 8.039853727938729e-07,
+      "loss": 0.2898,
+      "num_input_tokens_seen": 438878856,
+      "step": 7960,
+      "train_runtime": 48065.4144,
+      "train_tokens_per_second": 9130.866
+    },
+    {
+      "epoch": 1.635187846438103,
+      "grad_norm": 0.7135882987866788,
+      "learning_rate": 7.9960657568802e-07,
+      "loss": 0.2803,
+      "num_input_tokens_seen": 439146056,
+      "step": 7965,
+      "train_runtime": 48094.3225,
+      "train_tokens_per_second": 9130.933
+    },
+    {
+      "epoch": 1.6362143297064258,
+      "grad_norm": 0.7243551722256891,
+      "learning_rate": 7.952386988522476e-07,
+      "loss": 0.2702,
+      "num_input_tokens_seen": 439409096,
+      "step": 7970,
+      "train_runtime": 48130.3652,
+      "train_tokens_per_second": 9129.561
+    },
+    {
+      "epoch": 1.6372408129747487,
+      "grad_norm": 0.7168130535419368,
+      "learning_rate": 7.908817536422509e-07,
+      "loss": 0.2959,
+      "num_input_tokens_seen": 439706608,
+      "step": 7975,
+      "train_runtime": 48159.9179,
+      "train_tokens_per_second": 9130.136
+    },
+    {
+      "epoch": 1.6382672962430713,
+      "grad_norm": 0.6549228890069749,
+      "learning_rate": 7.865357513853067e-07,
+      "loss": 0.2805,
+      "num_input_tokens_seen": 439956128,
+      "step": 7980,
+      "train_runtime": 48190.3443,
+      "train_tokens_per_second": 9129.549
+    },
+    {
+      "epoch": 1.639293779511394,
+      "grad_norm": 0.7141654113479357,
+      "learning_rate": 7.822007033802415e-07,
+      "loss": 0.3024,
+      "num_input_tokens_seen": 440237176,
+      "step": 7985,
+      "train_runtime": 48224.5313,
+      "train_tokens_per_second": 9128.905
+    },
+    {
+      "epoch": 1.6403202627797167,
+      "grad_norm": 0.7396312107144063,
+      "learning_rate": 7.778766208974037e-07,
+      "loss": 0.2707,
+      "num_input_tokens_seen": 440476000,
+      "step": 7990,
+      "train_runtime": 48257.1226,
+      "train_tokens_per_second": 9127.689
+    },
+    {
+      "epoch": 1.6413467460480393,
+      "grad_norm": 0.6904963035267855,
+      "learning_rate": 7.735635151786325e-07,
+      "loss": 0.2752,
+      "num_input_tokens_seen": 440732912,
+      "step": 7995,
+      "train_runtime": 48284.6905,
+      "train_tokens_per_second": 9127.798
+    },
+    {
+      "epoch": 1.642373229316362,
+      "grad_norm": 0.7280509114437805,
+      "learning_rate": 7.692613974372287e-07,
+      "loss": 0.3079,
+      "num_input_tokens_seen": 441022776,
+      "step": 8000,
+      "train_runtime": 48311.7845,
+      "train_tokens_per_second": 9128.679
+    },
+    {
+      "epoch": 1.6433997125846849,
+      "grad_norm": 0.7764320396383488,
+      "learning_rate": 7.649702788579305e-07,
+      "loss": 0.295,
+      "num_input_tokens_seen": 441303984,
+      "step": 8005,
+      "train_runtime": 48344.4523,
+      "train_tokens_per_second": 9128.327
+    },
+    {
+      "epoch": 1.6444261958530078,
+      "grad_norm": 0.7470181460888216,
+      "learning_rate": 7.606901705968745e-07,
+      "loss": 0.2839,
+      "num_input_tokens_seen": 441582864,
+      "step": 8010,
+      "train_runtime": 48375.0668,
+      "train_tokens_per_second": 9128.315
+    },
+    {
+      "epoch": 1.6454526791213304,
+      "grad_norm": 0.6559060029494648,
+      "learning_rate": 7.564210837815745e-07,
+      "loss": 0.3052,
+      "num_input_tokens_seen": 441860888,
+      "step": 8015,
+      "train_runtime": 48406.0176,
+      "train_tokens_per_second": 9128.222
+    },
+    {
+      "epoch": 1.646479162389653,
+      "grad_norm": 0.6394420600233549,
+      "learning_rate": 7.521630295108928e-07,
+      "loss": 0.2758,
+      "num_input_tokens_seen": 442114304,
+      "step": 8020,
+      "train_runtime": 48439.1317,
+      "train_tokens_per_second": 9127.214
+    },
+    {
+      "epoch": 1.6475056456579757,
+      "grad_norm": 0.7029581114730538,
+      "learning_rate": 7.479160188550077e-07,
+      "loss": 0.2688,
+      "num_input_tokens_seen": 442386416,
+      "step": 8025,
+      "train_runtime": 48471.6936,
+      "train_tokens_per_second": 9126.696
+    },
+    {
+      "epoch": 1.6485321289262984,
+      "grad_norm": 0.6805178385107817,
+      "learning_rate": 7.436800628553853e-07,
+      "loss": 0.2825,
+      "num_input_tokens_seen": 442677408,
+      "step": 8030,
+      "train_runtime": 48495.5379,
+      "train_tokens_per_second": 9128.209
+    },
+    {
+      "epoch": 1.649558612194621,
+      "grad_norm": 0.705249793924718,
+      "learning_rate": 7.394551725247495e-07,
+      "loss": 0.2699,
+      "num_input_tokens_seen": 442947224,
+      "step": 8035,
+      "train_runtime": 48529.5085,
+      "train_tokens_per_second": 9127.379
+    },
+    {
+      "epoch": 1.650585095462944,
+      "grad_norm": 0.6361337825975473,
+      "learning_rate": 7.352413588470614e-07,
+      "loss": 0.2706,
+      "num_input_tokens_seen": 443239648,
+      "step": 8040,
+      "train_runtime": 48560.2695,
+      "train_tokens_per_second": 9127.619
+    },
+    {
+      "epoch": 1.6516115787312666,
+      "grad_norm": 0.6306702184090788,
+      "learning_rate": 7.310386327774793e-07,
+      "loss": 0.2977,
+      "num_input_tokens_seen": 443527512,
+      "step": 8045,
+      "train_runtime": 48593.8979,
+      "train_tokens_per_second": 9127.226
+    },
+    {
+      "epoch": 1.6526380619995895,
+      "grad_norm": 0.7303603989321179,
+      "learning_rate": 7.268470052423388e-07,
+      "loss": 0.3002,
+      "num_input_tokens_seen": 443806984,
+      "step": 8050,
+      "train_runtime": 48622.3349,
+      "train_tokens_per_second": 9127.636
+    },
+    {
+      "epoch": 1.6536645452679122,
+      "grad_norm": 0.6742290975936446,
+      "learning_rate": 7.226664871391198e-07,
+      "loss": 0.253,
+      "num_input_tokens_seen": 444064456,
+      "step": 8055,
+      "train_runtime": 48655.5125,
+      "train_tokens_per_second": 9126.704
+    },
+    {
+      "epoch": 1.6546910285362348,
+      "grad_norm": 0.7246836087614845,
+      "learning_rate": 7.184970893364201e-07,
+      "loss": 0.3024,
+      "num_input_tokens_seen": 444337952,
+      "step": 8060,
+      "train_runtime": 48689.0219,
+      "train_tokens_per_second": 9126.04
+    },
+    {
+      "epoch": 1.6557175118045575,
+      "grad_norm": 0.6255398792423416,
+      "learning_rate": 7.143388226739267e-07,
+      "loss": 0.2917,
+      "num_input_tokens_seen": 444627176,
+      "step": 8065,
+      "train_runtime": 48718.0719,
+      "train_tokens_per_second": 9126.535
+    },
+    {
+      "epoch": 1.6567439950728802,
+      "grad_norm": 0.6527552939661799,
+      "learning_rate": 7.101916979623863e-07,
+      "loss": 0.3017,
+      "num_input_tokens_seen": 444859128,
+      "step": 8070,
+      "train_runtime": 48744.7029,
+      "train_tokens_per_second": 9126.307
+    },
+    {
+      "epoch": 1.657770478341203,
+      "grad_norm": 0.7412544114592524,
+      "learning_rate": 7.060557259835826e-07,
+      "loss": 0.2706,
+      "num_input_tokens_seen": 445127944,
+      "step": 8075,
+      "train_runtime": 48779.0557,
+      "train_tokens_per_second": 9125.391
+    },
+    {
+      "epoch": 1.6587969616095257,
+      "grad_norm": 0.7429655015257002,
+      "learning_rate": 7.01930917490301e-07,
+      "loss": 0.2758,
+      "num_input_tokens_seen": 445416664,
+      "step": 8080,
+      "train_runtime": 48810.8549,
+      "train_tokens_per_second": 9125.361
+    },
+    {
+      "epoch": 1.6598234448778486,
+      "grad_norm": 0.6842689891446042,
+      "learning_rate": 6.978172832063024e-07,
+      "loss": 0.2809,
+      "num_input_tokens_seen": 445683912,
+      "step": 8085,
+      "train_runtime": 48842.4114,
+      "train_tokens_per_second": 9124.937
+    },
+    {
+      "epoch": 1.6608499281461713,
+      "grad_norm": 0.7850896056904704,
+      "learning_rate": 6.937148338262989e-07,
+      "loss": 0.3188,
+      "num_input_tokens_seen": 445952928,
+      "step": 8090,
+      "train_runtime": 48874.5304,
+      "train_tokens_per_second": 9124.444
+    },
+    {
+      "epoch": 1.661876411414494,
+      "grad_norm": 0.6154617101850914,
+      "learning_rate": 6.896235800159251e-07,
+      "loss": 0.2913,
+      "num_input_tokens_seen": 446217616,
+      "step": 8095,
+      "train_runtime": 48905.1807,
+      "train_tokens_per_second": 9124.138
+    },
+    {
+      "epoch": 1.6629028946828166,
+      "grad_norm": 0.6462166766541324,
+      "learning_rate": 6.855435324117071e-07,
+      "loss": 0.3021,
+      "num_input_tokens_seen": 446509160,
+      "step": 8100,
+      "train_runtime": 48943.8046,
+      "train_tokens_per_second": 9122.894
+    },
+    {
+      "epoch": 1.6639293779511393,
+      "grad_norm": 0.6368642558232501,
+      "learning_rate": 6.814747016210371e-07,
+      "loss": 0.2704,
+      "num_input_tokens_seen": 446779192,
+      "step": 8105,
+      "train_runtime": 48971.2343,
+      "train_tokens_per_second": 9123.299
+    },
+    {
+      "epoch": 1.6649558612194622,
+      "grad_norm": 0.6594418050573767,
+      "learning_rate": 6.774170982221462e-07,
+      "loss": 0.3005,
+      "num_input_tokens_seen": 447069184,
+      "step": 8110,
+      "train_runtime": 49000.981,
+      "train_tokens_per_second": 9123.678
+    },
+    {
+      "epoch": 1.6659823444877848,
+      "grad_norm": 0.7913615491127663,
+      "learning_rate": 6.733707327640759e-07,
+      "loss": 0.2866,
+      "num_input_tokens_seen": 447341704,
+      "step": 8115,
+      "train_runtime": 49026.7545,
+      "train_tokens_per_second": 9124.44
+    },
+    {
+      "epoch": 1.6670088277561077,
+      "grad_norm": 0.6898955129751854,
+      "learning_rate": 6.693356157666514e-07,
+      "loss": 0.2763,
+      "num_input_tokens_seen": 447609544,
+      "step": 8120,
+      "train_runtime": 49051.442,
+      "train_tokens_per_second": 9125.309
+    },
+    {
+      "epoch": 1.6680353110244304,
+      "grad_norm": 0.7535095398274261,
+      "learning_rate": 6.653117577204515e-07,
+      "loss": 0.2792,
+      "num_input_tokens_seen": 447852464,
+      "step": 8125,
+      "train_runtime": 49083.0152,
+      "train_tokens_per_second": 9124.388
+    },
+    {
+      "epoch": 1.669061794292753,
+      "grad_norm": 0.7914291063150231,
+      "learning_rate": 6.612991690867898e-07,
+      "loss": 0.2668,
+      "num_input_tokens_seen": 448127808,
+      "step": 8130,
+      "train_runtime": 49111.4655,
+      "train_tokens_per_second": 9124.709
+    },
+    {
+      "epoch": 1.6700882775610757,
+      "grad_norm": 0.621272762445698,
+      "learning_rate": 6.572978602976765e-07,
+      "loss": 0.2795,
+      "num_input_tokens_seen": 448394480,
+      "step": 8135,
+      "train_runtime": 49145.4866,
+      "train_tokens_per_second": 9123.818
+    },
+    {
+      "epoch": 1.6711147608293984,
+      "grad_norm": 0.661095713582504,
+      "learning_rate": 6.533078417557953e-07,
+      "loss": 0.2915,
+      "num_input_tokens_seen": 448670016,
+      "step": 8140,
+      "train_runtime": 49171.2456,
+      "train_tokens_per_second": 9124.642
+    },
+    {
+      "epoch": 1.6721412440977212,
+      "grad_norm": 0.8059912713480164,
+      "learning_rate": 6.493291238344795e-07,
+      "loss": 0.3209,
+      "num_input_tokens_seen": 448924128,
+      "step": 8145,
+      "train_runtime": 49192.7204,
+      "train_tokens_per_second": 9125.824
+    },
+    {
+      "epoch": 1.673167727366044,
+      "grad_norm": 0.6210150868767165,
+      "learning_rate": 6.453617168776844e-07,
+      "loss": 0.2758,
+      "num_input_tokens_seen": 449201896,
+      "step": 8150,
+      "train_runtime": 49218.5279,
+      "train_tokens_per_second": 9126.683
+    },
+    {
+      "epoch": 1.6741942106343668,
+      "grad_norm": 0.7141964105514711,
+      "learning_rate": 6.414056311999561e-07,
+      "loss": 0.3048,
+      "num_input_tokens_seen": 449473960,
+      "step": 8155,
+      "train_runtime": 49242.3745,
+      "train_tokens_per_second": 9127.788
+    },
+    {
+      "epoch": 1.6752206939026895,
+      "grad_norm": 0.6536998272199167,
+      "learning_rate": 6.374608770864088e-07,
+      "loss": 0.2912,
+      "num_input_tokens_seen": 449755888,
+      "step": 8160,
+      "train_runtime": 49277.099,
+      "train_tokens_per_second": 9127.077
+    },
+    {
+      "epoch": 1.6762471771710121,
+      "grad_norm": 0.7970973150702851,
+      "learning_rate": 6.335274647926953e-07,
+      "loss": 0.3119,
+      "num_input_tokens_seen": 450031936,
+      "step": 8165,
+      "train_runtime": 49306.8875,
+      "train_tokens_per_second": 9127.162
+    },
+    {
+      "epoch": 1.6772736604393348,
+      "grad_norm": 0.5763886063576822,
+      "learning_rate": 6.296054045449834e-07,
+      "loss": 0.2736,
+      "num_input_tokens_seen": 450301664,
+      "step": 8170,
+      "train_runtime": 49336.8731,
+      "train_tokens_per_second": 9127.082
+    },
+    {
+      "epoch": 1.6783001437076575,
+      "grad_norm": 0.7896869134962191,
+      "learning_rate": 6.256947065399266e-07,
+      "loss": 0.2838,
+      "num_input_tokens_seen": 450547688,
+      "step": 8175,
+      "train_runtime": 49362.3209,
+      "train_tokens_per_second": 9127.36
+    },
+    {
+      "epoch": 1.6793266269759803,
+      "grad_norm": 0.6954533464873006,
+      "learning_rate": 6.217953809446375e-07,
+      "loss": 0.3027,
+      "num_input_tokens_seen": 450832408,
+      "step": 8180,
+      "train_runtime": 49396.0406,
+      "train_tokens_per_second": 9126.894
+    },
+    {
+      "epoch": 1.680353110244303,
+      "grad_norm": 0.6511275362609542,
+      "learning_rate": 6.179074378966671e-07,
+      "loss": 0.293,
+      "num_input_tokens_seen": 451126168,
+      "step": 8185,
+      "train_runtime": 49419.567,
+      "train_tokens_per_second": 9128.493
+    },
+    {
+      "epoch": 1.681379593512626,
+      "grad_norm": 0.7060915781714715,
+      "learning_rate": 6.140308875039669e-07,
+      "loss": 0.2992,
+      "num_input_tokens_seen": 451394600,
+      "step": 8190,
+      "train_runtime": 49446.0732,
+      "train_tokens_per_second": 9129.028
+    },
+    {
+      "epoch": 1.6824060767809486,
+      "grad_norm": 0.6501826908150334,
+      "learning_rate": 6.101657398448741e-07,
+      "loss": 0.2797,
+      "num_input_tokens_seen": 451652440,
+      "step": 8195,
+      "train_runtime": 49478.6559,
+      "train_tokens_per_second": 9128.228
+    },
+    {
+      "epoch": 1.6834325600492712,
+      "grad_norm": 0.7464325968150256,
+      "learning_rate": 6.06312004968081e-07,
+      "loss": 0.2907,
+      "num_input_tokens_seen": 451929240,
+      "step": 8200,
+      "train_runtime": 49508.2992,
+      "train_tokens_per_second": 9128.353
+    },
+    {
+      "epoch": 1.6844590433175939,
+      "grad_norm": 0.6865142566232638,
+      "learning_rate": 6.024696928926071e-07,
+      "loss": 0.2643,
+      "num_input_tokens_seen": 452211712,
+      "step": 8205,
+      "train_runtime": 49543.4248,
+      "train_tokens_per_second": 9127.583
+    },
+    {
+      "epoch": 1.6854855265859165,
+      "grad_norm": 0.6509076428630495,
+      "learning_rate": 5.986388136077765e-07,
+      "loss": 0.2835,
+      "num_input_tokens_seen": 452468200,
+      "step": 8210,
+      "train_runtime": 49580.0901,
+      "train_tokens_per_second": 9126.006
+    },
+    {
+      "epoch": 1.6865120098542392,
+      "grad_norm": 0.6653624436571545,
+      "learning_rate": 5.948193770731847e-07,
+      "loss": 0.3054,
+      "num_input_tokens_seen": 452742712,
+      "step": 8215,
+      "train_runtime": 49612.9143,
+      "train_tokens_per_second": 9125.501
+    },
+    {
+      "epoch": 1.687538493122562,
+      "grad_norm": 0.7185930946666558,
+      "learning_rate": 5.910113932186862e-07,
+      "loss": 0.2721,
+      "num_input_tokens_seen": 453024576,
+      "step": 8220,
+      "train_runtime": 49649.0588,
+      "train_tokens_per_second": 9124.535
+    },
+    {
+      "epoch": 1.688564976390885,
+      "grad_norm": 0.7140948725073726,
+      "learning_rate": 5.872148719443543e-07,
+      "loss": 0.2734,
+      "num_input_tokens_seen": 453314344,
+      "step": 8225,
+      "train_runtime": 49681.6629,
+      "train_tokens_per_second": 9124.379
+    },
+    {
+      "epoch": 1.6895914596592077,
+      "grad_norm": 0.6096238648132344,
+      "learning_rate": 5.834298231204644e-07,
+      "loss": 0.2719,
+      "num_input_tokens_seen": 453585208,
+      "step": 8230,
+      "train_runtime": 49713.9846,
+      "train_tokens_per_second": 9123.896
+    },
+    {
+      "epoch": 1.6906179429275303,
+      "grad_norm": 0.7770488295648955,
+      "learning_rate": 5.79656256587467e-07,
+      "loss": 0.3224,
+      "num_input_tokens_seen": 453851152,
+      "step": 8235,
+      "train_runtime": 49749.0912,
+      "train_tokens_per_second": 9122.803
+    },
+    {
+      "epoch": 1.691644426195853,
+      "grad_norm": 0.6029146935343492,
+      "learning_rate": 5.758941821559555e-07,
+      "loss": 0.2865,
+      "num_input_tokens_seen": 454120144,
+      "step": 8240,
+      "train_runtime": 49781.0487,
+      "train_tokens_per_second": 9122.35
+    },
+    {
+      "epoch": 1.6926709094641756,
+      "grad_norm": 0.6465428717977908,
+      "learning_rate": 5.721436096066512e-07,
+      "loss": 0.2978,
+      "num_input_tokens_seen": 454395008,
+      "step": 8245,
+      "train_runtime": 49814.5666,
+      "train_tokens_per_second": 9121.73
+    },
+    {
+      "epoch": 1.6936973927324983,
+      "grad_norm": 0.7183579062914792,
+      "learning_rate": 5.684045486903689e-07,
+      "loss": 0.285,
+      "num_input_tokens_seen": 454664088,
+      "step": 8250,
+      "train_runtime": 49847.891,
+      "train_tokens_per_second": 9121.03
+    },
+    {
+      "epoch": 1.6947238760008212,
+      "grad_norm": 0.913664217395064,
+      "learning_rate": 5.646770091279991e-07,
+      "loss": 0.2929,
+      "num_input_tokens_seen": 454963760,
+      "step": 8255,
+      "train_runtime": 49879.0606,
+      "train_tokens_per_second": 9121.338
+    },
+    {
+      "epoch": 1.6957503592691439,
+      "grad_norm": 0.7814156721220169,
+      "learning_rate": 5.609610006104754e-07,
+      "loss": 0.2969,
+      "num_input_tokens_seen": 455244344,
+      "step": 8260,
+      "train_runtime": 49910.2312,
+      "train_tokens_per_second": 9121.263
+    },
+    {
+      "epoch": 1.6967768425374667,
+      "grad_norm": 0.750962191531132,
+      "learning_rate": 5.572565327987544e-07,
+      "loss": 0.2851,
+      "num_input_tokens_seen": 455539440,
+      "step": 8265,
+      "train_runtime": 49943.9926,
+      "train_tokens_per_second": 9121.006
+    },
+    {
+      "epoch": 1.6978033258057894,
+      "grad_norm": 0.6445087585162058,
+      "learning_rate": 5.535636153237866e-07,
+      "loss": 0.2785,
+      "num_input_tokens_seen": 455805104,
+      "step": 8270,
+      "train_runtime": 49975.5632,
+      "train_tokens_per_second": 9120.56
+    },
+    {
+      "epoch": 1.698829809074112,
+      "grad_norm": 0.566815058525282,
+      "learning_rate": 5.498822577864982e-07,
+      "loss": 0.2869,
+      "num_input_tokens_seen": 456097216,
+      "step": 8275,
+      "train_runtime": 50005.2159,
+      "train_tokens_per_second": 9120.993
+    },
+    {
+      "epoch": 1.6998562923424347,
+      "grad_norm": 0.8065799498832253,
+      "learning_rate": 5.462124697577576e-07,
+      "loss": 0.2779,
+      "num_input_tokens_seen": 456353824,
+      "step": 8280,
+      "train_runtime": 50037.1002,
+      "train_tokens_per_second": 9120.309
+    },
+    {
+      "epoch": 1.7008827756107574,
+      "grad_norm": 0.6979829416003148,
+      "learning_rate": 5.425542607783557e-07,
+      "loss": 0.3005,
+      "num_input_tokens_seen": 456655304,
+      "step": 8285,
+      "train_runtime": 50068.6302,
+      "train_tokens_per_second": 9120.587
+    },
+    {
+      "epoch": 1.7019092588790803,
+      "grad_norm": 0.7358837612638733,
+      "learning_rate": 5.38907640358981e-07,
+      "loss": 0.3071,
+      "num_input_tokens_seen": 456909856,
+      "step": 8290,
+      "train_runtime": 50098.3684,
+      "train_tokens_per_second": 9120.254
+    },
+    {
+      "epoch": 1.702935742147403,
+      "grad_norm": 0.7155204420020516,
+      "learning_rate": 5.352726179801926e-07,
+      "loss": 0.2967,
+      "num_input_tokens_seen": 457171792,
+      "step": 8295,
+      "train_runtime": 50132.7529,
+      "train_tokens_per_second": 9119.224
+    },
+    {
+      "epoch": 1.7039622254157258,
+      "grad_norm": 0.7116829657922566,
+      "learning_rate": 5.316492030923975e-07,
+      "loss": 0.3088,
+      "num_input_tokens_seen": 457453272,
+      "step": 8300,
+      "train_runtime": 50160.9903,
+      "train_tokens_per_second": 9119.702
+    },
+    {
+      "epoch": 1.7049887086840485,
+      "grad_norm": 0.6803976603773354,
+      "learning_rate": 5.280374051158238e-07,
+      "loss": 0.2992,
+      "num_input_tokens_seen": 457749384,
+      "step": 8305,
+      "train_runtime": 50191.0426,
+      "train_tokens_per_second": 9120.141
+    },
+    {
+      "epoch": 1.7060151919523712,
+      "grad_norm": 0.6382422744003661,
+      "learning_rate": 5.244372334405001e-07,
+      "loss": 0.2712,
+      "num_input_tokens_seen": 458035256,
+      "step": 8310,
+      "train_runtime": 50220.9796,
+      "train_tokens_per_second": 9120.397
+    },
+    {
+      "epoch": 1.7070416752206938,
+      "grad_norm": 0.7544325332368536,
+      "learning_rate": 5.208486974262289e-07,
+      "loss": 0.3129,
+      "num_input_tokens_seen": 458300640,
+      "step": 8315,
+      "train_runtime": 50252.9907,
+      "train_tokens_per_second": 9119.868
+    },
+    {
+      "epoch": 1.7080681584890165,
+      "grad_norm": 0.8040445212177487,
+      "learning_rate": 5.172718064025573e-07,
+      "loss": 0.2904,
+      "num_input_tokens_seen": 458554496,
+      "step": 8320,
+      "train_runtime": 50281.2254,
+      "train_tokens_per_second": 9119.796
+    },
+    {
+      "epoch": 1.7090946417573394,
+      "grad_norm": 0.7121431294320586,
+      "learning_rate": 5.137065696687632e-07,
+      "loss": 0.2845,
+      "num_input_tokens_seen": 458802392,
+      "step": 8325,
+      "train_runtime": 50310.8109,
+      "train_tokens_per_second": 9119.36
+    },
+    {
+      "epoch": 1.710121125025662,
+      "grad_norm": 0.7438791146809279,
+      "learning_rate": 5.101529964938223e-07,
+      "loss": 0.3009,
+      "num_input_tokens_seen": 459084000,
+      "step": 8330,
+      "train_runtime": 50337.939,
+      "train_tokens_per_second": 9120.04
+    },
+    {
+      "epoch": 1.711147608293985,
+      "grad_norm": 0.7122869659810445,
+      "learning_rate": 5.066110961163883e-07,
+      "loss": 0.3073,
+      "num_input_tokens_seen": 459355248,
+      "step": 8335,
+      "train_runtime": 50365.9571,
+      "train_tokens_per_second": 9120.352
+    },
+    {
+      "epoch": 1.7121740915623076,
+      "grad_norm": 0.6641843589813261,
+      "learning_rate": 5.030808777447677e-07,
+      "loss": 0.2747,
+      "num_input_tokens_seen": 459633208,
+      "step": 8340,
+      "train_runtime": 50390.0007,
+      "train_tokens_per_second": 9121.516
+    },
+    {
+      "epoch": 1.7132005748306303,
+      "grad_norm": 0.7366598234750439,
+      "learning_rate": 4.995623505568947e-07,
+      "loss": 0.2857,
+      "num_input_tokens_seen": 459906536,
+      "step": 8345,
+      "train_runtime": 50423.6359,
+      "train_tokens_per_second": 9120.852
+    },
+    {
+      "epoch": 1.714227058098953,
+      "grad_norm": 0.6224357305229347,
+      "learning_rate": 4.9605552370031e-07,
+      "loss": 0.2576,
+      "num_input_tokens_seen": 460182472,
+      "step": 8350,
+      "train_runtime": 50450.1407,
+      "train_tokens_per_second": 9121.53
+    },
+    {
+      "epoch": 1.7152535413672756,
+      "grad_norm": 0.7159108603385342,
+      "learning_rate": 4.925604062921347e-07,
+      "loss": 0.2896,
+      "num_input_tokens_seen": 460454720,
+      "step": 8355,
+      "train_runtime": 50480.1545,
+      "train_tokens_per_second": 9121.5
+    },
+    {
+      "epoch": 1.7162800246355985,
+      "grad_norm": 0.651721334159365,
+      "learning_rate": 4.89077007419047e-07,
+      "loss": 0.2953,
+      "num_input_tokens_seen": 460741432,
+      "step": 8360,
+      "train_runtime": 50509.0689,
+      "train_tokens_per_second": 9121.955
+    },
+    {
+      "epoch": 1.7173065079039211,
+      "grad_norm": 0.7098220152645676,
+      "learning_rate": 4.85605336137262e-07,
+      "loss": 0.2929,
+      "num_input_tokens_seen": 460985616,
+      "step": 8365,
+      "train_runtime": 50536.388,
+      "train_tokens_per_second": 9121.855
+    },
+    {
+      "epoch": 1.718332991172244,
+      "grad_norm": 0.697734320161261,
+      "learning_rate": 4.821454014724997e-07,
+      "loss": 0.2947,
+      "num_input_tokens_seen": 461232824,
+      "step": 8370,
+      "train_runtime": 50569.5075,
+      "train_tokens_per_second": 9120.77
+    },
+    {
+      "epoch": 1.7193594744405667,
+      "grad_norm": 0.6992525107840528,
+      "learning_rate": 4.786972124199712e-07,
+      "loss": 0.2903,
+      "num_input_tokens_seen": 461531400,
+      "step": 8375,
+      "train_runtime": 50602.1352,
+      "train_tokens_per_second": 9120.789
+    },
+    {
+      "epoch": 1.7203859577088894,
+      "grad_norm": 0.7039630495597398,
+      "learning_rate": 4.752607779443513e-07,
+      "loss": 0.28,
+      "num_input_tokens_seen": 461809648,
+      "step": 8380,
+      "train_runtime": 50634.2158,
+      "train_tokens_per_second": 9120.506
+    },
+    {
+      "epoch": 1.721412440977212,
+      "grad_norm": 0.6112558746862606,
+      "learning_rate": 4.7183610697975224e-07,
+      "loss": 0.3036,
+      "num_input_tokens_seen": 462077336,
+      "step": 8385,
+      "train_runtime": 50660.9299,
+      "train_tokens_per_second": 9120.98
+    },
+    {
+      "epoch": 1.7224389242455347,
+      "grad_norm": 0.5960654283320361,
+      "learning_rate": 4.684232084297058e-07,
+      "loss": 0.2746,
+      "num_input_tokens_seen": 462366736,
+      "step": 8390,
+      "train_runtime": 50690.029,
+      "train_tokens_per_second": 9121.453
+    },
+    {
+      "epoch": 1.7234654075138576,
+      "grad_norm": 0.7162072878165934,
+      "learning_rate": 4.6502209116713604e-07,
+      "loss": 0.2978,
+      "num_input_tokens_seen": 462615776,
+      "step": 8395,
+      "train_runtime": 50721.0924,
+      "train_tokens_per_second": 9120.777
+    },
+    {
+      "epoch": 1.7244918907821802,
+      "grad_norm": 0.7135495618981791,
+      "learning_rate": 4.6163276403433766e-07,
+      "loss": 0.2895,
+      "num_input_tokens_seen": 462882144,
+      "step": 8400,
+      "train_runtime": 50747.6728,
+      "train_tokens_per_second": 9121.249
+    },
+    {
+      "epoch": 1.7255183740505031,
+      "grad_norm": 0.6392060821479338,
+      "learning_rate": 4.5825523584295486e-07,
+      "loss": 0.2753,
+      "num_input_tokens_seen": 463155944,
+      "step": 8405,
+      "train_runtime": 50778.6627,
+      "train_tokens_per_second": 9121.074
+    },
+    {
+      "epoch": 1.7265448573188258,
+      "grad_norm": 0.7356767280544871,
+      "learning_rate": 4.548895153739541e-07,
+      "loss": 0.2845,
+      "num_input_tokens_seen": 463404296,
+      "step": 8410,
+      "train_runtime": 50807.3005,
+      "train_tokens_per_second": 9120.821
+    },
+    {
+      "epoch": 1.7275713405871485,
+      "grad_norm": 0.6298401157012272,
+      "learning_rate": 4.5153561137760813e-07,
+      "loss": 0.2498,
+      "num_input_tokens_seen": 463677928,
+      "step": 8415,
+      "train_runtime": 50837.7267,
+      "train_tokens_per_second": 9120.745
+    },
+    {
+      "epoch": 1.7285978238554711,
+      "grad_norm": 0.6465160027761254,
+      "learning_rate": 4.4819353257346374e-07,
+      "loss": 0.2817,
+      "num_input_tokens_seen": 463965056,
+      "step": 8420,
+      "train_runtime": 50866.1812,
+      "train_tokens_per_second": 9121.287
+    },
+    {
+      "epoch": 1.7296243071237938,
+      "grad_norm": 0.7369143460747666,
+      "learning_rate": 4.4486328765032727e-07,
+      "loss": 0.2933,
+      "num_input_tokens_seen": 464234888,
+      "step": 8425,
+      "train_runtime": 50896.9228,
+      "train_tokens_per_second": 9121.08
+    },
+    {
+      "epoch": 1.7306507903921164,
+      "grad_norm": 0.78933659034833,
+      "learning_rate": 4.41544885266238e-07,
+      "loss": 0.2893,
+      "num_input_tokens_seen": 464504728,
+      "step": 8430,
+      "train_runtime": 50932.468,
+      "train_tokens_per_second": 9120.012
+    },
+    {
+      "epoch": 1.7316772736604393,
+      "grad_norm": 0.6889211194408403,
+      "learning_rate": 4.382383340484492e-07,
+      "loss": 0.2799,
+      "num_input_tokens_seen": 464757056,
+      "step": 8435,
+      "train_runtime": 50962.3415,
+      "train_tokens_per_second": 9119.617
+    },
+    {
+      "epoch": 1.7327037569287622,
+      "grad_norm": 0.5879039481288022,
+      "learning_rate": 4.349436425934006e-07,
+      "loss": 0.29,
+      "num_input_tokens_seen": 465024704,
+      "step": 8440,
+      "train_runtime": 50986.5834,
+      "train_tokens_per_second": 9120.531
+    },
+    {
+      "epoch": 1.7337302401970849,
+      "grad_norm": 0.6446191354903684,
+      "learning_rate": 4.316608194666988e-07,
+      "loss": 0.2956,
+      "num_input_tokens_seen": 465319944,
+      "step": 8445,
+      "train_runtime": 51013.2869,
+      "train_tokens_per_second": 9121.544
+    },
+    {
+      "epoch": 1.7347567234654075,
+      "grad_norm": 0.6163316392528332,
+      "learning_rate": 4.2838987320309664e-07,
+      "loss": 0.2743,
+      "num_input_tokens_seen": 465601328,
+      "step": 8450,
+      "train_runtime": 51046.2997,
+      "train_tokens_per_second": 9121.157
+    },
+    {
+      "epoch": 1.7357832067337302,
+      "grad_norm": 0.7335099627618706,
+      "learning_rate": 4.2513081230646793e-07,
+      "loss": 0.3232,
+      "num_input_tokens_seen": 465877176,
+      "step": 8455,
+      "train_runtime": 51075.2861,
+      "train_tokens_per_second": 9121.382
+    },
+    {
+      "epoch": 1.7368096900020529,
+      "grad_norm": 0.6343968589644199,
+      "learning_rate": 4.218836452497865e-07,
+      "loss": 0.2894,
+      "num_input_tokens_seen": 466145400,
+      "step": 8460,
+      "train_runtime": 51109.7428,
+      "train_tokens_per_second": 9120.48
+    },
+    {
+      "epoch": 1.7378361732703755,
+      "grad_norm": 0.633559033044209,
+      "learning_rate": 4.186483804751046e-07,
+      "loss": 0.269,
+      "num_input_tokens_seen": 466428272,
+      "step": 8465,
+      "train_runtime": 51142.1691,
+      "train_tokens_per_second": 9120.229
+    },
+    {
+      "epoch": 1.7388626565386984,
+      "grad_norm": 0.6785330422463682,
+      "learning_rate": 4.154250263935311e-07,
+      "loss": 0.2943,
+      "num_input_tokens_seen": 466696672,
+      "step": 8470,
+      "train_runtime": 51176.5374,
+      "train_tokens_per_second": 9119.348
+    },
+    {
+      "epoch": 1.739889139807021,
+      "grad_norm": 0.7382449453542737,
+      "learning_rate": 4.122135913852088e-07,
+      "loss": 0.2857,
+      "num_input_tokens_seen": 466981800,
+      "step": 8475,
+      "train_runtime": 51205.9379,
+      "train_tokens_per_second": 9119.681
+    },
+    {
+      "epoch": 1.740915623075344,
+      "grad_norm": 0.6040100459010636,
+      "learning_rate": 4.090140837992934e-07,
+      "loss": 0.2769,
+      "num_input_tokens_seen": 467244896,
+      "step": 8480,
+      "train_runtime": 51236.382,
+      "train_tokens_per_second": 9119.397
+    },
+    {
+      "epoch": 1.7419421063436666,
+      "grad_norm": 0.7048310433976974,
+      "learning_rate": 4.0582651195393006e-07,
+      "loss": 0.3118,
+      "num_input_tokens_seen": 467505352,
+      "step": 8485,
+      "train_runtime": 51268.9641,
+      "train_tokens_per_second": 9118.681
+    },
+    {
+      "epoch": 1.7429685896119893,
+      "grad_norm": 0.7106474596843056,
+      "learning_rate": 4.026508841362359e-07,
+      "loss": 0.2947,
+      "num_input_tokens_seen": 467795008,
+      "step": 8490,
+      "train_runtime": 51294.7375,
+      "train_tokens_per_second": 9119.747
+    },
+    {
+      "epoch": 1.743995072880312,
+      "grad_norm": 0.5917401857118647,
+      "learning_rate": 3.9948720860227444e-07,
+      "loss": 0.2757,
+      "num_input_tokens_seen": 468076280,
+      "step": 8495,
+      "train_runtime": 51328.2208,
+      "train_tokens_per_second": 9119.277
+    },
+    {
+      "epoch": 1.7450215561486346,
+      "grad_norm": 0.7672567632568269,
+      "learning_rate": 3.9633549357703195e-07,
+      "loss": 0.297,
+      "num_input_tokens_seen": 468334104,
+      "step": 8500,
+      "train_runtime": 51358.0347,
+      "train_tokens_per_second": 9119.004
+    },
+    {
+      "epoch": 1.7460480394169575,
+      "grad_norm": 1.206401724941346,
+      "learning_rate": 3.931957472544057e-07,
+      "loss": 0.293,
+      "num_input_tokens_seen": 468617520,
+      "step": 8505,
+      "train_runtime": 51389.668,
+      "train_tokens_per_second": 9118.905
+    },
+    {
+      "epoch": 1.7470745226852802,
+      "grad_norm": 0.7496922132405401,
+      "learning_rate": 3.9006797779717274e-07,
+      "loss": 0.2929,
+      "num_input_tokens_seen": 468872344,
+      "step": 8510,
+      "train_runtime": 51422.2272,
+      "train_tokens_per_second": 9118.087
+    },
+    {
+      "epoch": 1.748101005953603,
+      "grad_norm": 0.6471610218763972,
+      "learning_rate": 3.8695219333697186e-07,
+      "loss": 0.3108,
+      "num_input_tokens_seen": 469164288,
+      "step": 8515,
+      "train_runtime": 51449.0515,
+      "train_tokens_per_second": 9119.008
+    },
+    {
+      "epoch": 1.7491274892219257,
+      "grad_norm": 0.607186239145958,
+      "learning_rate": 3.838484019742855e-07,
+      "loss": 0.2714,
+      "num_input_tokens_seen": 469462176,
+      "step": 8520,
+      "train_runtime": 51486.8095,
+      "train_tokens_per_second": 9118.106
+    },
+    {
+      "epoch": 1.7501539724902484,
+      "grad_norm": 0.8838194549811959,
+      "learning_rate": 3.8075661177841504e-07,
+      "loss": 0.2876,
+      "num_input_tokens_seen": 469742104,
+      "step": 8525,
+      "train_runtime": 51513.585,
+      "train_tokens_per_second": 9118.8
+    },
+    {
+      "epoch": 1.751180455758571,
+      "grad_norm": 0.6862562843227747,
+      "learning_rate": 3.7767683078745955e-07,
+      "loss": 0.2851,
+      "num_input_tokens_seen": 470014576,
+      "step": 8530,
+      "train_runtime": 51547.7905,
+      "train_tokens_per_second": 9118.035
+    },
+    {
+      "epoch": 1.7522069390268937,
+      "grad_norm": 0.6712951723107288,
+      "learning_rate": 3.74609067008298e-07,
+      "loss": 0.285,
+      "num_input_tokens_seen": 470269344,
+      "step": 8535,
+      "train_runtime": 51574.8639,
+      "train_tokens_per_second": 9118.189
+    },
+    {
+      "epoch": 1.7532334222952166,
+      "grad_norm": 0.6507476937317647,
+      "learning_rate": 3.7155332841656733e-07,
+      "loss": 0.2805,
+      "num_input_tokens_seen": 470533248,
+      "step": 8540,
+      "train_runtime": 51602.3726,
+      "train_tokens_per_second": 9118.442
+    },
+    {
+      "epoch": 1.7542599055635393,
+      "grad_norm": 0.6874008408712002,
+      "learning_rate": 3.685096229566404e-07,
+      "loss": 0.2817,
+      "num_input_tokens_seen": 470803528,
+      "step": 8545,
+      "train_runtime": 51628.5969,
+      "train_tokens_per_second": 9119.046
+    },
+    {
+      "epoch": 1.7552863888318622,
+      "grad_norm": 0.7474750101342508,
+      "learning_rate": 3.6547795854160464e-07,
+      "loss": 0.2781,
+      "num_input_tokens_seen": 471087392,
+      "step": 8550,
+      "train_runtime": 51659.4664,
+      "train_tokens_per_second": 9119.091
+    },
+    {
+      "epoch": 1.7563128721001848,
+      "grad_norm": 0.6414620784626478,
+      "learning_rate": 3.6245834305324354e-07,
+      "loss": 0.2954,
+      "num_input_tokens_seen": 471360840,
+      "step": 8555,
+      "train_runtime": 51693.9354,
+      "train_tokens_per_second": 9118.301
+    },
+    {
+      "epoch": 1.7573393553685075,
+      "grad_norm": 0.8845764062731134,
+      "learning_rate": 3.594507843420181e-07,
+      "loss": 0.2617,
+      "num_input_tokens_seen": 471621744,
+      "step": 8560,
+      "train_runtime": 51728.6785,
+      "train_tokens_per_second": 9117.22
+    },
+    {
+      "epoch": 1.7583658386368302,
+      "grad_norm": 0.666850604228915,
+      "learning_rate": 3.5645529022704127e-07,
+      "loss": 0.2823,
+      "num_input_tokens_seen": 471880696,
+      "step": 8565,
+      "train_runtime": 51759.5028,
+      "train_tokens_per_second": 9116.793
+    },
+    {
+      "epoch": 1.7593923219051528,
+      "grad_norm": 0.6388741781698176,
+      "learning_rate": 3.534718684960609e-07,
+      "loss": 0.2898,
+      "num_input_tokens_seen": 472167656,
+      "step": 8570,
+      "train_runtime": 51789.8965,
+      "train_tokens_per_second": 9116.984
+    },
+    {
+      "epoch": 1.7604188051734757,
+      "grad_norm": 0.7409115261471508,
+      "learning_rate": 3.5050052690543843e-07,
+      "loss": 0.2976,
+      "num_input_tokens_seen": 472443288,
+      "step": 8575,
+      "train_runtime": 51821.1277,
+      "train_tokens_per_second": 9116.808
+    },
+    {
+      "epoch": 1.7614452884417984,
+      "grad_norm": 0.7101073585761154,
+      "learning_rate": 3.475412731801298e-07,
+      "loss": 0.2683,
+      "num_input_tokens_seen": 472710872,
+      "step": 8580,
+      "train_runtime": 51848.4719,
+      "train_tokens_per_second": 9117.161
+    },
+    {
+      "epoch": 1.7624717717101213,
+      "grad_norm": 0.7901911074387137,
+      "learning_rate": 3.445941150136639e-07,
+      "loss": 0.2599,
+      "num_input_tokens_seen": 472988336,
+      "step": 8585,
+      "train_runtime": 51884.4831,
+      "train_tokens_per_second": 9116.181
+    },
+    {
+      "epoch": 1.763498254978444,
+      "grad_norm": 0.6890940181649449,
+      "learning_rate": 3.41659060068123e-07,
+      "loss": 0.2686,
+      "num_input_tokens_seen": 473266224,
+      "step": 8590,
+      "train_runtime": 51917.2226,
+      "train_tokens_per_second": 9115.785
+    },
+    {
+      "epoch": 1.7645247382467666,
+      "grad_norm": 0.7789286825477516,
+      "learning_rate": 3.3873611597412703e-07,
+      "loss": 0.316,
+      "num_input_tokens_seen": 473540024,
+      "step": 8595,
+      "train_runtime": 51941.7245,
+      "train_tokens_per_second": 9116.756
+    },
+    {
+      "epoch": 1.7655512215150893,
+      "grad_norm": 0.6075398147615854,
+      "learning_rate": 3.3582529033080345e-07,
+      "loss": 0.3031,
+      "num_input_tokens_seen": 473818904,
+      "step": 8600,
+      "train_runtime": 51967.5229,
+      "train_tokens_per_second": 9117.596
+    },
+    {
+      "epoch": 1.766577704783412,
+      "grad_norm": 0.7197426369029827,
+      "learning_rate": 3.3292659070577884e-07,
+      "loss": 0.2658,
+      "num_input_tokens_seen": 474099912,
+      "step": 8605,
+      "train_runtime": 51999.4644,
+      "train_tokens_per_second": 9117.4
+    },
+    {
+      "epoch": 1.7676041880517348,
+      "grad_norm": 0.6413070992619195,
+      "learning_rate": 3.300400246351515e-07,
+      "loss": 0.3107,
+      "num_input_tokens_seen": 474385976,
+      "step": 8610,
+      "train_runtime": 52033.0802,
+      "train_tokens_per_second": 9117.007
+    },
+    {
+      "epoch": 1.7686306713200575,
+      "grad_norm": 0.7503421613535955,
+      "learning_rate": 3.271655996234774e-07,
+      "loss": 0.2835,
+      "num_input_tokens_seen": 474656992,
+      "step": 8615,
+      "train_runtime": 52063.469,
+      "train_tokens_per_second": 9116.891
+    },
+    {
+      "epoch": 1.7696571545883804,
+      "grad_norm": 0.6351151501912675,
+      "learning_rate": 3.2430332314374535e-07,
+      "loss": 0.2636,
+      "num_input_tokens_seen": 474941584,
+      "step": 8620,
+      "train_runtime": 52095.3333,
+      "train_tokens_per_second": 9116.778
+    },
+    {
+      "epoch": 1.770683637856703,
+      "grad_norm": 0.6857006872721823,
+      "learning_rate": 3.214532026373618e-07,
+      "loss": 0.3134,
+      "num_input_tokens_seen": 475220440,
+      "step": 8625,
+      "train_runtime": 52125.4997,
+      "train_tokens_per_second": 9116.851
+    },
+    {
+      "epoch": 1.7717101211250257,
+      "grad_norm": 0.6280945147927723,
+      "learning_rate": 3.186152455141289e-07,
+      "loss": 0.2908,
+      "num_input_tokens_seen": 475481912,
+      "step": 8630,
+      "train_runtime": 52162.0317,
+      "train_tokens_per_second": 9115.479
+    },
+    {
+      "epoch": 1.7727366043933483,
+      "grad_norm": 0.7546735638633493,
+      "learning_rate": 3.157894591522265e-07,
+      "loss": 0.2657,
+      "num_input_tokens_seen": 475746080,
+      "step": 8635,
+      "train_runtime": 52193.0979,
+      "train_tokens_per_second": 9115.115
+    },
+    {
+      "epoch": 1.773763087661671,
+      "grad_norm": 0.6930789554233696,
+      "learning_rate": 3.129758508981928e-07,
+      "loss": 0.2954,
+      "num_input_tokens_seen": 476007992,
+      "step": 8640,
+      "train_runtime": 52223.4551,
+      "train_tokens_per_second": 9114.831
+    },
+    {
+      "epoch": 1.7747895709299937,
+      "grad_norm": 0.6669086784712884,
+      "learning_rate": 3.10174428066905e-07,
+      "loss": 0.297,
+      "num_input_tokens_seen": 476274560,
+      "step": 8645,
+      "train_runtime": 52253.554,
+      "train_tokens_per_second": 9114.683
+    },
+    {
+      "epoch": 1.7758160541983166,
+      "grad_norm": 0.6192748821417294,
+      "learning_rate": 3.073851979415604e-07,
+      "loss": 0.3047,
+      "num_input_tokens_seen": 476552704,
+      "step": 8650,
+      "train_runtime": 52282.9412,
+      "train_tokens_per_second": 9114.879
+    },
+    {
+      "epoch": 1.7768425374666394,
+      "grad_norm": 0.6214905439437024,
+      "learning_rate": 3.0460816777365734e-07,
+      "loss": 0.271,
+      "num_input_tokens_seen": 476825768,
+      "step": 8655,
+      "train_runtime": 52313.421,
+      "train_tokens_per_second": 9114.789
+    },
+    {
+      "epoch": 1.7778690207349621,
+      "grad_norm": 0.6388480082411784,
+      "learning_rate": 3.018433447829755e-07,
+      "loss": 0.3169,
+      "num_input_tokens_seen": 477105744,
+      "step": 8660,
+      "train_runtime": 52341.7711,
+      "train_tokens_per_second": 9115.201
+    },
+    {
+      "epoch": 1.7788955040032848,
+      "grad_norm": 0.7031542157989186,
+      "learning_rate": 2.9909073615756124e-07,
+      "loss": 0.3118,
+      "num_input_tokens_seen": 477398800,
+      "step": 8665,
+      "train_runtime": 52371.901,
+      "train_tokens_per_second": 9115.552
+    },
+    {
+      "epoch": 1.7799219872716074,
+      "grad_norm": 0.7039588878971844,
+      "learning_rate": 2.963503490537023e-07,
+      "loss": 0.2791,
+      "num_input_tokens_seen": 477679464,
+      "step": 8670,
+      "train_runtime": 52403.2657,
+      "train_tokens_per_second": 9115.452
+    },
+    {
+      "epoch": 1.78094847053993,
+      "grad_norm": 0.7407572752457579,
+      "learning_rate": 2.936221905959152e-07,
+      "loss": 0.3024,
+      "num_input_tokens_seen": 477967024,
+      "step": 8675,
+      "train_runtime": 52430.6535,
+      "train_tokens_per_second": 9116.175
+    },
+    {
+      "epoch": 1.7819749538082528,
+      "grad_norm": 8.25397350132661,
+      "learning_rate": 2.909062678769209e-07,
+      "loss": 0.3165,
+      "num_input_tokens_seen": 478242648,
+      "step": 8680,
+      "train_runtime": 52465.9051,
+      "train_tokens_per_second": 9115.303
+    },
+    {
+      "epoch": 1.7830014370765757,
+      "grad_norm": 0.6854051415253068,
+      "learning_rate": 2.882025879576328e-07,
+      "loss": 0.2772,
+      "num_input_tokens_seen": 478512736,
+      "step": 8685,
+      "train_runtime": 52499.6101,
+      "train_tokens_per_second": 9114.596
+    },
+    {
+      "epoch": 1.7840279203448985,
+      "grad_norm": 0.6852207511259671,
+      "learning_rate": 2.8551115786713493e-07,
+      "loss": 0.2862,
+      "num_input_tokens_seen": 478774328,
+      "step": 8690,
+      "train_runtime": 52529.9805,
+      "train_tokens_per_second": 9114.306
+    },
+    {
+      "epoch": 1.7850544036132212,
+      "grad_norm": 0.7190752982883251,
+      "learning_rate": 2.828319846026617e-07,
+      "loss": 0.2863,
+      "num_input_tokens_seen": 479054776,
+      "step": 8695,
+      "train_runtime": 52558.5847,
+      "train_tokens_per_second": 9114.682
+    },
+    {
+      "epoch": 1.7860808868815439,
+      "grad_norm": 0.7139991556300591,
+      "learning_rate": 2.801650751295848e-07,
+      "loss": 0.2733,
+      "num_input_tokens_seen": 479327792,
+      "step": 8700,
+      "train_runtime": 52585.8541,
+      "train_tokens_per_second": 9115.147
+    },
+    {
+      "epoch": 1.7871073701498665,
+      "grad_norm": 0.6118494172505335,
+      "learning_rate": 2.7751043638138966e-07,
+      "loss": 0.2782,
+      "num_input_tokens_seen": 479615776,
+      "step": 8705,
+      "train_runtime": 52619.942,
+      "train_tokens_per_second": 9114.715
+    },
+    {
+      "epoch": 1.7881338534181892,
+      "grad_norm": 0.6837012062530519,
+      "learning_rate": 2.748680752596622e-07,
+      "loss": 0.2904,
+      "num_input_tokens_seen": 479896704,
+      "step": 8710,
+      "train_runtime": 52653.7604,
+      "train_tokens_per_second": 9114.196
+    },
+    {
+      "epoch": 1.7891603366865119,
+      "grad_norm": 0.7757376754472908,
+      "learning_rate": 2.722379986340662e-07,
+      "loss": 0.29,
+      "num_input_tokens_seen": 480157160,
+      "step": 8715,
+      "train_runtime": 52685.9093,
+      "train_tokens_per_second": 9113.578
+    },
+    {
+      "epoch": 1.7901868199548348,
+      "grad_norm": 0.6137246828208798,
+      "learning_rate": 2.6962021334233147e-07,
+      "loss": 0.2827,
+      "num_input_tokens_seen": 480444152,
+      "step": 8720,
+      "train_runtime": 52718.8224,
+      "train_tokens_per_second": 9113.332
+    },
+    {
+      "epoch": 1.7912133032231574,
+      "grad_norm": 0.6707111196804657,
+      "learning_rate": 2.6701472619023063e-07,
+      "loss": 0.2926,
+      "num_input_tokens_seen": 480743896,
+      "step": 8725,
+      "train_runtime": 52755.521,
+      "train_tokens_per_second": 9112.675
+    },
+    {
+      "epoch": 1.7922397864914803,
+      "grad_norm": 0.79094943763519,
+      "learning_rate": 2.644215439515613e-07,
+      "loss": 0.2938,
+      "num_input_tokens_seen": 480994880,
+      "step": 8730,
+      "train_runtime": 52784.5197,
+      "train_tokens_per_second": 9112.423
+    },
+    {
+      "epoch": 1.793266269759803,
+      "grad_norm": 0.6699945823679576,
+      "learning_rate": 2.6184067336813336e-07,
+      "loss": 0.2819,
+      "num_input_tokens_seen": 481288344,
+      "step": 8735,
+      "train_runtime": 52812.8274,
+      "train_tokens_per_second": 9113.096
+    },
+    {
+      "epoch": 1.7942927530281256,
+      "grad_norm": 0.6564710882603288,
+      "learning_rate": 2.592721211497484e-07,
+      "loss": 0.2621,
+      "num_input_tokens_seen": 481559960,
+      "step": 8740,
+      "train_runtime": 52843.5584,
+      "train_tokens_per_second": 9112.936
+    },
+    {
+      "epoch": 1.7953192362964483,
+      "grad_norm": 0.6739544516816274,
+      "learning_rate": 2.567158939741815e-07,
+      "loss": 0.2747,
+      "num_input_tokens_seen": 481838512,
+      "step": 8745,
+      "train_runtime": 52871.1335,
+      "train_tokens_per_second": 9113.452
+    },
+    {
+      "epoch": 1.796345719564771,
+      "grad_norm": 0.7796363310177192,
+      "learning_rate": 2.5417199848716544e-07,
+      "loss": 0.2834,
+      "num_input_tokens_seen": 482104992,
+      "step": 8750,
+      "train_runtime": 52903.088,
+      "train_tokens_per_second": 9112.984
+    },
+    {
+      "epoch": 1.7973722028330938,
+      "grad_norm": 0.751452012144373,
+      "learning_rate": 2.516404413023721e-07,
+      "loss": 0.2871,
+      "num_input_tokens_seen": 482368080,
+      "step": 8755,
+      "train_runtime": 52933.0397,
+      "train_tokens_per_second": 9112.798
+    },
+    {
+      "epoch": 1.7983986861014165,
+      "grad_norm": 0.6210143446296893,
+      "learning_rate": 2.4912122900139666e-07,
+      "loss": 0.2792,
+      "num_input_tokens_seen": 482647880,
+      "step": 8760,
+      "train_runtime": 52969.1055,
+      "train_tokens_per_second": 9111.875
+    },
+    {
+      "epoch": 1.7994251693697394,
+      "grad_norm": 0.6374572237908778,
+      "learning_rate": 2.4661436813373965e-07,
+      "loss": 0.2757,
+      "num_input_tokens_seen": 482948480,
+      "step": 8765,
+      "train_runtime": 53001.8237,
+      "train_tokens_per_second": 9111.922
+    },
+    {
+      "epoch": 1.800451652638062,
+      "grad_norm": 0.6795666190016642,
+      "learning_rate": 2.4411986521678864e-07,
+      "loss": 0.2712,
+      "num_input_tokens_seen": 483226520,
+      "step": 8770,
+      "train_runtime": 53034.4162,
+      "train_tokens_per_second": 9111.565
+    },
+    {
+      "epoch": 1.8014781359063847,
+      "grad_norm": 0.7159292277213949,
+      "learning_rate": 2.416377267358072e-07,
+      "loss": 0.2749,
+      "num_input_tokens_seen": 483483136,
+      "step": 8775,
+      "train_runtime": 53060.8058,
+      "train_tokens_per_second": 9111.869
+    },
+    {
+      "epoch": 1.8025046191747074,
+      "grad_norm": 0.7789272337306646,
+      "learning_rate": 2.3916795914390835e-07,
+      "loss": 0.2918,
+      "num_input_tokens_seen": 483760424,
+      "step": 8780,
+      "train_runtime": 53085.231,
+      "train_tokens_per_second": 9112.9
+    },
+    {
+      "epoch": 1.80353110244303,
+      "grad_norm": 0.6197327224565805,
+      "learning_rate": 2.3671056886204412e-07,
+      "loss": 0.2935,
+      "num_input_tokens_seen": 484055328,
+      "step": 8785,
+      "train_runtime": 53118.1225,
+      "train_tokens_per_second": 9112.809
+    },
+    {
+      "epoch": 1.804557585711353,
+      "grad_norm": 0.6015968784735113,
+      "learning_rate": 2.3426556227899133e-07,
+      "loss": 0.274,
+      "num_input_tokens_seen": 484354968,
+      "step": 8790,
+      "train_runtime": 53152.2132,
+      "train_tokens_per_second": 9112.602
+    },
+    {
+      "epoch": 1.8055840689796756,
+      "grad_norm": 0.72799736436494,
+      "learning_rate": 2.3183294575132799e-07,
+      "loss": 0.3058,
+      "num_input_tokens_seen": 484630960,
+      "step": 8795,
+      "train_runtime": 53182.8058,
+      "train_tokens_per_second": 9112.55
+    },
+    {
+      "epoch": 1.8066105522479985,
+      "grad_norm": 1.0073606556866674,
+      "learning_rate": 2.2941272560342066e-07,
+      "loss": 0.2811,
+      "num_input_tokens_seen": 484898608,
+      "step": 8800,
+      "train_runtime": 53215.0281,
+      "train_tokens_per_second": 9112.061
+    },
+    {
+      "epoch": 1.8076370355163212,
+      "grad_norm": 0.6788306622023298,
+      "learning_rate": 2.270049081274084e-07,
+      "loss": 0.2955,
+      "num_input_tokens_seen": 485180728,
+      "step": 8805,
+      "train_runtime": 53239.5199,
+      "train_tokens_per_second": 9113.169
+    },
+    {
+      "epoch": 1.8086635187846438,
+      "grad_norm": 0.6645382529601984,
+      "learning_rate": 2.2460949958318433e-07,
+      "loss": 0.2931,
+      "num_input_tokens_seen": 485428000,
+      "step": 8810,
+      "train_runtime": 53267.3367,
+      "train_tokens_per_second": 9113.052
+    },
+    {
+      "epoch": 1.8096900020529665,
+      "grad_norm": 0.7007323210174909,
+      "learning_rate": 2.2222650619838248e-07,
+      "loss": 0.2937,
+      "num_input_tokens_seen": 485720880,
+      "step": 8815,
+      "train_runtime": 53294.3949,
+      "train_tokens_per_second": 9113.921
+    },
+    {
+      "epoch": 1.8107164853212891,
+      "grad_norm": 0.6543077328331551,
+      "learning_rate": 2.198559341683576e-07,
+      "loss": 0.3038,
+      "num_input_tokens_seen": 486009664,
+      "step": 8820,
+      "train_runtime": 53326.01,
+      "train_tokens_per_second": 9113.933
+    },
+    {
+      "epoch": 1.811742968589612,
+      "grad_norm": 0.7750325279813436,
+      "learning_rate": 2.1749778965617253e-07,
+      "loss": 0.2721,
+      "num_input_tokens_seen": 486269096,
+      "step": 8825,
+      "train_runtime": 53354.3614,
+      "train_tokens_per_second": 9113.952
+    },
+    {
+      "epoch": 1.8127694518579347,
+      "grad_norm": 0.6915037359136516,
+      "learning_rate": 2.151520787925804e-07,
+      "loss": 0.2871,
+      "num_input_tokens_seen": 486542320,
+      "step": 8830,
+      "train_runtime": 53386.4841,
+      "train_tokens_per_second": 9113.586
+    },
+    {
+      "epoch": 1.8137959351262576,
+      "grad_norm": 0.6224957865747789,
+      "learning_rate": 2.1281880767600904e-07,
+      "loss": 0.278,
+      "num_input_tokens_seen": 486811360,
+      "step": 8835,
+      "train_runtime": 53416.3592,
+      "train_tokens_per_second": 9113.526
+    },
+    {
+      "epoch": 1.8148224183945803,
+      "grad_norm": 0.6639003337341373,
+      "learning_rate": 2.1049798237254393e-07,
+      "loss": 0.2821,
+      "num_input_tokens_seen": 487089464,
+      "step": 8840,
+      "train_runtime": 53446.9044,
+      "train_tokens_per_second": 9113.521
+    },
+    {
+      "epoch": 1.815848901662903,
+      "grad_norm": 0.6643382061272886,
+      "learning_rate": 2.081896089159169e-07,
+      "loss": 0.2714,
+      "num_input_tokens_seen": 487387304,
+      "step": 8845,
+      "train_runtime": 53472.3591,
+      "train_tokens_per_second": 9114.752
+    },
+    {
+      "epoch": 1.8168753849312256,
+      "grad_norm": 0.755660516575454,
+      "learning_rate": 2.0589369330748456e-07,
+      "loss": 0.3025,
+      "num_input_tokens_seen": 487677496,
+      "step": 8850,
+      "train_runtime": 53509.6266,
+      "train_tokens_per_second": 9113.827
+    },
+    {
+      "epoch": 1.8179018681995482,
+      "grad_norm": 0.6716356493221904,
+      "learning_rate": 2.0361024151621667e-07,
+      "loss": 0.2907,
+      "num_input_tokens_seen": 487968968,
+      "step": 8855,
+      "train_runtime": 53546.225,
+      "train_tokens_per_second": 9113.041
+    },
+    {
+      "epoch": 1.818928351467871,
+      "grad_norm": 0.6892375530973385,
+      "learning_rate": 2.0133925947867617e-07,
+      "loss": 0.3109,
+      "num_input_tokens_seen": 488233184,
+      "step": 8860,
+      "train_runtime": 53578.9064,
+      "train_tokens_per_second": 9112.414
+    },
+    {
+      "epoch": 1.8199548347361938,
+      "grad_norm": 0.6654233978578813,
+      "learning_rate": 1.990807530990113e-07,
+      "loss": 0.2953,
+      "num_input_tokens_seen": 488505112,
+      "step": 8865,
+      "train_runtime": 53609.1385,
+      "train_tokens_per_second": 9112.348
+    },
+    {
+      "epoch": 1.8209813180045167,
+      "grad_norm": 0.6542717332780107,
+      "learning_rate": 1.9683472824893358e-07,
+      "loss": 0.2841,
+      "num_input_tokens_seen": 488786672,
+      "step": 8870,
+      "train_runtime": 53643.3372,
+      "train_tokens_per_second": 9111.787
+    },
+    {
+      "epoch": 1.8220078012728393,
+      "grad_norm": 0.8054485604424609,
+      "learning_rate": 1.9460119076770533e-07,
+      "loss": 0.2954,
+      "num_input_tokens_seen": 489056256,
+      "step": 8875,
+      "train_runtime": 53674.9994,
+      "train_tokens_per_second": 9111.435
+    },
+    {
+      "epoch": 1.823034284541162,
+      "grad_norm": 0.7294892712061295,
+      "learning_rate": 1.9238014646212279e-07,
+      "loss": 0.31,
+      "num_input_tokens_seen": 489343400,
+      "step": 8880,
+      "train_runtime": 53704.7439,
+      "train_tokens_per_second": 9111.735
+    },
+    {
+      "epoch": 1.8240607678094847,
+      "grad_norm": 0.6421727392811899,
+      "learning_rate": 1.9017160110650412e-07,
+      "loss": 0.2707,
+      "num_input_tokens_seen": 489622456,
+      "step": 8885,
+      "train_runtime": 53738.1866,
+      "train_tokens_per_second": 9111.258
+    },
+    {
+      "epoch": 1.8250872510778073,
+      "grad_norm": 0.6392584771928639,
+      "learning_rate": 1.8797556044267028e-07,
+      "loss": 0.2687,
+      "num_input_tokens_seen": 489916496,
+      "step": 8890,
+      "train_runtime": 53771.0079,
+      "train_tokens_per_second": 9111.164
+    },
+    {
+      "epoch": 1.82611373434613,
+      "grad_norm": 0.6731751839240714,
+      "learning_rate": 1.857920301799343e-07,
+      "loss": 0.2847,
+      "num_input_tokens_seen": 490152272,
+      "step": 8895,
+      "train_runtime": 53806.2535,
+      "train_tokens_per_second": 9109.578
+    },
+    {
+      "epoch": 1.8271402176144529,
+      "grad_norm": 0.6874918193935095,
+      "learning_rate": 1.8362101599508298e-07,
+      "loss": 0.2933,
+      "num_input_tokens_seen": 490389512,
+      "step": 8900,
+      "train_runtime": 53833.0255,
+      "train_tokens_per_second": 9109.455
+    },
+    {
+      "epoch": 1.8281667008827758,
+      "grad_norm": 0.6104651465017185,
+      "learning_rate": 1.8146252353236582e-07,
+      "loss": 0.3028,
+      "num_input_tokens_seen": 490688944,
+      "step": 8905,
+      "train_runtime": 53864.6742,
+      "train_tokens_per_second": 9109.661
+    },
+    {
+      "epoch": 1.8291931841510984,
+      "grad_norm": 0.748630818427099,
+      "learning_rate": 1.7931655840347396e-07,
+      "loss": 0.3026,
+      "num_input_tokens_seen": 490968880,
+      "step": 8910,
+      "train_runtime": 53888.6873,
+      "train_tokens_per_second": 9110.797
+    },
+    {
+      "epoch": 1.830219667419421,
+      "grad_norm": 0.6910538222522046,
+      "learning_rate": 1.7718312618753231e-07,
+      "loss": 0.2967,
+      "num_input_tokens_seen": 491274008,
+      "step": 8915,
+      "train_runtime": 53916.0918,
+      "train_tokens_per_second": 9111.825
+    },
+    {
+      "epoch": 1.8312461506877438,
+      "grad_norm": 0.633018120319185,
+      "learning_rate": 1.7506223243108355e-07,
+      "loss": 0.2883,
+      "num_input_tokens_seen": 491570480,
+      "step": 8920,
+      "train_runtime": 53950.4825,
+      "train_tokens_per_second": 9111.512
+    },
+    {
+      "epoch": 1.8322726339560664,
+      "grad_norm": 0.6284976791846368,
+      "learning_rate": 1.729538826480709e-07,
+      "loss": 0.2972,
+      "num_input_tokens_seen": 491837720,
+      "step": 8925,
+      "train_runtime": 53984.6854,
+      "train_tokens_per_second": 9110.69
+    },
+    {
+      "epoch": 1.833299117224389,
+      "grad_norm": 0.6229258010016079,
+      "learning_rate": 1.708580823198258e-07,
+      "loss": 0.3173,
+      "num_input_tokens_seen": 492106872,
+      "step": 8930,
+      "train_runtime": 54016.1106,
+      "train_tokens_per_second": 9110.372
+    },
+    {
+      "epoch": 1.834325600492712,
+      "grad_norm": 0.6260333711677472,
+      "learning_rate": 1.687748368950537e-07,
+      "loss": 0.2719,
+      "num_input_tokens_seen": 492402384,
+      "step": 8935,
+      "train_runtime": 54038.2011,
+      "train_tokens_per_second": 9112.117
+    },
+    {
+      "epoch": 1.8353520837610346,
+      "grad_norm": 0.6969356828487079,
+      "learning_rate": 1.6670415178981937e-07,
+      "loss": 0.2925,
+      "num_input_tokens_seen": 492670896,
+      "step": 8940,
+      "train_runtime": 54071.5245,
+      "train_tokens_per_second": 9111.467
+    },
+    {
+      "epoch": 1.8363785670293575,
+      "grad_norm": 0.6626904328287986,
+      "learning_rate": 1.6464603238753207e-07,
+      "loss": 0.293,
+      "num_input_tokens_seen": 492951096,
+      "step": 8945,
+      "train_runtime": 54103.3598,
+      "train_tokens_per_second": 9111.284
+    },
+    {
+      "epoch": 1.8374050502976802,
+      "grad_norm": 0.7338407066912486,
+      "learning_rate": 1.6260048403893392e-07,
+      "loss": 0.2657,
+      "num_input_tokens_seen": 493217432,
+      "step": 8950,
+      "train_runtime": 54134.0376,
+      "train_tokens_per_second": 9111.041
+    },
+    {
+      "epoch": 1.8384315335660029,
+      "grad_norm": 0.6492159486144057,
+      "learning_rate": 1.605675120620853e-07,
+      "loss": 0.2419,
+      "num_input_tokens_seen": 493513656,
+      "step": 8955,
+      "train_runtime": 54160.6763,
+      "train_tokens_per_second": 9112.029
+    },
+    {
+      "epoch": 1.8394580168343255,
+      "grad_norm": 0.7163419529439299,
+      "learning_rate": 1.5854712174234675e-07,
+      "loss": 0.2707,
+      "num_input_tokens_seen": 493784176,
+      "step": 8960,
+      "train_runtime": 54196.8359,
+      "train_tokens_per_second": 9110.941
+    },
+    {
+      "epoch": 1.8404845001026482,
+      "grad_norm": 0.6360011001396503,
+      "learning_rate": 1.5653931833237211e-07,
+      "loss": 0.2977,
+      "num_input_tokens_seen": 494059136,
+      "step": 8965,
+      "train_runtime": 54224.1401,
+      "train_tokens_per_second": 9111.424
+    },
+    {
+      "epoch": 1.841510983370971,
+      "grad_norm": 0.6567995660692819,
+      "learning_rate": 1.5454410705209033e-07,
+      "loss": 0.2986,
+      "num_input_tokens_seen": 494331008,
+      "step": 8970,
+      "train_runtime": 54258.2366,
+      "train_tokens_per_second": 9110.709
+    },
+    {
+      "epoch": 1.8425374666392937,
+      "grad_norm": 0.6699250376754979,
+      "learning_rate": 1.5256149308869372e-07,
+      "loss": 0.2756,
+      "num_input_tokens_seen": 494617168,
+      "step": 8975,
+      "train_runtime": 54286.7505,
+      "train_tokens_per_second": 9111.195
+    },
+    {
+      "epoch": 1.8435639499076166,
+      "grad_norm": 0.6835332451242612,
+      "learning_rate": 1.505914815966225e-07,
+      "loss": 0.2865,
+      "num_input_tokens_seen": 494906696,
+      "step": 8980,
+      "train_runtime": 54317.121,
+      "train_tokens_per_second": 9111.431
+    },
+    {
+      "epoch": 1.8445904331759393,
+      "grad_norm": 0.6667934880263798,
+      "learning_rate": 1.4863407769755312e-07,
+      "loss": 0.2765,
+      "num_input_tokens_seen": 495190032,
+      "step": 8985,
+      "train_runtime": 54344.2251,
+      "train_tokens_per_second": 9112.1
+    },
+    {
+      "epoch": 1.845616916444262,
+      "grad_norm": 0.6550129113772728,
+      "learning_rate": 1.4668928648038595e-07,
+      "loss": 0.288,
+      "num_input_tokens_seen": 495460336,
+      "step": 8990,
+      "train_runtime": 54374.2162,
+      "train_tokens_per_second": 9112.046
+    },
+    {
+      "epoch": 1.8466433997125846,
+      "grad_norm": 0.7292394089918705,
+      "learning_rate": 1.4475711300122874e-07,
+      "loss": 0.2922,
+      "num_input_tokens_seen": 495727320,
+      "step": 8995,
+      "train_runtime": 54406.8177,
+      "train_tokens_per_second": 9111.493
+    },
+    {
+      "epoch": 1.8476698829809073,
+      "grad_norm": 0.7494546115197279,
+      "learning_rate": 1.4283756228338662e-07,
+      "loss": 0.2852,
+      "num_input_tokens_seen": 495973064,
+      "step": 9000,
+      "train_runtime": 54436.3295,
+      "train_tokens_per_second": 9111.067
+    },
+    {
+      "epoch": 1.8486963662492302,
+      "grad_norm": 0.6558338669480048,
+      "learning_rate": 1.4093063931734764e-07,
+      "loss": 0.2553,
+      "num_input_tokens_seen": 496262672,
+      "step": 9005,
+      "train_runtime": 54466.3299,
+      "train_tokens_per_second": 9111.366
+    },
+    {
+      "epoch": 1.8497228495175528,
+      "grad_norm": 0.6413709551003409,
+      "learning_rate": 1.3903634906076934e-07,
+      "loss": 0.2747,
+      "num_input_tokens_seen": 496536024,
+      "step": 9010,
+      "train_runtime": 54494.3737,
+      "train_tokens_per_second": 9111.693
+    },
+    {
+      "epoch": 1.8507493327858757,
+      "grad_norm": 0.6409344756360071,
+      "learning_rate": 1.3715469643846734e-07,
+      "loss": 0.3101,
+      "num_input_tokens_seen": 496802304,
+      "step": 9015,
+      "train_runtime": 54525.5554,
+      "train_tokens_per_second": 9111.366
+    },
+    {
+      "epoch": 1.8517758160541984,
+      "grad_norm": 0.637549968219776,
+      "learning_rate": 1.3528568634239958e-07,
+      "loss": 0.282,
+      "num_input_tokens_seen": 497057512,
+      "step": 9020,
+      "train_runtime": 54557.7251,
+      "train_tokens_per_second": 9110.672
+    },
+    {
+      "epoch": 1.852802299322521,
+      "grad_norm": 0.7690507607075756,
+      "learning_rate": 1.3342932363165974e-07,
+      "loss": 0.2818,
+      "num_input_tokens_seen": 497322704,
+      "step": 9025,
+      "train_runtime": 54590.8844,
+      "train_tokens_per_second": 9109.995
+    },
+    {
+      "epoch": 1.8538287825908437,
+      "grad_norm": 0.6145574284857961,
+      "learning_rate": 1.3158561313245665e-07,
+      "loss": 0.2966,
+      "num_input_tokens_seen": 497580400,
+      "step": 9030,
+      "train_runtime": 54620.8559,
+      "train_tokens_per_second": 9109.714
+    },
+    {
+      "epoch": 1.8548552658591664,
+      "grad_norm": 0.7105057068249231,
+      "learning_rate": 1.2975455963810835e-07,
+      "loss": 0.2728,
+      "num_input_tokens_seen": 497840456,
+      "step": 9035,
+      "train_runtime": 54652.2555,
+      "train_tokens_per_second": 9109.239
+    },
+    {
+      "epoch": 1.8558817491274893,
+      "grad_norm": 0.579319285231163,
+      "learning_rate": 1.2793616790902407e-07,
+      "loss": 0.2883,
+      "num_input_tokens_seen": 498149136,
+      "step": 9040,
+      "train_runtime": 54685.9191,
+      "train_tokens_per_second": 9109.276
+    },
+    {
+      "epoch": 1.856908232395812,
+      "grad_norm": 0.7319489268819573,
+      "learning_rate": 1.261304426726978e-07,
+      "loss": 0.2935,
+      "num_input_tokens_seen": 498437232,
+      "step": 9045,
+      "train_runtime": 54717.117,
+      "train_tokens_per_second": 9109.347
+    },
+    {
+      "epoch": 1.8579347156641348,
+      "grad_norm": 0.584050507174379,
+      "learning_rate": 1.2433738862369038e-07,
+      "loss": 0.2928,
+      "num_input_tokens_seen": 498708272,
+      "step": 9050,
+      "train_runtime": 54742.897,
+      "train_tokens_per_second": 9110.009
+    },
+    {
+      "epoch": 1.8589611989324575,
+      "grad_norm": 0.726595300291363,
+      "learning_rate": 1.2255701042362234e-07,
+      "loss": 0.3089,
+      "num_input_tokens_seen": 498979992,
+      "step": 9055,
+      "train_runtime": 54777.3499,
+      "train_tokens_per_second": 9109.239
+    },
+    {
+      "epoch": 1.8599876822007801,
+      "grad_norm": 0.6871595236168418,
+      "learning_rate": 1.2078931270115667e-07,
+      "loss": 0.3247,
+      "num_input_tokens_seen": 499269896,
+      "step": 9060,
+      "train_runtime": 54806.7736,
+      "train_tokens_per_second": 9109.639
+    },
+    {
+      "epoch": 1.8610141654691028,
+      "grad_norm": 0.6387877149987767,
+      "learning_rate": 1.1903430005199169e-07,
+      "loss": 0.2833,
+      "num_input_tokens_seen": 499567424,
+      "step": 9065,
+      "train_runtime": 54841.1556,
+      "train_tokens_per_second": 9109.353
+    },
+    {
+      "epoch": 1.8620406487374255,
+      "grad_norm": 0.7670108252575413,
+      "learning_rate": 1.1729197703884476e-07,
+      "loss": 0.2824,
+      "num_input_tokens_seen": 499845112,
+      "step": 9070,
+      "train_runtime": 54869.4109,
+      "train_tokens_per_second": 9109.723
+    },
+    {
+      "epoch": 1.8630671320057484,
+      "grad_norm": 0.6994504446415342,
+      "learning_rate": 1.1556234819144307e-07,
+      "loss": 0.3007,
+      "num_input_tokens_seen": 500118248,
+      "step": 9075,
+      "train_runtime": 54895.3633,
+      "train_tokens_per_second": 9110.391
+    },
+    {
+      "epoch": 1.864093615274071,
+      "grad_norm": 0.7109266058514189,
+      "learning_rate": 1.138454180065135e-07,
+      "loss": 0.314,
+      "num_input_tokens_seen": 500406360,
+      "step": 9080,
+      "train_runtime": 54921.4816,
+      "train_tokens_per_second": 9111.305
+    },
+    {
+      "epoch": 1.865120098542394,
+      "grad_norm": 0.6355672011618831,
+      "learning_rate": 1.1214119094776598e-07,
+      "loss": 0.2959,
+      "num_input_tokens_seen": 500684424,
+      "step": 9085,
+      "train_runtime": 54946.7398,
+      "train_tokens_per_second": 9112.177
+    },
+    {
+      "epoch": 1.8661465818107166,
+      "grad_norm": 0.6172551148421539,
+      "learning_rate": 1.1044967144588415e-07,
+      "loss": 0.2653,
+      "num_input_tokens_seen": 500947520,
+      "step": 9090,
+      "train_runtime": 54972.9052,
+      "train_tokens_per_second": 9112.626
+    },
+    {
+      "epoch": 1.8671730650790392,
+      "grad_norm": 0.5767474463774188,
+      "learning_rate": 1.0877086389851743e-07,
+      "loss": 0.2759,
+      "num_input_tokens_seen": 501244000,
+      "step": 9095,
+      "train_runtime": 54999.5442,
+      "train_tokens_per_second": 9113.603
+    },
+    {
+      "epoch": 1.868199548347362,
+      "grad_norm": 0.7466011208419218,
+      "learning_rate": 1.0710477267026343e-07,
+      "loss": 0.3085,
+      "num_input_tokens_seen": 501492024,
+      "step": 9100,
+      "train_runtime": 55028.8784,
+      "train_tokens_per_second": 9113.252
+    },
+    {
+      "epoch": 1.8692260316156846,
+      "grad_norm": 0.606578322285639,
+      "learning_rate": 1.0545140209266114e-07,
+      "loss": 0.2781,
+      "num_input_tokens_seen": 501772656,
+      "step": 9105,
+      "train_runtime": 55058.8048,
+      "train_tokens_per_second": 9113.395
+    },
+    {
+      "epoch": 1.8702525148840072,
+      "grad_norm": 0.7472953916491398,
+      "learning_rate": 1.0381075646417716e-07,
+      "loss": 0.3285,
+      "num_input_tokens_seen": 502028408,
+      "step": 9110,
+      "train_runtime": 55087.0379,
+      "train_tokens_per_second": 9113.367
+    },
+    {
+      "epoch": 1.8712789981523301,
+      "grad_norm": 0.6648156608282068,
+      "learning_rate": 1.0218284005019675e-07,
+      "loss": 0.3101,
+      "num_input_tokens_seen": 502319656,
+      "step": 9115,
+      "train_runtime": 55114.8774,
+      "train_tokens_per_second": 9114.048
+    },
+    {
+      "epoch": 1.872305481420653,
+      "grad_norm": 0.6926228215374325,
+      "learning_rate": 1.0056765708301053e-07,
+      "loss": 0.3132,
+      "num_input_tokens_seen": 502583120,
+      "step": 9120,
+      "train_runtime": 55147.3837,
+      "train_tokens_per_second": 9113.454
+    },
+    {
+      "epoch": 1.8733319646889757,
+      "grad_norm": 0.651865656012838,
+      "learning_rate": 9.896521176180507e-08,
+      "loss": 0.2856,
+      "num_input_tokens_seen": 502848680,
+      "step": 9125,
+      "train_runtime": 55179.2013,
+      "train_tokens_per_second": 9113.011
+    },
+    {
+      "epoch": 1.8743584479572983,
+      "grad_norm": 0.7145385061820912,
+      "learning_rate": 9.737550825265008e-08,
+      "loss": 0.2927,
+      "num_input_tokens_seen": 503118800,
+      "step": 9130,
+      "train_runtime": 55211.359,
+      "train_tokens_per_second": 9112.596
+    },
+    {
+      "epoch": 1.875384931225621,
+      "grad_norm": 0.683866325562885,
+      "learning_rate": 9.57985506884912e-08,
+      "loss": 0.2887,
+      "num_input_tokens_seen": 503384424,
+      "step": 9135,
+      "train_runtime": 55235.64,
+      "train_tokens_per_second": 9113.399
+    },
+    {
+      "epoch": 1.8764114144939437,
+      "grad_norm": 0.5592265472547767,
+      "learning_rate": 9.423434316913449e-08,
+      "loss": 0.2732,
+      "num_input_tokens_seen": 503681592,
+      "step": 9140,
+      "train_runtime": 55263.1215,
+      "train_tokens_per_second": 9114.244
+    },
+    {
+      "epoch": 1.8774378977622663,
+      "grad_norm": 0.6571774520065234,
+      "learning_rate": 9.268288976123863e-08,
+      "loss": 0.2925,
+      "num_input_tokens_seen": 503949056,
+      "step": 9145,
+      "train_runtime": 55292.568,
+      "train_tokens_per_second": 9114.228
+    },
+    {
+      "epoch": 1.8784643810305892,
+      "grad_norm": 0.659065634472469,
+      "learning_rate": 9.114419449830547e-08,
+      "loss": 0.2888,
+      "num_input_tokens_seen": 504199248,
+      "step": 9150,
+      "train_runtime": 55321.2268,
+      "train_tokens_per_second": 9114.029
+    },
+    {
+      "epoch": 1.8794908642989119,
+      "grad_norm": 0.6466669666163387,
+      "learning_rate": 8.961826138066675e-08,
+      "loss": 0.2846,
+      "num_input_tokens_seen": 504461712,
+      "step": 9155,
+      "train_runtime": 55356.1569,
+      "train_tokens_per_second": 9113.019
+    },
+    {
+      "epoch": 1.8805173475672348,
+      "grad_norm": 0.8836878902292502,
+      "learning_rate": 8.81050943754752e-08,
+      "loss": 0.2909,
+      "num_input_tokens_seen": 504738760,
+      "step": 9160,
+      "train_runtime": 55388.5112,
+      "train_tokens_per_second": 9112.698
+    },
+    {
+      "epoch": 1.8815438308355574,
+      "grad_norm": 0.636261722010579,
+      "learning_rate": 8.660469741669287e-08,
+      "loss": 0.3103,
+      "num_input_tokens_seen": 505001680,
+      "step": 9165,
+      "train_runtime": 55418.3496,
+      "train_tokens_per_second": 9112.536
+    },
+    {
+      "epoch": 1.88257031410388,
+      "grad_norm": 0.6011552551582583,
+      "learning_rate": 8.511707440508388e-08,
+      "loss": 0.2552,
+      "num_input_tokens_seen": 505276408,
+      "step": 9170,
+      "train_runtime": 55450.96,
+      "train_tokens_per_second": 9112.131
+    },
+    {
+      "epoch": 1.8835967973722028,
+      "grad_norm": 0.6137124486721656,
+      "learning_rate": 8.36422292082012e-08,
+      "loss": 0.2769,
+      "num_input_tokens_seen": 505517016,
+      "step": 9175,
+      "train_runtime": 55478.3112,
+      "train_tokens_per_second": 9111.976
+    },
+    {
+      "epoch": 1.8846232806405254,
+      "grad_norm": 0.6579323303705139,
+      "learning_rate": 8.218016566037767e-08,
+      "loss": 0.2729,
+      "num_input_tokens_seen": 505791416,
+      "step": 9180,
+      "train_runtime": 55510.3692,
+      "train_tokens_per_second": 9111.656
+    },
+    {
+      "epoch": 1.8856497639088483,
+      "grad_norm": 0.6255418720519567,
+      "learning_rate": 8.073088756271775e-08,
+      "loss": 0.2682,
+      "num_input_tokens_seen": 506088840,
+      "step": 9185,
+      "train_runtime": 55541.1497,
+      "train_tokens_per_second": 9111.962
+    },
+    {
+      "epoch": 1.886676247177171,
+      "grad_norm": 0.7509314553590535,
+      "learning_rate": 7.929439868308298e-08,
+      "loss": 0.3034,
+      "num_input_tokens_seen": 506376608,
+      "step": 9190,
+      "train_runtime": 55573.7828,
+      "train_tokens_per_second": 9111.789
+    },
+    {
+      "epoch": 1.8877027304454939,
+      "grad_norm": 0.7114292649128451,
+      "learning_rate": 7.787070275608766e-08,
+      "loss": 0.2748,
+      "num_input_tokens_seen": 506640416,
+      "step": 9195,
+      "train_runtime": 55605.9614,
+      "train_tokens_per_second": 9111.261
+    },
+    {
+      "epoch": 1.8887292137138165,
+      "grad_norm": 0.6186954787498993,
+      "learning_rate": 7.645980348308601e-08,
+      "loss": 0.2779,
+      "num_input_tokens_seen": 506923048,
+      "step": 9200,
+      "train_runtime": 55638.1709,
+      "train_tokens_per_second": 9111.066
+    },
+    {
+      "epoch": 1.8897556969821392,
+      "grad_norm": 0.7607773465064438,
+      "learning_rate": 7.506170453216388e-08,
+      "loss": 0.2582,
+      "num_input_tokens_seen": 507192448,
+      "step": 9205,
+      "train_runtime": 55666.4647,
+      "train_tokens_per_second": 9111.275
+    },
+    {
+      "epoch": 1.8907821802504619,
+      "grad_norm": 0.743827836907956,
+      "learning_rate": 7.367640953812761e-08,
+      "loss": 0.3042,
+      "num_input_tokens_seen": 507470808,
+      "step": 9210,
+      "train_runtime": 55696.3357,
+      "train_tokens_per_second": 9111.386
+    },
+    {
+      "epoch": 1.8918086635187845,
+      "grad_norm": 0.6917620865186774,
+      "learning_rate": 7.23039221024957e-08,
+      "loss": 0.268,
+      "num_input_tokens_seen": 507754808,
+      "step": 9215,
+      "train_runtime": 55727.783,
+      "train_tokens_per_second": 9111.341
+    },
+    {
+      "epoch": 1.8928351467871074,
+      "grad_norm": 0.6950796458561943,
+      "learning_rate": 7.094424579348946e-08,
+      "loss": 0.2637,
+      "num_input_tokens_seen": 508030640,
+      "step": 9220,
+      "train_runtime": 55757.4473,
+      "train_tokens_per_second": 9111.44
+    },
+    {
+      "epoch": 1.89386163005543,
+      "grad_norm": 0.6082906920436056,
+      "learning_rate": 6.959738414602347e-08,
+      "loss": 0.3054,
+      "num_input_tokens_seen": 508308144,
+      "step": 9225,
+      "train_runtime": 55785.6779,
+      "train_tokens_per_second": 9111.804
+    },
+    {
+      "epoch": 1.894888113323753,
+      "grad_norm": 0.6500646432084701,
+      "learning_rate": 6.826334066169726e-08,
+      "loss": 0.2816,
+      "num_input_tokens_seen": 508570872,
+      "step": 9230,
+      "train_runtime": 55816.0041,
+      "train_tokens_per_second": 9111.56
+    },
+    {
+      "epoch": 1.8959145965920756,
+      "grad_norm": 0.6400385365370455,
+      "learning_rate": 6.694211880878432e-08,
+      "loss": 0.2906,
+      "num_input_tokens_seen": 508850736,
+      "step": 9235,
+      "train_runtime": 55848.5144,
+      "train_tokens_per_second": 9111.267
+    },
+    {
+      "epoch": 1.8969410798603983,
+      "grad_norm": 0.6483540950670247,
+      "learning_rate": 6.563372202222473e-08,
+      "loss": 0.2987,
+      "num_input_tokens_seen": 509151672,
+      "step": 9240,
+      "train_runtime": 55875.4342,
+      "train_tokens_per_second": 9112.263
+    },
+    {
+      "epoch": 1.897967563128721,
+      "grad_norm": 0.699690312501339,
+      "learning_rate": 6.433815370361529e-08,
+      "loss": 0.2927,
+      "num_input_tokens_seen": 509433728,
+      "step": 9245,
+      "train_runtime": 55899.0316,
+      "train_tokens_per_second": 9113.463
+    },
+    {
+      "epoch": 1.8989940463970436,
+      "grad_norm": 0.6470253670429796,
+      "learning_rate": 6.305541722120223e-08,
+      "loss": 0.2948,
+      "num_input_tokens_seen": 509701736,
+      "step": 9250,
+      "train_runtime": 55927.3124,
+      "train_tokens_per_second": 9113.646
+    },
+    {
+      "epoch": 1.9000205296653665,
+      "grad_norm": 0.7086114655308235,
+      "learning_rate": 6.178551590986959e-08,
+      "loss": 0.2651,
+      "num_input_tokens_seen": 509952032,
+      "step": 9255,
+      "train_runtime": 55957.3787,
+      "train_tokens_per_second": 9113.222
+    },
+    {
+      "epoch": 1.9010470129336892,
+      "grad_norm": 0.6466709332820999,
+      "learning_rate": 6.05284530711342e-08,
+      "loss": 0.2801,
+      "num_input_tokens_seen": 510253696,
+      "step": 9260,
+      "train_runtime": 55992.5786,
+      "train_tokens_per_second": 9112.881
+    },
+    {
+      "epoch": 1.902073496202012,
+      "grad_norm": 0.7458730883258999,
+      "learning_rate": 5.9284231973134045e-08,
+      "loss": 0.2964,
+      "num_input_tokens_seen": 510548272,
+      "step": 9265,
+      "train_runtime": 56021.5058,
+      "train_tokens_per_second": 9113.434
+    },
+    {
+      "epoch": 1.9030999794703347,
+      "grad_norm": 0.6567662493423765,
+      "learning_rate": 5.805285585061937e-08,
+      "loss": 0.279,
+      "num_input_tokens_seen": 510814312,
+      "step": 9270,
+      "train_runtime": 56050.1716,
+      "train_tokens_per_second": 9113.519
+    },
+    {
+      "epoch": 1.9041264627386574,
+      "grad_norm": 0.7729282677760901,
+      "learning_rate": 5.683432790494936e-08,
+      "loss": 0.2842,
+      "num_input_tokens_seen": 511090296,
+      "step": 9275,
+      "train_runtime": 56084.9107,
+      "train_tokens_per_second": 9112.795
+    },
+    {
+      "epoch": 1.90515294600698,
+      "grad_norm": 0.6937012078776588,
+      "learning_rate": 5.562865130407713e-08,
+      "loss": 0.2804,
+      "num_input_tokens_seen": 511379304,
+      "step": 9280,
+      "train_runtime": 56112.7196,
+      "train_tokens_per_second": 9113.429
+    },
+    {
+      "epoch": 1.9061794292753027,
+      "grad_norm": 0.5859911011469947,
+      "learning_rate": 5.443582918254642e-08,
+      "loss": 0.2894,
+      "num_input_tokens_seen": 511657920,
+      "step": 9285,
+      "train_runtime": 56140.1076,
+      "train_tokens_per_second": 9113.946
+    },
+    {
+      "epoch": 1.9072059125436256,
+      "grad_norm": 0.877095942594919,
+      "learning_rate": 5.325586464148047e-08,
+      "loss": 0.2749,
+      "num_input_tokens_seen": 511943240,
+      "step": 9290,
+      "train_runtime": 56166.5595,
+      "train_tokens_per_second": 9114.734
+    },
+    {
+      "epoch": 1.9082323958119483,
+      "grad_norm": 0.64422198479806,
+      "learning_rate": 5.2088760748575386e-08,
+      "loss": 0.2922,
+      "num_input_tokens_seen": 512224176,
+      "step": 9295,
+      "train_runtime": 56196.9113,
+      "train_tokens_per_second": 9114.81
+    },
+    {
+      "epoch": 1.9092588790802711,
+      "grad_norm": 0.6195702570215084,
+      "learning_rate": 5.093452053809345e-08,
+      "loss": 0.2982,
+      "num_input_tokens_seen": 512495184,
+      "step": 9300,
+      "train_runtime": 56229.0227,
+      "train_tokens_per_second": 9114.425
+    },
+    {
+      "epoch": 1.9102853623485938,
+      "grad_norm": 0.7157754021637037,
+      "learning_rate": 4.9793147010850365e-08,
+      "loss": 0.3006,
+      "num_input_tokens_seen": 512768104,
+      "step": 9305,
+      "train_runtime": 56257.0789,
+      "train_tokens_per_second": 9114.73
+    },
+    {
+      "epoch": 1.9113118456169165,
+      "grad_norm": 0.7011164456642937,
+      "learning_rate": 4.866464313421415e-08,
+      "loss": 0.2979,
+      "num_input_tokens_seen": 513059352,
+      "step": 9310,
+      "train_runtime": 56290.9965,
+      "train_tokens_per_second": 9114.412
+    },
+    {
+      "epoch": 1.9123383288852391,
+      "grad_norm": 0.7020595956604242,
+      "learning_rate": 4.754901184209293e-08,
+      "loss": 0.2988,
+      "num_input_tokens_seen": 513343080,
+      "step": 9315,
+      "train_runtime": 56327.3162,
+      "train_tokens_per_second": 9113.572
+    },
+    {
+      "epoch": 1.9133648121535618,
+      "grad_norm": 0.7397659064274764,
+      "learning_rate": 4.644625603492658e-08,
+      "loss": 0.2786,
+      "num_input_tokens_seen": 513611320,
+      "step": 9320,
+      "train_runtime": 56357.8761,
+      "train_tokens_per_second": 9113.39
+    },
+    {
+      "epoch": 1.9143912954218845,
+      "grad_norm": 0.6635771137019438,
+      "learning_rate": 4.535637857968345e-08,
+      "loss": 0.2755,
+      "num_input_tokens_seen": 513896200,
+      "step": 9325,
+      "train_runtime": 56388.548,
+      "train_tokens_per_second": 9113.485
+    },
+    {
+      "epoch": 1.9154177786902074,
+      "grad_norm": 0.6311994826367897,
+      "learning_rate": 4.427938230984974e-08,
+      "loss": 0.307,
+      "num_input_tokens_seen": 514156488,
+      "step": 9330,
+      "train_runtime": 56418.3609,
+      "train_tokens_per_second": 9113.283
+    },
+    {
+      "epoch": 1.9164442619585302,
+      "grad_norm": 0.7963067842223635,
+      "learning_rate": 4.321527002542292e-08,
+      "loss": 0.2905,
+      "num_input_tokens_seen": 514451424,
+      "step": 9335,
+      "train_runtime": 56453.7203,
+      "train_tokens_per_second": 9112.799
+    },
+    {
+      "epoch": 1.917470745226853,
+      "grad_norm": 0.816528937373697,
+      "learning_rate": 4.216404449290334e-08,
+      "loss": 0.2957,
+      "num_input_tokens_seen": 514703760,
+      "step": 9340,
+      "train_runtime": 56486.7499,
+      "train_tokens_per_second": 9111.938
+    },
+    {
+      "epoch": 1.9184972284951756,
+      "grad_norm": 0.6869403124842147,
+      "learning_rate": 4.112570844528929e-08,
+      "loss": 0.2835,
+      "num_input_tokens_seen": 514993168,
+      "step": 9345,
+      "train_runtime": 56511.5673,
+      "train_tokens_per_second": 9113.058
+    },
+    {
+      "epoch": 1.9195237117634982,
+      "grad_norm": 0.6603914676499402,
+      "learning_rate": 4.010026458206917e-08,
+      "loss": 0.2963,
+      "num_input_tokens_seen": 515261360,
+      "step": 9350,
+      "train_runtime": 56541.2626,
+      "train_tokens_per_second": 9113.015
+    },
+    {
+      "epoch": 1.920550195031821,
+      "grad_norm": 0.5987141398508651,
+      "learning_rate": 3.9087715569212094e-08,
+      "loss": 0.2912,
+      "num_input_tokens_seen": 515548152,
+      "step": 9355,
+      "train_runtime": 56568.3505,
+      "train_tokens_per_second": 9113.721
+    },
+    {
+      "epoch": 1.9215766783001436,
+      "grad_norm": 0.7020222636516997,
+      "learning_rate": 3.808806403916454e-08,
+      "loss": 0.2819,
+      "num_input_tokens_seen": 515825992,
+      "step": 9360,
+      "train_runtime": 56599.6457,
+      "train_tokens_per_second": 9113.59
+    },
+    {
+      "epoch": 1.9226031615684664,
+      "grad_norm": 0.7106311510331623,
+      "learning_rate": 3.7101312590842045e-08,
+      "loss": 0.2946,
+      "num_input_tokens_seen": 516093168,
+      "step": 9365,
+      "train_runtime": 56632.4126,
+      "train_tokens_per_second": 9113.035
+    },
+    {
+      "epoch": 1.923629644836789,
+      "grad_norm": 0.7661414284452,
+      "learning_rate": 3.612746378962029e-08,
+      "loss": 0.3067,
+      "num_input_tokens_seen": 516376456,
+      "step": 9370,
+      "train_runtime": 56663.4522,
+      "train_tokens_per_second": 9113.043
+    },
+    {
+      "epoch": 1.924656128105112,
+      "grad_norm": 0.6848063411498132,
+      "learning_rate": 3.5166520167331795e-08,
+      "loss": 0.2898,
+      "num_input_tokens_seen": 516668944,
+      "step": 9375,
+      "train_runtime": 56693.6496,
+      "train_tokens_per_second": 9113.348
+    },
+    {
+      "epoch": 1.9256826113734347,
+      "grad_norm": 0.7016162217768939,
+      "learning_rate": 3.421848422225871e-08,
+      "loss": 0.2905,
+      "num_input_tokens_seen": 516951352,
+      "step": 9380,
+      "train_runtime": 56729.0263,
+      "train_tokens_per_second": 9112.643
+    },
+    {
+      "epoch": 1.9267090946417573,
+      "grad_norm": 0.7143628250978753,
+      "learning_rate": 3.328335841912333e-08,
+      "loss": 0.311,
+      "num_input_tokens_seen": 517241712,
+      "step": 9385,
+      "train_runtime": 56754.5827,
+      "train_tokens_per_second": 9113.655
+    },
+    {
+      "epoch": 1.92773557791008,
+      "grad_norm": 0.6368044505307717,
+      "learning_rate": 3.2361145189086504e-08,
+      "loss": 0.2928,
+      "num_input_tokens_seen": 517519568,
+      "step": 9390,
+      "train_runtime": 56783.5955,
+      "train_tokens_per_second": 9113.892
+    },
+    {
+      "epoch": 1.9287620611784027,
+      "grad_norm": 0.6800929312946785,
+      "learning_rate": 3.145184692973591e-08,
+      "loss": 0.2791,
+      "num_input_tokens_seen": 517787512,
+      "step": 9395,
+      "train_runtime": 56820.8465,
+      "train_tokens_per_second": 9112.633
+    },
+    {
+      "epoch": 1.9297885444467255,
+      "grad_norm": 0.6288739763376526,
+      "learning_rate": 3.055546600508552e-08,
+      "loss": 0.255,
+      "num_input_tokens_seen": 518056840,
+      "step": 9400,
+      "train_runtime": 56843.5896,
+      "train_tokens_per_second": 9113.725
+    },
+    {
+      "epoch": 1.9308150277150482,
+      "grad_norm": 0.655490554899647,
+      "learning_rate": 2.9672004745563976e-08,
+      "loss": 0.3015,
+      "num_input_tokens_seen": 518361144,
+      "step": 9405,
+      "train_runtime": 56877.638,
+      "train_tokens_per_second": 9113.619
+    },
+    {
+      "epoch": 1.931841510983371,
+      "grad_norm": 0.601737159777282,
+      "learning_rate": 2.8801465448012323e-08,
+      "loss": 0.2686,
+      "num_input_tokens_seen": 518645080,
+      "step": 9410,
+      "train_runtime": 56911.9092,
+      "train_tokens_per_second": 9113.12
+    },
+    {
+      "epoch": 1.9328679942516938,
+      "grad_norm": 0.6393932758317636,
+      "learning_rate": 2.7943850375677373e-08,
+      "loss": 0.2863,
+      "num_input_tokens_seen": 518937328,
+      "step": 9415,
+      "train_runtime": 56942.6548,
+      "train_tokens_per_second": 9113.332
+    },
+    {
+      "epoch": 1.9338944775200164,
+      "grad_norm": 0.7318580723661846,
+      "learning_rate": 2.709916175820393e-08,
+      "loss": 0.2863,
+      "num_input_tokens_seen": 519221168,
+      "step": 9420,
+      "train_runtime": 56980.076,
+      "train_tokens_per_second": 9112.328
+    },
+    {
+      "epoch": 1.934920960788339,
+      "grad_norm": 0.6852406399716597,
+      "learning_rate": 2.6267401791630896e-08,
+      "loss": 0.2601,
+      "num_input_tokens_seen": 519496968,
+      "step": 9425,
+      "train_runtime": 57008.6929,
+      "train_tokens_per_second": 9112.592
+    },
+    {
+      "epoch": 1.9359474440566617,
+      "grad_norm": 0.6720589982705765,
+      "learning_rate": 2.5448572638385738e-08,
+      "loss": 0.2789,
+      "num_input_tokens_seen": 519807008,
+      "step": 9430,
+      "train_runtime": 57044.3814,
+      "train_tokens_per_second": 9112.326
+    },
+    {
+      "epoch": 1.9369739273249846,
+      "grad_norm": 0.6625118905638145,
+      "learning_rate": 2.4642676427277244e-08,
+      "loss": 0.2814,
+      "num_input_tokens_seen": 520094224,
+      "step": 9435,
+      "train_runtime": 57075.4781,
+      "train_tokens_per_second": 9112.394
+    },
+    {
+      "epoch": 1.9380004105933073,
+      "grad_norm": 0.629884088643173,
+      "learning_rate": 2.3849715253491114e-08,
+      "loss": 0.2969,
+      "num_input_tokens_seen": 520358200,
+      "step": 9440,
+      "train_runtime": 57098.9266,
+      "train_tokens_per_second": 9113.275
+    },
+    {
+      "epoch": 1.9390268938616302,
+      "grad_norm": 0.6089221851206861,
+      "learning_rate": 2.306969117858493e-08,
+      "loss": 0.2858,
+      "num_input_tokens_seen": 520641584,
+      "step": 9445,
+      "train_runtime": 57133.5203,
+      "train_tokens_per_second": 9112.717
+    },
+    {
+      "epoch": 1.9400533771299528,
+      "grad_norm": 0.690461218936342,
+      "learning_rate": 2.2302606230479863e-08,
+      "loss": 0.2621,
+      "num_input_tokens_seen": 520939208,
+      "step": 9450,
+      "train_runtime": 57162.0476,
+      "train_tokens_per_second": 9113.376
+    },
+    {
+      "epoch": 1.9410798603982755,
+      "grad_norm": 0.7320254388212963,
+      "learning_rate": 2.1548462403460646e-08,
+      "loss": 0.3113,
+      "num_input_tokens_seen": 521218248,
+      "step": 9455,
+      "train_runtime": 57188.4535,
+      "train_tokens_per_second": 9114.047
+    },
+    {
+      "epoch": 1.9421063436665982,
+      "grad_norm": 0.7502424000072345,
+      "learning_rate": 2.08072616581656e-08,
+      "loss": 0.2975,
+      "num_input_tokens_seen": 521504424,
+      "step": 9460,
+      "train_runtime": 57220.2517,
+      "train_tokens_per_second": 9113.983
+    },
+    {
+      "epoch": 1.9431328269349208,
+      "grad_norm": 0.6312674561484174,
+      "learning_rate": 2.0079005921584405e-08,
+      "loss": 0.3034,
+      "num_input_tokens_seen": 521771032,
+      "step": 9465,
+      "train_runtime": 57254.7772,
+      "train_tokens_per_second": 9113.144
+    },
+    {
+      "epoch": 1.9441593102032437,
+      "grad_norm": 0.6433596987925244,
+      "learning_rate": 1.9363697087050327e-08,
+      "loss": 0.3077,
+      "num_input_tokens_seen": 522065352,
+      "step": 9470,
+      "train_runtime": 57278.6518,
+      "train_tokens_per_second": 9114.484
+    },
+    {
+      "epoch": 1.9451857934715664,
+      "grad_norm": 0.7956465936273058,
+      "learning_rate": 1.8661337014239113e-08,
+      "loss": 0.2963,
+      "num_input_tokens_seen": 522345744,
+      "step": 9475,
+      "train_runtime": 57308.0677,
+      "train_tokens_per_second": 9114.698
+    },
+    {
+      "epoch": 1.9462122767398893,
+      "grad_norm": 0.6764328517353161,
+      "learning_rate": 1.797192752916066e-08,
+      "loss": 0.2929,
+      "num_input_tokens_seen": 522632888,
+      "step": 9480,
+      "train_runtime": 57342.9189,
+      "train_tokens_per_second": 9114.166
+    },
+    {
+      "epoch": 1.947238760008212,
+      "grad_norm": 0.7859378619881348,
+      "learning_rate": 1.729547042415569e-08,
+      "loss": 0.3063,
+      "num_input_tokens_seen": 522908096,
+      "step": 9485,
+      "train_runtime": 57374.7678,
+      "train_tokens_per_second": 9113.903
+    },
+    {
+      "epoch": 1.9482652432765346,
+      "grad_norm": 0.6529420326088011,
+      "learning_rate": 1.6631967457892418e-08,
+      "loss": 0.288,
+      "num_input_tokens_seen": 523185288,
+      "step": 9490,
+      "train_runtime": 57402.5711,
+      "train_tokens_per_second": 9114.318
+    },
+    {
+      "epoch": 1.9492917265448573,
+      "grad_norm": 0.7966892799968799,
+      "learning_rate": 1.5981420355358767e-08,
+      "loss": 0.2959,
+      "num_input_tokens_seen": 523452472,
+      "step": 9495,
+      "train_runtime": 57430.9607,
+      "train_tokens_per_second": 9114.465
+    },
+    {
+      "epoch": 1.95031820981318,
+      "grad_norm": 0.6533801077528296,
+      "learning_rate": 1.534383080786128e-08,
+      "loss": 0.2826,
+      "num_input_tokens_seen": 523731240,
+      "step": 9500,
+      "train_runtime": 57462.5834,
+      "train_tokens_per_second": 9114.3
+    },
+    {
+      "epoch": 1.9513446930815028,
+      "grad_norm": 0.8092724184082158,
+      "learning_rate": 1.4719200473017892e-08,
+      "loss": 0.2933,
+      "num_input_tokens_seen": 524015464,
+      "step": 9505,
+      "train_runtime": 57490.6036,
+      "train_tokens_per_second": 9114.802
+    },
+    {
+      "epoch": 1.9523711763498255,
+      "grad_norm": 0.7259010123781818,
+      "learning_rate": 1.41075309747557e-08,
+      "loss": 0.2722,
+      "num_input_tokens_seen": 524282400,
+      "step": 9510,
+      "train_runtime": 57517.1515,
+      "train_tokens_per_second": 9115.236
+    },
+    {
+      "epoch": 1.9533976596181484,
+      "grad_norm": 0.7260629954129617,
+      "learning_rate": 1.350882390330599e-08,
+      "loss": 0.2928,
+      "num_input_tokens_seen": 524560704,
+      "step": 9515,
+      "train_runtime": 57547.6662,
+      "train_tokens_per_second": 9115.239
+    },
+    {
+      "epoch": 1.954424142886471,
+      "grad_norm": 0.9717714899064241,
+      "learning_rate": 1.2923080815199773e-08,
+      "loss": 0.2794,
+      "num_input_tokens_seen": 524848128,
+      "step": 9520,
+      "train_runtime": 57575.2552,
+      "train_tokens_per_second": 9115.863
+    },
+    {
+      "epoch": 1.9554506261547937,
+      "grad_norm": 0.6518778091712374,
+      "learning_rate": 1.2350303233263916e-08,
+      "loss": 0.2847,
+      "num_input_tokens_seen": 525132800,
+      "step": 9525,
+      "train_runtime": 57608.6121,
+      "train_tokens_per_second": 9115.526
+    },
+    {
+      "epoch": 1.9564771094231164,
+      "grad_norm": 0.720570002471073,
+      "learning_rate": 1.1790492646617801e-08,
+      "loss": 0.3086,
+      "num_input_tokens_seen": 525394200,
+      "step": 9530,
+      "train_runtime": 57644.7368,
+      "train_tokens_per_second": 9114.348
+    },
+    {
+      "epoch": 1.957503592691439,
+      "grad_norm": 0.6570790068706449,
+      "learning_rate": 1.1243650510668336e-08,
+      "loss": 0.2838,
+      "num_input_tokens_seen": 525695504,
+      "step": 9535,
+      "train_runtime": 57672.7666,
+      "train_tokens_per_second": 9115.143
+    },
+    {
+      "epoch": 1.9585300759597617,
+      "grad_norm": 0.7456689173264177,
+      "learning_rate": 1.0709778247107172e-08,
+      "loss": 0.2783,
+      "num_input_tokens_seen": 525973336,
+      "step": 9540,
+      "train_runtime": 57701.5206,
+      "train_tokens_per_second": 9115.416
+    },
+    {
+      "epoch": 1.9595565592280846,
+      "grad_norm": 0.6310247362463759,
+      "learning_rate": 1.018887724390627e-08,
+      "loss": 0.307,
+      "num_input_tokens_seen": 526252112,
+      "step": 9545,
+      "train_runtime": 57731.3718,
+      "train_tokens_per_second": 9115.531
+    },
+    {
+      "epoch": 1.9605830424964075,
+      "grad_norm": 0.654456030588889,
+      "learning_rate": 9.680948855315675e-09,
+      "loss": 0.2865,
+      "num_input_tokens_seen": 526520928,
+      "step": 9550,
+      "train_runtime": 57762.7392,
+      "train_tokens_per_second": 9115.235
+    },
+    {
+      "epoch": 1.9616095257647301,
+      "grad_norm": 0.6321689639451823,
+      "learning_rate": 9.185994401856857e-09,
+      "loss": 0.2667,
+      "num_input_tokens_seen": 526805032,
+      "step": 9555,
+      "train_runtime": 57797.5174,
+      "train_tokens_per_second": 9114.665
+    },
+    {
+      "epoch": 1.9626360090330528,
+      "grad_norm": 0.6818705334335124,
+      "learning_rate": 8.704015170323821e-09,
+      "loss": 0.2821,
+      "num_input_tokens_seen": 527081496,
+      "step": 9560,
+      "train_runtime": 57829.2074,
+      "train_tokens_per_second": 9114.451
+    },
+    {
+      "epoch": 1.9636624923013755,
+      "grad_norm": 0.6435235949441107,
+      "learning_rate": 8.235012413775334e-09,
+      "loss": 0.2779,
+      "num_input_tokens_seen": 527378920,
+      "step": 9565,
+      "train_runtime": 57857.1997,
+      "train_tokens_per_second": 9115.182
+    },
+    {
+      "epoch": 1.9646889755696981,
+      "grad_norm": 0.6504897607750524,
+      "learning_rate": 7.778987351535484e-09,
+      "loss": 0.2838,
+      "num_input_tokens_seen": 527649736,
+      "step": 9570,
+      "train_runtime": 57883.5805,
+      "train_tokens_per_second": 9115.707
+    },
+    {
+      "epoch": 1.9657154588380208,
+      "grad_norm": 0.681801684505555,
+      "learning_rate": 7.335941169187011e-09,
+      "loss": 0.3049,
+      "num_input_tokens_seen": 527946376,
+      "step": 9575,
+      "train_runtime": 57917.4646,
+      "train_tokens_per_second": 9115.495
+    },
+    {
+      "epoch": 1.9667419421063437,
+      "grad_norm": 0.7133616820030736,
+      "learning_rate": 6.905875018570762e-09,
+      "loss": 0.2838,
+      "num_input_tokens_seen": 528233288,
+      "step": 9580,
+      "train_runtime": 57945.4337,
+      "train_tokens_per_second": 9116.047
+    },
+    {
+      "epoch": 1.9677684253746663,
+      "grad_norm": 0.6078797158347238,
+      "learning_rate": 6.4887900177812435e-09,
+      "loss": 0.2623,
+      "num_input_tokens_seen": 528531528,
+      "step": 9585,
+      "train_runtime": 57980.541,
+      "train_tokens_per_second": 9115.671
+    },
+    {
+      "epoch": 1.9687949086429892,
+      "grad_norm": 0.7819374506219917,
+      "learning_rate": 6.084687251165511e-09,
+      "loss": 0.2853,
+      "num_input_tokens_seen": 528778992,
+      "step": 9590,
+      "train_runtime": 58003.9701,
+      "train_tokens_per_second": 9116.255
+    },
+    {
+      "epoch": 1.969821391911312,
+      "grad_norm": 0.7877024826192885,
+      "learning_rate": 5.69356776931762e-09,
+      "loss": 0.2928,
+      "num_input_tokens_seen": 529038952,
+      "step": 9595,
+      "train_runtime": 58029.6026,
+      "train_tokens_per_second": 9116.708
+    },
+    {
+      "epoch": 1.9708478751796346,
+      "grad_norm": 0.8056159209699786,
+      "learning_rate": 5.315432589078628e-09,
+      "loss": 0.3106,
+      "num_input_tokens_seen": 529312392,
+      "step": 9600,
+      "train_runtime": 58060.934,
+      "train_tokens_per_second": 9116.498
+    },
+    {
+      "epoch": 1.9718743584479572,
+      "grad_norm": 0.801488290798269,
+      "learning_rate": 4.950282693532149e-09,
+      "loss": 0.2788,
+      "num_input_tokens_seen": 529594024,
+      "step": 9605,
+      "train_runtime": 58092.451,
+      "train_tokens_per_second": 9116.4
+    },
+    {
+      "epoch": 1.9729008417162799,
+      "grad_norm": 0.6675829345462915,
+      "learning_rate": 4.59811903200158e-09,
+      "loss": 0.2917,
+      "num_input_tokens_seen": 529883040,
+      "step": 9610,
+      "train_runtime": 58117.5639,
+      "train_tokens_per_second": 9117.434
+    },
+    {
+      "epoch": 1.9739273249846028,
+      "grad_norm": 0.7212386161127493,
+      "learning_rate": 4.258942520050657e-09,
+      "loss": 0.3042,
+      "num_input_tokens_seen": 530139240,
+      "step": 9615,
+      "train_runtime": 58147.668,
+      "train_tokens_per_second": 9117.12
+    },
+    {
+      "epoch": 1.9749538082529254,
+      "grad_norm": 0.6455260055345227,
+      "learning_rate": 3.932754039476794e-09,
+      "loss": 0.3,
+      "num_input_tokens_seen": 530419088,
+      "step": 9620,
+      "train_runtime": 58177.3164,
+      "train_tokens_per_second": 9117.284
+    },
+    {
+      "epoch": 1.9759802915212483,
+      "grad_norm": 0.6163665804529724,
+      "learning_rate": 3.6195544383121896e-09,
+      "loss": 0.271,
+      "num_input_tokens_seen": 530723232,
+      "step": 9625,
+      "train_runtime": 58208.4688,
+      "train_tokens_per_second": 9117.629
+    },
+    {
+      "epoch": 1.977006774789571,
+      "grad_norm": 0.6317219285171051,
+      "learning_rate": 3.3193445308188356e-09,
+      "loss": 0.3086,
+      "num_input_tokens_seen": 531016456,
+      "step": 9630,
+      "train_runtime": 58238.9654,
+      "train_tokens_per_second": 9117.89
+    },
+    {
+      "epoch": 1.9780332580578937,
+      "grad_norm": 0.6787497431595687,
+      "learning_rate": 3.0321250974890693e-09,
+      "loss": 0.296,
+      "num_input_tokens_seen": 531279256,
+      "step": 9635,
+      "train_runtime": 58271.49,
+      "train_tokens_per_second": 9117.31
+    },
+    {
+      "epoch": 1.9790597413262163,
+      "grad_norm": 0.6677854431057147,
+      "learning_rate": 2.757896885041689e-09,
+      "loss": 0.2985,
+      "num_input_tokens_seen": 531514264,
+      "step": 9640,
+      "train_runtime": 58292.7719,
+      "train_tokens_per_second": 9118.013
+    },
+    {
+      "epoch": 1.980086224594539,
+      "grad_norm": 0.8502043376636467,
+      "learning_rate": 2.496660606421397e-09,
+      "loss": 0.2992,
+      "num_input_tokens_seen": 531784648,
+      "step": 9645,
+      "train_runtime": 58321.0579,
+      "train_tokens_per_second": 9118.227
+    },
+    {
+      "epoch": 1.9811127078628619,
+      "grad_norm": 0.6453819055522446,
+      "learning_rate": 2.2484169407954727e-09,
+      "loss": 0.2955,
+      "num_input_tokens_seen": 532076072,
+      "step": 9650,
+      "train_runtime": 58354.033,
+      "train_tokens_per_second": 9118.069
+    },
+    {
+      "epoch": 1.9821391911311845,
+      "grad_norm": 0.5953601148461377,
+      "learning_rate": 2.0131665335526572e-09,
+      "loss": 0.2892,
+      "num_input_tokens_seen": 532368344,
+      "step": 9655,
+      "train_runtime": 58378.8041,
+      "train_tokens_per_second": 9119.206
+    },
+    {
+      "epoch": 1.9831656743995074,
+      "grad_norm": 0.6344066071096935,
+      "learning_rate": 1.7909099963020483e-09,
+      "loss": 0.2644,
+      "num_input_tokens_seen": 532639944,
+      "step": 9660,
+      "train_runtime": 58402.7125,
+      "train_tokens_per_second": 9120.123
+    },
+    {
+      "epoch": 1.98419215766783,
+      "grad_norm": 0.6418152224212895,
+      "learning_rate": 1.5816479068714307e-09,
+      "loss": 0.2895,
+      "num_input_tokens_seen": 532938552,
+      "step": 9665,
+      "train_runtime": 58437.4058,
+      "train_tokens_per_second": 9119.819
+    },
+    {
+      "epoch": 1.9852186409361527,
+      "grad_norm": 0.8153042232644806,
+      "learning_rate": 1.385380809303949e-09,
+      "loss": 0.2911,
+      "num_input_tokens_seen": 533201104,
+      "step": 9670,
+      "train_runtime": 58471.963,
+      "train_tokens_per_second": 9118.919
+    },
+    {
+      "epoch": 1.9862451242044754,
+      "grad_norm": 0.6067223459507053,
+      "learning_rate": 1.20210921385866e-09,
+      "loss": 0.2833,
+      "num_input_tokens_seen": 533471776,
+      "step": 9675,
+      "train_runtime": 58503.6622,
+      "train_tokens_per_second": 9118.605
+    },
+    {
+      "epoch": 1.987271607472798,
+      "grad_norm": 0.7957783059246915,
+      "learning_rate": 1.0318335970099791e-09,
+      "loss": 0.2918,
+      "num_input_tokens_seen": 533740120,
+      "step": 9680,
+      "train_runtime": 58531.7178,
+      "train_tokens_per_second": 9118.819
+    },
+    {
+      "epoch": 1.988298090741121,
+      "grad_norm": 0.8078959121515734,
+      "learning_rate": 8.745544014426844e-10,
+      "loss": 0.2839,
+      "num_input_tokens_seen": 534015856,
+      "step": 9685,
+      "train_runtime": 58562.1508,
+      "train_tokens_per_second": 9118.788
+    },
+    {
+      "epoch": 1.9893245740094436,
+      "grad_norm": 0.7359948479204134,
+      "learning_rate": 7.302720360552462e-10,
+      "loss": 0.2975,
+      "num_input_tokens_seen": 534276760,
+      "step": 9690,
+      "train_runtime": 58588.8359,
+      "train_tokens_per_second": 9119.088
+    },
+    {
+      "epoch": 1.9903510572777665,
+      "grad_norm": 0.5793264326865378,
+      "learning_rate": 5.989868759559425e-10,
+      "loss": 0.2597,
+      "num_input_tokens_seen": 534558760,
+      "step": 9695,
+      "train_runtime": 58615.4618,
+      "train_tokens_per_second": 9119.757
+    },
+    {
+      "epoch": 1.9913775405460892,
+      "grad_norm": 0.7533520629058247,
+      "learning_rate": 4.806992624628581e-10,
+      "loss": 0.2847,
+      "num_input_tokens_seen": 534814336,
+      "step": 9700,
+      "train_runtime": 58645.8303,
+      "train_tokens_per_second": 9119.392
+    },
+    {
+      "epoch": 1.9924040238144118,
+      "grad_norm": 0.679450800667022,
+      "learning_rate": 3.7540950310166466e-10,
+      "loss": 0.2894,
+      "num_input_tokens_seen": 535062656,
+      "step": 9705,
+      "train_runtime": 58674.3112,
+      "train_tokens_per_second": 9119.198
+    },
+    {
+      "epoch": 1.9934305070827345,
+      "grad_norm": 0.6586268502996842,
+      "learning_rate": 2.8311787160784087e-10,
+      "loss": 0.3159,
+      "num_input_tokens_seen": 535331224,
+      "step": 9710,
+      "train_runtime": 58703.17,
+      "train_tokens_per_second": 9119.29
+    },
+    {
+      "epoch": 1.9944569903510572,
+      "grad_norm": 0.6080625476203282,
+      "learning_rate": 2.038246079227868e-10,
+      "loss": 0.3062,
+      "num_input_tokens_seen": 535606440,
+      "step": 9715,
+      "train_runtime": 58733.7546,
+      "train_tokens_per_second": 9119.227
+    },
+    {
+      "epoch": 1.99548347361938,
+      "grad_norm": 0.6811932961389373,
+      "learning_rate": 1.3752991819493411e-10,
+      "loss": 0.2926,
+      "num_input_tokens_seen": 535869248,
+      "step": 9720,
+      "train_runtime": 58762.8206,
+      "train_tokens_per_second": 9119.189
+    },
+    {
+      "epoch": 1.9965099568877027,
+      "grad_norm": 0.6864643695296379,
+      "learning_rate": 8.42339747780807e-11,
+      "loss": 0.2711,
+      "num_input_tokens_seen": 536114296,
+      "step": 9725,
+      "train_runtime": 58789.2028,
+      "train_tokens_per_second": 9119.265
+    },
+    {
+      "epoch": 1.9975364401560256,
+      "grad_norm": 0.6015993388509988,
+      "learning_rate": 4.3936916232500956e-11,
+      "loss": 0.2773,
+      "num_input_tokens_seen": 536421928,
+      "step": 9730,
+      "train_runtime": 58815.9799,
+      "train_tokens_per_second": 9120.343
+    },
+    {
+      "epoch": 1.9985629234243483,
+      "grad_norm": 0.7092101363718902,
+      "learning_rate": 1.663884732272525e-11,
+      "loss": 0.281,
+      "num_input_tokens_seen": 536687712,
+      "step": 9735,
+      "train_runtime": 58847.6181,
+      "train_tokens_per_second": 9119.956
+    },
+    {
+      "epoch": 1.999589406692671,
+      "grad_norm": 0.6579909231361643,
+      "learning_rate": 2.339839019760426e-12,
+      "loss": 0.2846,
+      "num_input_tokens_seen": 536964728,
+      "step": 9740,
+      "train_runtime": 58875.4329,
+      "train_tokens_per_second": 9120.353
+    },
+    {
+      "epoch": 2.0,
+      "num_input_tokens_seen": 537080160,
+      "step": 9742,
+      "total_flos": 993245970989056.0,
+      "train_loss": 0.3348774538836032,
+      "train_runtime": 59003.2702,
+      "train_samples_per_second": 0.66,
+      "train_steps_per_second": 0.165
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 9742,
+  "num_input_tokens_seen": 537080160,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 993245970989056.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}