diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.2711810543175663,
+  "eval_steps": 500,
+  "global_step": 200000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006355905271587832,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 2.3296,
+      "step": 100
+    },
+    {
+      "epoch": 0.0012711810543175664,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 2.3266,
+      "step": 200
+    },
+    {
+      "epoch": 0.0019067715814763497,
+      "grad_norm": 3.625,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 2.3437,
+      "step": 300
+    },
+    {
+      "epoch": 0.002542362108635133,
+      "grad_norm": 3.640625,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 2.3306,
+      "step": 400
+    },
+    {
+      "epoch": 0.003177952635793916,
+      "grad_norm": 3.25,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 2.3232,
+      "step": 500
+    },
+    {
+      "epoch": 0.0038135431629526995,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 2.3369,
+      "step": 600
+    },
+    {
+      "epoch": 0.004449133690111483,
+      "grad_norm": 4.125,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 2.3436,
+      "step": 700
+    },
+    {
+      "epoch": 0.005084724217270266,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 2.3334,
+      "step": 800
+    },
+    {
+      "epoch": 0.0057203147444290494,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 2.3337,
+      "step": 900
+    },
+    {
+      "epoch": 0.006355905271587832,
+      "grad_norm": 4.65625,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 2.3052,
+      "step": 1000
+    },
+    {
+      "epoch": 0.006991495798746615,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.2e-06,
+      "loss": 2.3237,
+      "step": 1100
+    },
+    {
+      "epoch": 0.007627086325905399,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 2.3015,
+      "step": 1200
+    },
+    {
+      "epoch": 0.008262676853064182,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.6e-06,
+      "loss": 2.3166,
+      "step": 1300
+    },
+    {
+      "epoch": 0.008898267380222966,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 2.2848,
+      "step": 1400
+    },
+    {
+      "epoch": 0.009533857907381748,
+      "grad_norm": 2.453125,
+      "learning_rate": 3e-06,
+      "loss": 2.2821,
+      "step": 1500
+    },
+    {
+      "epoch": 0.010169448434540531,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 2.2781,
+      "step": 1600
+    },
+    {
+      "epoch": 0.010805038961699315,
+      "grad_norm": 3.0,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 2.2595,
+      "step": 1700
+    },
+    {
+      "epoch": 0.011440629488858099,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 2.2409,
+      "step": 1800
+    },
+    {
+      "epoch": 0.012076220016016881,
+      "grad_norm": 2.9375,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 2.2393,
+      "step": 1900
+    },
+    {
+      "epoch": 0.012711810543175665,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.2361,
+      "step": 2000
+    },
+    {
+      "epoch": 0.013347401070334448,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 2.2238,
+      "step": 2100
+    },
+    {
+      "epoch": 0.01398299159749323,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.4e-06,
+      "loss": 2.2031,
+      "step": 2200
+    },
+    {
+      "epoch": 0.014618582124652014,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 2.1762,
+      "step": 2300
+    },
+    {
+      "epoch": 0.015254172651810798,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 2.1772,
+      "step": 2400
+    },
+    {
+      "epoch": 0.01588976317896958,
+      "grad_norm": 2.265625,
+      "learning_rate": 5e-06,
+      "loss": 2.147,
+      "step": 2500
+    },
+    {
+      "epoch": 0.016525353706128364,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.2e-06,
+      "loss": 2.1295,
+      "step": 2600
+    },
+    {
+      "epoch": 0.017160944233287147,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 2.0829,
+      "step": 2700
+    },
+    {
+      "epoch": 0.01779653476044593,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 2.082,
+      "step": 2800
+    },
+    {
+      "epoch": 0.018432125287604715,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.8e-06,
+      "loss": 2.02,
+      "step": 2900
+    },
+    {
+      "epoch": 0.019067715814763495,
+      "grad_norm": 3.53125,
+      "learning_rate": 6e-06,
+      "loss": 2.003,
+      "step": 3000
+    },
+    {
+      "epoch": 0.01970330634192228,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 1.9711,
+      "step": 3100
+    },
+    {
+      "epoch": 0.020338896869081063,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 1.9499,
+      "step": 3200
+    },
+    {
+      "epoch": 0.020974487396239846,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 1.9166,
+      "step": 3300
+    },
+    {
+      "epoch": 0.02161007792339863,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 1.9366,
+      "step": 3400
+    },
+    {
+      "epoch": 0.022245668450557414,
+      "grad_norm": 1.015625,
+      "learning_rate": 7e-06,
+      "loss": 1.9282,
+      "step": 3500
+    },
+    {
+      "epoch": 0.022881258977716198,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 1.9043,
+      "step": 3600
+    },
+    {
+      "epoch": 0.023516849504874978,
+      "grad_norm": 1.421875,
+      "learning_rate": 7.4e-06,
+      "loss": 1.8715,
+      "step": 3700
+    },
+    {
+      "epoch": 0.024152440032033762,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 1.8862,
+      "step": 3800
+    },
+    {
+      "epoch": 0.024788030559192546,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 1.8584,
+      "step": 3900
+    },
+    {
+      "epoch": 0.02542362108635133,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.8515,
+      "step": 4000
+    },
+    {
+      "epoch": 0.026059211613510113,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.2e-06,
+      "loss": 1.8317,
+      "step": 4100
+    },
+    {
+      "epoch": 0.026694802140668897,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 1.8488,
+      "step": 4200
+    },
+    {
+      "epoch": 0.027330392667827677,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.6e-06,
+      "loss": 1.8256,
+      "step": 4300
+    },
+    {
+      "epoch": 0.02796598319498646,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.8e-06,
+      "loss": 1.8079,
+      "step": 4400
+    },
+    {
+      "epoch": 0.028601573722145245,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9e-06,
+      "loss": 1.8301,
+      "step": 4500
+    },
+    {
+      "epoch": 0.02923716424930403,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 1.8207,
+      "step": 4600
+    },
+    {
+      "epoch": 0.029872754776462812,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.4e-06,
+      "loss": 1.8153,
+      "step": 4700
+    },
+    {
+      "epoch": 0.030508345303621596,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 1.7974,
+      "step": 4800
+    },
+    {
+      "epoch": 0.03114393583078038,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 1.7669,
+      "step": 4900
+    },
+    {
+      "epoch": 0.03177952635793916,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1e-05,
+      "loss": 1.7712,
+      "step": 5000
+    },
+    {
+      "epoch": 0.03241511688509795,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.994871794871795e-06,
+      "loss": 1.7679,
+      "step": 5100
+    },
+    {
+      "epoch": 0.03305070741225673,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.98974358974359e-06,
+      "loss": 1.7683,
+      "step": 5200
+    },
+    {
+      "epoch": 0.03368629793941551,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.984615384615386e-06,
+      "loss": 1.777,
+      "step": 5300
+    },
+    {
+      "epoch": 0.034321888466574295,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.97948717948718e-06,
+      "loss": 1.7545,
+      "step": 5400
+    },
+    {
+      "epoch": 0.034957478993733075,
+      "grad_norm": 0.4765625,
+      "learning_rate": 9.974358974358974e-06,
+      "loss": 1.7718,
+      "step": 5500
+    },
+    {
+      "epoch": 0.03559306952089186,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.96923076923077e-06,
+      "loss": 1.7687,
+      "step": 5600
+    },
+    {
+      "epoch": 0.03622866004805064,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.964102564102564e-06,
+      "loss": 1.7664,
+      "step": 5700
+    },
+    {
+      "epoch": 0.03686425057520943,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.95897435897436e-06,
+      "loss": 1.768,
+      "step": 5800
+    },
+    {
+      "epoch": 0.03749984110236821,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.953846153846156e-06,
+      "loss": 1.7759,
+      "step": 5900
+    },
+    {
+      "epoch": 0.03813543162952699,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.94871794871795e-06,
+      "loss": 1.7379,
+      "step": 6000
+    },
+    {
+      "epoch": 0.03877102215668578,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.943589743589744e-06,
+      "loss": 1.7611,
+      "step": 6100
+    },
+    {
+      "epoch": 0.03940661268384456,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.93846153846154e-06,
+      "loss": 1.7453,
+      "step": 6200
+    },
+    {
+      "epoch": 0.040042203211003345,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.933333333333334e-06,
+      "loss": 1.7493,
+      "step": 6300
+    },
+    {
+      "epoch": 0.040677793738162125,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.92820512820513e-06,
+      "loss": 1.7308,
+      "step": 6400
+    },
+    {
+      "epoch": 0.04131338426532091,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.923076923076923e-06,
+      "loss": 1.7557,
+      "step": 6500
+    },
+    {
+      "epoch": 0.04194897479247969,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.91794871794872e-06,
+      "loss": 1.7244,
+      "step": 6600
+    },
+    {
+      "epoch": 0.04258456531963847,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.912820512820513e-06,
+      "loss": 1.7365,
+      "step": 6700
+    },
+    {
+      "epoch": 0.04322015584679726,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.907692307692309e-06,
+      "loss": 1.7416,
+      "step": 6800
+    },
+    {
+      "epoch": 0.04385574637395604,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.902564102564103e-06,
+      "loss": 1.744,
+      "step": 6900
+    },
+    {
+      "epoch": 0.04449133690111483,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.897435897435899e-06,
+      "loss": 1.7343,
+      "step": 7000
+    },
+    {
+      "epoch": 0.04512692742827361,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.892307692307693e-06,
+      "loss": 1.7476,
+      "step": 7100
+    },
+    {
+      "epoch": 0.045762517955432395,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.887179487179489e-06,
+      "loss": 1.7382,
+      "step": 7200
+    },
+    {
+      "epoch": 0.046398108482591176,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.882051282051283e-06,
+      "loss": 1.7376,
+      "step": 7300
+    },
+    {
+      "epoch": 0.047033699009749956,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.876923076923077e-06,
+      "loss": 1.73,
+      "step": 7400
+    },
+    {
+      "epoch": 0.04766928953690874,
+      "grad_norm": 0.875,
+      "learning_rate": 9.871794871794872e-06,
+      "loss": 1.7139,
+      "step": 7500
+    },
+    {
+      "epoch": 0.048304880064067524,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.866666666666668e-06,
+      "loss": 1.7271,
+      "step": 7600
+    },
+    {
+      "epoch": 0.04894047059122631,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.861538461538462e-06,
+      "loss": 1.728,
+      "step": 7700
+    },
+    {
+      "epoch": 0.04957606111838509,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.856410256410256e-06,
+      "loss": 1.7238,
+      "step": 7800
+    },
+    {
+      "epoch": 0.05021165164554388,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.851282051282052e-06,
+      "loss": 1.7245,
+      "step": 7900
+    },
+    {
+      "epoch": 0.05084724217270266,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.846153846153848e-06,
+      "loss": 1.7213,
+      "step": 8000
+    },
+    {
+      "epoch": 0.05148283269986144,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.841025641025642e-06,
+      "loss": 1.7318,
+      "step": 8100
+    },
+    {
+      "epoch": 0.052118423227020226,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.835897435897438e-06,
+      "loss": 1.7245,
+      "step": 8200
+    },
+    {
+      "epoch": 0.052754013754179006,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.830769230769232e-06,
+      "loss": 1.7085,
+      "step": 8300
+    },
+    {
+      "epoch": 0.053389604281337794,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.825641025641026e-06,
+      "loss": 1.714,
+      "step": 8400
+    },
+    {
+      "epoch": 0.054025194808496574,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.820512820512821e-06,
+      "loss": 1.7158,
+      "step": 8500
+    },
+    {
+      "epoch": 0.054660785335655354,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.815384615384617e-06,
+      "loss": 1.7205,
+      "step": 8600
+    },
+    {
+      "epoch": 0.05529637586281414,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.810256410256411e-06,
+      "loss": 1.72,
+      "step": 8700
+    },
+    {
+      "epoch": 0.05593196638997292,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.805128205128205e-06,
+      "loss": 1.7039,
+      "step": 8800
+    },
+    {
+      "epoch": 0.05656755691713171,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 1.7238,
+      "step": 8900
+    },
+    {
+      "epoch": 0.05720314744429049,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.794871794871795e-06,
+      "loss": 1.7165,
+      "step": 9000
+    },
+    {
+      "epoch": 0.057838737971449276,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.78974358974359e-06,
+      "loss": 1.7248,
+      "step": 9100
+    },
+    {
+      "epoch": 0.05847432849860806,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.784615384615387e-06,
+      "loss": 1.7327,
+      "step": 9200
+    },
+    {
+      "epoch": 0.05910991902576684,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.77948717948718e-06,
+      "loss": 1.726,
+      "step": 9300
+    },
+    {
+      "epoch": 0.059745509552925624,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.774358974358975e-06,
+      "loss": 1.7145,
+      "step": 9400
+    },
+    {
+      "epoch": 0.060381100080084404,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.76923076923077e-06,
+      "loss": 1.7061,
+      "step": 9500
+    },
+    {
+      "epoch": 0.06101669060724319,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.764102564102564e-06,
+      "loss": 1.7186,
+      "step": 9600
+    },
+    {
+      "epoch": 0.06165228113440197,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.75897435897436e-06,
+      "loss": 1.7153,
+      "step": 9700
+    },
+    {
+      "epoch": 0.06228787166156076,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.753846153846154e-06,
+      "loss": 1.7038,
+      "step": 9800
+    },
+    {
+      "epoch": 0.06292346218871954,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.74871794871795e-06,
+      "loss": 1.7135,
+      "step": 9900
+    },
+    {
+      "epoch": 0.06355905271587832,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 1.6987,
+      "step": 10000
+    },
+    {
+      "epoch": 0.0641946432430371,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.738461538461538e-06,
+      "loss": 1.6867,
+      "step": 10100
+    },
+    {
+      "epoch": 0.0648302337701959,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.733333333333334e-06,
+      "loss": 1.7,
+      "step": 10200
+    },
+    {
+      "epoch": 0.06546582429735467,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.72820512820513e-06,
+      "loss": 1.7057,
+      "step": 10300
+    },
+    {
+      "epoch": 0.06610141482451345,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.723076923076924e-06,
+      "loss": 1.7206,
+      "step": 10400
+    },
+    {
+      "epoch": 0.06673700535167224,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.71794871794872e-06,
+      "loss": 1.694,
+      "step": 10500
+    },
+    {
+      "epoch": 0.06737259587883102,
+      "grad_norm": 0.49609375,
+      "learning_rate": 9.712820512820513e-06,
+      "loss": 1.7058,
+      "step": 10600
+    },
+    {
+      "epoch": 0.06800818640598981,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.707692307692308e-06,
+      "loss": 1.7093,
+      "step": 10700
+    },
+    {
+      "epoch": 0.06864377693314859,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.702564102564103e-06,
+      "loss": 1.6968,
+      "step": 10800
+    },
+    {
+      "epoch": 0.06927936746030737,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.697435897435899e-06,
+      "loss": 1.6991,
+      "step": 10900
+    },
+    {
+      "epoch": 0.06991495798746615,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.692307692307693e-06,
+      "loss": 1.7002,
+      "step": 11000
+    },
+    {
+      "epoch": 0.07055054851462494,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.687179487179487e-06,
+      "loss": 1.7096,
+      "step": 11100
+    },
+    {
+      "epoch": 0.07118613904178372,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.682051282051283e-06,
+      "loss": 1.6926,
+      "step": 11200
+    },
+    {
+      "epoch": 0.0718217295689425,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.676923076923079e-06,
+      "loss": 1.6882,
+      "step": 11300
+    },
+    {
+      "epoch": 0.07245732009610129,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.671794871794873e-06,
+      "loss": 1.6937,
+      "step": 11400
+    },
+    {
+      "epoch": 0.07309291062326007,
+      "grad_norm": 1.2578125,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 1.7164,
+      "step": 11500
+    },
+    {
+      "epoch": 0.07372850115041886,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.661538461538462e-06,
+      "loss": 1.6886,
+      "step": 11600
+    },
+    {
+      "epoch": 0.07436409167757764,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.656410256410257e-06,
+      "loss": 1.7014,
+      "step": 11700
+    },
+    {
+      "epoch": 0.07499968220473642,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.651282051282052e-06,
+      "loss": 1.7044,
+      "step": 11800
+    },
+    {
+      "epoch": 0.0756352727318952,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.646153846153848e-06,
+      "loss": 1.7096,
+      "step": 11900
+    },
+    {
+      "epoch": 0.07627086325905398,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.641025641025642e-06,
+      "loss": 1.7007,
+      "step": 12000
+    },
+    {
+      "epoch": 0.07690645378621278,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.635897435897436e-06,
+      "loss": 1.6985,
+      "step": 12100
+    },
+    {
+      "epoch": 0.07754204431337156,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.630769230769232e-06,
+      "loss": 1.6963,
+      "step": 12200
+    },
+    {
+      "epoch": 0.07817763484053034,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.625641025641026e-06,
+      "loss": 1.6804,
+      "step": 12300
+    },
+    {
+      "epoch": 0.07881322536768912,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.620512820512822e-06,
+      "loss": 1.6936,
+      "step": 12400
+    },
+    {
+      "epoch": 0.07944881589484791,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 1.697,
+      "step": 12500
+    },
+    {
+      "epoch": 0.08008440642200669,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.610256410256411e-06,
+      "loss": 1.6945,
+      "step": 12600
+    },
+    {
+      "epoch": 0.08071999694916547,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.605128205128206e-06,
+      "loss": 1.6923,
+      "step": 12700
+    },
+    {
+      "epoch": 0.08135558747632425,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 1.6909,
+      "step": 12800
+    },
+    {
+      "epoch": 0.08199117800348303,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.594871794871797e-06,
+      "loss": 1.688,
+      "step": 12900
+    },
+    {
+      "epoch": 0.08262676853064183,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.589743589743591e-06,
+      "loss": 1.6987,
+      "step": 13000
+    },
+    {
+      "epoch": 0.0832623590578006,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.584615384615385e-06,
+      "loss": 1.6974,
+      "step": 13100
+    },
+    {
+      "epoch": 0.08389794958495939,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.579487179487181e-06,
+      "loss": 1.6831,
+      "step": 13200
+    },
+    {
+      "epoch": 0.08453354011211817,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.574358974358975e-06,
+      "loss": 1.6862,
+      "step": 13300
+    },
+    {
+      "epoch": 0.08516913063927695,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.569230769230769e-06,
+      "loss": 1.6805,
+      "step": 13400
+    },
+    {
+      "epoch": 0.08580472116643574,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.564102564102565e-06,
+      "loss": 1.6963,
+      "step": 13500
+    },
+    {
+      "epoch": 0.08644031169359452,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.55897435897436e-06,
+      "loss": 1.6835,
+      "step": 13600
+    },
+    {
+      "epoch": 0.0870759022207533,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.553846153846155e-06,
+      "loss": 1.6947,
+      "step": 13700
+    },
+    {
+      "epoch": 0.08771149274791208,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.548717948717949e-06,
+      "loss": 1.6866,
+      "step": 13800
+    },
+    {
+      "epoch": 0.08834708327507086,
+      "grad_norm": 0.625,
+      "learning_rate": 9.543589743589744e-06,
+      "loss": 1.6842,
+      "step": 13900
+    },
+    {
+      "epoch": 0.08898267380222966,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.53846153846154e-06,
+      "loss": 1.6714,
+      "step": 14000
+    },
+    {
+      "epoch": 0.08961826432938844,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.533333333333334e-06,
+      "loss": 1.679,
+      "step": 14100
+    },
+    {
+      "epoch": 0.09025385485654722,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.52820512820513e-06,
+      "loss": 1.6923,
+      "step": 14200
+    },
+    {
+      "epoch": 0.090889445383706,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.523076923076924e-06,
+      "loss": 1.6745,
+      "step": 14300
+    },
+    {
+      "epoch": 0.09152503591086479,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.517948717948718e-06,
+      "loss": 1.6871,
+      "step": 14400
+    },
+    {
+      "epoch": 0.09216062643802357,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.512820512820514e-06,
+      "loss": 1.6795,
+      "step": 14500
+    },
+    {
+      "epoch": 0.09279621696518235,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.50769230769231e-06,
+      "loss": 1.6779,
+      "step": 14600
+    },
+    {
+      "epoch": 0.09343180749234113,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.502564102564103e-06,
+      "loss": 1.6846,
+      "step": 14700
+    },
+    {
+      "epoch": 0.09406739801949991,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.497435897435898e-06,
+      "loss": 1.6762,
+      "step": 14800
+    },
+    {
+      "epoch": 0.0947029885466587,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.492307692307693e-06,
+      "loss": 1.6929,
+      "step": 14900
+    },
+    {
+      "epoch": 0.09533857907381749,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.487179487179487e-06,
+      "loss": 1.6791,
+      "step": 15000
+    },
+    {
+      "epoch": 0.09597416960097627,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.482051282051283e-06,
+      "loss": 1.677,
+      "step": 15100
+    },
+    {
+      "epoch": 0.09660976012813505,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.476923076923079e-06,
+      "loss": 1.6817,
+      "step": 15200
+    },
+    {
+      "epoch": 0.09724535065529383,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.471794871794873e-06,
+      "loss": 1.6632,
+      "step": 15300
+    },
+    {
+      "epoch": 0.09788094118245262,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.466666666666667e-06,
+      "loss": 1.681,
+      "step": 15400
+    },
+    {
+      "epoch": 0.0985165317096114,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.461538461538463e-06,
+      "loss": 1.681,
+      "step": 15500
+    },
+    {
+      "epoch": 0.09915212223677018,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.456410256410257e-06,
+      "loss": 1.6915,
+      "step": 15600
+    },
+    {
+      "epoch": 0.09978771276392896,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.451282051282052e-06,
+      "loss": 1.6797,
+      "step": 15700
+    },
+    {
+      "epoch": 0.10042330329108776,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.446153846153847e-06,
+      "loss": 1.664,
+      "step": 15800
+    },
+    {
+      "epoch": 0.10105889381824654,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.441025641025642e-06,
+      "loss": 1.6738,
+      "step": 15900
+    },
+    {
+      "epoch": 0.10169448434540532,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.435897435897436e-06,
+      "loss": 1.6822,
+      "step": 16000
+    },
+    {
+      "epoch": 0.1023300748725641,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.43076923076923e-06,
+      "loss": 1.6828,
+      "step": 16100
+    },
+    {
+      "epoch": 0.10296566539972288,
+      "grad_norm": 0.498046875,
+      "learning_rate": 9.425641025641026e-06,
+      "loss": 1.6714,
+      "step": 16200
+    },
+    {
+      "epoch": 0.10360125592688167,
+      "grad_norm": 0.46484375,
+      "learning_rate": 9.420512820512822e-06,
+      "loss": 1.692,
+      "step": 16300
+    },
+    {
+      "epoch": 0.10423684645404045,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.415384615384616e-06,
+      "loss": 1.6833,
+      "step": 16400
+    },
+    {
+      "epoch": 0.10487243698119923,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.410256410256412e-06,
+      "loss": 1.6835,
+      "step": 16500
+    },
+    {
+      "epoch": 0.10550802750835801,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.405128205128206e-06,
+      "loss": 1.6738,
+      "step": 16600
+    },
+    {
+      "epoch": 0.10614361803551679,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.4e-06,
+      "loss": 1.6769,
+      "step": 16700
+    },
+    {
+      "epoch": 0.10677920856267559,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.394871794871796e-06,
+      "loss": 1.677,
+      "step": 16800
+    },
+    {
+      "epoch": 0.10741479908983437,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.389743589743591e-06,
+      "loss": 1.6887,
+      "step": 16900
+    },
+    {
+      "epoch": 0.10805038961699315,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.384615384615385e-06,
+      "loss": 1.6889,
+      "step": 17000
+    },
+    {
+      "epoch": 0.10868598014415193,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.37948717948718e-06,
+      "loss": 1.6689,
+      "step": 17100
+    },
+    {
+      "epoch": 0.10932157067131071,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.374358974358975e-06,
+      "loss": 1.6851,
+      "step": 17200
+    },
+    {
+      "epoch": 0.1099571611984695,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.369230769230771e-06,
+      "loss": 1.6878,
+      "step": 17300
+    },
+    {
+      "epoch": 0.11059275172562828,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.364102564102565e-06,
+      "loss": 1.6609,
+      "step": 17400
+    },
+    {
+      "epoch": 0.11122834225278706,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.358974358974359e-06,
+      "loss": 1.6778,
+      "step": 17500
+    },
+    {
+      "epoch": 0.11186393277994584,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.353846153846155e-06,
+      "loss": 1.674,
+      "step": 17600
+    },
+    {
+      "epoch": 0.11249952330710464,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.348717948717949e-06,
+      "loss": 1.6686,
+      "step": 17700
+    },
+    {
+      "epoch": 0.11313511383426342,
+      "grad_norm": 0.474609375,
+      "learning_rate": 9.343589743589745e-06,
+      "loss": 1.6728,
+      "step": 17800
+    },
+    {
+      "epoch": 0.1137707043614222,
+      "grad_norm": 0.482421875,
+      "learning_rate": 9.33846153846154e-06,
+      "loss": 1.6664,
+      "step": 17900
+    },
+    {
+      "epoch": 0.11440629488858098,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 1.6746,
+      "step": 18000
+    },
+    {
+      "epoch": 0.11504188541573976,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.328205128205128e-06,
+      "loss": 1.6738,
+      "step": 18100
+    },
+    {
+      "epoch": 0.11567747594289855,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.323076923076924e-06,
+      "loss": 1.6628,
+      "step": 18200
+    },
+    {
+      "epoch": 0.11631306647005733,
+      "grad_norm": 0.48828125,
+      "learning_rate": 9.317948717948718e-06,
+      "loss": 1.6718,
+      "step": 18300
+    },
+    {
+      "epoch": 0.11694865699721611,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.312820512820514e-06,
+      "loss": 1.6789,
+      "step": 18400
+    },
+    {
+      "epoch": 0.1175842475243749,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.307692307692308e-06,
+      "loss": 1.6618,
+      "step": 18500
+    },
+    {
+      "epoch": 0.11821983805153367,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.302564102564104e-06,
+      "loss": 1.6648,
+      "step": 18600
+    },
+    {
+      "epoch": 0.11885542857869247,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.297435897435898e-06,
+      "loss": 1.683,
+      "step": 18700
+    },
+    {
+      "epoch": 0.11949101910585125,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.292307692307694e-06,
+      "loss": 1.6457,
+      "step": 18800
+    },
+    {
+      "epoch": 0.12012660963301003,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.28717948717949e-06,
+      "loss": 1.6728,
+      "step": 18900
+    },
+    {
+      "epoch": 0.12076220016016881,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.282051282051283e-06,
+      "loss": 1.683,
+      "step": 19000
+    },
+    {
+      "epoch": 0.12139779068732759,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.276923076923077e-06,
+      "loss": 1.6626,
+      "step": 19100
+    },
+    {
+      "epoch": 0.12203338121448638,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.271794871794873e-06,
+      "loss": 1.6704,
+      "step": 19200
+    },
+    {
+      "epoch": 0.12266897174164516,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.266666666666667e-06,
+      "loss": 1.6608,
+      "step": 19300
+    },
+    {
+      "epoch": 0.12330456226880394,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.261538461538461e-06,
+      "loss": 1.6761,
+      "step": 19400
+    },
+    {
+      "epoch": 0.12394015279596272,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.256410256410257e-06,
+      "loss": 1.6537,
+      "step": 19500
+    },
+    {
+      "epoch": 0.12457574332312152,
+      "grad_norm": 0.478515625,
+      "learning_rate": 9.251282051282053e-06,
+      "loss": 1.6732,
+      "step": 19600
+    },
+    {
+      "epoch": 0.1252113338502803,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.246153846153847e-06,
+      "loss": 1.6708,
+      "step": 19700
+    },
+    {
+      "epoch": 0.12584692437743908,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.24102564102564e-06,
+      "loss": 1.6619,
+      "step": 19800
+    },
+    {
+      "epoch": 0.12648251490459786,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.235897435897437e-06,
+      "loss": 1.6666,
+      "step": 19900
+    },
+    {
+      "epoch": 0.12711810543175664,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 1.6725,
+      "step": 20000
+    },
+    {
+      "epoch": 0.12775369595891542,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.225641025641026e-06,
+      "loss": 1.6689,
+      "step": 20100
+    },
+    {
+      "epoch": 0.1283892864860742,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.220512820512822e-06,
+      "loss": 1.6735,
+      "step": 20200
+    },
+    {
+      "epoch": 0.129024877013233,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.215384615384616e-06,
+      "loss": 1.6652,
+      "step": 20300
+    },
+    {
+      "epoch": 0.1296604675403918,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.21025641025641e-06,
+      "loss": 1.6662,
+      "step": 20400
+    },
+    {
+      "epoch": 0.13029605806755057,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.205128205128206e-06,
+      "loss": 1.6553,
+      "step": 20500
+    },
+    {
+      "epoch": 0.13093164859470935,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 1.6571,
+      "step": 20600
+    },
+    {
+      "epoch": 0.13156723912186813,
+      "grad_norm": 0.478515625,
+      "learning_rate": 9.194871794871796e-06,
+      "loss": 1.6626,
+      "step": 20700
+    },
+    {
+      "epoch": 0.1322028296490269,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.18974358974359e-06,
+      "loss": 1.6703,
+      "step": 20800
+    },
+    {
+      "epoch": 0.1328384201761857,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.184615384615386e-06,
+      "loss": 1.6563,
+      "step": 20900
+    },
+    {
+      "epoch": 0.13347401070334447,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.17948717948718e-06,
+      "loss": 1.6549,
+      "step": 21000
+    },
+    {
+      "epoch": 0.13410960123050325,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.174358974358975e-06,
+      "loss": 1.6777,
+      "step": 21100
+    },
+    {
+      "epoch": 0.13474519175766203,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.169230769230771e-06,
+      "loss": 1.6478,
+      "step": 21200
+    },
+    {
+      "epoch": 0.13538078228482084,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.164102564102565e-06,
+      "loss": 1.6671,
+      "step": 21300
+    },
+    {
+      "epoch": 0.13601637281197962,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.15897435897436e-06,
+      "loss": 1.6846,
+      "step": 21400
+    },
+    {
+      "epoch": 0.1366519633391384,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.153846153846155e-06,
+      "loss": 1.6869,
+      "step": 21500
+    },
+    {
+      "epoch": 0.13728755386629718,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.148717948717949e-06,
+      "loss": 1.6531,
+      "step": 21600
+    },
+    {
+      "epoch": 0.13792314439345596,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.143589743589745e-06,
+      "loss": 1.669,
+      "step": 21700
+    },
+    {
+      "epoch": 0.13855873492061474,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.138461538461539e-06,
+      "loss": 1.6553,
+      "step": 21800
+    },
+    {
+      "epoch": 0.13919432544777352,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.133333333333335e-06,
+      "loss": 1.6653,
+      "step": 21900
+    },
+    {
+      "epoch": 0.1398299159749323,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.128205128205129e-06,
+      "loss": 1.6617,
+      "step": 22000
+    },
+    {
+      "epoch": 0.14046550650209108,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.123076923076923e-06,
+      "loss": 1.6819,
+      "step": 22100
+    },
+    {
+      "epoch": 0.1411010970292499,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.117948717948718e-06,
+      "loss": 1.6546,
+      "step": 22200
+    },
+    {
+      "epoch": 0.14173668755640867,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.112820512820514e-06,
+      "loss": 1.6635,
+      "step": 22300
+    },
+    {
+      "epoch": 0.14237227808356745,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.107692307692308e-06,
+      "loss": 1.6673,
+      "step": 22400
+    },
+    {
+      "epoch": 0.14300786861072623,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.102564102564104e-06,
+      "loss": 1.6683,
+      "step": 22500
+    },
+    {
+      "epoch": 0.143643459137885,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.097435897435898e-06,
+      "loss": 1.666,
+      "step": 22600
+    },
+    {
+      "epoch": 0.1442790496650438,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.092307692307692e-06,
+      "loss": 1.6557,
+      "step": 22700
+    },
+    {
+      "epoch": 0.14491464019220257,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.087179487179488e-06,
+      "loss": 1.6566,
+      "step": 22800
+    },
+    {
+      "epoch": 0.14555023071936135,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.082051282051284e-06,
+      "loss": 1.6515,
+      "step": 22900
+    },
+    {
+      "epoch": 0.14618582124652013,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.076923076923078e-06,
+      "loss": 1.6468,
+      "step": 23000
+    },
+    {
+      "epoch": 0.1468214117736789,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.071794871794872e-06,
+      "loss": 1.6619,
+      "step": 23100
+    },
+    {
+      "epoch": 0.14745700230083772,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.066666666666667e-06,
+      "loss": 1.6479,
+      "step": 23200
+    },
+    {
+      "epoch": 0.1480925928279965,
+      "grad_norm": 0.625,
+      "learning_rate": 9.061538461538463e-06,
+      "loss": 1.6581,
+      "step": 23300
+    },
+    {
+      "epoch": 0.14872818335515528,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.056410256410257e-06,
+      "loss": 1.6668,
+      "step": 23400
+    },
+    {
+      "epoch": 0.14936377388231406,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.051282051282051e-06,
+      "loss": 1.664,
+      "step": 23500
+    },
+    {
+      "epoch": 0.14999936440947284,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.046153846153847e-06,
+      "loss": 1.6472,
+      "step": 23600
+    },
+    {
+      "epoch": 0.15063495493663162,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.041025641025641e-06,
+      "loss": 1.6529,
+      "step": 23700
+    },
+    {
+      "epoch": 0.1512705454637904,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.035897435897437e-06,
+      "loss": 1.6544,
+      "step": 23800
+    },
+    {
+      "epoch": 0.15190613599094918,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.030769230769233e-06,
+      "loss": 1.6549,
+      "step": 23900
+    },
+    {
+      "epoch": 0.15254172651810796,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.025641025641027e-06,
+      "loss": 1.6627,
+      "step": 24000
+    },
+    {
+      "epoch": 0.15317731704526677,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.02051282051282e-06,
+      "loss": 1.646,
+      "step": 24100
+    },
+    {
+      "epoch": 0.15381290757242555,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.015384615384616e-06,
+      "loss": 1.6602,
+      "step": 24200
+    },
+    {
+      "epoch": 0.15444849809958433,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.01025641025641e-06,
+      "loss": 1.6532,
+      "step": 24300
+    },
+    {
+      "epoch": 0.1550840886267431,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.005128205128206e-06,
+      "loss": 1.6411,
+      "step": 24400
+    },
+    {
+      "epoch": 0.1557196791539019,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9e-06,
+      "loss": 1.6684,
+      "step": 24500
+    },
+    {
+      "epoch": 0.15635526968106067,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.994871794871796e-06,
+      "loss": 1.6517,
+      "step": 24600
+    },
+    {
+      "epoch": 0.15699086020821945,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.98974358974359e-06,
+      "loss": 1.6509,
+      "step": 24700
+    },
+    {
+      "epoch": 0.15762645073537823,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.984615384615386e-06,
+      "loss": 1.6603,
+      "step": 24800
+    },
+    {
+      "epoch": 0.158262041262537,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.979487179487182e-06,
+      "loss": 1.661,
+      "step": 24900
+    },
+    {
+      "epoch": 0.15889763178969582,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 1.6679,
+      "step": 25000
+    },
+    {
+      "epoch": 0.1595332223168546,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.96923076923077e-06,
+      "loss": 1.6555,
+      "step": 25100
+    },
+    {
+      "epoch": 0.16016881284401338,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.964102564102565e-06,
+      "loss": 1.6652,
+      "step": 25200
+    },
+    {
+      "epoch": 0.16080440337117216,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.95897435897436e-06,
+      "loss": 1.6393,
+      "step": 25300
+    },
+    {
+      "epoch": 0.16143999389833094,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.953846153846153e-06,
+      "loss": 1.6479,
+      "step": 25400
+    },
+    {
+      "epoch": 0.16207558442548972,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.94871794871795e-06,
+      "loss": 1.6509,
+      "step": 25500
+    },
+    {
+      "epoch": 0.1627111749526485,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.943589743589745e-06,
+      "loss": 1.66,
+      "step": 25600
+    },
+    {
+      "epoch": 0.16334676547980728,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.938461538461539e-06,
+      "loss": 1.6697,
+      "step": 25700
+    },
+    {
+      "epoch": 0.16398235600696606,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.933333333333333e-06,
+      "loss": 1.6493,
+      "step": 25800
+    },
+    {
+      "epoch": 0.16461794653412484,
+      "grad_norm": 0.5,
+      "learning_rate": 8.928205128205129e-06,
+      "loss": 1.6503,
+      "step": 25900
+    },
+    {
+      "epoch": 0.16525353706128365,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.923076923076925e-06,
+      "loss": 1.65,
+      "step": 26000
+    },
+    {
+      "epoch": 0.16588912758844243,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.917948717948719e-06,
+      "loss": 1.6532,
+      "step": 26100
+    },
+    {
+      "epoch": 0.1665247181156012,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.912820512820514e-06,
+      "loss": 1.6583,
+      "step": 26200
+    },
+    {
+      "epoch": 0.16716030864276,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.907692307692308e-06,
+      "loss": 1.6423,
+      "step": 26300
+    },
+    {
+      "epoch": 0.16779589916991877,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.902564102564102e-06,
+      "loss": 1.6699,
+      "step": 26400
+    },
+    {
+      "epoch": 0.16843148969707755,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.897435897435898e-06,
+      "loss": 1.6522,
+      "step": 26500
+    },
+    {
+      "epoch": 0.16906708022423633,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.892307692307694e-06,
+      "loss": 1.6531,
+      "step": 26600
+    },
+    {
+      "epoch": 0.1697026707513951,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.887179487179488e-06,
+      "loss": 1.6512,
+      "step": 26700
+    },
+    {
+      "epoch": 0.1703382612785539,
+      "grad_norm": 0.4765625,
+      "learning_rate": 8.882051282051282e-06,
+      "loss": 1.6158,
+      "step": 26800
+    },
+    {
+      "epoch": 0.1709738518057127,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.876923076923078e-06,
+      "loss": 1.6516,
+      "step": 26900
+    },
+    {
+      "epoch": 0.17160944233287148,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.871794871794872e-06,
+      "loss": 1.6409,
+      "step": 27000
+    },
+    {
+      "epoch": 0.17224503286003026,
+      "grad_norm": 0.40234375,
+      "learning_rate": 8.866666666666668e-06,
+      "loss": 1.6526,
+      "step": 27100
+    },
+    {
+      "epoch": 0.17288062338718904,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.861538461538463e-06,
+      "loss": 1.6387,
+      "step": 27200
+    },
+    {
+      "epoch": 0.17351621391434782,
+      "grad_norm": 0.625,
+      "learning_rate": 8.856410256410257e-06,
+      "loss": 1.6583,
+      "step": 27300
+    },
+    {
+      "epoch": 0.1741518044415066,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.851282051282051e-06,
+      "loss": 1.64,
+      "step": 27400
+    },
+    {
+      "epoch": 0.17478739496866538,
+      "grad_norm": 1.2578125,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 1.6365,
+      "step": 27500
+    },
+    {
+      "epoch": 0.17542298549582416,
+      "grad_norm": 0.5,
+      "learning_rate": 8.841025641025641e-06,
+      "loss": 1.6424,
+      "step": 27600
+    },
+    {
+      "epoch": 0.17605857602298294,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.835897435897437e-06,
+      "loss": 1.646,
+      "step": 27700
+    },
+    {
+      "epoch": 0.17669416655014172,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.830769230769231e-06,
+      "loss": 1.6514,
+      "step": 27800
+    },
+    {
+      "epoch": 0.17732975707730053,
+      "grad_norm": 0.470703125,
+      "learning_rate": 8.825641025641027e-06,
+      "loss": 1.6687,
+      "step": 27900
+    },
+    {
+      "epoch": 0.1779653476044593,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.820512820512821e-06,
+      "loss": 1.6496,
+      "step": 28000
+    },
+    {
+      "epoch": 0.1786009381316181,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.815384615384615e-06,
+      "loss": 1.6357,
+      "step": 28100
+    },
+    {
+      "epoch": 0.17923652865877687,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.81025641025641e-06,
+      "loss": 1.6453,
+      "step": 28200
+    },
+    {
+      "epoch": 0.17987211918593565,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.805128205128206e-06,
+      "loss": 1.659,
+      "step": 28300
+    },
+    {
+      "epoch": 0.18050770971309443,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.8e-06,
+      "loss": 1.6527,
+      "step": 28400
+    },
+    {
+      "epoch": 0.1811433002402532,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.794871794871796e-06,
+      "loss": 1.6669,
+      "step": 28500
+    },
+    {
+      "epoch": 0.181778890767412,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.78974358974359e-06,
+      "loss": 1.6514,
+      "step": 28600
+    },
+    {
+      "epoch": 0.18241448129457077,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.784615384615386e-06,
+      "loss": 1.6655,
+      "step": 28700
+    },
+    {
+      "epoch": 0.18305007182172958,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.77948717948718e-06,
+      "loss": 1.6524,
+      "step": 28800
+    },
+    {
+      "epoch": 0.18368566234888836,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.774358974358976e-06,
+      "loss": 1.6286,
+      "step": 28900
+    },
+    {
+      "epoch": 0.18432125287604714,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.76923076923077e-06,
+      "loss": 1.639,
+      "step": 29000
+    },
+    {
+      "epoch": 0.18495684340320592,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.764102564102564e-06,
+      "loss": 1.6487,
+      "step": 29100
+    },
+    {
+      "epoch": 0.1855924339303647,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.75897435897436e-06,
+      "loss": 1.6576,
+      "step": 29200
+    },
+    {
+      "epoch": 0.18622802445752348,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.753846153846155e-06,
+      "loss": 1.6483,
+      "step": 29300
+    },
+    {
+      "epoch": 0.18686361498468226,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.74871794871795e-06,
+      "loss": 1.6399,
+      "step": 29400
+    },
+    {
+      "epoch": 0.18749920551184104,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.743589743589743e-06,
+      "loss": 1.6383,
+      "step": 29500
+    },
+    {
+      "epoch": 0.18813479603899982,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.73846153846154e-06,
+      "loss": 1.6486,
+      "step": 29600
+    },
+    {
+      "epoch": 0.1887703865661586,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.733333333333333e-06,
+      "loss": 1.6503,
+      "step": 29700
+    },
+    {
+      "epoch": 0.1894059770933174,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.728205128205129e-06,
+      "loss": 1.6361,
+      "step": 29800
+    },
+    {
+      "epoch": 0.1900415676204762,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.723076923076925e-06,
+      "loss": 1.6402,
+      "step": 29900
+    },
+    {
+      "epoch": 0.19067715814763497,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 1.6462,
+      "step": 30000
+    },
+    {
+      "epoch": 0.19131274867479375,
+      "grad_norm": 0.482421875,
+      "learning_rate": 8.712820512820513e-06,
+      "loss": 1.6481,
+      "step": 30100
+    },
+    {
+      "epoch": 0.19194833920195253,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.707692307692309e-06,
+      "loss": 1.6497,
+      "step": 30200
+    },
+    {
+      "epoch": 0.19258392972911131,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.702564102564103e-06,
+      "loss": 1.6486,
+      "step": 30300
+    },
+    {
+      "epoch": 0.1932195202562701,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.697435897435898e-06,
+      "loss": 1.6494,
+      "step": 30400
+    },
+    {
+      "epoch": 0.19385511078342887,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.692307692307692e-06,
+      "loss": 1.6529,
+      "step": 30500
+    },
+    {
+      "epoch": 0.19449070131058765,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.687179487179488e-06,
+      "loss": 1.642,
+      "step": 30600
+    },
+    {
+      "epoch": 0.19512629183774646,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.682051282051282e-06,
+      "loss": 1.6206,
+      "step": 30700
+    },
+    {
+      "epoch": 0.19576188236490524,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.676923076923078e-06,
+      "loss": 1.6438,
+      "step": 30800
+    },
+    {
+      "epoch": 0.19639747289206402,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.671794871794874e-06,
+      "loss": 1.6431,
+      "step": 30900
+    },
+    {
+      "epoch": 0.1970330634192228,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 1.6529,
+      "step": 31000
+    },
+    {
+      "epoch": 0.19766865394638158,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.661538461538462e-06,
+      "loss": 1.6476,
+      "step": 31100
+    },
+    {
+      "epoch": 0.19830424447354036,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.656410256410258e-06,
+      "loss": 1.6645,
+      "step": 31200
+    },
+    {
+      "epoch": 0.19893983500069914,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.651282051282052e-06,
+      "loss": 1.6336,
+      "step": 31300
+    },
+    {
+      "epoch": 0.19957542552785792,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.646153846153846e-06,
+      "loss": 1.6391,
+      "step": 31400
+    },
+    {
+      "epoch": 0.2002110160550167,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.641025641025641e-06,
+      "loss": 1.6344,
+      "step": 31500
+    },
+    {
+      "epoch": 0.2008466065821755,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.635897435897437e-06,
+      "loss": 1.6469,
+      "step": 31600
+    },
+    {
+      "epoch": 0.2014821971093343,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.630769230769231e-06,
+      "loss": 1.6422,
+      "step": 31700
+    },
+    {
+      "epoch": 0.20211778763649307,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.625641025641025e-06,
+      "loss": 1.6404,
+      "step": 31800
+    },
+    {
+      "epoch": 0.20275337816365185,
+      "grad_norm": 0.75,
+      "learning_rate": 8.620512820512821e-06,
+      "loss": 1.6453,
+      "step": 31900
+    },
+    {
+      "epoch": 0.20338896869081063,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.615384615384617e-06,
+      "loss": 1.6445,
+      "step": 32000
+    },
+    {
+      "epoch": 0.20402455921796941,
+      "grad_norm": 0.44140625,
+      "learning_rate": 8.610256410256411e-06,
+      "loss": 1.6565,
+      "step": 32100
+    },
+    {
+      "epoch": 0.2046601497451282,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.605128205128207e-06,
+      "loss": 1.6426,
+      "step": 32200
+    },
+    {
+      "epoch": 0.20529574027228698,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.6e-06,
+      "loss": 1.6288,
+      "step": 32300
+    },
+    {
+      "epoch": 0.20593133079944576,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.594871794871795e-06,
+      "loss": 1.6402,
+      "step": 32400
+    },
+    {
+      "epoch": 0.20656692132660454,
+      "grad_norm": 0.482421875,
+      "learning_rate": 8.58974358974359e-06,
+      "loss": 1.6308,
+      "step": 32500
+    },
+    {
+      "epoch": 0.20720251185376334,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.584615384615386e-06,
+      "loss": 1.6371,
+      "step": 32600
+    },
+    {
+      "epoch": 0.20783810238092212,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.57948717948718e-06,
+      "loss": 1.6449,
+      "step": 32700
+    },
+    {
+      "epoch": 0.2084736929080809,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.574358974358974e-06,
+      "loss": 1.6521,
+      "step": 32800
+    },
+    {
+      "epoch": 0.20910928343523968,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.56923076923077e-06,
+      "loss": 1.6418,
+      "step": 32900
+    },
+    {
+      "epoch": 0.20974487396239846,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.564102564102564e-06,
+      "loss": 1.6342,
+      "step": 33000
+    },
+    {
+      "epoch": 0.21038046448955725,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.55897435897436e-06,
+      "loss": 1.6413,
+      "step": 33100
+    },
+    {
+      "epoch": 0.21101605501671603,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.553846153846156e-06,
+      "loss": 1.6432,
+      "step": 33200
+    },
+    {
+      "epoch": 0.2116516455438748,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.54871794871795e-06,
+      "loss": 1.6419,
+      "step": 33300
+    },
+    {
+      "epoch": 0.21228723607103359,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.543589743589744e-06,
+      "loss": 1.6508,
+      "step": 33400
+    },
+    {
+      "epoch": 0.2129228265981924,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.53846153846154e-06,
+      "loss": 1.6431,
+      "step": 33500
+    },
+    {
+      "epoch": 0.21355841712535117,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.533333333333335e-06,
+      "loss": 1.6333,
+      "step": 33600
+    },
+    {
+      "epoch": 0.21419400765250995,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.52820512820513e-06,
+      "loss": 1.659,
+      "step": 33700
+    },
+    {
+      "epoch": 0.21482959817966873,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.523076923076923e-06,
+      "loss": 1.6344,
+      "step": 33800
+    },
+    {
+      "epoch": 0.21546518870682752,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.517948717948719e-06,
+      "loss": 1.6409,
+      "step": 33900
+    },
+    {
+      "epoch": 0.2161007792339863,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.512820512820513e-06,
+      "loss": 1.6304,
+      "step": 34000
+    },
+    {
+      "epoch": 0.21673636976114508,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.507692307692307e-06,
+      "loss": 1.6352,
+      "step": 34100
+    },
+    {
+      "epoch": 0.21737196028830386,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.502564102564103e-06,
+      "loss": 1.6341,
+      "step": 34200
+    },
+    {
+      "epoch": 0.21800755081546264,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.497435897435899e-06,
+      "loss": 1.6448,
+      "step": 34300
+    },
+    {
+      "epoch": 0.21864314134262142,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.492307692307693e-06,
+      "loss": 1.6371,
+      "step": 34400
+    },
+    {
+      "epoch": 0.21927873186978022,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.487179487179488e-06,
+      "loss": 1.6548,
+      "step": 34500
+    },
+    {
+      "epoch": 0.219914322396939,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.482051282051283e-06,
+      "loss": 1.6575,
+      "step": 34600
+    },
+    {
+      "epoch": 0.22054991292409779,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.476923076923078e-06,
+      "loss": 1.652,
+      "step": 34700
+    },
+    {
+      "epoch": 0.22118550345125657,
+      "grad_norm": 0.470703125,
+      "learning_rate": 8.471794871794872e-06,
+      "loss": 1.6636,
+      "step": 34800
+    },
+    {
+      "epoch": 0.22182109397841535,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.466666666666668e-06,
+      "loss": 1.6483,
+      "step": 34900
+    },
+    {
+      "epoch": 0.22245668450557413,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 1.6532,
+      "step": 35000
+    },
+    {
+      "epoch": 0.2230922750327329,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.456410256410256e-06,
+      "loss": 1.6306,
+      "step": 35100
+    },
+    {
+      "epoch": 0.2237278655598917,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.451282051282052e-06,
+      "loss": 1.6137,
+      "step": 35200
+    },
+    {
+      "epoch": 0.22436345608705047,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.446153846153848e-06,
+      "loss": 1.6424,
+      "step": 35300
+    },
+    {
+      "epoch": 0.22499904661420927,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.441025641025642e-06,
+      "loss": 1.6641,
+      "step": 35400
+    },
+    {
+      "epoch": 0.22563463714136806,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.435897435897436e-06,
+      "loss": 1.6388,
+      "step": 35500
+    },
+    {
+      "epoch": 0.22627022766852684,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.430769230769231e-06,
+      "loss": 1.6259,
+      "step": 35600
+    },
+    {
+      "epoch": 0.22690581819568562,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.425641025641026e-06,
+      "loss": 1.639,
+      "step": 35700
+    },
+    {
+      "epoch": 0.2275414087228444,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.420512820512821e-06,
+      "loss": 1.6384,
+      "step": 35800
+    },
+    {
+      "epoch": 0.22817699925000318,
+      "grad_norm": 0.5,
+      "learning_rate": 8.415384615384617e-06,
+      "loss": 1.6446,
+      "step": 35900
+    },
+    {
+      "epoch": 0.22881258977716196,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.410256410256411e-06,
+      "loss": 1.6237,
+      "step": 36000
+    },
+    {
+      "epoch": 0.22944818030432074,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.405128205128205e-06,
+      "loss": 1.6413,
+      "step": 36100
+    },
+    {
+      "epoch": 0.23008377083147952,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 1.6305,
+      "step": 36200
+    },
+    {
+      "epoch": 0.2307193613586383,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.394871794871795e-06,
+      "loss": 1.6586,
+      "step": 36300
+    },
+    {
+      "epoch": 0.2313549518857971,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.38974358974359e-06,
+      "loss": 1.6388,
+      "step": 36400
+    },
+    {
+      "epoch": 0.23199054241295589,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.384615384615385e-06,
+      "loss": 1.6322,
+      "step": 36500
+    },
+    {
+      "epoch": 0.23262613294011467,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.37948717948718e-06,
+      "loss": 1.6377,
+      "step": 36600
+    },
+    {
+      "epoch": 0.23326172346727345,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.374358974358975e-06,
+      "loss": 1.6323,
+      "step": 36700
+    },
+    {
+      "epoch": 0.23389731399443223,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.36923076923077e-06,
+      "loss": 1.6332,
+      "step": 36800
+    },
+    {
+      "epoch": 0.234532904521591,
+      "grad_norm": 0.466796875,
+      "learning_rate": 8.364102564102566e-06,
+      "loss": 1.6308,
+      "step": 36900
+    },
+    {
+      "epoch": 0.2351684950487498,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.35897435897436e-06,
+      "loss": 1.6235,
+      "step": 37000
+    },
+    {
+      "epoch": 0.23580408557590857,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.353846153846154e-06,
+      "loss": 1.6393,
+      "step": 37100
+    },
+    {
+      "epoch": 0.23643967610306735,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.34871794871795e-06,
+      "loss": 1.6371,
+      "step": 37200
+    },
+    {
+      "epoch": 0.23707526663022616,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.343589743589744e-06,
+      "loss": 1.635,
+      "step": 37300
+    },
+    {
+      "epoch": 0.23771085715738494,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.338461538461538e-06,
+      "loss": 1.6295,
+      "step": 37400
+    },
+    {
+      "epoch": 0.23834644768454372,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.6413,
+      "step": 37500
+    },
+    {
+      "epoch": 0.2389820382117025,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.32820512820513e-06,
+      "loss": 1.6512,
+      "step": 37600
+    },
+    {
+      "epoch": 0.23961762873886128,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.323076923076924e-06,
+      "loss": 1.6558,
+      "step": 37700
+    },
+    {
+      "epoch": 0.24025321926602006,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.317948717948718e-06,
+      "loss": 1.6632,
+      "step": 37800
+    },
+    {
+      "epoch": 0.24088880979317884,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.312820512820513e-06,
+      "loss": 1.6344,
+      "step": 37900
+    },
+    {
+      "epoch": 0.24152440032033762,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.307692307692309e-06,
+      "loss": 1.6351,
+      "step": 38000
+    },
+    {
+      "epoch": 0.2421599908474964,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.302564102564103e-06,
+      "loss": 1.6342,
+      "step": 38100
+    },
+    {
+      "epoch": 0.24279558137465518,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.297435897435899e-06,
+      "loss": 1.632,
+      "step": 38200
+    },
+    {
+      "epoch": 0.243431171901814,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.292307692307693e-06,
+      "loss": 1.6476,
+      "step": 38300
+    },
+    {
+      "epoch": 0.24406676242897277,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.287179487179487e-06,
+      "loss": 1.6275,
+      "step": 38400
+    },
+    {
+      "epoch": 0.24470235295613155,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.282051282051283e-06,
+      "loss": 1.6432,
+      "step": 38500
+    },
+    {
+      "epoch": 0.24533794348329033,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.276923076923078e-06,
+      "loss": 1.6506,
+      "step": 38600
+    },
+    {
+      "epoch": 0.2459735340104491,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.271794871794873e-06,
+      "loss": 1.6399,
+      "step": 38700
+    },
+    {
+      "epoch": 0.2466091245376079,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.266666666666667e-06,
+      "loss": 1.6384,
+      "step": 38800
+    },
+    {
+      "epoch": 0.24724471506476667,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.261538461538462e-06,
+      "loss": 1.6286,
+      "step": 38900
+    },
+    {
+      "epoch": 0.24788030559192545,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.256410256410256e-06,
+      "loss": 1.6442,
+      "step": 39000
+    },
+    {
+      "epoch": 0.24851589611908423,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.251282051282052e-06,
+      "loss": 1.6467,
+      "step": 39100
+    },
+    {
+      "epoch": 0.24915148664624304,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.246153846153848e-06,
+      "loss": 1.6507,
+      "step": 39200
+    },
+    {
+      "epoch": 0.24978707717340182,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.241025641025642e-06,
+      "loss": 1.63,
+      "step": 39300
+    },
+    {
+      "epoch": 0.2504226677005606,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.235897435897436e-06,
+      "loss": 1.6294,
+      "step": 39400
+    },
+    {
+      "epoch": 0.2510582582277194,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.230769230769232e-06,
+      "loss": 1.6233,
+      "step": 39500
+    },
+    {
+      "epoch": 0.25169384875487816,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.225641025641027e-06,
+      "loss": 1.6528,
+      "step": 39600
+    },
+    {
+      "epoch": 0.25232943928203694,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.220512820512822e-06,
+      "loss": 1.6369,
+      "step": 39700
+    },
+    {
+      "epoch": 0.2529650298091957,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.215384615384616e-06,
+      "loss": 1.6193,
+      "step": 39800
+    },
+    {
+      "epoch": 0.2536006203363545,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.210256410256411e-06,
+      "loss": 1.646,
+      "step": 39900
+    },
+    {
+      "epoch": 0.2542362108635133,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 1.6195,
+      "step": 40000
+    },
+    {
+      "epoch": 0.25487180139067206,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.2e-06,
+      "loss": 1.6344,
+      "step": 40100
+    },
+    {
+      "epoch": 0.25550739191783084,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.194871794871795e-06,
+      "loss": 1.647,
+      "step": 40200
+    },
+    {
+      "epoch": 0.2561429824449896,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.189743589743591e-06,
+      "loss": 1.6453,
+      "step": 40300
+    },
+    {
+      "epoch": 0.2567785729721484,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.184615384615385e-06,
+      "loss": 1.6263,
+      "step": 40400
+    },
+    {
+      "epoch": 0.2574141634993072,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.17948717948718e-06,
+      "loss": 1.6456,
+      "step": 40500
+    },
+    {
+      "epoch": 0.258049754026466,
+      "grad_norm": 0.625,
+      "learning_rate": 8.174358974358975e-06,
+      "loss": 1.6372,
+      "step": 40600
+    },
+    {
+      "epoch": 0.2586853445536248,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.16923076923077e-06,
+      "loss": 1.6321,
+      "step": 40700
+    },
+    {
+      "epoch": 0.2593209350807836,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.164102564102565e-06,
+      "loss": 1.63,
+      "step": 40800
+    },
+    {
+      "epoch": 0.25995652560794236,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.15897435897436e-06,
+      "loss": 1.6338,
+      "step": 40900
+    },
+    {
+      "epoch": 0.26059211613510114,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.153846153846154e-06,
+      "loss": 1.6366,
+      "step": 41000
+    },
+    {
+      "epoch": 0.2612277066622599,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.148717948717948e-06,
+      "loss": 1.639,
+      "step": 41100
+    },
+    {
+      "epoch": 0.2618632971894187,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.143589743589744e-06,
+      "loss": 1.6464,
+      "step": 41200
+    },
+    {
+      "epoch": 0.2624988877165775,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.13846153846154e-06,
+      "loss": 1.6243,
+      "step": 41300
+    },
+    {
+      "epoch": 0.26313447824373626,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.133333333333334e-06,
+      "loss": 1.6245,
+      "step": 41400
+    },
+    {
+      "epoch": 0.26377006877089504,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.12820512820513e-06,
+      "loss": 1.6267,
+      "step": 41500
+    },
+    {
+      "epoch": 0.2644056592980538,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.123076923076924e-06,
+      "loss": 1.6351,
+      "step": 41600
+    },
+    {
+      "epoch": 0.2650412498252126,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.117948717948718e-06,
+      "loss": 1.6419,
+      "step": 41700
+    },
+    {
+      "epoch": 0.2656768403523714,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.112820512820514e-06,
+      "loss": 1.6504,
+      "step": 41800
+    },
+    {
+      "epoch": 0.26631243087953016,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.10769230769231e-06,
+      "loss": 1.6221,
+      "step": 41900
+    },
+    {
+      "epoch": 0.26694802140668894,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.102564102564103e-06,
+      "loss": 1.6303,
+      "step": 42000
+    },
+    {
+      "epoch": 0.2675836119338477,
+      "grad_norm": 0.486328125,
+      "learning_rate": 8.097435897435897e-06,
+      "loss": 1.6425,
+      "step": 42100
+    },
+    {
+      "epoch": 0.2682192024610065,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.092307692307693e-06,
+      "loss": 1.6439,
+      "step": 42200
+    },
+    {
+      "epoch": 0.2688547929881653,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.087179487179487e-06,
+      "loss": 1.6236,
+      "step": 42300
+    },
+    {
+      "epoch": 0.26949038351532406,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.082051282051283e-06,
+      "loss": 1.6187,
+      "step": 42400
+    },
+    {
+      "epoch": 0.2701259740424829,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 1.6479,
+      "step": 42500
+    },
+    {
+      "epoch": 0.2707615645696417,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.071794871794873e-06,
+      "loss": 1.6291,
+      "step": 42600
+    },
+    {
+      "epoch": 0.27139715509680046,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.066666666666667e-06,
+      "loss": 1.6303,
+      "step": 42700
+    },
+    {
+      "epoch": 0.27203274562395924,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.061538461538463e-06,
+      "loss": 1.644,
+      "step": 42800
+    },
+    {
+      "epoch": 0.272668336151118,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.056410256410258e-06,
+      "loss": 1.618,
+      "step": 42900
+    },
+    {
+      "epoch": 0.2733039266782768,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.051282051282052e-06,
+      "loss": 1.6366,
+      "step": 43000
+    },
+    {
+      "epoch": 0.2739395172054356,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.046153846153846e-06,
+      "loss": 1.639,
+      "step": 43100
+    },
+    {
+      "epoch": 0.27457510773259436,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.041025641025642e-06,
+      "loss": 1.6251,
+      "step": 43200
+    },
+    {
+      "epoch": 0.27521069825975314,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.035897435897436e-06,
+      "loss": 1.636,
+      "step": 43300
+    },
+    {
+      "epoch": 0.2758462887869119,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.03076923076923e-06,
+      "loss": 1.6392,
+      "step": 43400
+    },
+    {
+      "epoch": 0.2764818793140707,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.025641025641026e-06,
+      "loss": 1.6438,
+      "step": 43500
+    },
+    {
+      "epoch": 0.2771174698412295,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.020512820512822e-06,
+      "loss": 1.6318,
+      "step": 43600
+    },
+    {
+      "epoch": 0.27775306036838826,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.015384615384616e-06,
+      "loss": 1.6582,
+      "step": 43700
+    },
+    {
+      "epoch": 0.27838865089554704,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.01025641025641e-06,
+      "loss": 1.6218,
+      "step": 43800
+    },
+    {
+      "epoch": 0.2790242414227058,
+      "grad_norm": 1.421875,
+      "learning_rate": 8.005128205128206e-06,
+      "loss": 1.6256,
+      "step": 43900
+    },
+    {
+      "epoch": 0.2796598319498646,
+      "grad_norm": 1.625,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.6213,
+      "step": 44000
+    },
+    {
+      "epoch": 0.2802954224770234,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.994871794871795e-06,
+      "loss": 1.6373,
+      "step": 44100
+    },
+    {
+      "epoch": 0.28093101300418216,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.989743589743591e-06,
+      "loss": 1.6333,
+      "step": 44200
+    },
+    {
+      "epoch": 0.28156660353134094,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.984615384615385e-06,
+      "loss": 1.6335,
+      "step": 44300
+    },
+    {
+      "epoch": 0.2822021940584998,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.97948717948718e-06,
+      "loss": 1.6411,
+      "step": 44400
+    },
+    {
+      "epoch": 0.28283778458565856,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.974358974358975e-06,
+      "loss": 1.635,
+      "step": 44500
+    },
+    {
+      "epoch": 0.28347337511281734,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.96923076923077e-06,
+      "loss": 1.6425,
+      "step": 44600
+    },
+    {
+      "epoch": 0.2841089656399761,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.964102564102565e-06,
+      "loss": 1.623,
+      "step": 44700
+    },
+    {
+      "epoch": 0.2847445561671349,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.958974358974359e-06,
+      "loss": 1.6395,
+      "step": 44800
+    },
+    {
+      "epoch": 0.2853801466942937,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.953846153846155e-06,
+      "loss": 1.6331,
+      "step": 44900
+    },
+    {
+      "epoch": 0.28601573722145246,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.948717948717949e-06,
+      "loss": 1.6358,
+      "step": 45000
+    },
+    {
+      "epoch": 0.28665132774861124,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.943589743589744e-06,
+      "loss": 1.6369,
+      "step": 45100
+    },
+    {
+      "epoch": 0.28728691827577,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.93846153846154e-06,
+      "loss": 1.6279,
+      "step": 45200
+    },
+    {
+      "epoch": 0.2879225088029288,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.933333333333334e-06,
+      "loss": 1.6299,
+      "step": 45300
+    },
+    {
+      "epoch": 0.2885580993300876,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.928205128205128e-06,
+      "loss": 1.6322,
+      "step": 45400
+    },
+    {
+      "epoch": 0.28919368985724636,
+      "grad_norm": 0.875,
+      "learning_rate": 7.923076923076924e-06,
+      "loss": 1.6241,
+      "step": 45500
+    },
+    {
+      "epoch": 0.28982928038440514,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.91794871794872e-06,
+      "loss": 1.6197,
+      "step": 45600
+    },
+    {
+      "epoch": 0.2904648709115639,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.912820512820514e-06,
+      "loss": 1.6213,
+      "step": 45700
+    },
+    {
+      "epoch": 0.2911004614387227,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.907692307692308e-06,
+      "loss": 1.6283,
+      "step": 45800
+    },
+    {
+      "epoch": 0.2917360519658815,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.902564102564104e-06,
+      "loss": 1.6331,
+      "step": 45900
+    },
+    {
+      "epoch": 0.29237164249304026,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.897435897435898e-06,
+      "loss": 1.6354,
+      "step": 46000
+    },
+    {
+      "epoch": 0.29300723302019904,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.892307692307692e-06,
+      "loss": 1.6243,
+      "step": 46100
+    },
+    {
+      "epoch": 0.2936428235473578,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.887179487179487e-06,
+      "loss": 1.6225,
+      "step": 46200
+    },
+    {
+      "epoch": 0.29427841407451666,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.882051282051283e-06,
+      "loss": 1.626,
+      "step": 46300
+    },
+    {
+      "epoch": 0.29491400460167544,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.876923076923077e-06,
+      "loss": 1.6093,
+      "step": 46400
+    },
+    {
+      "epoch": 0.2955495951288342,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.871794871794873e-06,
+      "loss": 1.632,
+      "step": 46500
+    },
+    {
+      "epoch": 0.296185185655993,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.866666666666667e-06,
+      "loss": 1.6322,
+      "step": 46600
+    },
+    {
+      "epoch": 0.2968207761831518,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.861538461538463e-06,
+      "loss": 1.6305,
+      "step": 46700
+    },
+    {
+      "epoch": 0.29745636671031056,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.856410256410257e-06,
+      "loss": 1.6245,
+      "step": 46800
+    },
+    {
+      "epoch": 0.29809195723746934,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.851282051282053e-06,
+      "loss": 1.6194,
+      "step": 46900
+    },
+    {
+      "epoch": 0.2987275477646281,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.846153846153847e-06,
+      "loss": 1.6272,
+      "step": 47000
+    },
+    {
+      "epoch": 0.2993631382917869,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.84102564102564e-06,
+      "loss": 1.6377,
+      "step": 47100
+    },
+    {
+      "epoch": 0.2999987288189457,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.835897435897436e-06,
+      "loss": 1.6276,
+      "step": 47200
+    },
+    {
+      "epoch": 0.30063431934610446,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.830769230769232e-06,
+      "loss": 1.6217,
+      "step": 47300
+    },
+    {
+      "epoch": 0.30126990987326324,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.825641025641026e-06,
+      "loss": 1.6226,
+      "step": 47400
+    },
+    {
+      "epoch": 0.301905500400422,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.820512820512822e-06,
+      "loss": 1.6194,
+      "step": 47500
+    },
+    {
+      "epoch": 0.3025410909275808,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.815384615384616e-06,
+      "loss": 1.6229,
+      "step": 47600
+    },
+    {
+      "epoch": 0.3031766814547396,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.81025641025641e-06,
+      "loss": 1.6346,
+      "step": 47700
+    },
+    {
+      "epoch": 0.30381227198189836,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.805128205128206e-06,
+      "loss": 1.6313,
+      "step": 47800
+    },
+    {
+      "epoch": 0.30444786250905714,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 1.6418,
+      "step": 47900
+    },
+    {
+      "epoch": 0.3050834530362159,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.794871794871796e-06,
+      "loss": 1.616,
+      "step": 48000
+    },
+    {
+      "epoch": 0.3057190435633747,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.78974358974359e-06,
+      "loss": 1.6425,
+      "step": 48100
+    },
+    {
+      "epoch": 0.30635463409053354,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.784615384615385e-06,
+      "loss": 1.6228,
+      "step": 48200
+    },
+    {
+      "epoch": 0.3069902246176923,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.77948717948718e-06,
+      "loss": 1.6098,
+      "step": 48300
+    },
+    {
+      "epoch": 0.3076258151448511,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.774358974358975e-06,
+      "loss": 1.6483,
+      "step": 48400
+    },
+    {
+      "epoch": 0.3082614056720099,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.76923076923077e-06,
+      "loss": 1.6251,
+      "step": 48500
+    },
+    {
+      "epoch": 0.30889699619916866,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.764102564102565e-06,
+      "loss": 1.6371,
+      "step": 48600
+    },
+    {
+      "epoch": 0.30953258672632744,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.758974358974359e-06,
+      "loss": 1.6277,
+      "step": 48700
+    },
+    {
+      "epoch": 0.3101681772534862,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.753846153846155e-06,
+      "loss": 1.6302,
+      "step": 48800
+    },
+    {
+      "epoch": 0.310803767780645,
+      "grad_norm": 0.46484375,
+      "learning_rate": 7.74871794871795e-06,
+      "loss": 1.6416,
+      "step": 48900
+    },
+    {
+      "epoch": 0.3114393583078038,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.743589743589745e-06,
+      "loss": 1.6144,
+      "step": 49000
+    },
+    {
+      "epoch": 0.31207494883496256,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.738461538461539e-06,
+      "loss": 1.6316,
+      "step": 49100
+    },
+    {
+      "epoch": 0.31271053936212134,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.733333333333334e-06,
+      "loss": 1.6118,
+      "step": 49200
+    },
+    {
+      "epoch": 0.3133461298892801,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.728205128205128e-06,
+      "loss": 1.6303,
+      "step": 49300
+    },
+    {
+      "epoch": 0.3139817204164389,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.723076923076924e-06,
+      "loss": 1.6304,
+      "step": 49400
+    },
+    {
+      "epoch": 0.3146173109435977,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.717948717948718e-06,
+      "loss": 1.6298,
+      "step": 49500
+    },
+    {
+      "epoch": 0.31525290147075646,
+      "grad_norm": 0.474609375,
+      "learning_rate": 7.712820512820514e-06,
+      "loss": 1.6223,
+      "step": 49600
+    },
+    {
+      "epoch": 0.31588849199791524,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.707692307692308e-06,
+      "loss": 1.6282,
+      "step": 49700
+    },
+    {
+      "epoch": 0.316524082525074,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.702564102564102e-06,
+      "loss": 1.6254,
+      "step": 49800
+    },
+    {
+      "epoch": 0.3171596730522328,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.697435897435898e-06,
+      "loss": 1.6377,
+      "step": 49900
+    },
+    {
+      "epoch": 0.31779526357939164,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 1.6068,
+      "step": 50000
+    },
+    {
+      "epoch": 0.3184308541065504,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.687179487179488e-06,
+      "loss": 1.6144,
+      "step": 50100
+    },
+    {
+      "epoch": 0.3190664446337092,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.682051282051283e-06,
+      "loss": 1.6314,
+      "step": 50200
+    },
+    {
+      "epoch": 0.319702035160868,
+      "grad_norm": 1.125,
+      "learning_rate": 7.676923076923077e-06,
+      "loss": 1.6279,
+      "step": 50300
+    },
+    {
+      "epoch": 0.32033762568802676,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.671794871794871e-06,
+      "loss": 1.649,
+      "step": 50400
+    },
+    {
+      "epoch": 0.32097321621518554,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.666666666666667e-06,
+      "loss": 1.6387,
+      "step": 50500
+    },
+    {
+      "epoch": 0.3216088067423443,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.661538461538463e-06,
+      "loss": 1.627,
+      "step": 50600
+    },
+    {
+      "epoch": 0.3222443972695031,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.656410256410257e-06,
+      "loss": 1.6261,
+      "step": 50700
+    },
+    {
+      "epoch": 0.3228799877966619,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.651282051282051e-06,
+      "loss": 1.627,
+      "step": 50800
+    },
+    {
+      "epoch": 0.32351557832382066,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.646153846153847e-06,
+      "loss": 1.6219,
+      "step": 50900
+    },
+    {
+      "epoch": 0.32415116885097944,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.641025641025641e-06,
+      "loss": 1.6318,
+      "step": 51000
+    },
+    {
+      "epoch": 0.3247867593781382,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.635897435897437e-06,
+      "loss": 1.6109,
+      "step": 51100
+    },
+    {
+      "epoch": 0.325422349905297,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.630769230769232e-06,
+      "loss": 1.6273,
+      "step": 51200
+    },
+    {
+      "epoch": 0.3260579404324558,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.6256410256410264e-06,
+      "loss": 1.6264,
+      "step": 51300
+    },
+    {
+      "epoch": 0.32669353095961456,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.620512820512821e-06,
+      "loss": 1.6306,
+      "step": 51400
+    },
+    {
+      "epoch": 0.32732912148677334,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.615384615384615e-06,
+      "loss": 1.6276,
+      "step": 51500
+    },
+    {
+      "epoch": 0.3279647120139321,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.610256410256411e-06,
+      "loss": 1.623,
+      "step": 51600
+    },
+    {
+      "epoch": 0.3286003025410909,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.605128205128206e-06,
+      "loss": 1.641,
+      "step": 51700
+    },
+    {
+      "epoch": 0.3292358930682497,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 1.6255,
+      "step": 51800
+    },
+    {
+      "epoch": 0.3298714835954085,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.594871794871795e-06,
+      "loss": 1.6192,
+      "step": 51900
+    },
+    {
+      "epoch": 0.3305070741225673,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.58974358974359e-06,
+      "loss": 1.6209,
+      "step": 52000
+    },
+    {
+      "epoch": 0.3311426646497261,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.584615384615385e-06,
+      "loss": 1.6305,
+      "step": 52100
+    },
+    {
+      "epoch": 0.33177825517688486,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.5794871794871805e-06,
+      "loss": 1.6112,
+      "step": 52200
+    },
+    {
+      "epoch": 0.33241384570404364,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.574358974358975e-06,
+      "loss": 1.6185,
+      "step": 52300
+    },
+    {
+      "epoch": 0.3330494362312024,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.5692307692307695e-06,
+      "loss": 1.6255,
+      "step": 52400
+    },
+    {
+      "epoch": 0.3336850267583612,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 1.6124,
+      "step": 52500
+    },
+    {
+      "epoch": 0.33432061728552,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.558974358974359e-06,
+      "loss": 1.648,
+      "step": 52600
+    },
+    {
+      "epoch": 0.33495620781267876,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.553846153846155e-06,
+      "loss": 1.6192,
+      "step": 52700
+    },
+    {
+      "epoch": 0.33559179833983754,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.54871794871795e-06,
+      "loss": 1.6265,
+      "step": 52800
+    },
+    {
+      "epoch": 0.3362273888669963,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.543589743589744e-06,
+      "loss": 1.6227,
+      "step": 52900
+    },
+    {
+      "epoch": 0.3368629793941551,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.538461538461539e-06,
+      "loss": 1.6165,
+      "step": 53000
+    },
+    {
+      "epoch": 0.3374985699213139,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.533333333333334e-06,
+      "loss": 1.6203,
+      "step": 53100
+    },
+    {
+      "epoch": 0.33813416044847266,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.528205128205129e-06,
+      "loss": 1.6253,
+      "step": 53200
+    },
+    {
+      "epoch": 0.33876975097563145,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.523076923076924e-06,
+      "loss": 1.6359,
+      "step": 53300
+    },
+    {
+      "epoch": 0.3394053415027902,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.5179487179487185e-06,
+      "loss": 1.6145,
+      "step": 53400
+    },
+    {
+      "epoch": 0.340040932029949,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.512820512820513e-06,
+      "loss": 1.6269,
+      "step": 53500
+    },
+    {
+      "epoch": 0.3406765225571078,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.507692307692308e-06,
+      "loss": 1.6113,
+      "step": 53600
+    },
+    {
+      "epoch": 0.34131211308426657,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.502564102564102e-06,
+      "loss": 1.6376,
+      "step": 53700
+    },
+    {
+      "epoch": 0.3419477036114254,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.497435897435899e-06,
+      "loss": 1.634,
+      "step": 53800
+    },
+    {
+      "epoch": 0.3425832941385842,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.492307692307693e-06,
+      "loss": 1.6335,
+      "step": 53900
+    },
+    {
+      "epoch": 0.34321888466574296,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.487179487179488e-06,
+      "loss": 1.6334,
+      "step": 54000
+    },
+    {
+      "epoch": 0.34385447519290174,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.482051282051283e-06,
+      "loss": 1.6347,
+      "step": 54100
+    },
+    {
+      "epoch": 0.3444900657200605,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.476923076923077e-06,
+      "loss": 1.6233,
+      "step": 54200
+    },
+    {
+      "epoch": 0.3451256562472193,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.4717948717948726e-06,
+      "loss": 1.6291,
+      "step": 54300
+    },
+    {
+      "epoch": 0.3457612467743781,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.4666666666666675e-06,
+      "loss": 1.6182,
+      "step": 54400
+    },
+    {
+      "epoch": 0.34639683730153686,
+      "grad_norm": 1.4609375,
+      "learning_rate": 7.461538461538462e-06,
+      "loss": 1.6365,
+      "step": 54500
+    },
+    {
+      "epoch": 0.34703242782869564,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.456410256410257e-06,
+      "loss": 1.623,
+      "step": 54600
+    },
+    {
+      "epoch": 0.3476680183558544,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.451282051282051e-06,
+      "loss": 1.6104,
+      "step": 54700
+    },
+    {
+      "epoch": 0.3483036088830132,
+      "grad_norm": 1.2890625,
+      "learning_rate": 7.446153846153846e-06,
+      "loss": 1.6203,
+      "step": 54800
+    },
+    {
+      "epoch": 0.348939199410172,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.441025641025642e-06,
+      "loss": 1.6089,
+      "step": 54900
+    },
+    {
+      "epoch": 0.34957478993733077,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.435897435897437e-06,
+      "loss": 1.6191,
+      "step": 55000
+    },
+    {
+      "epoch": 0.35021038046448955,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.430769230769232e-06,
+      "loss": 1.626,
+      "step": 55100
+    },
+    {
+      "epoch": 0.3508459709916483,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.425641025641026e-06,
+      "loss": 1.6189,
+      "step": 55200
+    },
+    {
+      "epoch": 0.3514815615188071,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.420512820512821e-06,
+      "loss": 1.6273,
+      "step": 55300
+    },
+    {
+      "epoch": 0.3521171520459659,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.4153846153846164e-06,
+      "loss": 1.6278,
+      "step": 55400
+    },
+    {
+      "epoch": 0.35275274257312467,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.410256410256411e-06,
+      "loss": 1.6254,
+      "step": 55500
+    },
+    {
+      "epoch": 0.35338833310028345,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.405128205128206e-06,
+      "loss": 1.6146,
+      "step": 55600
+    },
+    {
+      "epoch": 0.3540239236274423,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.4e-06,
+      "loss": 1.6295,
+      "step": 55700
+    },
+    {
+      "epoch": 0.35465951415460106,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.394871794871795e-06,
+      "loss": 1.6237,
+      "step": 55800
+    },
+    {
+      "epoch": 0.35529510468175984,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.38974358974359e-06,
+      "loss": 1.619,
+      "step": 55900
+    },
+    {
+      "epoch": 0.3559306952089186,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.384615384615386e-06,
+      "loss": 1.631,
+      "step": 56000
+    },
+    {
+      "epoch": 0.3565662857360774,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.37948717948718e-06,
+      "loss": 1.6142,
+      "step": 56100
+    },
+    {
+      "epoch": 0.3572018762632362,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.374358974358975e-06,
+      "loss": 1.619,
+      "step": 56200
+    },
+    {
+      "epoch": 0.35783746679039496,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.36923076923077e-06,
+      "loss": 1.6294,
+      "step": 56300
+    },
+    {
+      "epoch": 0.35847305731755374,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.364102564102565e-06,
+      "loss": 1.6206,
+      "step": 56400
+    },
+    {
+      "epoch": 0.3591086478447125,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.35897435897436e-06,
+      "loss": 1.646,
+      "step": 56500
+    },
+    {
+      "epoch": 0.3597442383718713,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.353846153846154e-06,
+      "loss": 1.6298,
+      "step": 56600
+    },
+    {
+      "epoch": 0.3603798288990301,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.348717948717949e-06,
+      "loss": 1.6214,
+      "step": 56700
+    },
+    {
+      "epoch": 0.36101541942618887,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.343589743589744e-06,
+      "loss": 1.6305,
+      "step": 56800
+    },
+    {
+      "epoch": 0.36165100995334765,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.338461538461539e-06,
+      "loss": 1.6268,
+      "step": 56900
+    },
+    {
+      "epoch": 0.3622866004805064,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 1.6267,
+      "step": 57000
+    },
+    {
+      "epoch": 0.3629221910076652,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.328205128205129e-06,
+      "loss": 1.6278,
+      "step": 57100
+    },
+    {
+      "epoch": 0.363557781534824,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.323076923076924e-06,
+      "loss": 1.612,
+      "step": 57200
+    },
+    {
+      "epoch": 0.36419337206198277,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.317948717948719e-06,
+      "loss": 1.6283,
+      "step": 57300
+    },
+    {
+      "epoch": 0.36482896258914155,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.312820512820514e-06,
+      "loss": 1.6327,
+      "step": 57400
+    },
+    {
+      "epoch": 0.36546455311630033,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 1.6393,
+      "step": 57500
+    },
+    {
+      "epoch": 0.36610014364345916,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.302564102564103e-06,
+      "loss": 1.6296,
+      "step": 57600
+    },
+    {
+      "epoch": 0.36673573417061794,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.297435897435898e-06,
+      "loss": 1.6276,
+      "step": 57700
+    },
+    {
+      "epoch": 0.3673713246977767,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.292307692307693e-06,
+      "loss": 1.6174,
+      "step": 57800
+    },
+    {
+      "epoch": 0.3680069152249355,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.287179487179487e-06,
+      "loss": 1.6387,
+      "step": 57900
+    },
+    {
+      "epoch": 0.3686425057520943,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.282051282051282e-06,
+      "loss": 1.6251,
+      "step": 58000
+    },
+    {
+      "epoch": 0.36927809627925307,
+      "grad_norm": 0.458984375,
+      "learning_rate": 7.276923076923077e-06,
+      "loss": 1.6321,
+      "step": 58100
+    },
+    {
+      "epoch": 0.36991368680641185,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.271794871794873e-06,
+      "loss": 1.6283,
+      "step": 58200
+    },
+    {
+      "epoch": 0.3705492773335706,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.266666666666668e-06,
+      "loss": 1.6319,
+      "step": 58300
+    },
+    {
+      "epoch": 0.3711848678607294,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.261538461538462e-06,
+      "loss": 1.6195,
+      "step": 58400
+    },
+    {
+      "epoch": 0.3718204583878882,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.256410256410257e-06,
+      "loss": 1.6246,
+      "step": 58500
+    },
+    {
+      "epoch": 0.37245604891504697,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.2512820512820515e-06,
+      "loss": 1.61,
+      "step": 58600
+    },
+    {
+      "epoch": 0.37309163944220575,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.246153846153847e-06,
+      "loss": 1.6302,
+      "step": 58700
+    },
+    {
+      "epoch": 0.3737272299693645,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.241025641025642e-06,
+      "loss": 1.6367,
+      "step": 58800
+    },
+    {
+      "epoch": 0.3743628204965233,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.235897435897436e-06,
+      "loss": 1.6333,
+      "step": 58900
+    },
+    {
+      "epoch": 0.3749984110236821,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.230769230769231e-06,
+      "loss": 1.6266,
+      "step": 59000
+    },
+    {
+      "epoch": 0.37563400155084087,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.225641025641026e-06,
+      "loss": 1.6174,
+      "step": 59100
+    },
+    {
+      "epoch": 0.37626959207799965,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.220512820512822e-06,
+      "loss": 1.6243,
+      "step": 59200
+    },
+    {
+      "epoch": 0.37690518260515843,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.215384615384617e-06,
+      "loss": 1.6376,
+      "step": 59300
+    },
+    {
+      "epoch": 0.3775407731323172,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.210256410256411e-06,
+      "loss": 1.6324,
+      "step": 59400
+    },
+    {
+      "epoch": 0.37817636365947604,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.205128205128206e-06,
+      "loss": 1.6226,
+      "step": 59500
+    },
+    {
+      "epoch": 0.3788119541866348,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 1.6299,
+      "step": 59600
+    },
+    {
+      "epoch": 0.3794475447137936,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.1948717948717946e-06,
+      "loss": 1.6221,
+      "step": 59700
+    },
+    {
+      "epoch": 0.3800831352409524,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.189743589743591e-06,
+      "loss": 1.6129,
+      "step": 59800
+    },
+    {
+      "epoch": 0.38071872576811117,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.184615384615385e-06,
+      "loss": 1.626,
+      "step": 59900
+    },
+    {
+      "epoch": 0.38135431629526995,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 1.6091,
+      "step": 60000
+    },
+    {
+      "epoch": 0.3819899068224287,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.174358974358975e-06,
+      "loss": 1.619,
+      "step": 60100
+    },
+    {
+      "epoch": 0.3826254973495875,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.169230769230769e-06,
+      "loss": 1.624,
+      "step": 60200
+    },
+    {
+      "epoch": 0.3832610878767463,
+      "grad_norm": 0.4921875,
+      "learning_rate": 7.164102564102565e-06,
+      "loss": 1.6376,
+      "step": 60300
+    },
+    {
+      "epoch": 0.38389667840390507,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.15897435897436e-06,
+      "loss": 1.6373,
+      "step": 60400
+    },
+    {
+      "epoch": 0.38453226893106385,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.153846153846155e-06,
+      "loss": 1.6022,
+      "step": 60500
+    },
+    {
+      "epoch": 0.38516785945822263,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.1487179487179495e-06,
+      "loss": 1.626,
+      "step": 60600
+    },
+    {
+      "epoch": 0.3858034499853814,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.1435897435897436e-06,
+      "loss": 1.6359,
+      "step": 60700
+    },
+    {
+      "epoch": 0.3864390405125402,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.1384615384615385e-06,
+      "loss": 1.6121,
+      "step": 60800
+    },
+    {
+      "epoch": 0.38707463103969897,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.133333333333334e-06,
+      "loss": 1.6308,
+      "step": 60900
+    },
+    {
+      "epoch": 0.38771022156685775,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.128205128205129e-06,
+      "loss": 1.6218,
+      "step": 61000
+    },
+    {
+      "epoch": 0.38834581209401653,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.123076923076924e-06,
+      "loss": 1.6201,
+      "step": 61100
+    },
+    {
+      "epoch": 0.3889814026211753,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.117948717948718e-06,
+      "loss": 1.6238,
+      "step": 61200
+    },
+    {
+      "epoch": 0.3896169931483341,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.112820512820513e-06,
+      "loss": 1.6237,
+      "step": 61300
+    },
+    {
+      "epoch": 0.3902525836754929,
+      "grad_norm": 0.4609375,
+      "learning_rate": 7.107692307692309e-06,
+      "loss": 1.6169,
+      "step": 61400
+    },
+    {
+      "epoch": 0.3908881742026517,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.102564102564104e-06,
+      "loss": 1.6405,
+      "step": 61500
+    },
+    {
+      "epoch": 0.3915237647298105,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.0974358974358985e-06,
+      "loss": 1.6304,
+      "step": 61600
+    },
+    {
+      "epoch": 0.39215935525696927,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.0923076923076926e-06,
+      "loss": 1.6279,
+      "step": 61700
+    },
+    {
+      "epoch": 0.39279494578412805,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.0871794871794875e-06,
+      "loss": 1.6094,
+      "step": 61800
+    },
+    {
+      "epoch": 0.3934305363112868,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.082051282051282e-06,
+      "loss": 1.6176,
+      "step": 61900
+    },
+    {
+      "epoch": 0.3940661268384456,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.076923076923078e-06,
+      "loss": 1.6207,
+      "step": 62000
+    },
+    {
+      "epoch": 0.3947017173656044,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.071794871794872e-06,
+      "loss": 1.6209,
+      "step": 62100
+    },
+    {
+      "epoch": 0.39533730789276317,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.066666666666667e-06,
+      "loss": 1.6333,
+      "step": 62200
+    },
+    {
+      "epoch": 0.39597289841992195,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.061538461538462e-06,
+      "loss": 1.6369,
+      "step": 62300
+    },
+    {
+      "epoch": 0.39660848894708073,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.056410256410257e-06,
+      "loss": 1.6412,
+      "step": 62400
+    },
+    {
+      "epoch": 0.3972440794742395,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.051282051282053e-06,
+      "loss": 1.6256,
+      "step": 62500
+    },
+    {
+      "epoch": 0.3978796700013983,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.046153846153847e-06,
+      "loss": 1.6118,
+      "step": 62600
+    },
+    {
+      "epoch": 0.39851526052855707,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.0410256410256415e-06,
+      "loss": 1.6338,
+      "step": 62700
+    },
+    {
+      "epoch": 0.39915085105571585,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.0358974358974364e-06,
+      "loss": 1.6091,
+      "step": 62800
+    },
+    {
+      "epoch": 0.39978644158287463,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.030769230769231e-06,
+      "loss": 1.6314,
+      "step": 62900
+    },
+    {
+      "epoch": 0.4004220321100334,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.025641025641025e-06,
+      "loss": 1.6059,
+      "step": 63000
+    },
+    {
+      "epoch": 0.4010576226371922,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.020512820512821e-06,
+      "loss": 1.6246,
+      "step": 63100
+    },
+    {
+      "epoch": 0.401693213164351,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.015384615384616e-06,
+      "loss": 1.6295,
+      "step": 63200
+    },
+    {
+      "epoch": 0.4023288036915098,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.010256410256411e-06,
+      "loss": 1.6275,
+      "step": 63300
+    },
+    {
+      "epoch": 0.4029643942186686,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.005128205128206e-06,
+      "loss": 1.6077,
+      "step": 63400
+    },
+    {
+      "epoch": 0.40359998474582737,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7e-06,
+      "loss": 1.6033,
+      "step": 63500
+    },
+    {
+      "epoch": 0.40423557527298615,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.994871794871796e-06,
+      "loss": 1.6069,
+      "step": 63600
+    },
+    {
+      "epoch": 0.4048711658001449,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.9897435897435905e-06,
+      "loss": 1.6142,
+      "step": 63700
+    },
+    {
+      "epoch": 0.4055067563273037,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.9846153846153854e-06,
+      "loss": 1.6277,
+      "step": 63800
+    },
+    {
+      "epoch": 0.4061423468544625,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.9794871794871795e-06,
+      "loss": 1.6277,
+      "step": 63900
+    },
+    {
+      "epoch": 0.40677793738162127,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.974358974358974e-06,
+      "loss": 1.6231,
+      "step": 64000
+    },
+    {
+      "epoch": 0.40741352790878005,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.96923076923077e-06,
+      "loss": 1.6271,
+      "step": 64100
+    },
+    {
+      "epoch": 0.40804911843593883,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.964102564102565e-06,
+      "loss": 1.619,
+      "step": 64200
+    },
+    {
+      "epoch": 0.4086847089630976,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.95897435897436e-06,
+      "loss": 1.6364,
+      "step": 64300
+    },
+    {
+      "epoch": 0.4093202994902564,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.953846153846154e-06,
+      "loss": 1.6403,
+      "step": 64400
+    },
+    {
+      "epoch": 0.40995589001741517,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.948717948717949e-06,
+      "loss": 1.6099,
+      "step": 64500
+    },
+    {
+      "epoch": 0.41059148054457395,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.943589743589744e-06,
+      "loss": 1.6325,
+      "step": 64600
+    },
+    {
+      "epoch": 0.41122707107173273,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.9384615384615395e-06,
+      "loss": 1.6174,
+      "step": 64700
+    },
+    {
+      "epoch": 0.4118626615988915,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.9333333333333344e-06,
+      "loss": 1.6233,
+      "step": 64800
+    },
+    {
+      "epoch": 0.4124982521260503,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.9282051282051285e-06,
+      "loss": 1.6225,
+      "step": 64900
+    },
+    {
+      "epoch": 0.41313384265320907,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 1.6309,
+      "step": 65000
+    },
+    {
+      "epoch": 0.4137694331803679,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.917948717948718e-06,
+      "loss": 1.6209,
+      "step": 65100
+    },
+    {
+      "epoch": 0.4144050237075267,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.912820512820514e-06,
+      "loss": 1.6248,
+      "step": 65200
+    },
+    {
+      "epoch": 0.41504061423468547,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.907692307692309e-06,
+      "loss": 1.624,
+      "step": 65300
+    },
+    {
+      "epoch": 0.41567620476184425,
+      "grad_norm": 1.125,
+      "learning_rate": 6.902564102564103e-06,
+      "loss": 1.6319,
+      "step": 65400
+    },
+    {
+      "epoch": 0.41631179528900303,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.897435897435898e-06,
+      "loss": 1.6336,
+      "step": 65500
+    },
+    {
+      "epoch": 0.4169473858161618,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.892307692307693e-06,
+      "loss": 1.5998,
+      "step": 65600
+    },
+    {
+      "epoch": 0.4175829763433206,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.887179487179488e-06,
+      "loss": 1.6258,
+      "step": 65700
+    },
+    {
+      "epoch": 0.41821856687047937,
+      "grad_norm": 0.8515625,
+      "learning_rate": 6.882051282051283e-06,
+      "loss": 1.62,
+      "step": 65800
+    },
+    {
+      "epoch": 0.41885415739763815,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6.8769230769230775e-06,
+      "loss": 1.6249,
+      "step": 65900
+    },
+    {
+      "epoch": 0.41948974792479693,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.871794871794872e-06,
+      "loss": 1.6178,
+      "step": 66000
+    },
+    {
+      "epoch": 0.4201253384519557,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.866666666666667e-06,
+      "loss": 1.623,
+      "step": 66100
+    },
+    {
+      "epoch": 0.4207609289791145,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.861538461538461e-06,
+      "loss": 1.6172,
+      "step": 66200
+    },
+    {
+      "epoch": 0.42139651950627327,
+      "grad_norm": 0.8671875,
+      "learning_rate": 6.856410256410257e-06,
+      "loss": 1.6335,
+      "step": 66300
+    },
+    {
+      "epoch": 0.42203211003343205,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.851282051282052e-06,
+      "loss": 1.6238,
+      "step": 66400
+    },
+    {
+      "epoch": 0.42266770056059083,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.846153846153847e-06,
+      "loss": 1.6213,
+      "step": 66500
+    },
+    {
+      "epoch": 0.4233032910877496,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.841025641025642e-06,
+      "loss": 1.608,
+      "step": 66600
+    },
+    {
+      "epoch": 0.4239388816149084,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.835897435897436e-06,
+      "loss": 1.6241,
+      "step": 66700
+    },
+    {
+      "epoch": 0.42457447214206717,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.830769230769231e-06,
+      "loss": 1.6229,
+      "step": 66800
+    },
+    {
+      "epoch": 0.42521006266922595,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.8256410256410265e-06,
+      "loss": 1.5989,
+      "step": 66900
+    },
+    {
+      "epoch": 0.4258456531963848,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.820512820512821e-06,
+      "loss": 1.616,
+      "step": 67000
+    },
+    {
+      "epoch": 0.42648124372354357,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.815384615384616e-06,
+      "loss": 1.6209,
+      "step": 67100
+    },
+    {
+      "epoch": 0.42711683425070235,
+      "grad_norm": 0.84375,
+      "learning_rate": 6.81025641025641e-06,
+      "loss": 1.6152,
+      "step": 67200
+    },
+    {
+      "epoch": 0.42775242477786113,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.805128205128205e-06,
+      "loss": 1.6277,
+      "step": 67300
+    },
+    {
+      "epoch": 0.4283880153050199,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 1.6131,
+      "step": 67400
+    },
+    {
+      "epoch": 0.4290236058321787,
+      "grad_norm": 0.625,
+      "learning_rate": 6.794871794871796e-06,
+      "loss": 1.6216,
+      "step": 67500
+    },
+    {
+      "epoch": 0.42965919635933747,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.789743589743591e-06,
+      "loss": 1.641,
+      "step": 67600
+    },
+    {
+      "epoch": 0.43029478688649625,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.784615384615385e-06,
+      "loss": 1.6123,
+      "step": 67700
+    },
+    {
+      "epoch": 0.43093037741365503,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.77948717948718e-06,
+      "loss": 1.6135,
+      "step": 67800
+    },
+    {
+      "epoch": 0.4315659679408138,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.774358974358975e-06,
+      "loss": 1.6264,
+      "step": 67900
+    },
+    {
+      "epoch": 0.4322015584679726,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.76923076923077e-06,
+      "loss": 1.625,
+      "step": 68000
+    },
+    {
+      "epoch": 0.43283714899513137,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.764102564102564e-06,
+      "loss": 1.6164,
+      "step": 68100
+    },
+    {
+      "epoch": 0.43347273952229015,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.758974358974359e-06,
+      "loss": 1.6189,
+      "step": 68200
+    },
+    {
+      "epoch": 0.43410833004944893,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.753846153846154e-06,
+      "loss": 1.6089,
+      "step": 68300
+    },
+    {
+      "epoch": 0.4347439205766077,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.748717948717949e-06,
+      "loss": 1.624,
+      "step": 68400
+    },
+    {
+      "epoch": 0.4353795111037665,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.743589743589745e-06,
+      "loss": 1.6255,
+      "step": 68500
+    },
+    {
+      "epoch": 0.4360151016309253,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.738461538461539e-06,
+      "loss": 1.6231,
+      "step": 68600
+    },
+    {
+      "epoch": 0.43665069215808405,
+      "grad_norm": 0.80859375,
+      "learning_rate": 6.733333333333334e-06,
+      "loss": 1.6272,
+      "step": 68700
+    },
+    {
+      "epoch": 0.43728628268524283,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.728205128205129e-06,
+      "loss": 1.615,
+      "step": 68800
+    },
+    {
+      "epoch": 0.43792187321240167,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.723076923076924e-06,
+      "loss": 1.602,
+      "step": 68900
+    },
+    {
+      "epoch": 0.43855746373956045,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.717948717948718e-06,
+      "loss": 1.6188,
+      "step": 69000
+    },
+    {
+      "epoch": 0.43919305426671923,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.712820512820513e-06,
+      "loss": 1.6165,
+      "step": 69100
+    },
+    {
+      "epoch": 0.439828644793878,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.707692307692308e-06,
+      "loss": 1.6149,
+      "step": 69200
+    },
+    {
+      "epoch": 0.4404642353210368,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.702564102564103e-06,
+      "loss": 1.6197,
+      "step": 69300
+    },
+    {
+      "epoch": 0.44109982584819557,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.697435897435898e-06,
+      "loss": 1.6023,
+      "step": 69400
+    },
+    {
+      "epoch": 0.44173541637535435,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.692307692307692e-06,
+      "loss": 1.6237,
+      "step": 69500
+    },
+    {
+      "epoch": 0.44237100690251313,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.687179487179488e-06,
+      "loss": 1.6323,
+      "step": 69600
+    },
+    {
+      "epoch": 0.4430065974296719,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.682051282051283e-06,
+      "loss": 1.6327,
+      "step": 69700
+    },
+    {
+      "epoch": 0.4436421879568307,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.676923076923078e-06,
+      "loss": 1.6022,
+      "step": 69800
+    },
+    {
+      "epoch": 0.44427777848398947,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.671794871794873e-06,
+      "loss": 1.612,
+      "step": 69900
+    },
+    {
+      "epoch": 0.44491336901114825,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.6142,
+      "step": 70000
+    },
+    {
+      "epoch": 0.44554895953830703,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.661538461538462e-06,
+      "loss": 1.6161,
+      "step": 70100
+    },
+    {
+      "epoch": 0.4461845500654658,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.656410256410257e-06,
+      "loss": 1.6224,
+      "step": 70200
+    },
+    {
+      "epoch": 0.4468201405926246,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.651282051282052e-06,
+      "loss": 1.6251,
+      "step": 70300
+    },
+    {
+      "epoch": 0.4474557311197834,
+      "grad_norm": 0.85546875,
+      "learning_rate": 6.646153846153846e-06,
+      "loss": 1.6293,
+      "step": 70400
+    },
+    {
+      "epoch": 0.44809132164694215,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.641025641025641e-06,
+      "loss": 1.6059,
+      "step": 70500
+    },
+    {
+      "epoch": 0.44872691217410093,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.635897435897436e-06,
+      "loss": 1.6126,
+      "step": 70600
+    },
+    {
+      "epoch": 0.4493625027012597,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.630769230769232e-06,
+      "loss": 1.6285,
+      "step": 70700
+    },
+    {
+      "epoch": 0.44999809322841855,
+      "grad_norm": 0.8515625,
+      "learning_rate": 6.625641025641027e-06,
+      "loss": 1.6192,
+      "step": 70800
+    },
+    {
+      "epoch": 0.45063368375557733,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.620512820512821e-06,
+      "loss": 1.6384,
+      "step": 70900
+    },
+    {
+      "epoch": 0.4512692742827361,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.615384615384616e-06,
+      "loss": 1.6262,
+      "step": 71000
+    },
+    {
+      "epoch": 0.4519048648098949,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.6102564102564105e-06,
+      "loss": 1.6247,
+      "step": 71100
+    },
+    {
+      "epoch": 0.45254045533705367,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.605128205128206e-06,
+      "loss": 1.6194,
+      "step": 71200
+    },
+    {
+      "epoch": 0.45317604586421245,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 1.6058,
+      "step": 71300
+    },
+    {
+      "epoch": 0.45381163639137123,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.594871794871795e-06,
+      "loss": 1.6114,
+      "step": 71400
+    },
+    {
+      "epoch": 0.45444722691853,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.58974358974359e-06,
+      "loss": 1.6256,
+      "step": 71500
+    },
+    {
+      "epoch": 0.4550828174456888,
+      "grad_norm": 0.78125,
+      "learning_rate": 6.584615384615385e-06,
+      "loss": 1.6191,
+      "step": 71600
+    },
+    {
+      "epoch": 0.4557184079728476,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.57948717948718e-06,
+      "loss": 1.6058,
+      "step": 71700
+    },
+    {
+      "epoch": 0.45635399850000635,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.574358974358976e-06,
+      "loss": 1.6137,
+      "step": 71800
+    },
+    {
+      "epoch": 0.45698958902716513,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.56923076923077e-06,
+      "loss": 1.6105,
+      "step": 71900
+    },
+    {
+      "epoch": 0.4576251795543239,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.564102564102565e-06,
+      "loss": 1.6335,
+      "step": 72000
+    },
+    {
+      "epoch": 0.4582607700814827,
+      "grad_norm": 0.8203125,
+      "learning_rate": 6.5589743589743595e-06,
+      "loss": 1.6285,
+      "step": 72100
+    },
+    {
+      "epoch": 0.4588963606086415,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.553846153846154e-06,
+      "loss": 1.6156,
+      "step": 72200
+    },
+    {
+      "epoch": 0.45953195113580025,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.548717948717949e-06,
+      "loss": 1.605,
+      "step": 72300
+    },
+    {
+      "epoch": 0.46016754166295903,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.543589743589744e-06,
+      "loss": 1.6195,
+      "step": 72400
+    },
+    {
+      "epoch": 0.4608031321901178,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 1.6225,
+      "step": 72500
+    },
+    {
+      "epoch": 0.4614387227172766,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.533333333333334e-06,
+      "loss": 1.6157,
+      "step": 72600
+    },
+    {
+      "epoch": 0.46207431324443543,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.528205128205128e-06,
+      "loss": 1.6234,
+      "step": 72700
+    },
+    {
+      "epoch": 0.4627099037715942,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.523076923076923e-06,
+      "loss": 1.6382,
+      "step": 72800
+    },
+    {
+      "epoch": 0.463345494298753,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.517948717948719e-06,
+      "loss": 1.6067,
+      "step": 72900
+    },
+    {
+      "epoch": 0.46398108482591177,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.512820512820514e-06,
+      "loss": 1.6205,
+      "step": 73000
+    },
+    {
+      "epoch": 0.46461667535307055,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.5076923076923085e-06,
+      "loss": 1.6172,
+      "step": 73100
+    },
+    {
+      "epoch": 0.46525226588022933,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.5025641025641026e-06,
+      "loss": 1.6187,
+      "step": 73200
+    },
+    {
+      "epoch": 0.4658878564073881,
+      "grad_norm": 0.859375,
+      "learning_rate": 6.4974358974358975e-06,
+      "loss": 1.6047,
+      "step": 73300
+    },
+    {
+      "epoch": 0.4665234469345469,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.492307692307693e-06,
+      "loss": 1.6084,
+      "step": 73400
+    },
+    {
+      "epoch": 0.4671590374617057,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.487179487179488e-06,
+      "loss": 1.6262,
+      "step": 73500
+    },
+    {
+      "epoch": 0.46779462798886445,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.482051282051283e-06,
+      "loss": 1.6148,
+      "step": 73600
+    },
+    {
+      "epoch": 0.46843021851602323,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.476923076923077e-06,
+      "loss": 1.6225,
+      "step": 73700
+    },
+    {
+      "epoch": 0.469065809043182,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.471794871794872e-06,
+      "loss": 1.6238,
+      "step": 73800
+    },
+    {
+      "epoch": 0.4697013995703408,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.466666666666667e-06,
+      "loss": 1.6082,
+      "step": 73900
+    },
+    {
+      "epoch": 0.4703369900974996,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.461538461538463e-06,
+      "loss": 1.6232,
+      "step": 74000
+    },
+    {
+      "epoch": 0.47097258062465835,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.4564102564102575e-06,
+      "loss": 1.6253,
+      "step": 74100
+    },
+    {
+      "epoch": 0.47160817115181713,
+      "grad_norm": 0.86328125,
+      "learning_rate": 6.4512820512820516e-06,
+      "loss": 1.618,
+      "step": 74200
+    },
+    {
+      "epoch": 0.4722437616789759,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.4461538461538465e-06,
+      "loss": 1.6153,
+      "step": 74300
+    },
+    {
+      "epoch": 0.4728793522061347,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.441025641025641e-06,
+      "loss": 1.6182,
+      "step": 74400
+    },
+    {
+      "epoch": 0.4735149427332935,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.435897435897437e-06,
+      "loss": 1.6308,
+      "step": 74500
+    },
+    {
+      "epoch": 0.4741505332604523,
+      "grad_norm": 0.859375,
+      "learning_rate": 6.430769230769231e-06,
+      "loss": 1.6259,
+      "step": 74600
+    },
+    {
+      "epoch": 0.4747861237876111,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.425641025641026e-06,
+      "loss": 1.6183,
+      "step": 74700
+    },
+    {
+      "epoch": 0.47542171431476987,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.420512820512821e-06,
+      "loss": 1.6124,
+      "step": 74800
+    },
+    {
+      "epoch": 0.47605730484192865,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.415384615384616e-06,
+      "loss": 1.6082,
+      "step": 74900
+    },
+    {
+      "epoch": 0.47669289536908743,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.410256410256412e-06,
+      "loss": 1.6288,
+      "step": 75000
+    },
+    {
+      "epoch": 0.4773284858962462,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.405128205128206e-06,
+      "loss": 1.6089,
+      "step": 75100
+    },
+    {
+      "epoch": 0.477964076423405,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 1.6039,
+      "step": 75200
+    },
+    {
+      "epoch": 0.4785996669505638,
+      "grad_norm": 0.81640625,
+      "learning_rate": 6.3948717948717955e-06,
+      "loss": 1.6228,
+      "step": 75300
+    },
+    {
+      "epoch": 0.47923525747772255,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.38974358974359e-06,
+      "loss": 1.6295,
+      "step": 75400
+    },
+    {
+      "epoch": 0.47987084800488133,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.384615384615384e-06,
+      "loss": 1.6375,
+      "step": 75500
+    },
+    {
+      "epoch": 0.4805064385320401,
+      "grad_norm": 0.8515625,
+      "learning_rate": 6.37948717948718e-06,
+      "loss": 1.621,
+      "step": 75600
+    },
+    {
+      "epoch": 0.4811420290591989,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.374358974358975e-06,
+      "loss": 1.6176,
+      "step": 75700
+    },
+    {
+      "epoch": 0.4817776195863577,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.36923076923077e-06,
+      "loss": 1.6037,
+      "step": 75800
+    },
+    {
+      "epoch": 0.48241321011351646,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.364102564102565e-06,
+      "loss": 1.6202,
+      "step": 75900
+    },
+    {
+      "epoch": 0.48304880064067524,
+      "grad_norm": 0.84375,
+      "learning_rate": 6.358974358974359e-06,
+      "loss": 1.6072,
+      "step": 76000
+    },
+    {
+      "epoch": 0.483684391167834,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.353846153846155e-06,
+      "loss": 1.6253,
+      "step": 76100
+    },
+    {
+      "epoch": 0.4843199816949928,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.3487179487179495e-06,
+      "loss": 1.6154,
+      "step": 76200
+    },
+    {
+      "epoch": 0.4849555722221516,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.3435897435897444e-06,
+      "loss": 1.6045,
+      "step": 76300
+    },
+    {
+      "epoch": 0.48559116274931036,
+      "grad_norm": 0.86328125,
+      "learning_rate": 6.3384615384615385e-06,
+      "loss": 1.63,
+      "step": 76400
+    },
+    {
+      "epoch": 0.4862267532764692,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.333333333333333e-06,
+      "loss": 1.615,
+      "step": 76500
+    },
+    {
+      "epoch": 0.486862343803628,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.328205128205128e-06,
+      "loss": 1.6177,
+      "step": 76600
+    },
+    {
+      "epoch": 0.48749793433078675,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.323076923076924e-06,
+      "loss": 1.6147,
+      "step": 76700
+    },
+    {
+      "epoch": 0.48813352485794553,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.317948717948719e-06,
+      "loss": 1.6187,
+      "step": 76800
+    },
+    {
+      "epoch": 0.4887691153851043,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.312820512820513e-06,
+      "loss": 1.6278,
+      "step": 76900
+    },
+    {
+      "epoch": 0.4894047059122631,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.307692307692308e-06,
+      "loss": 1.613,
+      "step": 77000
+    },
+    {
+      "epoch": 0.4900402964394219,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.302564102564103e-06,
+      "loss": 1.6121,
+      "step": 77100
+    },
+    {
+      "epoch": 0.49067588696658065,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.2974358974358985e-06,
+      "loss": 1.6129,
+      "step": 77200
+    },
+    {
+      "epoch": 0.49131147749373943,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.2923076923076934e-06,
+      "loss": 1.6027,
+      "step": 77300
+    },
+    {
+      "epoch": 0.4919470680208982,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.2871794871794875e-06,
+      "loss": 1.6215,
+      "step": 77400
+    },
+    {
+      "epoch": 0.492582658548057,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 1.6139,
+      "step": 77500
+    },
+    {
+      "epoch": 0.4932182490752158,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.276923076923077e-06,
+      "loss": 1.6079,
+      "step": 77600
+    },
+    {
+      "epoch": 0.49385383960237456,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.271794871794872e-06,
+      "loss": 1.6265,
+      "step": 77700
+    },
+    {
+      "epoch": 0.49448943012953334,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.266666666666668e-06,
+      "loss": 1.6138,
+      "step": 77800
+    },
+    {
+      "epoch": 0.4951250206566921,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.261538461538462e-06,
+      "loss": 1.6229,
+      "step": 77900
+    },
+    {
+      "epoch": 0.4957606111838509,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.256410256410257e-06,
+      "loss": 1.6074,
+      "step": 78000
+    },
+    {
+      "epoch": 0.4963962017110097,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.251282051282052e-06,
+      "loss": 1.61,
+      "step": 78100
+    },
+    {
+      "epoch": 0.49703179223816846,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.246153846153846e-06,
+      "loss": 1.6369,
+      "step": 78200
+    },
+    {
+      "epoch": 0.4976673827653273,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.2410256410256424e-06,
+      "loss": 1.6319,
+      "step": 78300
+    },
+    {
+      "epoch": 0.4983029732924861,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.2358974358974365e-06,
+      "loss": 1.6078,
+      "step": 78400
+    },
+    {
+      "epoch": 0.49893856381964485,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.230769230769231e-06,
+      "loss": 1.6118,
+      "step": 78500
+    },
+    {
+      "epoch": 0.49957415434680363,
+      "grad_norm": 0.80859375,
+      "learning_rate": 6.225641025641026e-06,
+      "loss": 1.6093,
+      "step": 78600
+    },
+    {
+      "epoch": 0.5002097448739624,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.22051282051282e-06,
+      "loss": 1.6282,
+      "step": 78700
+    },
+    {
+      "epoch": 0.5008453354011212,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.215384615384615e-06,
+      "loss": 1.6049,
+      "step": 78800
+    },
+    {
+      "epoch": 0.50148092592828,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.210256410256411e-06,
+      "loss": 1.6071,
+      "step": 78900
+    },
+    {
+      "epoch": 0.5021165164554388,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.205128205128206e-06,
+      "loss": 1.6211,
+      "step": 79000
+    },
+    {
+      "epoch": 0.5027521069825975,
+      "grad_norm": 0.84375,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 1.6098,
+      "step": 79100
+    },
+    {
+      "epoch": 0.5033876975097563,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.194871794871795e-06,
+      "loss": 1.6244,
+      "step": 79200
+    },
+    {
+      "epoch": 0.5040232880369151,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.18974358974359e-06,
+      "loss": 1.6257,
+      "step": 79300
+    },
+    {
+      "epoch": 0.5046588785640739,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.1846153846153855e-06,
+      "loss": 1.6124,
+      "step": 79400
+    },
+    {
+      "epoch": 0.5052944690912327,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.17948717948718e-06,
+      "loss": 1.6267,
+      "step": 79500
+    },
+    {
+      "epoch": 0.5059300596183914,
+      "grad_norm": 1.2734375,
+      "learning_rate": 6.174358974358975e-06,
+      "loss": 1.6178,
+      "step": 79600
+    },
+    {
+      "epoch": 0.5065656501455502,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.169230769230769e-06,
+      "loss": 1.6265,
+      "step": 79700
+    },
+    {
+      "epoch": 0.507201240672709,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.164102564102564e-06,
+      "loss": 1.6248,
+      "step": 79800
+    },
+    {
+      "epoch": 0.5078368311998678,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.15897435897436e-06,
+      "loss": 1.6117,
+      "step": 79900
+    },
+    {
+      "epoch": 0.5084724217270266,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 1.6164,
+      "step": 80000
+    },
+    {
+      "epoch": 0.5091080122541853,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.14871794871795e-06,
+      "loss": 1.6257,
+      "step": 80100
+    },
+    {
+      "epoch": 0.5097436027813441,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.143589743589744e-06,
+      "loss": 1.6203,
+      "step": 80200
+    },
+    {
+      "epoch": 0.5103791933085029,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.138461538461539e-06,
+      "loss": 1.6308,
+      "step": 80300
+    },
+    {
+      "epoch": 0.5110147838356617,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.133333333333334e-06,
+      "loss": 1.6082,
+      "step": 80400
+    },
+    {
+      "epoch": 0.5116503743628205,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.128205128205129e-06,
+      "loss": 1.6221,
+      "step": 80500
+    },
+    {
+      "epoch": 0.5122859648899792,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.123076923076923e-06,
+      "loss": 1.6169,
+      "step": 80600
+    },
+    {
+      "epoch": 0.512921555417138,
+      "grad_norm": 0.81640625,
+      "learning_rate": 6.117948717948718e-06,
+      "loss": 1.5958,
+      "step": 80700
+    },
+    {
+      "epoch": 0.5135571459442968,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.112820512820513e-06,
+      "loss": 1.6164,
+      "step": 80800
+    },
+    {
+      "epoch": 0.5141927364714556,
+      "grad_norm": 0.77734375,
+      "learning_rate": 6.107692307692308e-06,
+      "loss": 1.6204,
+      "step": 80900
+    },
+    {
+      "epoch": 0.5148283269986144,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.102564102564104e-06,
+      "loss": 1.6124,
+      "step": 81000
+    },
+    {
+      "epoch": 0.5154639175257731,
+      "grad_norm": 0.828125,
+      "learning_rate": 6.097435897435898e-06,
+      "loss": 1.6359,
+      "step": 81100
+    },
+    {
+      "epoch": 0.516099508052932,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.092307692307693e-06,
+      "loss": 1.625,
+      "step": 81200
+    },
+    {
+      "epoch": 0.5167350985800908,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.087179487179488e-06,
+      "loss": 1.6147,
+      "step": 81300
+    },
+    {
+      "epoch": 0.5173706891072496,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.082051282051283e-06,
+      "loss": 1.6221,
+      "step": 81400
+    },
+    {
+      "epoch": 0.5180062796344084,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.076923076923077e-06,
+      "loss": 1.6185,
+      "step": 81500
+    },
+    {
+      "epoch": 0.5186418701615672,
+      "grad_norm": 1.296875,
+      "learning_rate": 6.071794871794872e-06,
+      "loss": 1.6126,
+      "step": 81600
+    },
+    {
+      "epoch": 0.5192774606887259,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.066666666666667e-06,
+      "loss": 1.6242,
+      "step": 81700
+    },
+    {
+      "epoch": 0.5199130512158847,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.061538461538462e-06,
+      "loss": 1.6178,
+      "step": 81800
+    },
+    {
+      "epoch": 0.5205486417430435,
+      "grad_norm": 0.875,
+      "learning_rate": 6.056410256410257e-06,
+      "loss": 1.6235,
+      "step": 81900
+    },
+    {
+      "epoch": 0.5211842322702023,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.051282051282051e-06,
+      "loss": 1.6243,
+      "step": 82000
+    },
+    {
+      "epoch": 0.521819822797361,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.046153846153847e-06,
+      "loss": 1.637,
+      "step": 82100
+    },
+    {
+      "epoch": 0.5224554133245198,
+      "grad_norm": 0.83203125,
+      "learning_rate": 6.041025641025642e-06,
+      "loss": 1.5933,
+      "step": 82200
+    },
+    {
+      "epoch": 0.5230910038516786,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.035897435897437e-06,
+      "loss": 1.6261,
+      "step": 82300
+    },
+    {
+      "epoch": 0.5237265943788374,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.030769230769231e-06,
+      "loss": 1.629,
+      "step": 82400
+    },
+    {
+      "epoch": 0.5243621849059962,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.025641025641026e-06,
+      "loss": 1.616,
+      "step": 82500
+    },
+    {
+      "epoch": 0.524997775433155,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.0205128205128206e-06,
+      "loss": 1.6094,
+      "step": 82600
+    },
+    {
+      "epoch": 0.5256333659603137,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.015384615384616e-06,
+      "loss": 1.6166,
+      "step": 82700
+    },
+    {
+      "epoch": 0.5262689564874725,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.010256410256411e-06,
+      "loss": 1.6264,
+      "step": 82800
+    },
+    {
+      "epoch": 0.5269045470146313,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.005128205128205e-06,
+      "loss": 1.6136,
+      "step": 82900
+    },
+    {
+      "epoch": 0.5275401375417901,
+      "grad_norm": 0.8125,
+      "learning_rate": 6e-06,
+      "loss": 1.6208,
+      "step": 83000
+    },
+    {
+      "epoch": 0.5281757280689489,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.994871794871795e-06,
+      "loss": 1.6156,
+      "step": 83100
+    },
+    {
+      "epoch": 0.5288113185961076,
+      "grad_norm": 1.0,
+      "learning_rate": 5.989743589743591e-06,
+      "loss": 1.6168,
+      "step": 83200
+    },
+    {
+      "epoch": 0.5294469091232664,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.984615384615386e-06,
+      "loss": 1.6293,
+      "step": 83300
+    },
+    {
+      "epoch": 0.5300824996504252,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.97948717948718e-06,
+      "loss": 1.6071,
+      "step": 83400
+    },
+    {
+      "epoch": 0.530718090177584,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.974358974358975e-06,
+      "loss": 1.6244,
+      "step": 83500
+    },
+    {
+      "epoch": 0.5313536807047428,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.9692307692307695e-06,
+      "loss": 1.6089,
+      "step": 83600
+    },
+    {
+      "epoch": 0.5319892712319015,
+      "grad_norm": 1.0,
+      "learning_rate": 5.9641025641025644e-06,
+      "loss": 1.6285,
+      "step": 83700
+    },
+    {
+      "epoch": 0.5326248617590603,
+      "grad_norm": 0.84375,
+      "learning_rate": 5.95897435897436e-06,
+      "loss": 1.6131,
+      "step": 83800
+    },
+    {
+      "epoch": 0.5332604522862191,
+      "grad_norm": 0.91015625,
+      "learning_rate": 5.953846153846154e-06,
+      "loss": 1.6274,
+      "step": 83900
+    },
+    {
+      "epoch": 0.5338960428133779,
+      "grad_norm": 0.75,
+      "learning_rate": 5.948717948717949e-06,
+      "loss": 1.6138,
+      "step": 84000
+    },
+    {
+      "epoch": 0.5345316333405367,
+      "grad_norm": 0.85546875,
+      "learning_rate": 5.943589743589744e-06,
+      "loss": 1.6348,
+      "step": 84100
+    },
+    {
+      "epoch": 0.5351672238676954,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.938461538461538e-06,
+      "loss": 1.6115,
+      "step": 84200
+    },
+    {
+      "epoch": 0.5358028143948542,
+      "grad_norm": 0.84765625,
+      "learning_rate": 5.933333333333335e-06,
+      "loss": 1.6196,
+      "step": 84300
+    },
+    {
+      "epoch": 0.536438404922013,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.928205128205129e-06,
+      "loss": 1.6257,
+      "step": 84400
+    },
+    {
+      "epoch": 0.5370739954491718,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.923076923076924e-06,
+      "loss": 1.6163,
+      "step": 84500
+    },
+    {
+      "epoch": 0.5377095859763306,
+      "grad_norm": 0.82421875,
+      "learning_rate": 5.9179487179487185e-06,
+      "loss": 1.6273,
+      "step": 84600
+    },
+    {
+      "epoch": 0.5383451765034893,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.912820512820513e-06,
+      "loss": 1.634,
+      "step": 84700
+    },
+    {
+      "epoch": 0.5389807670306481,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.907692307692308e-06,
+      "loss": 1.6107,
+      "step": 84800
+    },
+    {
+      "epoch": 0.539616357557807,
+      "grad_norm": 0.828125,
+      "learning_rate": 5.902564102564103e-06,
+      "loss": 1.6259,
+      "step": 84900
+    },
+    {
+      "epoch": 0.5402519480849658,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.897435897435898e-06,
+      "loss": 1.6044,
+      "step": 85000
+    },
+    {
+      "epoch": 0.5408875386121246,
+      "grad_norm": 0.83203125,
+      "learning_rate": 5.892307692307693e-06,
+      "loss": 1.6064,
+      "step": 85100
+    },
+    {
+      "epoch": 0.5415231291392834,
+      "grad_norm": 0.79296875,
+      "learning_rate": 5.887179487179487e-06,
+      "loss": 1.6197,
+      "step": 85200
+    },
+    {
+      "epoch": 0.5421587196664421,
+      "grad_norm": 0.84375,
+      "learning_rate": 5.882051282051282e-06,
+      "loss": 1.619,
+      "step": 85300
+    },
+    {
+      "epoch": 0.5427943101936009,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.876923076923078e-06,
+      "loss": 1.5951,
+      "step": 85400
+    },
+    {
+      "epoch": 0.5434299007207597,
+      "grad_norm": 0.8828125,
+      "learning_rate": 5.871794871794873e-06,
+      "loss": 1.6088,
+      "step": 85500
+    },
+    {
+      "epoch": 0.5440654912479185,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.8666666666666675e-06,
+      "loss": 1.6291,
+      "step": 85600
+    },
+    {
+      "epoch": 0.5447010817750773,
+      "grad_norm": 1.203125,
+      "learning_rate": 5.861538461538462e-06,
+      "loss": 1.6296,
+      "step": 85700
+    },
+    {
+      "epoch": 0.545336672302236,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.8564102564102565e-06,
+      "loss": 1.6253,
+      "step": 85800
+    },
+    {
+      "epoch": 0.5459722628293948,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.851282051282052e-06,
+      "loss": 1.6294,
+      "step": 85900
+    },
+    {
+      "epoch": 0.5466078533565536,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.846153846153847e-06,
+      "loss": 1.6221,
+      "step": 86000
+    },
+    {
+      "epoch": 0.5472434438837124,
+      "grad_norm": 0.79296875,
+      "learning_rate": 5.841025641025642e-06,
+      "loss": 1.6114,
+      "step": 86100
+    },
+    {
+      "epoch": 0.5478790344108712,
+      "grad_norm": 0.92578125,
+      "learning_rate": 5.835897435897436e-06,
+      "loss": 1.6158,
+      "step": 86200
+    },
+    {
+      "epoch": 0.5485146249380299,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.830769230769231e-06,
+      "loss": 1.6054,
+      "step": 86300
+    },
+    {
+      "epoch": 0.5491502154651887,
+      "grad_norm": 0.828125,
+      "learning_rate": 5.825641025641026e-06,
+      "loss": 1.623,
+      "step": 86400
+    },
+    {
+      "epoch": 0.5497858059923475,
+      "grad_norm": 0.85546875,
+      "learning_rate": 5.820512820512822e-06,
+      "loss": 1.6108,
+      "step": 86500
+    },
+    {
+      "epoch": 0.5504213965195063,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.815384615384616e-06,
+      "loss": 1.6228,
+      "step": 86600
+    },
+    {
+      "epoch": 0.5510569870466651,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.8102564102564106e-06,
+      "loss": 1.633,
+      "step": 86700
+    },
+    {
+      "epoch": 0.5516925775738238,
+      "grad_norm": 0.78515625,
+      "learning_rate": 5.8051282051282055e-06,
+      "loss": 1.6187,
+      "step": 86800
+    },
+    {
+      "epoch": 0.5523281681009826,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.8e-06,
+      "loss": 1.6125,
+      "step": 86900
+    },
+    {
+      "epoch": 0.5529637586281414,
+      "grad_norm": 0.8203125,
+      "learning_rate": 5.794871794871796e-06,
+      "loss": 1.6183,
+      "step": 87000
+    },
+    {
+      "epoch": 0.5535993491553002,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.78974358974359e-06,
+      "loss": 1.6264,
+      "step": 87100
+    },
+    {
+      "epoch": 0.554234939682459,
+      "grad_norm": 0.80078125,
+      "learning_rate": 5.784615384615385e-06,
+      "loss": 1.6289,
+      "step": 87200
+    },
+    {
+      "epoch": 0.5548705302096177,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.77948717948718e-06,
+      "loss": 1.6042,
+      "step": 87300
+    },
+    {
+      "epoch": 0.5555061207367765,
+      "grad_norm": 0.8671875,
+      "learning_rate": 5.774358974358975e-06,
+      "loss": 1.6234,
+      "step": 87400
+    },
+    {
+      "epoch": 0.5561417112639353,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 1.6233,
+      "step": 87500
+    },
+    {
+      "epoch": 0.5567773017910941,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.764102564102565e-06,
+      "loss": 1.6176,
+      "step": 87600
+    },
+    {
+      "epoch": 0.5574128923182529,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.7589743589743596e-06,
+      "loss": 1.6216,
+      "step": 87700
+    },
+    {
+      "epoch": 0.5580484828454116,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.7538461538461545e-06,
+      "loss": 1.6321,
+      "step": 87800
+    },
+    {
+      "epoch": 0.5586840733725704,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.748717948717949e-06,
+      "loss": 1.6245,
+      "step": 87900
+    },
+    {
+      "epoch": 0.5593196638997292,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.743589743589743e-06,
+      "loss": 1.6279,
+      "step": 88000
+    },
+    {
+      "epoch": 0.559955254426888,
+      "grad_norm": 0.85546875,
+      "learning_rate": 5.738461538461539e-06,
+      "loss": 1.6187,
+      "step": 88100
+    },
+    {
+      "epoch": 0.5605908449540468,
+      "grad_norm": 0.625,
+      "learning_rate": 5.733333333333334e-06,
+      "loss": 1.6266,
+      "step": 88200
+    },
+    {
+      "epoch": 0.5612264354812055,
+      "grad_norm": 0.625,
+      "learning_rate": 5.728205128205129e-06,
+      "loss": 1.6191,
+      "step": 88300
+    },
+    {
+      "epoch": 0.5618620260083643,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.723076923076923e-06,
+      "loss": 1.609,
+      "step": 88400
+    },
+    {
+      "epoch": 0.5624976165355231,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.717948717948718e-06,
+      "loss": 1.6182,
+      "step": 88500
+    },
+    {
+      "epoch": 0.5631332070626819,
+      "grad_norm": 0.8359375,
+      "learning_rate": 5.712820512820513e-06,
+      "loss": 1.6317,
+      "step": 88600
+    },
+    {
+      "epoch": 0.5637687975898408,
+      "grad_norm": 0.91796875,
+      "learning_rate": 5.7076923076923086e-06,
+      "loss": 1.6025,
+      "step": 88700
+    },
+    {
+      "epoch": 0.5644043881169996,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.7025641025641035e-06,
+      "loss": 1.6301,
+      "step": 88800
+    },
+    {
+      "epoch": 0.5650399786441583,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.6974358974358975e-06,
+      "loss": 1.6327,
+      "step": 88900
+    },
+    {
+      "epoch": 0.5656755691713171,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.692307692307692e-06,
+      "loss": 1.6271,
+      "step": 89000
+    },
+    {
+      "epoch": 0.5663111596984759,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.687179487179487e-06,
+      "loss": 1.6137,
+      "step": 89100
+    },
+    {
+      "epoch": 0.5669467502256347,
+      "grad_norm": 1.2109375,
+      "learning_rate": 5.682051282051283e-06,
+      "loss": 1.6217,
+      "step": 89200
+    },
+    {
+      "epoch": 0.5675823407527935,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.676923076923078e-06,
+      "loss": 1.6228,
+      "step": 89300
+    },
+    {
+      "epoch": 0.5682179312799522,
+      "grad_norm": 1.203125,
+      "learning_rate": 5.671794871794872e-06,
+      "loss": 1.6288,
+      "step": 89400
+    },
+    {
+      "epoch": 0.568853521807111,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.666666666666667e-06,
+      "loss": 1.6124,
+      "step": 89500
+    },
+    {
+      "epoch": 0.5694891123342698,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.661538461538462e-06,
+      "loss": 1.6081,
+      "step": 89600
+    },
+    {
+      "epoch": 0.5701247028614286,
+      "grad_norm": 0.77734375,
+      "learning_rate": 5.6564102564102575e-06,
+      "loss": 1.625,
+      "step": 89700
+    },
+    {
+      "epoch": 0.5707602933885874,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.6512820512820524e-06,
+      "loss": 1.6106,
+      "step": 89800
+    },
+    {
+      "epoch": 0.5713958839157461,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.6461538461538465e-06,
+      "loss": 1.6322,
+      "step": 89900
+    },
+    {
+      "epoch": 0.5720314744429049,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 1.633,
+      "step": 90000
+    },
+    {
+      "epoch": 0.5726670649700637,
+      "grad_norm": 0.625,
+      "learning_rate": 5.635897435897436e-06,
+      "loss": 1.6196,
+      "step": 90100
+    },
+    {
+      "epoch": 0.5733026554972225,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.63076923076923e-06,
+      "loss": 1.645,
+      "step": 90200
+    },
+    {
+      "epoch": 0.5739382460243813,
+      "grad_norm": 0.89453125,
+      "learning_rate": 5.625641025641027e-06,
+      "loss": 1.6166,
+      "step": 90300
+    },
+    {
+      "epoch": 0.57457383655154,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.620512820512821e-06,
+      "loss": 1.6207,
+      "step": 90400
+    },
+    {
+      "epoch": 0.5752094270786988,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.615384615384616e-06,
+      "loss": 1.6138,
+      "step": 90500
+    },
+    {
+      "epoch": 0.5758450176058576,
+      "grad_norm": 1.0,
+      "learning_rate": 5.610256410256411e-06,
+      "loss": 1.6108,
+      "step": 90600
+    },
+    {
+      "epoch": 0.5764806081330164,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.605128205128205e-06,
+      "loss": 1.6291,
+      "step": 90700
+    },
+    {
+      "epoch": 0.5771161986601752,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.6296,
+      "step": 90800
+    },
+    {
+      "epoch": 0.5777517891873339,
+      "grad_norm": 0.75,
+      "learning_rate": 5.5948717948717955e-06,
+      "loss": 1.6183,
+      "step": 90900
+    },
+    {
+      "epoch": 0.5783873797144927,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.58974358974359e-06,
+      "loss": 1.6069,
+      "step": 91000
+    },
+    {
+      "epoch": 0.5790229702416515,
+      "grad_norm": 0.87109375,
+      "learning_rate": 5.584615384615385e-06,
+      "loss": 1.608,
+      "step": 91100
+    },
+    {
+      "epoch": 0.5796585607688103,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.579487179487179e-06,
+      "loss": 1.6259,
+      "step": 91200
+    },
+    {
+      "epoch": 0.5802941512959691,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.574358974358974e-06,
+      "loss": 1.6293,
+      "step": 91300
+    },
+    {
+      "epoch": 0.5809297418231278,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.56923076923077e-06,
+      "loss": 1.6288,
+      "step": 91400
+    },
+    {
+      "epoch": 0.5815653323502866,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.564102564102565e-06,
+      "loss": 1.619,
+      "step": 91500
+    },
+    {
+      "epoch": 0.5822009228774454,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.55897435897436e-06,
+      "loss": 1.6338,
+      "step": 91600
+    },
+    {
+      "epoch": 0.5828365134046042,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.553846153846154e-06,
+      "loss": 1.6173,
+      "step": 91700
+    },
+    {
+      "epoch": 0.583472103931763,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.548717948717949e-06,
+      "loss": 1.6255,
+      "step": 91800
+    },
+    {
+      "epoch": 0.5841076944589217,
+      "grad_norm": 0.490234375,
+      "learning_rate": 5.5435897435897445e-06,
+      "loss": 1.5943,
+      "step": 91900
+    },
+    {
+      "epoch": 0.5847432849860805,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.538461538461539e-06,
+      "loss": 1.6178,
+      "step": 92000
+    },
+    {
+      "epoch": 0.5853788755132393,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.533333333333334e-06,
+      "loss": 1.6184,
+      "step": 92100
+    },
+    {
+      "epoch": 0.5860144660403981,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.528205128205128e-06,
+      "loss": 1.6053,
+      "step": 92200
+    },
+    {
+      "epoch": 0.5866500565675569,
+      "grad_norm": 1.0,
+      "learning_rate": 5.523076923076923e-06,
+      "loss": 1.6308,
+      "step": 92300
+    },
+    {
+      "epoch": 0.5872856470947156,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.517948717948718e-06,
+      "loss": 1.626,
+      "step": 92400
+    },
+    {
+      "epoch": 0.5879212376218745,
+      "grad_norm": 1.25,
+      "learning_rate": 5.512820512820514e-06,
+      "loss": 1.623,
+      "step": 92500
+    },
+    {
+      "epoch": 0.5885568281490333,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.507692307692308e-06,
+      "loss": 1.631,
+      "step": 92600
+    },
+    {
+      "epoch": 0.5891924186761921,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.502564102564103e-06,
+      "loss": 1.6301,
+      "step": 92700
+    },
+    {
+      "epoch": 0.5898280092033509,
+      "grad_norm": 0.76171875,
+      "learning_rate": 5.497435897435898e-06,
+      "loss": 1.6266,
+      "step": 92800
+    },
+    {
+      "epoch": 0.5904635997305097,
+      "grad_norm": 0.78515625,
+      "learning_rate": 5.492307692307693e-06,
+      "loss": 1.6179,
+      "step": 92900
+    },
+    {
+      "epoch": 0.5910991902576684,
+      "grad_norm": 0.8125,
+      "learning_rate": 5.487179487179488e-06,
+      "loss": 1.625,
+      "step": 93000
+    },
+    {
+      "epoch": 0.5917347807848272,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.4820512820512824e-06,
+      "loss": 1.6181,
+      "step": 93100
+    },
+    {
+      "epoch": 0.592370371311986,
+      "grad_norm": 0.75,
+      "learning_rate": 5.476923076923077e-06,
+      "loss": 1.6304,
+      "step": 93200
+    },
+    {
+      "epoch": 0.5930059618391448,
+      "grad_norm": 0.79296875,
+      "learning_rate": 5.471794871794872e-06,
+      "loss": 1.6004,
+      "step": 93300
+    },
+    {
+      "epoch": 0.5936415523663036,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.466666666666667e-06,
+      "loss": 1.6335,
+      "step": 93400
+    },
+    {
+      "epoch": 0.5942771428934623,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.461538461538461e-06,
+      "loss": 1.616,
+      "step": 93500
+    },
+    {
+      "epoch": 0.5949127334206211,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.456410256410257e-06,
+      "loss": 1.6155,
+      "step": 93600
+    },
+    {
+      "epoch": 0.5955483239477799,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.451282051282052e-06,
+      "loss": 1.6241,
+      "step": 93700
+    },
+    {
+      "epoch": 0.5961839144749387,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.446153846153847e-06,
+      "loss": 1.6173,
+      "step": 93800
+    },
+    {
+      "epoch": 0.5968195050020975,
+      "grad_norm": 0.7578125,
+      "learning_rate": 5.441025641025642e-06,
+      "loss": 1.6389,
+      "step": 93900
+    },
+    {
+      "epoch": 0.5974550955292562,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.435897435897436e-06,
+      "loss": 1.6153,
+      "step": 94000
+    },
+    {
+      "epoch": 0.598090686056415,
+      "grad_norm": 0.83984375,
+      "learning_rate": 5.430769230769231e-06,
+      "loss": 1.6275,
+      "step": 94100
+    },
+    {
+      "epoch": 0.5987262765835738,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.425641025641026e-06,
+      "loss": 1.6286,
+      "step": 94200
+    },
+    {
+      "epoch": 0.5993618671107326,
+      "grad_norm": 0.84765625,
+      "learning_rate": 5.420512820512821e-06,
+      "loss": 1.617,
+      "step": 94300
+    },
+    {
+      "epoch": 0.5999974576378914,
+      "grad_norm": 0.8359375,
+      "learning_rate": 5.415384615384615e-06,
+      "loss": 1.6307,
+      "step": 94400
+    },
+    {
+      "epoch": 0.6006330481650501,
+      "grad_norm": 0.7734375,
+      "learning_rate": 5.41025641025641e-06,
+      "loss": 1.6397,
+      "step": 94500
+    },
+    {
+      "epoch": 0.6012686386922089,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.405128205128205e-06,
+      "loss": 1.6146,
+      "step": 94600
+    },
+    {
+      "epoch": 0.6019042292193677,
+      "grad_norm": 0.9609375,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 1.628,
+      "step": 94700
+    },
+    {
+      "epoch": 0.6025398197465265,
+      "grad_norm": 0.859375,
+      "learning_rate": 5.394871794871796e-06,
+      "loss": 1.6315,
+      "step": 94800
+    },
+    {
+      "epoch": 0.6031754102736853,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.38974358974359e-06,
+      "loss": 1.6189,
+      "step": 94900
+    },
+    {
+      "epoch": 0.603811000800844,
+      "grad_norm": 0.91015625,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 1.6219,
+      "step": 95000
+    },
+    {
+      "epoch": 0.6044465913280028,
+      "grad_norm": 0.92578125,
+      "learning_rate": 5.3794871794871796e-06,
+      "loss": 1.613,
+      "step": 95100
+    },
+    {
+      "epoch": 0.6050821818551616,
+      "grad_norm": 1.328125,
+      "learning_rate": 5.374358974358975e-06,
+      "loss": 1.6054,
+      "step": 95200
+    },
+    {
+      "epoch": 0.6057177723823204,
+      "grad_norm": 1.203125,
+      "learning_rate": 5.36923076923077e-06,
+      "loss": 1.6211,
+      "step": 95300
+    },
+    {
+      "epoch": 0.6063533629094792,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.364102564102564e-06,
+      "loss": 1.6168,
+      "step": 95400
+    },
+    {
+      "epoch": 0.606988953436638,
+      "grad_norm": 0.85546875,
+      "learning_rate": 5.358974358974359e-06,
+      "loss": 1.6289,
+      "step": 95500
+    },
+    {
+      "epoch": 0.6076245439637967,
+      "grad_norm": 0.83984375,
+      "learning_rate": 5.353846153846154e-06,
+      "loss": 1.612,
+      "step": 95600
+    },
+    {
+      "epoch": 0.6082601344909555,
+      "grad_norm": 0.8046875,
+      "learning_rate": 5.34871794871795e-06,
+      "loss": 1.6154,
+      "step": 95700
+    },
+    {
+      "epoch": 0.6088957250181143,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.343589743589745e-06,
+      "loss": 1.6178,
+      "step": 95800
+    },
+    {
+      "epoch": 0.6095313155452731,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.338461538461539e-06,
+      "loss": 1.6332,
+      "step": 95900
+    },
+    {
+      "epoch": 0.6101669060724318,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.6216,
+      "step": 96000
+    },
+    {
+      "epoch": 0.6108024965995906,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.3282051282051286e-06,
+      "loss": 1.6099,
+      "step": 96100
+    },
+    {
+      "epoch": 0.6114380871267494,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.323076923076923e-06,
+      "loss": 1.636,
+      "step": 96200
+    },
+    {
+      "epoch": 0.6120736776539083,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.317948717948719e-06,
+      "loss": 1.6261,
+      "step": 96300
+    },
+    {
+      "epoch": 0.6127092681810671,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.312820512820513e-06,
+      "loss": 1.6299,
+      "step": 96400
+    },
+    {
+      "epoch": 0.6133448587082259,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.307692307692308e-06,
+      "loss": 1.6277,
+      "step": 96500
+    },
+    {
+      "epoch": 0.6139804492353846,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.302564102564103e-06,
+      "loss": 1.6248,
+      "step": 96600
+    },
+    {
+      "epoch": 0.6146160397625434,
+      "grad_norm": 0.9453125,
+      "learning_rate": 5.297435897435897e-06,
+      "loss": 1.6107,
+      "step": 96700
+    },
+    {
+      "epoch": 0.6152516302897022,
+      "grad_norm": 0.875,
+      "learning_rate": 5.292307692307693e-06,
+      "loss": 1.6303,
+      "step": 96800
+    },
+    {
+      "epoch": 0.615887220816861,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.287179487179488e-06,
+      "loss": 1.6134,
+      "step": 96900
+    },
+    {
+      "epoch": 0.6165228113440198,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.282051282051283e-06,
+      "loss": 1.6326,
+      "step": 97000
+    },
+    {
+      "epoch": 0.6171584018711785,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.2769230769230775e-06,
+      "loss": 1.6081,
+      "step": 97100
+    },
+    {
+      "epoch": 0.6177939923983373,
+      "grad_norm": 1.265625,
+      "learning_rate": 5.271794871794872e-06,
+      "loss": 1.6208,
+      "step": 97200
+    },
+    {
+      "epoch": 0.6184295829254961,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.2666666666666665e-06,
+      "loss": 1.6238,
+      "step": 97300
+    },
+    {
+      "epoch": 0.6190651734526549,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.261538461538462e-06,
+      "loss": 1.6219,
+      "step": 97400
+    },
+    {
+      "epoch": 0.6197007639798137,
+      "grad_norm": 0.875,
+      "learning_rate": 5.256410256410257e-06,
+      "loss": 1.6219,
+      "step": 97500
+    },
+    {
+      "epoch": 0.6203363545069724,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.251282051282052e-06,
+      "loss": 1.6101,
+      "step": 97600
+    },
+    {
+      "epoch": 0.6209719450341312,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.246153846153846e-06,
+      "loss": 1.6172,
+      "step": 97700
+    },
+    {
+      "epoch": 0.62160753556129,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.241025641025641e-06,
+      "loss": 1.6167,
+      "step": 97800
+    },
+    {
+      "epoch": 0.6222431260884488,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.235897435897437e-06,
+      "loss": 1.616,
+      "step": 97900
+    },
+    {
+      "epoch": 0.6228787166156076,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.230769230769232e-06,
+      "loss": 1.6145,
+      "step": 98000
+    },
+    {
+      "epoch": 0.6235143071427663,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.2256410256410265e-06,
+      "loss": 1.6299,
+      "step": 98100
+    },
+    {
+      "epoch": 0.6241498976699251,
+      "grad_norm": 0.80078125,
+      "learning_rate": 5.220512820512821e-06,
+      "loss": 1.6195,
+      "step": 98200
+    },
+    {
+      "epoch": 0.6247854881970839,
+      "grad_norm": 0.859375,
+      "learning_rate": 5.2153846153846155e-06,
+      "loss": 1.6244,
+      "step": 98300
+    },
+    {
+      "epoch": 0.6254210787242427,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.21025641025641e-06,
+      "loss": 1.6223,
+      "step": 98400
+    },
+    {
+      "epoch": 0.6260566692514015,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.205128205128206e-06,
+      "loss": 1.6134,
+      "step": 98500
+    },
+    {
+      "epoch": 0.6266922597785602,
+      "grad_norm": 0.87109375,
+      "learning_rate": 5.2e-06,
+      "loss": 1.6261,
+      "step": 98600
+    },
+    {
+      "epoch": 0.627327850305719,
+      "grad_norm": 0.75,
+      "learning_rate": 5.194871794871795e-06,
+      "loss": 1.6227,
+      "step": 98700
+    },
+    {
+      "epoch": 0.6279634408328778,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.18974358974359e-06,
+      "loss": 1.6215,
+      "step": 98800
+    },
+    {
+      "epoch": 0.6285990313600366,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.184615384615385e-06,
+      "loss": 1.6231,
+      "step": 98900
+    },
+    {
+      "epoch": 0.6292346218871954,
+      "grad_norm": 0.82421875,
+      "learning_rate": 5.179487179487181e-06,
+      "loss": 1.6289,
+      "step": 99000
+    },
+    {
+      "epoch": 0.6298702124143541,
+      "grad_norm": 0.85546875,
+      "learning_rate": 5.174358974358975e-06,
+      "loss": 1.6127,
+      "step": 99100
+    },
+    {
+      "epoch": 0.6305058029415129,
+      "grad_norm": 1.28125,
+      "learning_rate": 5.16923076923077e-06,
+      "loss": 1.6208,
+      "step": 99200
+    },
+    {
+      "epoch": 0.6311413934686717,
+      "grad_norm": 0.78515625,
+      "learning_rate": 5.1641025641025645e-06,
+      "loss": 1.6161,
+      "step": 99300
+    },
+    {
+      "epoch": 0.6317769839958305,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.158974358974359e-06,
+      "loss": 1.6175,
+      "step": 99400
+    },
+    {
+      "epoch": 0.6324125745229893,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.1538461538461534e-06,
+      "loss": 1.6174,
+      "step": 99500
+    },
+    {
+      "epoch": 0.633048165050148,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.148717948717949e-06,
+      "loss": 1.6236,
+      "step": 99600
+    },
+    {
+      "epoch": 0.6336837555773068,
+      "grad_norm": 0.8515625,
+      "learning_rate": 5.143589743589744e-06,
+      "loss": 1.6216,
+      "step": 99700
+    },
+    {
+      "epoch": 0.6343193461044656,
+      "grad_norm": 0.75390625,
+      "learning_rate": 5.138461538461539e-06,
+      "loss": 1.6239,
+      "step": 99800
+    },
+    {
+      "epoch": 0.6349549366316244,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.133333333333334e-06,
+      "loss": 1.6215,
+      "step": 99900
+    },
+    {
+      "epoch": 0.6355905271587833,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 1.5993,
+      "step": 100000
+    },
+    {
+      "epoch": 0.6362261176859421,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.123076923076924e-06,
+      "loss": 1.608,
+      "step": 100100
+    },
+    {
+      "epoch": 0.6368617082131008,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.1179487179487186e-06,
+      "loss": 1.6131,
+      "step": 100200
+    },
+    {
+      "epoch": 0.6374972987402596,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.1128205128205135e-06,
+      "loss": 1.626,
+      "step": 100300
+    },
+    {
+      "epoch": 0.6381328892674184,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.1076923076923075e-06,
+      "loss": 1.6142,
+      "step": 100400
+    },
+    {
+      "epoch": 0.6387684797945772,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.1025641025641024e-06,
+      "loss": 1.6346,
+      "step": 100500
+    },
+    {
+      "epoch": 0.639404070321736,
+      "grad_norm": 0.828125,
+      "learning_rate": 5.097435897435898e-06,
+      "loss": 1.604,
+      "step": 100600
+    },
+    {
+      "epoch": 0.6400396608488947,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.092307692307693e-06,
+      "loss": 1.629,
+      "step": 100700
+    },
+    {
+      "epoch": 0.6406752513760535,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.087179487179488e-06,
+      "loss": 1.6069,
+      "step": 100800
+    },
+    {
+      "epoch": 0.6413108419032123,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.082051282051282e-06,
+      "loss": 1.6207,
+      "step": 100900
+    },
+    {
+      "epoch": 0.6419464324303711,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.076923076923077e-06,
+      "loss": 1.6052,
+      "step": 101000
+    },
+    {
+      "epoch": 0.6425820229575299,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.071794871794872e-06,
+      "loss": 1.6145,
+      "step": 101100
+    },
+    {
+      "epoch": 0.6432176134846886,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.0666666666666676e-06,
+      "loss": 1.6133,
+      "step": 101200
+    },
+    {
+      "epoch": 0.6438532040118474,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.0615384615384625e-06,
+      "loss": 1.6339,
+      "step": 101300
+    },
+    {
+      "epoch": 0.6444887945390062,
+      "grad_norm": 0.8046875,
+      "learning_rate": 5.0564102564102565e-06,
+      "loss": 1.5978,
+      "step": 101400
+    },
+    {
+      "epoch": 0.645124385066165,
+      "grad_norm": 0.93359375,
+      "learning_rate": 5.051282051282051e-06,
+      "loss": 1.6232,
+      "step": 101500
+    },
+    {
+      "epoch": 0.6457599755933238,
+      "grad_norm": 0.7890625,
+      "learning_rate": 5.046153846153846e-06,
+      "loss": 1.6328,
+      "step": 101600
+    },
+    {
+      "epoch": 0.6463955661204825,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.041025641025642e-06,
+      "loss": 1.6116,
+      "step": 101700
+    },
+    {
+      "epoch": 0.6470311566476413,
+      "grad_norm": 0.84765625,
+      "learning_rate": 5.035897435897437e-06,
+      "loss": 1.6376,
+      "step": 101800
+    },
+    {
+      "epoch": 0.6476667471748001,
+      "grad_norm": 0.85546875,
+      "learning_rate": 5.030769230769231e-06,
+      "loss": 1.614,
+      "step": 101900
+    },
+    {
+      "epoch": 0.6483023377019589,
+      "grad_norm": 0.7890625,
+      "learning_rate": 5.025641025641026e-06,
+      "loss": 1.616,
+      "step": 102000
+    },
+    {
+      "epoch": 0.6489379282291177,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.020512820512821e-06,
+      "loss": 1.6242,
+      "step": 102100
+    },
+    {
+      "epoch": 0.6495735187562764,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.015384615384616e-06,
+      "loss": 1.6139,
+      "step": 102200
+    },
+    {
+      "epoch": 0.6502091092834352,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.0102564102564115e-06,
+      "loss": 1.6071,
+      "step": 102300
+    },
+    {
+      "epoch": 0.650844699810594,
+      "grad_norm": 0.8671875,
+      "learning_rate": 5.0051282051282055e-06,
+      "loss": 1.6095,
+      "step": 102400
+    },
+    {
+      "epoch": 0.6514802903377528,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5e-06,
+      "loss": 1.6153,
+      "step": 102500
+    },
+    {
+      "epoch": 0.6521158808649116,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.994871794871795e-06,
+      "loss": 1.6096,
+      "step": 102600
+    },
+    {
+      "epoch": 0.6527514713920703,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.98974358974359e-06,
+      "loss": 1.6182,
+      "step": 102700
+    },
+    {
+      "epoch": 0.6533870619192291,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.984615384615385e-06,
+      "loss": 1.6078,
+      "step": 102800
+    },
+    {
+      "epoch": 0.6540226524463879,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.97948717948718e-06,
+      "loss": 1.6203,
+      "step": 102900
+    },
+    {
+      "epoch": 0.6546582429735467,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.974358974358975e-06,
+      "loss": 1.625,
+      "step": 103000
+    },
+    {
+      "epoch": 0.6552938335007055,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.96923076923077e-06,
+      "loss": 1.6202,
+      "step": 103100
+    },
+    {
+      "epoch": 0.6559294240278642,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.964102564102565e-06,
+      "loss": 1.6192,
+      "step": 103200
+    },
+    {
+      "epoch": 0.656565014555023,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.95897435897436e-06,
+      "loss": 1.6077,
+      "step": 103300
+    },
+    {
+      "epoch": 0.6572006050821818,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.9538461538461545e-06,
+      "loss": 1.6113,
+      "step": 103400
+    },
+    {
+      "epoch": 0.6578361956093406,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.948717948717949e-06,
+      "loss": 1.6292,
+      "step": 103500
+    },
+    {
+      "epoch": 0.6584717861364994,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.943589743589744e-06,
+      "loss": 1.616,
+      "step": 103600
+    },
+    {
+      "epoch": 0.6591073766636582,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.938461538461538e-06,
+      "loss": 1.6265,
+      "step": 103700
+    },
+    {
+      "epoch": 0.659742967190817,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.933333333333334e-06,
+      "loss": 1.6158,
+      "step": 103800
+    },
+    {
+      "epoch": 0.6603785577179758,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.928205128205128e-06,
+      "loss": 1.6274,
+      "step": 103900
+    },
+    {
+      "epoch": 0.6610141482451346,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.923076923076924e-06,
+      "loss": 1.6122,
+      "step": 104000
+    },
+    {
+      "epoch": 0.6616497387722934,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.917948717948719e-06,
+      "loss": 1.6128,
+      "step": 104100
+    },
+    {
+      "epoch": 0.6622853292994522,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.912820512820513e-06,
+      "loss": 1.6175,
+      "step": 104200
+    },
+    {
+      "epoch": 0.6629209198266109,
+      "grad_norm": 1.390625,
+      "learning_rate": 4.907692307692309e-06,
+      "loss": 1.6262,
+      "step": 104300
+    },
+    {
+      "epoch": 0.6635565103537697,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.902564102564103e-06,
+      "loss": 1.6097,
+      "step": 104400
+    },
+    {
+      "epoch": 0.6641921008809285,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.8974358974358975e-06,
+      "loss": 1.6119,
+      "step": 104500
+    },
+    {
+      "epoch": 0.6648276914080873,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.892307692307693e-06,
+      "loss": 1.6309,
+      "step": 104600
+    },
+    {
+      "epoch": 0.6654632819352461,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.887179487179487e-06,
+      "loss": 1.6189,
+      "step": 104700
+    },
+    {
+      "epoch": 0.6660988724624048,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.882051282051282e-06,
+      "loss": 1.6162,
+      "step": 104800
+    },
+    {
+      "epoch": 0.6667344629895636,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.876923076923077e-06,
+      "loss": 1.634,
+      "step": 104900
+    },
+    {
+      "epoch": 0.6673700535167224,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.871794871794872e-06,
+      "loss": 1.6159,
+      "step": 105000
+    },
+    {
+      "epoch": 0.6680056440438812,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.866666666666667e-06,
+      "loss": 1.6145,
+      "step": 105100
+    },
+    {
+      "epoch": 0.66864123457104,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.861538461538462e-06,
+      "loss": 1.6403,
+      "step": 105200
+    },
+    {
+      "epoch": 0.6692768250981987,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.856410256410257e-06,
+      "loss": 1.6177,
+      "step": 105300
+    },
+    {
+      "epoch": 0.6699124156253575,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.851282051282052e-06,
+      "loss": 1.6122,
+      "step": 105400
+    },
+    {
+      "epoch": 0.6705480061525163,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.8461538461538465e-06,
+      "loss": 1.6122,
+      "step": 105500
+    },
+    {
+      "epoch": 0.6711835966796751,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.8410256410256414e-06,
+      "loss": 1.6299,
+      "step": 105600
+    },
+    {
+      "epoch": 0.6718191872068339,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.835897435897436e-06,
+      "loss": 1.6227,
+      "step": 105700
+    },
+    {
+      "epoch": 0.6724547777339926,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.830769230769231e-06,
+      "loss": 1.6101,
+      "step": 105800
+    },
+    {
+      "epoch": 0.6730903682611514,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.825641025641026e-06,
+      "loss": 1.628,
+      "step": 105900
+    },
+    {
+      "epoch": 0.6737259587883102,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.820512820512821e-06,
+      "loss": 1.6062,
+      "step": 106000
+    },
+    {
+      "epoch": 0.674361549315469,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.815384615384616e-06,
+      "loss": 1.6278,
+      "step": 106100
+    },
+    {
+      "epoch": 0.6749971398426278,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.810256410256411e-06,
+      "loss": 1.6188,
+      "step": 106200
+    },
+    {
+      "epoch": 0.6756327303697865,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.805128205128206e-06,
+      "loss": 1.6195,
+      "step": 106300
+    },
+    {
+      "epoch": 0.6762683208969453,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.6191,
+      "step": 106400
+    },
+    {
+      "epoch": 0.6769039114241041,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.7948717948717955e-06,
+      "loss": 1.6235,
+      "step": 106500
+    },
+    {
+      "epoch": 0.6775395019512629,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.7897435897435904e-06,
+      "loss": 1.6166,
+      "step": 106600
+    },
+    {
+      "epoch": 0.6781750924784217,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.7846153846153845e-06,
+      "loss": 1.5958,
+      "step": 106700
+    },
+    {
+      "epoch": 0.6788106830055805,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.77948717948718e-06,
+      "loss": 1.622,
+      "step": 106800
+    },
+    {
+      "epoch": 0.6794462735327392,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.774358974358974e-06,
+      "loss": 1.6315,
+      "step": 106900
+    },
+    {
+      "epoch": 0.680081864059898,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.76923076923077e-06,
+      "loss": 1.6155,
+      "step": 107000
+    },
+    {
+      "epoch": 0.6807174545870568,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.764102564102565e-06,
+      "loss": 1.6098,
+      "step": 107100
+    },
+    {
+      "epoch": 0.6813530451142156,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.758974358974359e-06,
+      "loss": 1.6073,
+      "step": 107200
+    },
+    {
+      "epoch": 0.6819886356413744,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.753846153846155e-06,
+      "loss": 1.6177,
+      "step": 107300
+    },
+    {
+      "epoch": 0.6826242261685331,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.748717948717949e-06,
+      "loss": 1.6186,
+      "step": 107400
+    },
+    {
+      "epoch": 0.6832598166956919,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.743589743589744e-06,
+      "loss": 1.6383,
+      "step": 107500
+    },
+    {
+      "epoch": 0.6838954072228508,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.738461538461539e-06,
+      "loss": 1.6377,
+      "step": 107600
+    },
+    {
+      "epoch": 0.6845309977500096,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.7333333333333335e-06,
+      "loss": 1.616,
+      "step": 107700
+    },
+    {
+      "epoch": 0.6851665882771684,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.728205128205128e-06,
+      "loss": 1.6348,
+      "step": 107800
+    },
+    {
+      "epoch": 0.6858021788043271,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.723076923076923e-06,
+      "loss": 1.6176,
+      "step": 107900
+    },
+    {
+      "epoch": 0.6864377693314859,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.717948717948718e-06,
+      "loss": 1.6137,
+      "step": 108000
+    },
+    {
+      "epoch": 0.6870733598586447,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.712820512820513e-06,
+      "loss": 1.6239,
+      "step": 108100
+    },
+    {
+      "epoch": 0.6877089503858035,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.707692307692308e-06,
+      "loss": 1.6267,
+      "step": 108200
+    },
+    {
+      "epoch": 0.6883445409129623,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.702564102564103e-06,
+      "loss": 1.6296,
+      "step": 108300
+    },
+    {
+      "epoch": 0.688980131440121,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.697435897435898e-06,
+      "loss": 1.6124,
+      "step": 108400
+    },
+    {
+      "epoch": 0.6896157219672798,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.692307692307693e-06,
+      "loss": 1.6137,
+      "step": 108500
+    },
+    {
+      "epoch": 0.6902513124944386,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.6871794871794876e-06,
+      "loss": 1.6152,
+      "step": 108600
+    },
+    {
+      "epoch": 0.6908869030215974,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.6820512820512825e-06,
+      "loss": 1.6073,
+      "step": 108700
+    },
+    {
+      "epoch": 0.6915224935487562,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.676923076923077e-06,
+      "loss": 1.6117,
+      "step": 108800
+    },
+    {
+      "epoch": 0.692158084075915,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.671794871794872e-06,
+      "loss": 1.6196,
+      "step": 108900
+    },
+    {
+      "epoch": 0.6927936746030737,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 1.6081,
+      "step": 109000
+    },
+    {
+      "epoch": 0.6934292651302325,
+      "grad_norm": 1.921875,
+      "learning_rate": 4.661538461538462e-06,
+      "loss": 1.6223,
+      "step": 109100
+    },
+    {
+      "epoch": 0.6940648556573913,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.656410256410257e-06,
+      "loss": 1.6228,
+      "step": 109200
+    },
+    {
+      "epoch": 0.6947004461845501,
+      "grad_norm": 0.75,
+      "learning_rate": 4.651282051282052e-06,
+      "loss": 1.6186,
+      "step": 109300
+    },
+    {
+      "epoch": 0.6953360367117088,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.646153846153847e-06,
+      "loss": 1.6278,
+      "step": 109400
+    },
+    {
+      "epoch": 0.6959716272388676,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.641025641025642e-06,
+      "loss": 1.6158,
+      "step": 109500
+    },
+    {
+      "epoch": 0.6966072177660264,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.6358974358974366e-06,
+      "loss": 1.6168,
+      "step": 109600
+    },
+    {
+      "epoch": 0.6972428082931852,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.630769230769231e-06,
+      "loss": 1.629,
+      "step": 109700
+    },
+    {
+      "epoch": 0.697878398820344,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.625641025641026e-06,
+      "loss": 1.6176,
+      "step": 109800
+    },
+    {
+      "epoch": 0.6985139893475028,
+      "grad_norm": 0.625,
+      "learning_rate": 4.62051282051282e-06,
+      "loss": 1.6137,
+      "step": 109900
+    },
+    {
+      "epoch": 0.6991495798746615,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 1.609,
+      "step": 110000
+    },
+    {
+      "epoch": 0.6997851704018203,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.610256410256411e-06,
+      "loss": 1.6135,
+      "step": 110100
+    },
+    {
+      "epoch": 0.7004207609289791,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.605128205128205e-06,
+      "loss": 1.6305,
+      "step": 110200
+    },
+    {
+      "epoch": 0.7010563514561379,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 1.6075,
+      "step": 110300
+    },
+    {
+      "epoch": 0.7016919419832967,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.594871794871795e-06,
+      "loss": 1.6183,
+      "step": 110400
+    },
+    {
+      "epoch": 0.7023275325104554,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.58974358974359e-06,
+      "loss": 1.6249,
+      "step": 110500
+    },
+    {
+      "epoch": 0.7029631230376142,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.5846153846153855e-06,
+      "loss": 1.6048,
+      "step": 110600
+    },
+    {
+      "epoch": 0.703598713564773,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.57948717948718e-06,
+      "loss": 1.641,
+      "step": 110700
+    },
+    {
+      "epoch": 0.7042343040919318,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.5743589743589745e-06,
+      "loss": 1.6341,
+      "step": 110800
+    },
+    {
+      "epoch": 0.7048698946190906,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.569230769230769e-06,
+      "loss": 1.6187,
+      "step": 110900
+    },
+    {
+      "epoch": 0.7055054851462493,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.564102564102564e-06,
+      "loss": 1.6084,
+      "step": 111000
+    },
+    {
+      "epoch": 0.7061410756734081,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.558974358974359e-06,
+      "loss": 1.6216,
+      "step": 111100
+    },
+    {
+      "epoch": 0.7067766662005669,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.553846153846154e-06,
+      "loss": 1.6216,
+      "step": 111200
+    },
+    {
+      "epoch": 0.7074122567277257,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.548717948717949e-06,
+      "loss": 1.6371,
+      "step": 111300
+    },
+    {
+      "epoch": 0.7080478472548846,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.543589743589744e-06,
+      "loss": 1.6226,
+      "step": 111400
+    },
+    {
+      "epoch": 0.7086834377820433,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.538461538461539e-06,
+      "loss": 1.6266,
+      "step": 111500
+    },
+    {
+      "epoch": 0.7093190283092021,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.533333333333334e-06,
+      "loss": 1.6185,
+      "step": 111600
+    },
+    {
+      "epoch": 0.7099546188363609,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.528205128205129e-06,
+      "loss": 1.6345,
+      "step": 111700
+    },
+    {
+      "epoch": 0.7105902093635197,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.5230769230769235e-06,
+      "loss": 1.6351,
+      "step": 111800
+    },
+    {
+      "epoch": 0.7112257998906785,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.517948717948718e-06,
+      "loss": 1.6055,
+      "step": 111900
+    },
+    {
+      "epoch": 0.7118613904178372,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.512820512820513e-06,
+      "loss": 1.6265,
+      "step": 112000
+    },
+    {
+      "epoch": 0.712496980944996,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.507692307692308e-06,
+      "loss": 1.6411,
+      "step": 112100
+    },
+    {
+      "epoch": 0.7131325714721548,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.502564102564103e-06,
+      "loss": 1.627,
+      "step": 112200
+    },
+    {
+      "epoch": 0.7137681619993136,
+      "grad_norm": 1.125,
+      "learning_rate": 4.497435897435898e-06,
+      "loss": 1.618,
+      "step": 112300
+    },
+    {
+      "epoch": 0.7144037525264724,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.492307692307693e-06,
+      "loss": 1.6237,
+      "step": 112400
+    },
+    {
+      "epoch": 0.7150393430536311,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 1.6341,
+      "step": 112500
+    },
+    {
+      "epoch": 0.7156749335807899,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.482051282051283e-06,
+      "loss": 1.6358,
+      "step": 112600
+    },
+    {
+      "epoch": 0.7163105241079487,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.476923076923077e-06,
+      "loss": 1.6074,
+      "step": 112700
+    },
+    {
+      "epoch": 0.7169461146351075,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.4717948717948725e-06,
+      "loss": 1.624,
+      "step": 112800
+    },
+    {
+      "epoch": 0.7175817051622663,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.4666666666666665e-06,
+      "loss": 1.6122,
+      "step": 112900
+    },
+    {
+      "epoch": 0.718217295689425,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.461538461538462e-06,
+      "loss": 1.6184,
+      "step": 113000
+    },
+    {
+      "epoch": 0.7188528862165838,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.456410256410257e-06,
+      "loss": 1.6202,
+      "step": 113100
+    },
+    {
+      "epoch": 0.7194884767437426,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.451282051282051e-06,
+      "loss": 1.6271,
+      "step": 113200
+    },
+    {
+      "epoch": 0.7201240672709014,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.446153846153847e-06,
+      "loss": 1.6147,
+      "step": 113300
+    },
+    {
+      "epoch": 0.7207596577980602,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.441025641025641e-06,
+      "loss": 1.6253,
+      "step": 113400
+    },
+    {
+      "epoch": 0.721395248325219,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.435897435897436e-06,
+      "loss": 1.6078,
+      "step": 113500
+    },
+    {
+      "epoch": 0.7220308388523777,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.430769230769232e-06,
+      "loss": 1.633,
+      "step": 113600
+    },
+    {
+      "epoch": 0.7226664293795365,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.425641025641026e-06,
+      "loss": 1.6203,
+      "step": 113700
+    },
+    {
+      "epoch": 0.7233020199066953,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.420512820512821e-06,
+      "loss": 1.6073,
+      "step": 113800
+    },
+    {
+      "epoch": 0.7239376104338541,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.4153846153846155e-06,
+      "loss": 1.62,
+      "step": 113900
+    },
+    {
+      "epoch": 0.7245732009610129,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.4102564102564104e-06,
+      "loss": 1.6269,
+      "step": 114000
+    },
+    {
+      "epoch": 0.7252087914881716,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.405128205128205e-06,
+      "loss": 1.636,
+      "step": 114100
+    },
+    {
+      "epoch": 0.7258443820153304,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.4e-06,
+      "loss": 1.6242,
+      "step": 114200
+    },
+    {
+      "epoch": 0.7264799725424892,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.394871794871795e-06,
+      "loss": 1.6222,
+      "step": 114300
+    },
+    {
+      "epoch": 0.727115563069648,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.38974358974359e-06,
+      "loss": 1.6353,
+      "step": 114400
+    },
+    {
+      "epoch": 0.7277511535968068,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.384615384615385e-06,
+      "loss": 1.6359,
+      "step": 114500
+    },
+    {
+      "epoch": 0.7283867441239655,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.37948717948718e-06,
+      "loss": 1.62,
+      "step": 114600
+    },
+    {
+      "epoch": 0.7290223346511243,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.374358974358975e-06,
+      "loss": 1.6287,
+      "step": 114700
+    },
+    {
+      "epoch": 0.7296579251782831,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.36923076923077e-06,
+      "loss": 1.6362,
+      "step": 114800
+    },
+    {
+      "epoch": 0.7302935157054419,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.3641025641025645e-06,
+      "loss": 1.6261,
+      "step": 114900
+    },
+    {
+      "epoch": 0.7309291062326007,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.358974358974359e-06,
+      "loss": 1.6185,
+      "step": 115000
+    },
+    {
+      "epoch": 0.7315646967597595,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.353846153846154e-06,
+      "loss": 1.6173,
+      "step": 115100
+    },
+    {
+      "epoch": 0.7322002872869183,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.348717948717949e-06,
+      "loss": 1.627,
+      "step": 115200
+    },
+    {
+      "epoch": 0.7328358778140771,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.343589743589744e-06,
+      "loss": 1.643,
+      "step": 115300
+    },
+    {
+      "epoch": 0.7334714683412359,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.338461538461539e-06,
+      "loss": 1.6276,
+      "step": 115400
+    },
+    {
+      "epoch": 0.7341070588683947,
+      "grad_norm": 0.75,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 1.6176,
+      "step": 115500
+    },
+    {
+      "epoch": 0.7347426493955534,
+      "grad_norm": 1.2734375,
+      "learning_rate": 4.328205128205129e-06,
+      "loss": 1.6187,
+      "step": 115600
+    },
+    {
+      "epoch": 0.7353782399227122,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.323076923076923e-06,
+      "loss": 1.6196,
+      "step": 115700
+    },
+    {
+      "epoch": 0.736013830449871,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.317948717948719e-06,
+      "loss": 1.6119,
+      "step": 115800
+    },
+    {
+      "epoch": 0.7366494209770298,
+      "grad_norm": 0.875,
+      "learning_rate": 4.312820512820513e-06,
+      "loss": 1.6179,
+      "step": 115900
+    },
+    {
+      "epoch": 0.7372850115041886,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.307692307692308e-06,
+      "loss": 1.6227,
+      "step": 116000
+    },
+    {
+      "epoch": 0.7379206020313474,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.302564102564103e-06,
+      "loss": 1.6225,
+      "step": 116100
+    },
+    {
+      "epoch": 0.7385561925585061,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.297435897435897e-06,
+      "loss": 1.603,
+      "step": 116200
+    },
+    {
+      "epoch": 0.7391917830856649,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.292307692307693e-06,
+      "loss": 1.62,
+      "step": 116300
+    },
+    {
+      "epoch": 0.7398273736128237,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.287179487179487e-06,
+      "loss": 1.6205,
+      "step": 116400
+    },
+    {
+      "epoch": 0.7404629641399825,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.282051282051282e-06,
+      "loss": 1.6217,
+      "step": 116500
+    },
+    {
+      "epoch": 0.7410985546671413,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.276923076923078e-06,
+      "loss": 1.6287,
+      "step": 116600
+    },
+    {
+      "epoch": 0.7417341451943,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.271794871794872e-06,
+      "loss": 1.6273,
+      "step": 116700
+    },
+    {
+      "epoch": 0.7423697357214588,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.266666666666668e-06,
+      "loss": 1.644,
+      "step": 116800
+    },
+    {
+      "epoch": 0.7430053262486176,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.261538461538462e-06,
+      "loss": 1.6179,
+      "step": 116900
+    },
+    {
+      "epoch": 0.7436409167757764,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.2564102564102566e-06,
+      "loss": 1.625,
+      "step": 117000
+    },
+    {
+      "epoch": 0.7442765073029352,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.2512820512820515e-06,
+      "loss": 1.6082,
+      "step": 117100
+    },
+    {
+      "epoch": 0.7449120978300939,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.246153846153846e-06,
+      "loss": 1.6159,
+      "step": 117200
+    },
+    {
+      "epoch": 0.7455476883572527,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.241025641025641e-06,
+      "loss": 1.6112,
+      "step": 117300
+    },
+    {
+      "epoch": 0.7461832788844115,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.235897435897436e-06,
+      "loss": 1.6248,
+      "step": 117400
+    },
+    {
+      "epoch": 0.7468188694115703,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 1.6073,
+      "step": 117500
+    },
+    {
+      "epoch": 0.747454459938729,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.225641025641026e-06,
+      "loss": 1.6288,
+      "step": 117600
+    },
+    {
+      "epoch": 0.7480900504658878,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.220512820512821e-06,
+      "loss": 1.6191,
+      "step": 117700
+    },
+    {
+      "epoch": 0.7487256409930466,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.215384615384616e-06,
+      "loss": 1.622,
+      "step": 117800
+    },
+    {
+      "epoch": 0.7493612315202054,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.210256410256411e-06,
+      "loss": 1.6162,
+      "step": 117900
+    },
+    {
+      "epoch": 0.7499968220473642,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.2051282051282055e-06,
+      "loss": 1.6305,
+      "step": 118000
+    },
+    {
+      "epoch": 0.750632412574523,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 1.6306,
+      "step": 118100
+    },
+    {
+      "epoch": 0.7512680031016817,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.194871794871795e-06,
+      "loss": 1.627,
+      "step": 118200
+    },
+    {
+      "epoch": 0.7519035936288405,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.18974358974359e-06,
+      "loss": 1.6304,
+      "step": 118300
+    },
+    {
+      "epoch": 0.7525391841559993,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.184615384615385e-06,
+      "loss": 1.6091,
+      "step": 118400
+    },
+    {
+      "epoch": 0.7531747746831581,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.17948717948718e-06,
+      "loss": 1.6114,
+      "step": 118500
+    },
+    {
+      "epoch": 0.7538103652103169,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.174358974358975e-06,
+      "loss": 1.6268,
+      "step": 118600
+    },
+    {
+      "epoch": 0.7544459557374756,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.169230769230769e-06,
+      "loss": 1.6143,
+      "step": 118700
+    },
+    {
+      "epoch": 0.7550815462646344,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.164102564102565e-06,
+      "loss": 1.6197,
+      "step": 118800
+    },
+    {
+      "epoch": 0.7557171367917933,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.158974358974359e-06,
+      "loss": 1.6161,
+      "step": 118900
+    },
+    {
+      "epoch": 0.7563527273189521,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.1538461538461545e-06,
+      "loss": 1.6146,
+      "step": 119000
+    },
+    {
+      "epoch": 0.7569883178461109,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.1487179487179494e-06,
+      "loss": 1.6216,
+      "step": 119100
+    },
+    {
+      "epoch": 0.7576239083732696,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.1435897435897435e-06,
+      "loss": 1.6253,
+      "step": 119200
+    },
+    {
+      "epoch": 0.7582594989004284,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.138461538461539e-06,
+      "loss": 1.6314,
+      "step": 119300
+    },
+    {
+      "epoch": 0.7588950894275872,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.133333333333333e-06,
+      "loss": 1.6238,
+      "step": 119400
+    },
+    {
+      "epoch": 0.759530679954746,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.128205128205128e-06,
+      "loss": 1.6102,
+      "step": 119500
+    },
+    {
+      "epoch": 0.7601662704819048,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.123076923076924e-06,
+      "loss": 1.6004,
+      "step": 119600
+    },
+    {
+      "epoch": 0.7608018610090636,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.117948717948718e-06,
+      "loss": 1.618,
+      "step": 119700
+    },
+    {
+      "epoch": 0.7614374515362223,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.112820512820514e-06,
+      "loss": 1.6198,
+      "step": 119800
+    },
+    {
+      "epoch": 0.7620730420633811,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.107692307692308e-06,
+      "loss": 1.6105,
+      "step": 119900
+    },
+    {
+      "epoch": 0.7627086325905399,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 1.61,
+      "step": 120000
+    },
+    {
+      "epoch": 0.7633442231176987,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.097435897435898e-06,
+      "loss": 1.6288,
+      "step": 120100
+    },
+    {
+      "epoch": 0.7639798136448575,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.0923076923076925e-06,
+      "loss": 1.6215,
+      "step": 120200
+    },
+    {
+      "epoch": 0.7646154041720162,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.087179487179487e-06,
+      "loss": 1.6368,
+      "step": 120300
+    },
+    {
+      "epoch": 0.765250994699175,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.082051282051282e-06,
+      "loss": 1.6292,
+      "step": 120400
+    },
+    {
+      "epoch": 0.7658865852263338,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.076923076923077e-06,
+      "loss": 1.6237,
+      "step": 120500
+    },
+    {
+      "epoch": 0.7665221757534926,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.071794871794872e-06,
+      "loss": 1.6285,
+      "step": 120600
+    },
+    {
+      "epoch": 0.7671577662806514,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.066666666666667e-06,
+      "loss": 1.6288,
+      "step": 120700
+    },
+    {
+      "epoch": 0.7677933568078101,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.061538461538462e-06,
+      "loss": 1.634,
+      "step": 120800
+    },
+    {
+      "epoch": 0.7684289473349689,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.056410256410257e-06,
+      "loss": 1.6295,
+      "step": 120900
+    },
+    {
+      "epoch": 0.7690645378621277,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.051282051282052e-06,
+      "loss": 1.6324,
+      "step": 121000
+    },
+    {
+      "epoch": 0.7697001283892865,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.0461538461538466e-06,
+      "loss": 1.6293,
+      "step": 121100
+    },
+    {
+      "epoch": 0.7703357189164453,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.0410256410256415e-06,
+      "loss": 1.6165,
+      "step": 121200
+    },
+    {
+      "epoch": 0.770971309443604,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.035897435897436e-06,
+      "loss": 1.6188,
+      "step": 121300
+    },
+    {
+      "epoch": 0.7716068999707628,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.030769230769231e-06,
+      "loss": 1.6288,
+      "step": 121400
+    },
+    {
+      "epoch": 0.7722424904979216,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.025641025641026e-06,
+      "loss": 1.6196,
+      "step": 121500
+    },
+    {
+      "epoch": 0.7728780810250804,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.020512820512821e-06,
+      "loss": 1.6457,
+      "step": 121600
+    },
+    {
+      "epoch": 0.7735136715522392,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.015384615384615e-06,
+      "loss": 1.6246,
+      "step": 121700
+    },
+    {
+      "epoch": 0.7741492620793979,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.010256410256411e-06,
+      "loss": 1.6151,
+      "step": 121800
+    },
+    {
+      "epoch": 0.7747848526065567,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.005128205128205e-06,
+      "loss": 1.6346,
+      "step": 121900
+    },
+    {
+      "epoch": 0.7754204431337155,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.6198,
+      "step": 122000
+    },
+    {
+      "epoch": 0.7760560336608743,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.9948717948717956e-06,
+      "loss": 1.6082,
+      "step": 122100
+    },
+    {
+      "epoch": 0.7766916241880331,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.98974358974359e-06,
+      "loss": 1.5981,
+      "step": 122200
+    },
+    {
+      "epoch": 0.7773272147151918,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.984615384615385e-06,
+      "loss": 1.6236,
+      "step": 122300
+    },
+    {
+      "epoch": 0.7779628052423506,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.979487179487179e-06,
+      "loss": 1.6074,
+      "step": 122400
+    },
+    {
+      "epoch": 0.7785983957695094,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.974358974358974e-06,
+      "loss": 1.6149,
+      "step": 122500
+    },
+    {
+      "epoch": 0.7792339862966682,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.96923076923077e-06,
+      "loss": 1.6157,
+      "step": 122600
+    },
+    {
+      "epoch": 0.7798695768238271,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.964102564102564e-06,
+      "loss": 1.6356,
+      "step": 122700
+    },
+    {
+      "epoch": 0.7805051673509859,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.95897435897436e-06,
+      "loss": 1.6314,
+      "step": 122800
+    },
+    {
+      "epoch": 0.7811407578781446,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.953846153846154e-06,
+      "loss": 1.6252,
+      "step": 122900
+    },
+    {
+      "epoch": 0.7817763484053034,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.948717948717949e-06,
+      "loss": 1.6209,
+      "step": 123000
+    },
+    {
+      "epoch": 0.7824119389324622,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.943589743589744e-06,
+      "loss": 1.6145,
+      "step": 123100
+    },
+    {
+      "epoch": 0.783047529459621,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.938461538461539e-06,
+      "loss": 1.6273,
+      "step": 123200
+    },
+    {
+      "epoch": 0.7836831199867798,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.9333333333333335e-06,
+      "loss": 1.6175,
+      "step": 123300
+    },
+    {
+      "epoch": 0.7843187105139385,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.928205128205128e-06,
+      "loss": 1.6289,
+      "step": 123400
+    },
+    {
+      "epoch": 0.7849543010410973,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.923076923076923e-06,
+      "loss": 1.6126,
+      "step": 123500
+    },
+    {
+      "epoch": 0.7855898915682561,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.917948717948718e-06,
+      "loss": 1.6222,
+      "step": 123600
+    },
+    {
+      "epoch": 0.7862254820954149,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.912820512820513e-06,
+      "loss": 1.6176,
+      "step": 123700
+    },
+    {
+      "epoch": 0.7868610726225737,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.907692307692308e-06,
+      "loss": 1.626,
+      "step": 123800
+    },
+    {
+      "epoch": 0.7874966631497324,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.902564102564103e-06,
+      "loss": 1.6348,
+      "step": 123900
+    },
+    {
+      "epoch": 0.7881322536768912,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.897435897435898e-06,
+      "loss": 1.6156,
+      "step": 124000
+    },
+    {
+      "epoch": 0.78876784420405,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.892307692307693e-06,
+      "loss": 1.6132,
+      "step": 124100
+    },
+    {
+      "epoch": 0.7894034347312088,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.887179487179488e-06,
+      "loss": 1.6178,
+      "step": 124200
+    },
+    {
+      "epoch": 0.7900390252583676,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.8820512820512825e-06,
+      "loss": 1.6131,
+      "step": 124300
+    },
+    {
+      "epoch": 0.7906746157855263,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.876923076923077e-06,
+      "loss": 1.6157,
+      "step": 124400
+    },
+    {
+      "epoch": 0.7913102063126851,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.871794871794872e-06,
+      "loss": 1.6223,
+      "step": 124500
+    },
+    {
+      "epoch": 0.7919457968398439,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.866666666666667e-06,
+      "loss": 1.6285,
+      "step": 124600
+    },
+    {
+      "epoch": 0.7925813873670027,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.861538461538462e-06,
+      "loss": 1.6237,
+      "step": 124700
+    },
+    {
+      "epoch": 0.7932169778941615,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.856410256410257e-06,
+      "loss": 1.6423,
+      "step": 124800
+    },
+    {
+      "epoch": 0.7938525684213202,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.851282051282051e-06,
+      "loss": 1.6336,
+      "step": 124900
+    },
+    {
+      "epoch": 0.794488158948479,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 1.6335,
+      "step": 125000
+    },
+    {
+      "epoch": 0.7951237494756378,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.841025641025642e-06,
+      "loss": 1.614,
+      "step": 125100
+    },
+    {
+      "epoch": 0.7957593400027966,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.835897435897436e-06,
+      "loss": 1.6191,
+      "step": 125200
+    },
+    {
+      "epoch": 0.7963949305299554,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.8307692307692315e-06,
+      "loss": 1.6222,
+      "step": 125300
+    },
+    {
+      "epoch": 0.7970305210571141,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.8256410256410255e-06,
+      "loss": 1.6125,
+      "step": 125400
+    },
+    {
+      "epoch": 0.7976661115842729,
+      "grad_norm": 1.1171875,
+      "learning_rate": 3.8205128205128204e-06,
+      "loss": 1.6322,
+      "step": 125500
+    },
+    {
+      "epoch": 0.7983017021114317,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.815384615384616e-06,
+      "loss": 1.6159,
+      "step": 125600
+    },
+    {
+      "epoch": 0.7989372926385905,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.8102564102564107e-06,
+      "loss": 1.6374,
+      "step": 125700
+    },
+    {
+      "epoch": 0.7995728831657493,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.8051282051282056e-06,
+      "loss": 1.618,
+      "step": 125800
+    },
+    {
+      "epoch": 0.800208473692908,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 1.6325,
+      "step": 125900
+    },
+    {
+      "epoch": 0.8008440642200668,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.794871794871795e-06,
+      "loss": 1.6121,
+      "step": 126000
+    },
+    {
+      "epoch": 0.8014796547472256,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.7897435897435903e-06,
+      "loss": 1.6173,
+      "step": 126100
+    },
+    {
+      "epoch": 0.8021152452743844,
+      "grad_norm": 1.1640625,
+      "learning_rate": 3.7846153846153847e-06,
+      "loss": 1.6166,
+      "step": 126200
+    },
+    {
+      "epoch": 0.8027508358015432,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.7794871794871796e-06,
+      "loss": 1.6234,
+      "step": 126300
+    },
+    {
+      "epoch": 0.803386426328702,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.774358974358975e-06,
+      "loss": 1.6333,
+      "step": 126400
+    },
+    {
+      "epoch": 0.8040220168558608,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.7692307692307694e-06,
+      "loss": 1.6345,
+      "step": 126500
+    },
+    {
+      "epoch": 0.8046576073830196,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.7641025641025643e-06,
+      "loss": 1.6255,
+      "step": 126600
+    },
+    {
+      "epoch": 0.8052931979101784,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.7589743589743592e-06,
+      "loss": 1.6247,
+      "step": 126700
+    },
+    {
+      "epoch": 0.8059287884373372,
+      "grad_norm": 1.3515625,
+      "learning_rate": 3.753846153846154e-06,
+      "loss": 1.6234,
+      "step": 126800
+    },
+    {
+      "epoch": 0.806564378964496,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.7487179487179495e-06,
+      "loss": 1.6475,
+      "step": 126900
+    },
+    {
+      "epoch": 0.8071999694916547,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.743589743589744e-06,
+      "loss": 1.6282,
+      "step": 127000
+    },
+    {
+      "epoch": 0.8078355600188135,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.7384615384615384e-06,
+      "loss": 1.6238,
+      "step": 127100
+    },
+    {
+      "epoch": 0.8084711505459723,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.7333333333333337e-06,
+      "loss": 1.6218,
+      "step": 127200
+    },
+    {
+      "epoch": 0.8091067410731311,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.7282051282051286e-06,
+      "loss": 1.6254,
+      "step": 127300
+    },
+    {
+      "epoch": 0.8097423316002899,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.723076923076923e-06,
+      "loss": 1.6161,
+      "step": 127400
+    },
+    {
+      "epoch": 0.8103779221274486,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.6167,
+      "step": 127500
+    },
+    {
+      "epoch": 0.8110135126546074,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.712820512820513e-06,
+      "loss": 1.6153,
+      "step": 127600
+    },
+    {
+      "epoch": 0.8116491031817662,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.7076923076923082e-06,
+      "loss": 1.6216,
+      "step": 127700
+    },
+    {
+      "epoch": 0.812284693708925,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.702564102564103e-06,
+      "loss": 1.6247,
+      "step": 127800
+    },
+    {
+      "epoch": 0.8129202842360838,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.6974358974358976e-06,
+      "loss": 1.6345,
+      "step": 127900
+    },
+    {
+      "epoch": 0.8135558747632425,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.692307692307693e-06,
+      "loss": 1.6206,
+      "step": 128000
+    },
+    {
+      "epoch": 0.8141914652904013,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.6871794871794874e-06,
+      "loss": 1.626,
+      "step": 128100
+    },
+    {
+      "epoch": 0.8148270558175601,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.6820512820512823e-06,
+      "loss": 1.6217,
+      "step": 128200
+    },
+    {
+      "epoch": 0.8154626463447189,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.676923076923077e-06,
+      "loss": 1.6246,
+      "step": 128300
+    },
+    {
+      "epoch": 0.8160982368718777,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.671794871794872e-06,
+      "loss": 1.6156,
+      "step": 128400
+    },
+    {
+      "epoch": 0.8167338273990364,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.6666666666666666e-06,
+      "loss": 1.6211,
+      "step": 128500
+    },
+    {
+      "epoch": 0.8173694179261952,
+      "grad_norm": 1.15625,
+      "learning_rate": 3.661538461538462e-06,
+      "loss": 1.6198,
+      "step": 128600
+    },
+    {
+      "epoch": 0.818005008453354,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.656410256410257e-06,
+      "loss": 1.6179,
+      "step": 128700
+    },
+    {
+      "epoch": 0.8186405989805128,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.6512820512820517e-06,
+      "loss": 1.6355,
+      "step": 128800
+    },
+    {
+      "epoch": 0.8192761895076716,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.6461538461538466e-06,
+      "loss": 1.6277,
+      "step": 128900
+    },
+    {
+      "epoch": 0.8199117800348303,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.641025641025641e-06,
+      "loss": 1.6202,
+      "step": 129000
+    },
+    {
+      "epoch": 0.8205473705619891,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.6358974358974364e-06,
+      "loss": 1.6254,
+      "step": 129100
+    },
+    {
+      "epoch": 0.8211829610891479,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.630769230769231e-06,
+      "loss": 1.6199,
+      "step": 129200
+    },
+    {
+      "epoch": 0.8218185516163067,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.6256410256410258e-06,
+      "loss": 1.615,
+      "step": 129300
+    },
+    {
+      "epoch": 0.8224541421434655,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.620512820512821e-06,
+      "loss": 1.6212,
+      "step": 129400
+    },
+    {
+      "epoch": 0.8230897326706242,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.6153846153846156e-06,
+      "loss": 1.6211,
+      "step": 129500
+    },
+    {
+      "epoch": 0.823725323197783,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.610256410256411e-06,
+      "loss": 1.6148,
+      "step": 129600
+    },
+    {
+      "epoch": 0.8243609137249418,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.6051282051282054e-06,
+      "loss": 1.6238,
+      "step": 129700
+    },
+    {
+      "epoch": 0.8249965042521006,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 1.6315,
+      "step": 129800
+    },
+    {
+      "epoch": 0.8256320947792594,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.5948717948717956e-06,
+      "loss": 1.6074,
+      "step": 129900
+    },
+    {
+      "epoch": 0.8262676853064181,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 1.6063,
+      "step": 130000
+    },
+    {
+      "epoch": 0.8269032758335769,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.5846153846153845e-06,
+      "loss": 1.6152,
+      "step": 130100
+    },
+    {
+      "epoch": 0.8275388663607358,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.57948717948718e-06,
+      "loss": 1.5949,
+      "step": 130200
+    },
+    {
+      "epoch": 0.8281744568878946,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.5743589743589748e-06,
+      "loss": 1.6273,
+      "step": 130300
+    },
+    {
+      "epoch": 0.8288100474150534,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.5692307692307692e-06,
+      "loss": 1.6031,
+      "step": 130400
+    },
+    {
+      "epoch": 0.8294456379422122,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.5641025641025646e-06,
+      "loss": 1.625,
+      "step": 130500
+    },
+    {
+      "epoch": 0.8300812284693709,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.558974358974359e-06,
+      "loss": 1.6359,
+      "step": 130600
+    },
+    {
+      "epoch": 0.8307168189965297,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.5538461538461544e-06,
+      "loss": 1.626,
+      "step": 130700
+    },
+    {
+      "epoch": 0.8313524095236885,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.5487179487179493e-06,
+      "loss": 1.6101,
+      "step": 130800
+    },
+    {
+      "epoch": 0.8319880000508473,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.5435897435897437e-06,
+      "loss": 1.6109,
+      "step": 130900
+    },
+    {
+      "epoch": 0.8326235905780061,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.538461538461539e-06,
+      "loss": 1.6194,
+      "step": 131000
+    },
+    {
+      "epoch": 0.8332591811051648,
+      "grad_norm": 0.5,
+      "learning_rate": 3.5333333333333335e-06,
+      "loss": 1.6146,
+      "step": 131100
+    },
+    {
+      "epoch": 0.8338947716323236,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.5282051282051284e-06,
+      "loss": 1.6276,
+      "step": 131200
+    },
+    {
+      "epoch": 0.8345303621594824,
+      "grad_norm": 1.265625,
+      "learning_rate": 3.5230769230769233e-06,
+      "loss": 1.6218,
+      "step": 131300
+    },
+    {
+      "epoch": 0.8351659526866412,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.5179487179487182e-06,
+      "loss": 1.6442,
+      "step": 131400
+    },
+    {
+      "epoch": 0.8358015432138,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.5128205128205127e-06,
+      "loss": 1.6282,
+      "step": 131500
+    },
+    {
+      "epoch": 0.8364371337409587,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.507692307692308e-06,
+      "loss": 1.6294,
+      "step": 131600
+    },
+    {
+      "epoch": 0.8370727242681175,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.502564102564103e-06,
+      "loss": 1.6143,
+      "step": 131700
+    },
+    {
+      "epoch": 0.8377083147952763,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.497435897435898e-06,
+      "loss": 1.6447,
+      "step": 131800
+    },
+    {
+      "epoch": 0.8383439053224351,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.4923076923076927e-06,
+      "loss": 1.6045,
+      "step": 131900
+    },
+    {
+      "epoch": 0.8389794958495939,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.487179487179487e-06,
+      "loss": 1.6307,
+      "step": 132000
+    },
+    {
+      "epoch": 0.8396150863767526,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.4820512820512825e-06,
+      "loss": 1.6028,
+      "step": 132100
+    },
+    {
+      "epoch": 0.8402506769039114,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.476923076923077e-06,
+      "loss": 1.6198,
+      "step": 132200
+    },
+    {
+      "epoch": 0.8408862674310702,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.471794871794872e-06,
+      "loss": 1.634,
+      "step": 132300
+    },
+    {
+      "epoch": 0.841521857958229,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.4666666666666672e-06,
+      "loss": 1.6291,
+      "step": 132400
+    },
+    {
+      "epoch": 0.8421574484853878,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 1.631,
+      "step": 132500
+    },
+    {
+      "epoch": 0.8427930390125465,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.456410256410257e-06,
+      "loss": 1.6228,
+      "step": 132600
+    },
+    {
+      "epoch": 0.8434286295397053,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.4512820512820515e-06,
+      "loss": 1.6166,
+      "step": 132700
+    },
+    {
+      "epoch": 0.8440642200668641,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.4461538461538464e-06,
+      "loss": 1.6259,
+      "step": 132800
+    },
+    {
+      "epoch": 0.8446998105940229,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.4410256410256417e-06,
+      "loss": 1.6147,
+      "step": 132900
+    },
+    {
+      "epoch": 0.8453354011211817,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.435897435897436e-06,
+      "loss": 1.6124,
+      "step": 133000
+    },
+    {
+      "epoch": 0.8459709916483404,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.4307692307692307e-06,
+      "loss": 1.6137,
+      "step": 133100
+    },
+    {
+      "epoch": 0.8466065821754992,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.425641025641026e-06,
+      "loss": 1.622,
+      "step": 133200
+    },
+    {
+      "epoch": 0.847242172702658,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.420512820512821e-06,
+      "loss": 1.6324,
+      "step": 133300
+    },
+    {
+      "epoch": 0.8478777632298168,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.4153846153846154e-06,
+      "loss": 1.6176,
+      "step": 133400
+    },
+    {
+      "epoch": 0.8485133537569756,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.4102564102564107e-06,
+      "loss": 1.6249,
+      "step": 133500
+    },
+    {
+      "epoch": 0.8491489442841343,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.405128205128205e-06,
+      "loss": 1.635,
+      "step": 133600
+    },
+    {
+      "epoch": 0.8497845348112931,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 1.6369,
+      "step": 133700
+    },
+    {
+      "epoch": 0.8504201253384519,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.3948717948717954e-06,
+      "loss": 1.6082,
+      "step": 133800
+    },
+    {
+      "epoch": 0.8510557158656107,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.38974358974359e-06,
+      "loss": 1.6106,
+      "step": 133900
+    },
+    {
+      "epoch": 0.8516913063927696,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.384615384615385e-06,
+      "loss": 1.6174,
+      "step": 134000
+    },
+    {
+      "epoch": 0.8523268969199284,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.3794871794871797e-06,
+      "loss": 1.6293,
+      "step": 134100
+    },
+    {
+      "epoch": 0.8529624874470871,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.3743589743589746e-06,
+      "loss": 1.6205,
+      "step": 134200
+    },
+    {
+      "epoch": 0.8535980779742459,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.3692307692307695e-06,
+      "loss": 1.611,
+      "step": 134300
+    },
+    {
+      "epoch": 0.8542336685014047,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.3641025641025644e-06,
+      "loss": 1.648,
+      "step": 134400
+    },
+    {
+      "epoch": 0.8548692590285635,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.358974358974359e-06,
+      "loss": 1.621,
+      "step": 134500
+    },
+    {
+      "epoch": 0.8555048495557223,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.353846153846154e-06,
+      "loss": 1.6387,
+      "step": 134600
+    },
+    {
+      "epoch": 0.856140440082881,
+      "grad_norm": 1.1171875,
+      "learning_rate": 3.348717948717949e-06,
+      "loss": 1.6229,
+      "step": 134700
+    },
+    {
+      "epoch": 0.8567760306100398,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.343589743589744e-06,
+      "loss": 1.616,
+      "step": 134800
+    },
+    {
+      "epoch": 0.8574116211371986,
+      "grad_norm": 1.25,
+      "learning_rate": 3.338461538461539e-06,
+      "loss": 1.6196,
+      "step": 134900
+    },
+    {
+      "epoch": 0.8580472116643574,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.6187,
+      "step": 135000
+    },
+    {
+      "epoch": 0.8586828021915162,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.3282051282051286e-06,
+      "loss": 1.6255,
+      "step": 135100
+    },
+    {
+      "epoch": 0.8593183927186749,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.323076923076923e-06,
+      "loss": 1.6298,
+      "step": 135200
+    },
+    {
+      "epoch": 0.8599539832458337,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.317948717948718e-06,
+      "loss": 1.6271,
+      "step": 135300
+    },
+    {
+      "epoch": 0.8605895737729925,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.3128205128205133e-06,
+      "loss": 1.6132,
+      "step": 135400
+    },
+    {
+      "epoch": 0.8612251643001513,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.307692307692308e-06,
+      "loss": 1.6222,
+      "step": 135500
+    },
+    {
+      "epoch": 0.8618607548273101,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.302564102564103e-06,
+      "loss": 1.6203,
+      "step": 135600
+    },
+    {
+      "epoch": 0.8624963453544688,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.2974358974358976e-06,
+      "loss": 1.6274,
+      "step": 135700
+    },
+    {
+      "epoch": 0.8631319358816276,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.2923076923076925e-06,
+      "loss": 1.6089,
+      "step": 135800
+    },
+    {
+      "epoch": 0.8637675264087864,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.287179487179488e-06,
+      "loss": 1.621,
+      "step": 135900
+    },
+    {
+      "epoch": 0.8644031169359452,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.2820512820512823e-06,
+      "loss": 1.604,
+      "step": 136000
+    },
+    {
+      "epoch": 0.865038707463104,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.276923076923077e-06,
+      "loss": 1.6205,
+      "step": 136100
+    },
+    {
+      "epoch": 0.8656742979902627,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.271794871794872e-06,
+      "loss": 1.6287,
+      "step": 136200
+    },
+    {
+      "epoch": 0.8663098885174215,
+      "grad_norm": 1.2890625,
+      "learning_rate": 3.266666666666667e-06,
+      "loss": 1.6314,
+      "step": 136300
+    },
+    {
+      "epoch": 0.8669454790445803,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.2615384615384615e-06,
+      "loss": 1.6229,
+      "step": 136400
+    },
+    {
+      "epoch": 0.8675810695717391,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.256410256410257e-06,
+      "loss": 1.6234,
+      "step": 136500
+    },
+    {
+      "epoch": 0.8682166600988979,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.2512820512820513e-06,
+      "loss": 1.6114,
+      "step": 136600
+    },
+    {
+      "epoch": 0.8688522506260566,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.2461538461538466e-06,
+      "loss": 1.6299,
+      "step": 136700
+    },
+    {
+      "epoch": 0.8694878411532154,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.2410256410256415e-06,
+      "loss": 1.5946,
+      "step": 136800
+    },
+    {
+      "epoch": 0.8701234316803742,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.235897435897436e-06,
+      "loss": 1.6175,
+      "step": 136900
+    },
+    {
+      "epoch": 0.870759022207533,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.2307692307692313e-06,
+      "loss": 1.616,
+      "step": 137000
+    },
+    {
+      "epoch": 0.8713946127346918,
+      "grad_norm": 1.328125,
+      "learning_rate": 3.2256410256410258e-06,
+      "loss": 1.6083,
+      "step": 137100
+    },
+    {
+      "epoch": 0.8720302032618505,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.2205128205128207e-06,
+      "loss": 1.6048,
+      "step": 137200
+    },
+    {
+      "epoch": 0.8726657937890093,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.2153846153846156e-06,
+      "loss": 1.6351,
+      "step": 137300
+    },
+    {
+      "epoch": 0.8733013843161681,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.2102564102564105e-06,
+      "loss": 1.6187,
+      "step": 137400
+    },
+    {
+      "epoch": 0.8739369748433269,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 1.616,
+      "step": 137500
+    },
+    {
+      "epoch": 0.8745725653704857,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.6311,
+      "step": 137600
+    },
+    {
+      "epoch": 0.8752081558976444,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.194871794871795e-06,
+      "loss": 1.6204,
+      "step": 137700
+    },
+    {
+      "epoch": 0.8758437464248033,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.18974358974359e-06,
+      "loss": 1.6455,
+      "step": 137800
+    },
+    {
+      "epoch": 0.8764793369519621,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.184615384615385e-06,
+      "loss": 1.6207,
+      "step": 137900
+    },
+    {
+      "epoch": 0.8771149274791209,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.1794871794871795e-06,
+      "loss": 1.6087,
+      "step": 138000
+    },
+    {
+      "epoch": 0.8777505180062797,
+      "grad_norm": 1.1328125,
+      "learning_rate": 3.1743589743589748e-06,
+      "loss": 1.6088,
+      "step": 138100
+    },
+    {
+      "epoch": 0.8783861085334385,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.1692307692307693e-06,
+      "loss": 1.6146,
+      "step": 138200
+    },
+    {
+      "epoch": 0.8790216990605972,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.164102564102564e-06,
+      "loss": 1.609,
+      "step": 138300
+    },
+    {
+      "epoch": 0.879657289587756,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.1589743589743595e-06,
+      "loss": 1.6187,
+      "step": 138400
+    },
+    {
+      "epoch": 0.8802928801149148,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.153846153846154e-06,
+      "loss": 1.6228,
+      "step": 138500
+    },
+    {
+      "epoch": 0.8809284706420736,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.1487179487179493e-06,
+      "loss": 1.6435,
+      "step": 138600
+    },
+    {
+      "epoch": 0.8815640611692324,
+      "grad_norm": 1.3203125,
+      "learning_rate": 3.1435897435897437e-06,
+      "loss": 1.6138,
+      "step": 138700
+    },
+    {
+      "epoch": 0.8821996516963911,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.1384615384615386e-06,
+      "loss": 1.619,
+      "step": 138800
+    },
+    {
+      "epoch": 0.8828352422235499,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.133333333333334e-06,
+      "loss": 1.629,
+      "step": 138900
+    },
+    {
+      "epoch": 0.8834708327507087,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.1282051282051284e-06,
+      "loss": 1.6192,
+      "step": 139000
+    },
+    {
+      "epoch": 0.8841064232778675,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.123076923076923e-06,
+      "loss": 1.6226,
+      "step": 139100
+    },
+    {
+      "epoch": 0.8847420138050263,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.1179487179487182e-06,
+      "loss": 1.6317,
+      "step": 139200
+    },
+    {
+      "epoch": 0.885377604332185,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.112820512820513e-06,
+      "loss": 1.6209,
+      "step": 139300
+    },
+    {
+      "epoch": 0.8860131948593438,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.1076923076923076e-06,
+      "loss": 1.6208,
+      "step": 139400
+    },
+    {
+      "epoch": 0.8866487853865026,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.102564102564103e-06,
+      "loss": 1.626,
+      "step": 139500
+    },
+    {
+      "epoch": 0.8872843759136614,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.0974358974358974e-06,
+      "loss": 1.6158,
+      "step": 139600
+    },
+    {
+      "epoch": 0.8879199664408202,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.0923076923076927e-06,
+      "loss": 1.6279,
+      "step": 139700
+    },
+    {
+      "epoch": 0.8885555569679789,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.0871794871794876e-06,
+      "loss": 1.628,
+      "step": 139800
+    },
+    {
+      "epoch": 0.8891911474951377,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.082051282051282e-06,
+      "loss": 1.6184,
+      "step": 139900
+    },
+    {
+      "epoch": 0.8898267380222965,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 1.6352,
+      "step": 140000
+    },
+    {
+      "epoch": 0.8904623285494553,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.071794871794872e-06,
+      "loss": 1.6027,
+      "step": 140100
+    },
+    {
+      "epoch": 0.8910979190766141,
+      "grad_norm": 1.7421875,
+      "learning_rate": 3.066666666666667e-06,
+      "loss": 1.6102,
+      "step": 140200
+    },
+    {
+      "epoch": 0.8917335096037728,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.0615384615384617e-06,
+      "loss": 1.6218,
+      "step": 140300
+    },
+    {
+      "epoch": 0.8923691001309316,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.0564102564102566e-06,
+      "loss": 1.6068,
+      "step": 140400
+    },
+    {
+      "epoch": 0.8930046906580904,
+      "grad_norm": 0.75,
+      "learning_rate": 3.051282051282052e-06,
+      "loss": 1.6191,
+      "step": 140500
+    },
+    {
+      "epoch": 0.8936402811852492,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.0461538461538464e-06,
+      "loss": 1.6149,
+      "step": 140600
+    },
+    {
+      "epoch": 0.894275871712408,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.0410256410256413e-06,
+      "loss": 1.6343,
+      "step": 140700
+    },
+    {
+      "epoch": 0.8949114622395667,
+      "grad_norm": 1.1640625,
+      "learning_rate": 3.035897435897436e-06,
+      "loss": 1.6081,
+      "step": 140800
+    },
+    {
+      "epoch": 0.8955470527667255,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.030769230769231e-06,
+      "loss": 1.6139,
+      "step": 140900
+    },
+    {
+      "epoch": 0.8961826432938843,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.0256410256410256e-06,
+      "loss": 1.6054,
+      "step": 141000
+    },
+    {
+      "epoch": 0.8968182338210431,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.020512820512821e-06,
+      "loss": 1.6071,
+      "step": 141100
+    },
+    {
+      "epoch": 0.8974538243482019,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.0153846153846154e-06,
+      "loss": 1.6072,
+      "step": 141200
+    },
+    {
+      "epoch": 0.8980894148753606,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.0102564102564103e-06,
+      "loss": 1.6221,
+      "step": 141300
+    },
+    {
+      "epoch": 0.8987250054025194,
+      "grad_norm": 0.875,
+      "learning_rate": 3.0051282051282056e-06,
+      "loss": 1.6284,
+      "step": 141400
+    },
+    {
+      "epoch": 0.8993605959296783,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3e-06,
+      "loss": 1.6133,
+      "step": 141500
+    },
+    {
+      "epoch": 0.8999961864568371,
+      "grad_norm": 0.75,
+      "learning_rate": 2.9948717948717954e-06,
+      "loss": 1.6361,
+      "step": 141600
+    },
+    {
+      "epoch": 0.9006317769839959,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.98974358974359e-06,
+      "loss": 1.613,
+      "step": 141700
+    },
+    {
+      "epoch": 0.9012673675111547,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.9846153846153848e-06,
+      "loss": 1.6331,
+      "step": 141800
+    },
+    {
+      "epoch": 0.9019029580383134,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.97948717948718e-06,
+      "loss": 1.6272,
+      "step": 141900
+    },
+    {
+      "epoch": 0.9025385485654722,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.9743589743589746e-06,
+      "loss": 1.6234,
+      "step": 142000
+    },
+    {
+      "epoch": 0.903174139092631,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.969230769230769e-06,
+      "loss": 1.6222,
+      "step": 142100
+    },
+    {
+      "epoch": 0.9038097296197898,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.9641025641025644e-06,
+      "loss": 1.6143,
+      "step": 142200
+    },
+    {
+      "epoch": 0.9044453201469486,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.9589743589743593e-06,
+      "loss": 1.6141,
+      "step": 142300
+    },
+    {
+      "epoch": 0.9050809106741073,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.953846153846154e-06,
+      "loss": 1.6266,
+      "step": 142400
+    },
+    {
+      "epoch": 0.9057165012012661,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.948717948717949e-06,
+      "loss": 1.6443,
+      "step": 142500
+    },
+    {
+      "epoch": 0.9063520917284249,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.9435897435897435e-06,
+      "loss": 1.6212,
+      "step": 142600
+    },
+    {
+      "epoch": 0.9069876822555837,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.938461538461539e-06,
+      "loss": 1.6373,
+      "step": 142700
+    },
+    {
+      "epoch": 0.9076232727827425,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.9333333333333338e-06,
+      "loss": 1.6184,
+      "step": 142800
+    },
+    {
+      "epoch": 0.9082588633099012,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.9282051282051282e-06,
+      "loss": 1.6146,
+      "step": 142900
+    },
+    {
+      "epoch": 0.90889445383706,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.9230769230769236e-06,
+      "loss": 1.6159,
+      "step": 143000
+    },
+    {
+      "epoch": 0.9095300443642188,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.917948717948718e-06,
+      "loss": 1.6432,
+      "step": 143100
+    },
+    {
+      "epoch": 0.9101656348913776,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.912820512820513e-06,
+      "loss": 1.6271,
+      "step": 143200
+    },
+    {
+      "epoch": 0.9108012254185364,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.907692307692308e-06,
+      "loss": 1.6198,
+      "step": 143300
+    },
+    {
+      "epoch": 0.9114368159456951,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.9025641025641027e-06,
+      "loss": 1.638,
+      "step": 143400
+    },
+    {
+      "epoch": 0.9120724064728539,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.897435897435898e-06,
+      "loss": 1.6263,
+      "step": 143500
+    },
+    {
+      "epoch": 0.9127079970000127,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.8923076923076925e-06,
+      "loss": 1.6231,
+      "step": 143600
+    },
+    {
+      "epoch": 0.9133435875271715,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.8871794871794874e-06,
+      "loss": 1.6218,
+      "step": 143700
+    },
+    {
+      "epoch": 0.9139791780543303,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.8820512820512823e-06,
+      "loss": 1.6251,
+      "step": 143800
+    },
+    {
+      "epoch": 0.914614768581489,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.8769230769230772e-06,
+      "loss": 1.6289,
+      "step": 143900
+    },
+    {
+      "epoch": 0.9152503591086478,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.8717948717948717e-06,
+      "loss": 1.6282,
+      "step": 144000
+    },
+    {
+      "epoch": 0.9158859496358066,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.866666666666667e-06,
+      "loss": 1.6281,
+      "step": 144100
+    },
+    {
+      "epoch": 0.9165215401629654,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.8615384615384615e-06,
+      "loss": 1.6153,
+      "step": 144200
+    },
+    {
+      "epoch": 0.9171571306901242,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.8564102564102564e-06,
+      "loss": 1.6339,
+      "step": 144300
+    },
+    {
+      "epoch": 0.917792721217283,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.8512820512820517e-06,
+      "loss": 1.6205,
+      "step": 144400
+    },
+    {
+      "epoch": 0.9184283117444417,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.846153846153846e-06,
+      "loss": 1.6251,
+      "step": 144500
+    },
+    {
+      "epoch": 0.9190639022716005,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.8410256410256415e-06,
+      "loss": 1.6175,
+      "step": 144600
+    },
+    {
+      "epoch": 0.9196994927987593,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.835897435897436e-06,
+      "loss": 1.6274,
+      "step": 144700
+    },
+    {
+      "epoch": 0.9203350833259181,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.830769230769231e-06,
+      "loss": 1.6275,
+      "step": 144800
+    },
+    {
+      "epoch": 0.9209706738530768,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.8256410256410262e-06,
+      "loss": 1.6317,
+      "step": 144900
+    },
+    {
+      "epoch": 0.9216062643802356,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.8205128205128207e-06,
+      "loss": 1.6217,
+      "step": 145000
+    },
+    {
+      "epoch": 0.9222418549073944,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.815384615384615e-06,
+      "loss": 1.6317,
+      "step": 145100
+    },
+    {
+      "epoch": 0.9228774454345532,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.8102564102564105e-06,
+      "loss": 1.6154,
+      "step": 145200
+    },
+    {
+      "epoch": 0.9235130359617121,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.8051282051282054e-06,
+      "loss": 1.6229,
+      "step": 145300
+    },
+    {
+      "epoch": 0.9241486264888709,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 1.6387,
+      "step": 145400
+    },
+    {
+      "epoch": 0.9247842170160296,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.794871794871795e-06,
+      "loss": 1.6295,
+      "step": 145500
+    },
+    {
+      "epoch": 0.9254198075431884,
+      "grad_norm": 1.140625,
+      "learning_rate": 2.7897435897435897e-06,
+      "loss": 1.5936,
+      "step": 145600
+    },
+    {
+      "epoch": 0.9260553980703472,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.784615384615385e-06,
+      "loss": 1.6272,
+      "step": 145700
+    },
+    {
+      "epoch": 0.926690988597506,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.77948717948718e-06,
+      "loss": 1.6146,
+      "step": 145800
+    },
+    {
+      "epoch": 0.9273265791246648,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.7743589743589744e-06,
+      "loss": 1.6311,
+      "step": 145900
+    },
+    {
+      "epoch": 0.9279621696518235,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.7692307692307697e-06,
+      "loss": 1.6133,
+      "step": 146000
+    },
+    {
+      "epoch": 0.9285977601789823,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.764102564102564e-06,
+      "loss": 1.6253,
+      "step": 146100
+    },
+    {
+      "epoch": 0.9292333507061411,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.758974358974359e-06,
+      "loss": 1.6415,
+      "step": 146200
+    },
+    {
+      "epoch": 0.9298689412332999,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.753846153846154e-06,
+      "loss": 1.6187,
+      "step": 146300
+    },
+    {
+      "epoch": 0.9305045317604587,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.748717948717949e-06,
+      "loss": 1.6169,
+      "step": 146400
+    },
+    {
+      "epoch": 0.9311401222876174,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.743589743589744e-06,
+      "loss": 1.6248,
+      "step": 146500
+    },
+    {
+      "epoch": 0.9317757128147762,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.7384615384615387e-06,
+      "loss": 1.6086,
+      "step": 146600
+    },
+    {
+      "epoch": 0.932411303341935,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.7333333333333336e-06,
+      "loss": 1.6358,
+      "step": 146700
+    },
+    {
+      "epoch": 0.9330468938690938,
+      "grad_norm": 0.875,
+      "learning_rate": 2.7282051282051285e-06,
+      "loss": 1.6129,
+      "step": 146800
+    },
+    {
+      "epoch": 0.9336824843962526,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.7230769230769234e-06,
+      "loss": 1.6228,
+      "step": 146900
+    },
+    {
+      "epoch": 0.9343180749234113,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.717948717948718e-06,
+      "loss": 1.623,
+      "step": 147000
+    },
+    {
+      "epoch": 0.9349536654505701,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.712820512820513e-06,
+      "loss": 1.633,
+      "step": 147100
+    },
+    {
+      "epoch": 0.9355892559777289,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.7076923076923076e-06,
+      "loss": 1.6282,
+      "step": 147200
+    },
+    {
+      "epoch": 0.9362248465048877,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.7025641025641025e-06,
+      "loss": 1.6231,
+      "step": 147300
+    },
+    {
+      "epoch": 0.9368604370320465,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.697435897435898e-06,
+      "loss": 1.6084,
+      "step": 147400
+    },
+    {
+      "epoch": 0.9374960275592052,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 1.6344,
+      "step": 147500
+    },
+    {
+      "epoch": 0.938131618086364,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.6871794871794877e-06,
+      "loss": 1.6374,
+      "step": 147600
+    },
+    {
+      "epoch": 0.9387672086135228,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.682051282051282e-06,
+      "loss": 1.6311,
+      "step": 147700
+    },
+    {
+      "epoch": 0.9394027991406816,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.676923076923077e-06,
+      "loss": 1.6174,
+      "step": 147800
+    },
+    {
+      "epoch": 0.9400383896678404,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.6717948717948724e-06,
+      "loss": 1.6077,
+      "step": 147900
+    },
+    {
+      "epoch": 0.9406739801949991,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.6211,
+      "step": 148000
+    },
+    {
+      "epoch": 0.9413095707221579,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.6615384615384613e-06,
+      "loss": 1.6337,
+      "step": 148100
+    },
+    {
+      "epoch": 0.9419451612493167,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.6564102564102566e-06,
+      "loss": 1.6298,
+      "step": 148200
+    },
+    {
+      "epoch": 0.9425807517764755,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.6512820512820515e-06,
+      "loss": 1.6167,
+      "step": 148300
+    },
+    {
+      "epoch": 0.9432163423036343,
+      "grad_norm": 1.5625,
+      "learning_rate": 2.6461538461538464e-06,
+      "loss": 1.6251,
+      "step": 148400
+    },
+    {
+      "epoch": 0.943851932830793,
+      "grad_norm": 1.28125,
+      "learning_rate": 2.6410256410256413e-06,
+      "loss": 1.631,
+      "step": 148500
+    },
+    {
+      "epoch": 0.9444875233579518,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.635897435897436e-06,
+      "loss": 1.6133,
+      "step": 148600
+    },
+    {
+      "epoch": 0.9451231138851106,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.630769230769231e-06,
+      "loss": 1.6298,
+      "step": 148700
+    },
+    {
+      "epoch": 0.9457587044122694,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.625641025641026e-06,
+      "loss": 1.6176,
+      "step": 148800
+    },
+    {
+      "epoch": 0.9463942949394282,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.6205128205128205e-06,
+      "loss": 1.6192,
+      "step": 148900
+    },
+    {
+      "epoch": 0.947029885466587,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.615384615384616e-06,
+      "loss": 1.6072,
+      "step": 149000
+    },
+    {
+      "epoch": 0.9476654759937458,
+      "grad_norm": 1.140625,
+      "learning_rate": 2.6102564102564103e-06,
+      "loss": 1.6265,
+      "step": 149100
+    },
+    {
+      "epoch": 0.9483010665209046,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.605128205128205e-06,
+      "loss": 1.6223,
+      "step": 149200
+    },
+    {
+      "epoch": 0.9489366570480634,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.6e-06,
+      "loss": 1.6233,
+      "step": 149300
+    },
+    {
+      "epoch": 0.9495722475752222,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.594871794871795e-06,
+      "loss": 1.6216,
+      "step": 149400
+    },
+    {
+      "epoch": 0.950207838102381,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.5897435897435903e-06,
+      "loss": 1.6383,
+      "step": 149500
+    },
+    {
+      "epoch": 0.9508434286295397,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.584615384615385e-06,
+      "loss": 1.627,
+      "step": 149600
+    },
+    {
+      "epoch": 0.9514790191566985,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.5794871794871797e-06,
+      "loss": 1.6252,
+      "step": 149700
+    },
+    {
+      "epoch": 0.9521146096838573,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.5743589743589746e-06,
+      "loss": 1.6155,
+      "step": 149800
+    },
+    {
+      "epoch": 0.9527502002110161,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.5692307692307695e-06,
+      "loss": 1.6213,
+      "step": 149900
+    },
+    {
+      "epoch": 0.9533857907381749,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 1.6184,
+      "step": 150000
+    },
+    {
+      "epoch": 0.9540213812653336,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.5589743589743593e-06,
+      "loss": 1.6158,
+      "step": 150100
+    },
+    {
+      "epoch": 0.9546569717924924,
+      "grad_norm": 0.875,
+      "learning_rate": 2.5538461538461538e-06,
+      "loss": 1.6308,
+      "step": 150200
+    },
+    {
+      "epoch": 0.9552925623196512,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.548717948717949e-06,
+      "loss": 1.6245,
+      "step": 150300
+    },
+    {
+      "epoch": 0.95592815284681,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.543589743589744e-06,
+      "loss": 1.6166,
+      "step": 150400
+    },
+    {
+      "epoch": 0.9565637433739688,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.5384615384615385e-06,
+      "loss": 1.6234,
+      "step": 150500
+    },
+    {
+      "epoch": 0.9571993339011275,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.5333333333333338e-06,
+      "loss": 1.6191,
+      "step": 150600
+    },
+    {
+      "epoch": 0.9578349244282863,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.5282051282051283e-06,
+      "loss": 1.6152,
+      "step": 150700
+    },
+    {
+      "epoch": 0.9584705149554451,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.523076923076923e-06,
+      "loss": 1.6097,
+      "step": 150800
+    },
+    {
+      "epoch": 0.9591061054826039,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.5179487179487185e-06,
+      "loss": 1.6461,
+      "step": 150900
+    },
+    {
+      "epoch": 0.9597416960097627,
+      "grad_norm": 1.3046875,
+      "learning_rate": 2.512820512820513e-06,
+      "loss": 1.6146,
+      "step": 151000
+    },
+    {
+      "epoch": 0.9603772865369214,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.507692307692308e-06,
+      "loss": 1.6435,
+      "step": 151100
+    },
+    {
+      "epoch": 0.9610128770640802,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.5025641025641028e-06,
+      "loss": 1.6163,
+      "step": 151200
+    },
+    {
+      "epoch": 0.961648467591239,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.4974358974358977e-06,
+      "loss": 1.6316,
+      "step": 151300
+    },
+    {
+      "epoch": 0.9622840581183978,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.4923076923076926e-06,
+      "loss": 1.6194,
+      "step": 151400
+    },
+    {
+      "epoch": 0.9629196486455566,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.4871794871794875e-06,
+      "loss": 1.6299,
+      "step": 151500
+    },
+    {
+      "epoch": 0.9635552391727153,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.4820512820512824e-06,
+      "loss": 1.6017,
+      "step": 151600
+    },
+    {
+      "epoch": 0.9641908296998741,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.4769230769230773e-06,
+      "loss": 1.6184,
+      "step": 151700
+    },
+    {
+      "epoch": 0.9648264202270329,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.471794871794872e-06,
+      "loss": 1.6238,
+      "step": 151800
+    },
+    {
+      "epoch": 0.9654620107541917,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.466666666666667e-06,
+      "loss": 1.6227,
+      "step": 151900
+    },
+    {
+      "epoch": 0.9660976012813505,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.461538461538462e-06,
+      "loss": 1.6209,
+      "step": 152000
+    },
+    {
+      "epoch": 0.9667331918085093,
+      "grad_norm": 1.0,
+      "learning_rate": 2.4564102564102564e-06,
+      "loss": 1.6293,
+      "step": 152100
+    },
+    {
+      "epoch": 0.967368782335668,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.4512820512820513e-06,
+      "loss": 1.6269,
+      "step": 152200
+    },
+    {
+      "epoch": 0.9680043728628268,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.4461538461538466e-06,
+      "loss": 1.6321,
+      "step": 152300
+    },
+    {
+      "epoch": 0.9686399633899856,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.441025641025641e-06,
+      "loss": 1.621,
+      "step": 152400
+    },
+    {
+      "epoch": 0.9692755539171444,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 1.6389,
+      "step": 152500
+    },
+    {
+      "epoch": 0.9699111444443032,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.430769230769231e-06,
+      "loss": 1.6327,
+      "step": 152600
+    },
+    {
+      "epoch": 0.9705467349714619,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.425641025641026e-06,
+      "loss": 1.6313,
+      "step": 152700
+    },
+    {
+      "epoch": 0.9711823254986207,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.4205128205128207e-06,
+      "loss": 1.6189,
+      "step": 152800
+    },
+    {
+      "epoch": 0.9718179160257796,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.4153846153846156e-06,
+      "loss": 1.6292,
+      "step": 152900
+    },
+    {
+      "epoch": 0.9724535065529384,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.4102564102564105e-06,
+      "loss": 1.6358,
+      "step": 153000
+    },
+    {
+      "epoch": 0.9730890970800972,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.4051282051282054e-06,
+      "loss": 1.6183,
+      "step": 153100
+    },
+    {
+      "epoch": 0.973724687607256,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.6149,
+      "step": 153200
+    },
+    {
+      "epoch": 0.9743602781344147,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.3948717948717952e-06,
+      "loss": 1.612,
+      "step": 153300
+    },
+    {
+      "epoch": 0.9749958686615735,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.38974358974359e-06,
+      "loss": 1.6341,
+      "step": 153400
+    },
+    {
+      "epoch": 0.9756314591887323,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.384615384615385e-06,
+      "loss": 1.621,
+      "step": 153500
+    },
+    {
+      "epoch": 0.9762670497158911,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.3794871794871795e-06,
+      "loss": 1.6103,
+      "step": 153600
+    },
+    {
+      "epoch": 0.9769026402430498,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.3743589743589744e-06,
+      "loss": 1.6274,
+      "step": 153700
+    },
+    {
+      "epoch": 0.9775382307702086,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.3692307692307697e-06,
+      "loss": 1.6163,
+      "step": 153800
+    },
+    {
+      "epoch": 0.9781738212973674,
+      "grad_norm": 1.140625,
+      "learning_rate": 2.364102564102564e-06,
+      "loss": 1.6313,
+      "step": 153900
+    },
+    {
+      "epoch": 0.9788094118245262,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.358974358974359e-06,
+      "loss": 1.613,
+      "step": 154000
+    },
+    {
+      "epoch": 0.979445002351685,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.353846153846154e-06,
+      "loss": 1.6266,
+      "step": 154100
+    },
+    {
+      "epoch": 0.9800805928788437,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.348717948717949e-06,
+      "loss": 1.6425,
+      "step": 154200
+    },
+    {
+      "epoch": 0.9807161834060025,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.3435897435897438e-06,
+      "loss": 1.6238,
+      "step": 154300
+    },
+    {
+      "epoch": 0.9813517739331613,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.3384615384615387e-06,
+      "loss": 1.6389,
+      "step": 154400
+    },
+    {
+      "epoch": 0.9819873644603201,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": 1.6078,
+      "step": 154500
+    },
+    {
+      "epoch": 0.9826229549874789,
+      "grad_norm": 1.3203125,
+      "learning_rate": 2.3282051282051285e-06,
+      "loss": 1.6403,
+      "step": 154600
+    },
+    {
+      "epoch": 0.9832585455146376,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.3230769230769234e-06,
+      "loss": 1.6259,
+      "step": 154700
+    },
+    {
+      "epoch": 0.9838941360417964,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.3179487179487183e-06,
+      "loss": 1.6166,
+      "step": 154800
+    },
+    {
+      "epoch": 0.9845297265689552,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.312820512820513e-06,
+      "loss": 1.6286,
+      "step": 154900
+    },
+    {
+      "epoch": 0.985165317096114,
+      "grad_norm": 1.125,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 1.6285,
+      "step": 155000
+    },
+    {
+      "epoch": 0.9858009076232728,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.3025641025641026e-06,
+      "loss": 1.6191,
+      "step": 155100
+    },
+    {
+      "epoch": 0.9864364981504316,
+      "grad_norm": 1.125,
+      "learning_rate": 2.2974358974358975e-06,
+      "loss": 1.6229,
+      "step": 155200
+    },
+    {
+      "epoch": 0.9870720886775903,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.2923076923076928e-06,
+      "loss": 1.6099,
+      "step": 155300
+    },
+    {
+      "epoch": 0.9877076792047491,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.2871794871794872e-06,
+      "loss": 1.621,
+      "step": 155400
+    },
+    {
+      "epoch": 0.9883432697319079,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.282051282051282e-06,
+      "loss": 1.6241,
+      "step": 155500
+    },
+    {
+      "epoch": 0.9889788602590667,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.276923076923077e-06,
+      "loss": 1.6201,
+      "step": 155600
+    },
+    {
+      "epoch": 0.9896144507862255,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.271794871794872e-06,
+      "loss": 1.6286,
+      "step": 155700
+    },
+    {
+      "epoch": 0.9902500413133842,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.266666666666667e-06,
+      "loss": 1.6277,
+      "step": 155800
+    },
+    {
+      "epoch": 0.990885631840543,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.2615384615384617e-06,
+      "loss": 1.6249,
+      "step": 155900
+    },
+    {
+      "epoch": 0.9915212223677018,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.2564102564102566e-06,
+      "loss": 1.6259,
+      "step": 156000
+    },
+    {
+      "epoch": 0.9921568128948606,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.2512820512820515e-06,
+      "loss": 1.6344,
+      "step": 156100
+    },
+    {
+      "epoch": 0.9927924034220194,
+      "grad_norm": 1.2421875,
+      "learning_rate": 2.2461538461538464e-06,
+      "loss": 1.6339,
+      "step": 156200
+    },
+    {
+      "epoch": 0.9934279939491781,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.2410256410256413e-06,
+      "loss": 1.6155,
+      "step": 156300
+    },
+    {
+      "epoch": 0.9940635844763369,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.2358974358974362e-06,
+      "loss": 1.6393,
+      "step": 156400
+    },
+    {
+      "epoch": 0.9946991750034957,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.230769230769231e-06,
+      "loss": 1.6203,
+      "step": 156500
+    },
+    {
+      "epoch": 0.9953347655306546,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.2256410256410256e-06,
+      "loss": 1.6214,
+      "step": 156600
+    },
+    {
+      "epoch": 0.9959703560578134,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.2205128205128205e-06,
+      "loss": 1.6092,
+      "step": 156700
+    },
+    {
+      "epoch": 0.9966059465849721,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.215384615384616e-06,
+      "loss": 1.6123,
+      "step": 156800
+    },
+    {
+      "epoch": 0.9972415371121309,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.2102564102564103e-06,
+      "loss": 1.6175,
+      "step": 156900
+    },
+    {
+      "epoch": 0.9978771276392897,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.2051282051282052e-06,
+      "loss": 1.5962,
+      "step": 157000
+    },
+    {
+      "epoch": 0.9985127181664485,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.2e-06,
+      "loss": 1.6391,
+      "step": 157100
+    },
+    {
+      "epoch": 0.9991483086936073,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.194871794871795e-06,
+      "loss": 1.6408,
+      "step": 157200
+    },
+    {
+      "epoch": 0.999783899220766,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.18974358974359e-06,
+      "loss": 1.6303,
+      "step": 157300
+    },
+    {
+      "epoch": 1.0004194897479248,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.184615384615385e-06,
+      "loss": 1.6148,
+      "step": 157400
+    },
+    {
+      "epoch": 1.0010550802750835,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.1794871794871797e-06,
+      "loss": 1.6173,
+      "step": 157500
+    },
+    {
+      "epoch": 1.0016906708022424,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.1743589743589746e-06,
+      "loss": 1.6077,
+      "step": 157600
+    },
+    {
+      "epoch": 1.002326261329401,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.1692307692307695e-06,
+      "loss": 1.6165,
+      "step": 157700
+    },
+    {
+      "epoch": 1.00296185185656,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.1641025641025644e-06,
+      "loss": 1.6246,
+      "step": 157800
+    },
+    {
+      "epoch": 1.0035974423837186,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.1589743589743593e-06,
+      "loss": 1.6128,
+      "step": 157900
+    },
+    {
+      "epoch": 1.0042330329108775,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.153846153846154e-06,
+      "loss": 1.6114,
+      "step": 158000
+    },
+    {
+      "epoch": 1.0048686234380362,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.1487179487179487e-06,
+      "loss": 1.6056,
+      "step": 158100
+    },
+    {
+      "epoch": 1.005504213965195,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.1435897435897436e-06,
+      "loss": 1.6178,
+      "step": 158200
+    },
+    {
+      "epoch": 1.0061398044923537,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.138461538461539e-06,
+      "loss": 1.6287,
+      "step": 158300
+    },
+    {
+      "epoch": 1.0067753950195126,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.133333333333334e-06,
+      "loss": 1.6419,
+      "step": 158400
+    },
+    {
+      "epoch": 1.0074109855466715,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.1282051282051283e-06,
+      "loss": 1.6179,
+      "step": 158500
+    },
+    {
+      "epoch": 1.0080465760738302,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.123076923076923e-06,
+      "loss": 1.6242,
+      "step": 158600
+    },
+    {
+      "epoch": 1.008682166600989,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.117948717948718e-06,
+      "loss": 1.6114,
+      "step": 158700
+    },
+    {
+      "epoch": 1.0093177571281478,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.112820512820513e-06,
+      "loss": 1.6039,
+      "step": 158800
+    },
+    {
+      "epoch": 1.0099533476553066,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.107692307692308e-06,
+      "loss": 1.6231,
+      "step": 158900
+    },
+    {
+      "epoch": 1.0105889381824653,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.1025641025641028e-06,
+      "loss": 1.6244,
+      "step": 159000
+    },
+    {
+      "epoch": 1.0112245287096242,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.0974358974358977e-06,
+      "loss": 1.629,
+      "step": 159100
+    },
+    {
+      "epoch": 1.0118601192367829,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.0923076923076926e-06,
+      "loss": 1.6334,
+      "step": 159200
+    },
+    {
+      "epoch": 1.0124957097639418,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.0871794871794875e-06,
+      "loss": 1.6162,
+      "step": 159300
+    },
+    {
+      "epoch": 1.0131313002911004,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.0820512820512824e-06,
+      "loss": 1.622,
+      "step": 159400
+    },
+    {
+      "epoch": 1.0137668908182593,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.0769230769230773e-06,
+      "loss": 1.6202,
+      "step": 159500
+    },
+    {
+      "epoch": 1.014402481345418,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.0717948717948717e-06,
+      "loss": 1.6231,
+      "step": 159600
+    },
+    {
+      "epoch": 1.0150380718725769,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.0666666666666666e-06,
+      "loss": 1.6075,
+      "step": 159700
+    },
+    {
+      "epoch": 1.0156736623997356,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.061538461538462e-06,
+      "loss": 1.6256,
+      "step": 159800
+    },
+    {
+      "epoch": 1.0163092529268944,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.056410256410257e-06,
+      "loss": 1.6224,
+      "step": 159900
+    },
+    {
+      "epoch": 1.0169448434540531,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 1.6256,
+      "step": 160000
+    },
+    {
+      "epoch": 1.017580433981212,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.0461538461538462e-06,
+      "loss": 1.6318,
+      "step": 160100
+    },
+    {
+      "epoch": 1.0182160245083707,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.041025641025641e-06,
+      "loss": 1.621,
+      "step": 160200
+    },
+    {
+      "epoch": 1.0188516150355296,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.035897435897436e-06,
+      "loss": 1.6265,
+      "step": 160300
+    },
+    {
+      "epoch": 1.0194872055626882,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.030769230769231e-06,
+      "loss": 1.6318,
+      "step": 160400
+    },
+    {
+      "epoch": 1.0201227960898471,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.025641025641026e-06,
+      "loss": 1.6279,
+      "step": 160500
+    },
+    {
+      "epoch": 1.0207583866170058,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.0205128205128207e-06,
+      "loss": 1.6012,
+      "step": 160600
+    },
+    {
+      "epoch": 1.0213939771441647,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.0153846153846156e-06,
+      "loss": 1.6238,
+      "step": 160700
+    },
+    {
+      "epoch": 1.0220295676713234,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.0102564102564105e-06,
+      "loss": 1.6219,
+      "step": 160800
+    },
+    {
+      "epoch": 1.0226651581984822,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.0051282051282054e-06,
+      "loss": 1.6225,
+      "step": 160900
+    },
+    {
+      "epoch": 1.023300748725641,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.6197,
+      "step": 161000
+    },
+    {
+      "epoch": 1.0239363392527998,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.994871794871795e-06,
+      "loss": 1.6083,
+      "step": 161100
+    },
+    {
+      "epoch": 1.0245719297799585,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.9897435897435897e-06,
+      "loss": 1.641,
+      "step": 161200
+    },
+    {
+      "epoch": 1.0252075203071174,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.984615384615385e-06,
+      "loss": 1.6369,
+      "step": 161300
+    },
+    {
+      "epoch": 1.025843110834276,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.97948717948718e-06,
+      "loss": 1.6109,
+      "step": 161400
+    },
+    {
+      "epoch": 1.026478701361435,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.9743589743589744e-06,
+      "loss": 1.6108,
+      "step": 161500
+    },
+    {
+      "epoch": 1.0271142918885936,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.9692307692307693e-06,
+      "loss": 1.6292,
+      "step": 161600
+    },
+    {
+      "epoch": 1.0277498824157525,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.964102564102564e-06,
+      "loss": 1.6273,
+      "step": 161700
+    },
+    {
+      "epoch": 1.0283854729429112,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.958974358974359e-06,
+      "loss": 1.6217,
+      "step": 161800
+    },
+    {
+      "epoch": 1.02902106347007,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.953846153846154e-06,
+      "loss": 1.6362,
+      "step": 161900
+    },
+    {
+      "epoch": 1.0296566539972287,
+      "grad_norm": 1.125,
+      "learning_rate": 1.948717948717949e-06,
+      "loss": 1.6055,
+      "step": 162000
+    },
+    {
+      "epoch": 1.0302922445243876,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.943589743589744e-06,
+      "loss": 1.6192,
+      "step": 162100
+    },
+    {
+      "epoch": 1.0309278350515463,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9384615384615387e-06,
+      "loss": 1.6168,
+      "step": 162200
+    },
+    {
+      "epoch": 1.0315634255787052,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.9333333333333336e-06,
+      "loss": 1.6188,
+      "step": 162300
+    },
+    {
+      "epoch": 1.032199016105864,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9282051282051285e-06,
+      "loss": 1.6285,
+      "step": 162400
+    },
+    {
+      "epoch": 1.0328346066330227,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 1.6258,
+      "step": 162500
+    },
+    {
+      "epoch": 1.0334701971601816,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.917948717948718e-06,
+      "loss": 1.6148,
+      "step": 162600
+    },
+    {
+      "epoch": 1.0341057876873403,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.9128205128205128e-06,
+      "loss": 1.6059,
+      "step": 162700
+    },
+    {
+      "epoch": 1.0347413782144992,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.907692307692308e-06,
+      "loss": 1.635,
+      "step": 162800
+    },
+    {
+      "epoch": 1.0353769687416579,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.9025641025641028e-06,
+      "loss": 1.6108,
+      "step": 162900
+    },
+    {
+      "epoch": 1.0360125592688167,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.8974358974358975e-06,
+      "loss": 1.6288,
+      "step": 163000
+    },
+    {
+      "epoch": 1.0366481497959754,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.8923076923076924e-06,
+      "loss": 1.6269,
+      "step": 163100
+    },
+    {
+      "epoch": 1.0372837403231343,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.8871794871794875e-06,
+      "loss": 1.6147,
+      "step": 163200
+    },
+    {
+      "epoch": 1.037919330850293,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.8820512820512822e-06,
+      "loss": 1.6123,
+      "step": 163300
+    },
+    {
+      "epoch": 1.0385549213774519,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.876923076923077e-06,
+      "loss": 1.6011,
+      "step": 163400
+    },
+    {
+      "epoch": 1.0391905119046105,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.871794871794872e-06,
+      "loss": 1.6291,
+      "step": 163500
+    },
+    {
+      "epoch": 1.0398261024317694,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.8666666666666669e-06,
+      "loss": 1.62,
+      "step": 163600
+    },
+    {
+      "epoch": 1.040461692958928,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.8615384615384616e-06,
+      "loss": 1.6134,
+      "step": 163700
+    },
+    {
+      "epoch": 1.041097283486087,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8564102564102565e-06,
+      "loss": 1.6198,
+      "step": 163800
+    },
+    {
+      "epoch": 1.0417328740132457,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.8512820512820516e-06,
+      "loss": 1.624,
+      "step": 163900
+    },
+    {
+      "epoch": 1.0423684645404045,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.8461538461538465e-06,
+      "loss": 1.6289,
+      "step": 164000
+    },
+    {
+      "epoch": 1.0430040550675632,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.8410256410256411e-06,
+      "loss": 1.6283,
+      "step": 164100
+    },
+    {
+      "epoch": 1.043639645594722,
+      "grad_norm": 0.75,
+      "learning_rate": 1.835897435897436e-06,
+      "loss": 1.63,
+      "step": 164200
+    },
+    {
+      "epoch": 1.0442752361218808,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.830769230769231e-06,
+      "loss": 1.6104,
+      "step": 164300
+    },
+    {
+      "epoch": 1.0449108266490397,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.8256410256410258e-06,
+      "loss": 1.6221,
+      "step": 164400
+    },
+    {
+      "epoch": 1.0455464171761983,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.8205128205128205e-06,
+      "loss": 1.6318,
+      "step": 164500
+    },
+    {
+      "epoch": 1.0461820077033572,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.8153846153846154e-06,
+      "loss": 1.6349,
+      "step": 164600
+    },
+    {
+      "epoch": 1.046817598230516,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.8102564102564105e-06,
+      "loss": 1.6375,
+      "step": 164700
+    },
+    {
+      "epoch": 1.0474531887576748,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.8051282051282054e-06,
+      "loss": 1.6247,
+      "step": 164800
+    },
+    {
+      "epoch": 1.0480887792848335,
+      "grad_norm": 0.75,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 1.6145,
+      "step": 164900
+    },
+    {
+      "epoch": 1.0487243698119924,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.794871794871795e-06,
+      "loss": 1.6266,
+      "step": 165000
+    },
+    {
+      "epoch": 1.049359960339151,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.78974358974359e-06,
+      "loss": 1.6338,
+      "step": 165100
+    },
+    {
+      "epoch": 1.04999555086631,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.7846153846153846e-06,
+      "loss": 1.6246,
+      "step": 165200
+    },
+    {
+      "epoch": 1.0506311413934686,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.7794871794871795e-06,
+      "loss": 1.6204,
+      "step": 165300
+    },
+    {
+      "epoch": 1.0512667319206275,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.7743589743589746e-06,
+      "loss": 1.6184,
+      "step": 165400
+    },
+    {
+      "epoch": 1.0519023224477861,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.7692307692307695e-06,
+      "loss": 1.6243,
+      "step": 165500
+    },
+    {
+      "epoch": 1.052537912974945,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.7641025641025642e-06,
+      "loss": 1.623,
+      "step": 165600
+    },
+    {
+      "epoch": 1.0531735035021037,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.7589743589743591e-06,
+      "loss": 1.6189,
+      "step": 165700
+    },
+    {
+      "epoch": 1.0538090940292626,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.753846153846154e-06,
+      "loss": 1.6277,
+      "step": 165800
+    },
+    {
+      "epoch": 1.0544446845564215,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.748717948717949e-06,
+      "loss": 1.6033,
+      "step": 165900
+    },
+    {
+      "epoch": 1.0550802750835802,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.7435897435897436e-06,
+      "loss": 1.6138,
+      "step": 166000
+    },
+    {
+      "epoch": 1.0557158656107388,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.7384615384615385e-06,
+      "loss": 1.6237,
+      "step": 166100
+    },
+    {
+      "epoch": 1.0563514561378977,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.7333333333333336e-06,
+      "loss": 1.6208,
+      "step": 166200
+    },
+    {
+      "epoch": 1.0569870466650566,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.7282051282051285e-06,
+      "loss": 1.6214,
+      "step": 166300
+    },
+    {
+      "epoch": 1.0576226371922153,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.7230769230769232e-06,
+      "loss": 1.6138,
+      "step": 166400
+    },
+    {
+      "epoch": 1.0582582277193742,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.717948717948718e-06,
+      "loss": 1.6093,
+      "step": 166500
+    },
+    {
+      "epoch": 1.0588938182465328,
+      "grad_norm": 1.0,
+      "learning_rate": 1.712820512820513e-06,
+      "loss": 1.6212,
+      "step": 166600
+    },
+    {
+      "epoch": 1.0595294087736917,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.7076923076923077e-06,
+      "loss": 1.6342,
+      "step": 166700
+    },
+    {
+      "epoch": 1.0601649993008504,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7025641025641026e-06,
+      "loss": 1.6263,
+      "step": 166800
+    },
+    {
+      "epoch": 1.0608005898280093,
+      "grad_norm": 1.125,
+      "learning_rate": 1.6974358974358977e-06,
+      "loss": 1.6238,
+      "step": 166900
+    },
+    {
+      "epoch": 1.061436180355168,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.6923076923076926e-06,
+      "loss": 1.5956,
+      "step": 167000
+    },
+    {
+      "epoch": 1.0620717708823268,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.6871794871794873e-06,
+      "loss": 1.6477,
+      "step": 167100
+    },
+    {
+      "epoch": 1.0627073614094855,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.6820512820512822e-06,
+      "loss": 1.6159,
+      "step": 167200
+    },
+    {
+      "epoch": 1.0633429519366444,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.676923076923077e-06,
+      "loss": 1.6345,
+      "step": 167300
+    },
+    {
+      "epoch": 1.063978542463803,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.671794871794872e-06,
+      "loss": 1.6316,
+      "step": 167400
+    },
+    {
+      "epoch": 1.064614132990962,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.6155,
+      "step": 167500
+    },
+    {
+      "epoch": 1.0652497235181206,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.6615384615384616e-06,
+      "loss": 1.6145,
+      "step": 167600
+    },
+    {
+      "epoch": 1.0658853140452795,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6564102564102567e-06,
+      "loss": 1.626,
+      "step": 167700
+    },
+    {
+      "epoch": 1.0665209045724382,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.6512820512820516e-06,
+      "loss": 1.6213,
+      "step": 167800
+    },
+    {
+      "epoch": 1.067156495099597,
+      "grad_norm": 0.875,
+      "learning_rate": 1.6461538461538463e-06,
+      "loss": 1.6201,
+      "step": 167900
+    },
+    {
+      "epoch": 1.0677920856267558,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.6410256410256412e-06,
+      "loss": 1.6231,
+      "step": 168000
+    },
+    {
+      "epoch": 1.0684276761539147,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.635897435897436e-06,
+      "loss": 1.6285,
+      "step": 168100
+    },
+    {
+      "epoch": 1.0690632666810733,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.6307692307692307e-06,
+      "loss": 1.628,
+      "step": 168200
+    },
+    {
+      "epoch": 1.0696988572082322,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.6256410256410256e-06,
+      "loss": 1.625,
+      "step": 168300
+    },
+    {
+      "epoch": 1.0703344477353909,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.6205128205128208e-06,
+      "loss": 1.6231,
+      "step": 168400
+    },
+    {
+      "epoch": 1.0709700382625498,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.6153846153846157e-06,
+      "loss": 1.6204,
+      "step": 168500
+    },
+    {
+      "epoch": 1.0716056287897084,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.6102564102564103e-06,
+      "loss": 1.6307,
+      "step": 168600
+    },
+    {
+      "epoch": 1.0722412193168673,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.6051282051282052e-06,
+      "loss": 1.6235,
+      "step": 168700
+    },
+    {
+      "epoch": 1.072876809844026,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.6222,
+      "step": 168800
+    },
+    {
+      "epoch": 1.073512400371185,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.594871794871795e-06,
+      "loss": 1.6281,
+      "step": 168900
+    },
+    {
+      "epoch": 1.0741479908983436,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.5897435897435897e-06,
+      "loss": 1.6216,
+      "step": 169000
+    },
+    {
+      "epoch": 1.0747835814255025,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.5846153846153846e-06,
+      "loss": 1.63,
+      "step": 169100
+    },
+    {
+      "epoch": 1.0754191719526611,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.5794871794871797e-06,
+      "loss": 1.617,
+      "step": 169200
+    },
+    {
+      "epoch": 1.07605476247982,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.5743589743589746e-06,
+      "loss": 1.6236,
+      "step": 169300
+    },
+    {
+      "epoch": 1.0766903530069787,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.5692307692307693e-06,
+      "loss": 1.6239,
+      "step": 169400
+    },
+    {
+      "epoch": 1.0773259435341376,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.5641025641025642e-06,
+      "loss": 1.6126,
+      "step": 169500
+    },
+    {
+      "epoch": 1.0779615340612962,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5589743589743591e-06,
+      "loss": 1.6349,
+      "step": 169600
+    },
+    {
+      "epoch": 1.0785971245884551,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.5538461538461538e-06,
+      "loss": 1.6318,
+      "step": 169700
+    },
+    {
+      "epoch": 1.079232715115614,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.5487179487179487e-06,
+      "loss": 1.6238,
+      "step": 169800
+    },
+    {
+      "epoch": 1.0798683056427727,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.5435897435897438e-06,
+      "loss": 1.6289,
+      "step": 169900
+    },
+    {
+      "epoch": 1.0805038961699314,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 1.6279,
+      "step": 170000
+    },
+    {
+      "epoch": 1.0811394866970903,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5333333333333334e-06,
+      "loss": 1.6226,
+      "step": 170100
+    },
+    {
+      "epoch": 1.0817750772242491,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.5282051282051283e-06,
+      "loss": 1.621,
+      "step": 170200
+    },
+    {
+      "epoch": 1.0824106677514078,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5230769230769232e-06,
+      "loss": 1.6347,
+      "step": 170300
+    },
+    {
+      "epoch": 1.0830462582785667,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.517948717948718e-06,
+      "loss": 1.6362,
+      "step": 170400
+    },
+    {
+      "epoch": 1.0836818488057254,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.5128205128205128e-06,
+      "loss": 1.6452,
+      "step": 170500
+    },
+    {
+      "epoch": 1.0843174393328843,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.5076923076923077e-06,
+      "loss": 1.6087,
+      "step": 170600
+    },
+    {
+      "epoch": 1.084953029860043,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.5025641025641028e-06,
+      "loss": 1.614,
+      "step": 170700
+    },
+    {
+      "epoch": 1.0855886203872018,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.4974358974358977e-06,
+      "loss": 1.6244,
+      "step": 170800
+    },
+    {
+      "epoch": 1.0862242109143605,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4923076923076924e-06,
+      "loss": 1.6118,
+      "step": 170900
+    },
+    {
+      "epoch": 1.0868598014415194,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4871794871794873e-06,
+      "loss": 1.62,
+      "step": 171000
+    },
+    {
+      "epoch": 1.087495391968678,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4820512820512822e-06,
+      "loss": 1.6146,
+      "step": 171100
+    },
+    {
+      "epoch": 1.088130982495837,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.476923076923077e-06,
+      "loss": 1.6233,
+      "step": 171200
+    },
+    {
+      "epoch": 1.0887665730229956,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.4717948717948718e-06,
+      "loss": 1.607,
+      "step": 171300
+    },
+    {
+      "epoch": 1.0894021635501545,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.4666666666666669e-06,
+      "loss": 1.6358,
+      "step": 171400
+    },
+    {
+      "epoch": 1.0900377540773132,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.4615384615384618e-06,
+      "loss": 1.6165,
+      "step": 171500
+    },
+    {
+      "epoch": 1.090673344604472,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.4564102564102565e-06,
+      "loss": 1.624,
+      "step": 171600
+    },
+    {
+      "epoch": 1.0913089351316307,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4512820512820514e-06,
+      "loss": 1.6289,
+      "step": 171700
+    },
+    {
+      "epoch": 1.0919445256587896,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.4461538461538463e-06,
+      "loss": 1.6245,
+      "step": 171800
+    },
+    {
+      "epoch": 1.0925801161859483,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.4410256410256412e-06,
+      "loss": 1.635,
+      "step": 171900
+    },
+    {
+      "epoch": 1.0932157067131072,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.4358974358974359e-06,
+      "loss": 1.6379,
+      "step": 172000
+    },
+    {
+      "epoch": 1.0938512972402659,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.4307692307692308e-06,
+      "loss": 1.6384,
+      "step": 172100
+    },
+    {
+      "epoch": 1.0944868877674248,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4256410256410259e-06,
+      "loss": 1.6085,
+      "step": 172200
+    },
+    {
+      "epoch": 1.0951224782945834,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.4205128205128208e-06,
+      "loss": 1.6252,
+      "step": 172300
+    },
+    {
+      "epoch": 1.0957580688217423,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4153846153846155e-06,
+      "loss": 1.6271,
+      "step": 172400
+    },
+    {
+      "epoch": 1.096393659348901,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.4102564102564104e-06,
+      "loss": 1.6031,
+      "step": 172500
+    },
+    {
+      "epoch": 1.0970292498760599,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.4051282051282052e-06,
+      "loss": 1.6219,
+      "step": 172600
+    },
+    {
+      "epoch": 1.0976648404032185,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 1.629,
+      "step": 172700
+    },
+    {
+      "epoch": 1.0983004309303774,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.3948717948717948e-06,
+      "loss": 1.6137,
+      "step": 172800
+    },
+    {
+      "epoch": 1.098936021457536,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.38974358974359e-06,
+      "loss": 1.6134,
+      "step": 172900
+    },
+    {
+      "epoch": 1.099571611984695,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3846153846153848e-06,
+      "loss": 1.6247,
+      "step": 173000
+    },
+    {
+      "epoch": 1.1002072025118537,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.3794871794871795e-06,
+      "loss": 1.6231,
+      "step": 173100
+    },
+    {
+      "epoch": 1.1008427930390126,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.3743589743589744e-06,
+      "loss": 1.6254,
+      "step": 173200
+    },
+    {
+      "epoch": 1.1014783835661712,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.3692307692307693e-06,
+      "loss": 1.6358,
+      "step": 173300
+    },
+    {
+      "epoch": 1.1021139740933301,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.3641025641025642e-06,
+      "loss": 1.6345,
+      "step": 173400
+    },
+    {
+      "epoch": 1.1027495646204888,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.358974358974359e-06,
+      "loss": 1.6374,
+      "step": 173500
+    },
+    {
+      "epoch": 1.1033851551476477,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.3538461538461538e-06,
+      "loss": 1.6277,
+      "step": 173600
+    },
+    {
+      "epoch": 1.1040207456748066,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.348717948717949e-06,
+      "loss": 1.6246,
+      "step": 173700
+    },
+    {
+      "epoch": 1.1046563362019652,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.3435897435897438e-06,
+      "loss": 1.6221,
+      "step": 173800
+    },
+    {
+      "epoch": 1.1052919267291241,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3384615384615385e-06,
+      "loss": 1.61,
+      "step": 173900
+    },
+    {
+      "epoch": 1.1059275172562828,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 1.633,
+      "step": 174000
+    },
+    {
+      "epoch": 1.1065631077834417,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.3282051282051283e-06,
+      "loss": 1.6234,
+      "step": 174100
+    },
+    {
+      "epoch": 1.1071986983106004,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.3230769230769232e-06,
+      "loss": 1.6222,
+      "step": 174200
+    },
+    {
+      "epoch": 1.1078342888377593,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.317948717948718e-06,
+      "loss": 1.6104,
+      "step": 174300
+    },
+    {
+      "epoch": 1.108469879364918,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.312820512820513e-06,
+      "loss": 1.6357,
+      "step": 174400
+    },
+    {
+      "epoch": 1.1091054698920768,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.307692307692308e-06,
+      "loss": 1.6394,
+      "step": 174500
+    },
+    {
+      "epoch": 1.1097410604192355,
+      "grad_norm": 0.875,
+      "learning_rate": 1.3025641025641026e-06,
+      "loss": 1.6243,
+      "step": 174600
+    },
+    {
+      "epoch": 1.1103766509463944,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.2974358974358975e-06,
+      "loss": 1.6249,
+      "step": 174700
+    },
+    {
+      "epoch": 1.111012241473553,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.2923076923076924e-06,
+      "loss": 1.6309,
+      "step": 174800
+    },
+    {
+      "epoch": 1.111647832000712,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.2871794871794873e-06,
+      "loss": 1.6047,
+      "step": 174900
+    },
+    {
+      "epoch": 1.1122834225278706,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 1.6151,
+      "step": 175000
+    },
+    {
+      "epoch": 1.1129190130550295,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.2769230769230769e-06,
+      "loss": 1.6192,
+      "step": 175100
+    },
+    {
+      "epoch": 1.1135546035821882,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.271794871794872e-06,
+      "loss": 1.6136,
+      "step": 175200
+    },
+    {
+      "epoch": 1.114190194109347,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.2666666666666669e-06,
+      "loss": 1.6255,
+      "step": 175300
+    },
+    {
+      "epoch": 1.1148257846365057,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.2615384615384616e-06,
+      "loss": 1.6317,
+      "step": 175400
+    },
+    {
+      "epoch": 1.1154613751636646,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.2564102564102565e-06,
+      "loss": 1.6191,
+      "step": 175500
+    },
+    {
+      "epoch": 1.1160969656908233,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2512820512820514e-06,
+      "loss": 1.6208,
+      "step": 175600
+    },
+    {
+      "epoch": 1.1167325562179822,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.2461538461538463e-06,
+      "loss": 1.6289,
+      "step": 175700
+    },
+    {
+      "epoch": 1.1173681467451408,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2410256410256412e-06,
+      "loss": 1.632,
+      "step": 175800
+    },
+    {
+      "epoch": 1.1180037372722997,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.235897435897436e-06,
+      "loss": 1.6146,
+      "step": 175900
+    },
+    {
+      "epoch": 1.1186393277994584,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.230769230769231e-06,
+      "loss": 1.626,
+      "step": 176000
+    },
+    {
+      "epoch": 1.1192749183266173,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2256410256410257e-06,
+      "loss": 1.6242,
+      "step": 176100
+    },
+    {
+      "epoch": 1.119910508853776,
+      "grad_norm": 1.25,
+      "learning_rate": 1.2205128205128206e-06,
+      "loss": 1.6396,
+      "step": 176200
+    },
+    {
+      "epoch": 1.1205460993809349,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.2153846153846155e-06,
+      "loss": 1.6206,
+      "step": 176300
+    },
+    {
+      "epoch": 1.1211816899080935,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2102564102564104e-06,
+      "loss": 1.6212,
+      "step": 176400
+    },
+    {
+      "epoch": 1.1218172804352524,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.2051282051282053e-06,
+      "loss": 1.6244,
+      "step": 176500
+    },
+    {
+      "epoch": 1.122452870962411,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.623,
+      "step": 176600
+    },
+    {
+      "epoch": 1.12308846148957,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.194871794871795e-06,
+      "loss": 1.6183,
+      "step": 176700
+    },
+    {
+      "epoch": 1.1237240520167286,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.1897435897435897e-06,
+      "loss": 1.622,
+      "step": 176800
+    },
+    {
+      "epoch": 1.1243596425438875,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.1846153846153849e-06,
+      "loss": 1.6249,
+      "step": 176900
+    },
+    {
+      "epoch": 1.1249952330710462,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.1794871794871795e-06,
+      "loss": 1.627,
+      "step": 177000
+    },
+    {
+      "epoch": 1.125630823598205,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.1743589743589744e-06,
+      "loss": 1.6355,
+      "step": 177100
+    },
+    {
+      "epoch": 1.126266414125364,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.1692307692307693e-06,
+      "loss": 1.6318,
+      "step": 177200
+    },
+    {
+      "epoch": 1.1269020046525227,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.1641025641025642e-06,
+      "loss": 1.6169,
+      "step": 177300
+    },
+    {
+      "epoch": 1.1275375951796813,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.1589743589743591e-06,
+      "loss": 1.6195,
+      "step": 177400
+    },
+    {
+      "epoch": 1.1281731857068402,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 1.6207,
+      "step": 177500
+    },
+    {
+      "epoch": 1.1288087762339991,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.1487179487179487e-06,
+      "loss": 1.6202,
+      "step": 177600
+    },
+    {
+      "epoch": 1.1294443667611578,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.1435897435897436e-06,
+      "loss": 1.6311,
+      "step": 177700
+    },
+    {
+      "epoch": 1.1300799572883165,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.1384615384615385e-06,
+      "loss": 1.6345,
+      "step": 177800
+    },
+    {
+      "epoch": 1.1307155478154753,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.1333333333333334e-06,
+      "loss": 1.623,
+      "step": 177900
+    },
+    {
+      "epoch": 1.1313511383426342,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.1282051282051283e-06,
+      "loss": 1.6246,
+      "step": 178000
+    },
+    {
+      "epoch": 1.131986728869793,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.1230769230769232e-06,
+      "loss": 1.6302,
+      "step": 178100
+    },
+    {
+      "epoch": 1.1326223193969518,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.1179487179487181e-06,
+      "loss": 1.6337,
+      "step": 178200
+    },
+    {
+      "epoch": 1.1332579099241105,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.1128205128205128e-06,
+      "loss": 1.6308,
+      "step": 178300
+    },
+    {
+      "epoch": 1.1338935004512694,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.107692307692308e-06,
+      "loss": 1.6263,
+      "step": 178400
+    },
+    {
+      "epoch": 1.134529090978428,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1025641025641026e-06,
+      "loss": 1.6238,
+      "step": 178500
+    },
+    {
+      "epoch": 1.135164681505587,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.0974358974358975e-06,
+      "loss": 1.6269,
+      "step": 178600
+    },
+    {
+      "epoch": 1.1358002720327456,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.0923076923076924e-06,
+      "loss": 1.624,
+      "step": 178700
+    },
+    {
+      "epoch": 1.1364358625599045,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.0871794871794873e-06,
+      "loss": 1.6185,
+      "step": 178800
+    },
+    {
+      "epoch": 1.1370714530870631,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.0820512820512822e-06,
+      "loss": 1.612,
+      "step": 178900
+    },
+    {
+      "epoch": 1.137707043614222,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.076923076923077e-06,
+      "loss": 1.6091,
+      "step": 179000
+    },
+    {
+      "epoch": 1.1383426341413807,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.0717948717948718e-06,
+      "loss": 1.6232,
+      "step": 179100
+    },
+    {
+      "epoch": 1.1389782246685396,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.066666666666667e-06,
+      "loss": 1.604,
+      "step": 179200
+    },
+    {
+      "epoch": 1.1396138151956983,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.0615384615384616e-06,
+      "loss": 1.6217,
+      "step": 179300
+    },
+    {
+      "epoch": 1.1402494057228572,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.0564102564102565e-06,
+      "loss": 1.6103,
+      "step": 179400
+    },
+    {
+      "epoch": 1.1408849962500158,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.0512820512820514e-06,
+      "loss": 1.6224,
+      "step": 179500
+    },
+    {
+      "epoch": 1.1415205867771747,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.0461538461538463e-06,
+      "loss": 1.6232,
+      "step": 179600
+    },
+    {
+      "epoch": 1.1421561773043334,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.0410256410256412e-06,
+      "loss": 1.6241,
+      "step": 179700
+    },
+    {
+      "epoch": 1.1427917678314923,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.0358974358974359e-06,
+      "loss": 1.616,
+      "step": 179800
+    },
+    {
+      "epoch": 1.143427358358651,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.030769230769231e-06,
+      "loss": 1.6291,
+      "step": 179900
+    },
+    {
+      "epoch": 1.1440629488858098,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 1.6228,
+      "step": 180000
+    },
+    {
+      "epoch": 1.1446985394129685,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.0205128205128206e-06,
+      "loss": 1.6178,
+      "step": 180100
+    },
+    {
+      "epoch": 1.1453341299401274,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.0153846153846155e-06,
+      "loss": 1.6336,
+      "step": 180200
+    },
+    {
+      "epoch": 1.145969720467286,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.0102564102564104e-06,
+      "loss": 1.6261,
+      "step": 180300
+    },
+    {
+      "epoch": 1.146605310994445,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.0051282051282053e-06,
+      "loss": 1.6065,
+      "step": 180400
+    },
+    {
+      "epoch": 1.1472409015216036,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.605,
+      "step": 180500
+    },
+    {
+      "epoch": 1.1478764920487625,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.948717948717949e-07,
+      "loss": 1.6426,
+      "step": 180600
+    },
+    {
+      "epoch": 1.1485120825759214,
+      "grad_norm": 1.2734375,
+      "learning_rate": 9.8974358974359e-07,
+      "loss": 1.6313,
+      "step": 180700
+    },
+    {
+      "epoch": 1.14914767310308,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.846153846153847e-07,
+      "loss": 1.6194,
+      "step": 180800
+    },
+    {
+      "epoch": 1.1497832636302387,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.794871794871796e-07,
+      "loss": 1.6249,
+      "step": 180900
+    },
+    {
+      "epoch": 1.1504188541573976,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.743589743589745e-07,
+      "loss": 1.6149,
+      "step": 181000
+    },
+    {
+      "epoch": 1.1510544446845565,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.692307692307693e-07,
+      "loss": 1.6307,
+      "step": 181100
+    },
+    {
+      "epoch": 1.1516900352117152,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.641025641025642e-07,
+      "loss": 1.6369,
+      "step": 181200
+    },
+    {
+      "epoch": 1.1523256257388739,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.58974358974359e-07,
+      "loss": 1.627,
+      "step": 181300
+    },
+    {
+      "epoch": 1.1529612162660328,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.53846153846154e-07,
+      "loss": 1.6305,
+      "step": 181400
+    },
+    {
+      "epoch": 1.1535968067931917,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.487179487179487e-07,
+      "loss": 1.6156,
+      "step": 181500
+    },
+    {
+      "epoch": 1.1542323973203503,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.435897435897437e-07,
+      "loss": 1.6276,
+      "step": 181600
+    },
+    {
+      "epoch": 1.1548679878475092,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.384615384615385e-07,
+      "loss": 1.6395,
+      "step": 181700
+    },
+    {
+      "epoch": 1.1555035783746679,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.333333333333334e-07,
+      "loss": 1.6233,
+      "step": 181800
+    },
+    {
+      "epoch": 1.1561391689018268,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.282051282051282e-07,
+      "loss": 1.6164,
+      "step": 181900
+    },
+    {
+      "epoch": 1.1567747594289854,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.230769230769232e-07,
+      "loss": 1.6231,
+      "step": 182000
+    },
+    {
+      "epoch": 1.1574103499561443,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.17948717948718e-07,
+      "loss": 1.6251,
+      "step": 182100
+    },
+    {
+      "epoch": 1.158045940483303,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.128205128205129e-07,
+      "loss": 1.6266,
+      "step": 182200
+    },
+    {
+      "epoch": 1.158681531010462,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.076923076923077e-07,
+      "loss": 1.6057,
+      "step": 182300
+    },
+    {
+      "epoch": 1.1593171215376206,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.025641025641027e-07,
+      "loss": 1.6329,
+      "step": 182400
+    },
+    {
+      "epoch": 1.1599527120647795,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.974358974358975e-07,
+      "loss": 1.6308,
+      "step": 182500
+    },
+    {
+      "epoch": 1.1605883025919381,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.923076923076923e-07,
+      "loss": 1.6234,
+      "step": 182600
+    },
+    {
+      "epoch": 1.161223893119097,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.871794871794873e-07,
+      "loss": 1.6105,
+      "step": 182700
+    },
+    {
+      "epoch": 1.1618594836462557,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.820512820512821e-07,
+      "loss": 1.627,
+      "step": 182800
+    },
+    {
+      "epoch": 1.1624950741734146,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.76923076923077e-07,
+      "loss": 1.622,
+      "step": 182900
+    },
+    {
+      "epoch": 1.1631306647005732,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.717948717948718e-07,
+      "loss": 1.6267,
+      "step": 183000
+    },
+    {
+      "epoch": 1.1637662552277321,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.666666666666668e-07,
+      "loss": 1.628,
+      "step": 183100
+    },
+    {
+      "epoch": 1.1644018457548908,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.615384615384616e-07,
+      "loss": 1.6225,
+      "step": 183200
+    },
+    {
+      "epoch": 1.1650374362820497,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.564102564102565e-07,
+      "loss": 1.6211,
+      "step": 183300
+    },
+    {
+      "epoch": 1.1656730268092084,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.512820512820513e-07,
+      "loss": 1.6209,
+      "step": 183400
+    },
+    {
+      "epoch": 1.1663086173363673,
+      "grad_norm": 1.3046875,
+      "learning_rate": 8.461538461538463e-07,
+      "loss": 1.6237,
+      "step": 183500
+    },
+    {
+      "epoch": 1.166944207863526,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.410256410256411e-07,
+      "loss": 1.6333,
+      "step": 183600
+    },
+    {
+      "epoch": 1.1675797983906848,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.35897435897436e-07,
+      "loss": 1.6243,
+      "step": 183700
+    },
+    {
+      "epoch": 1.1682153889178435,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.307692307692308e-07,
+      "loss": 1.6183,
+      "step": 183800
+    },
+    {
+      "epoch": 1.1688509794450024,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.256410256410258e-07,
+      "loss": 1.6256,
+      "step": 183900
+    },
+    {
+      "epoch": 1.169486569972161,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.205128205128206e-07,
+      "loss": 1.6199,
+      "step": 184000
+    },
+    {
+      "epoch": 1.17012216049932,
+      "grad_norm": 1.3359375,
+      "learning_rate": 8.153846153846154e-07,
+      "loss": 1.622,
+      "step": 184100
+    },
+    {
+      "epoch": 1.1707577510264786,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.102564102564104e-07,
+      "loss": 1.6335,
+      "step": 184200
+    },
+    {
+      "epoch": 1.1713933415536375,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.051282051282052e-07,
+      "loss": 1.6162,
+      "step": 184300
+    },
+    {
+      "epoch": 1.1720289320807962,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 1.6326,
+      "step": 184400
+    },
+    {
+      "epoch": 1.172664522607955,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.948717948717949e-07,
+      "loss": 1.6317,
+      "step": 184500
+    },
+    {
+      "epoch": 1.173300113135114,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.897435897435899e-07,
+      "loss": 1.626,
+      "step": 184600
+    },
+    {
+      "epoch": 1.1739357036622726,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.846153846153847e-07,
+      "loss": 1.6298,
+      "step": 184700
+    },
+    {
+      "epoch": 1.1745712941894313,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.794871794871796e-07,
+      "loss": 1.6062,
+      "step": 184800
+    },
+    {
+      "epoch": 1.1752068847165902,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.743589743589744e-07,
+      "loss": 1.6121,
+      "step": 184900
+    },
+    {
+      "epoch": 1.175842475243749,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 1.6162,
+      "step": 185000
+    },
+    {
+      "epoch": 1.1764780657709077,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.641025641025642e-07,
+      "loss": 1.6225,
+      "step": 185100
+    },
+    {
+      "epoch": 1.1771136562980664,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.58974358974359e-07,
+      "loss": 1.6196,
+      "step": 185200
+    },
+    {
+      "epoch": 1.1777492468252253,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.538461538461538e-07,
+      "loss": 1.6323,
+      "step": 185300
+    },
+    {
+      "epoch": 1.1783848373523842,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.487179487179488e-07,
+      "loss": 1.6396,
+      "step": 185400
+    },
+    {
+      "epoch": 1.1790204278795429,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.435897435897436e-07,
+      "loss": 1.6313,
+      "step": 185500
+    },
+    {
+      "epoch": 1.1796560184067018,
+      "grad_norm": 0.75,
+      "learning_rate": 7.384615384615385e-07,
+      "loss": 1.6339,
+      "step": 185600
+    },
+    {
+      "epoch": 1.1802916089338604,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.333333333333334e-07,
+      "loss": 1.6106,
+      "step": 185700
+    },
+    {
+      "epoch": 1.1809271994610193,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.282051282051282e-07,
+      "loss": 1.6374,
+      "step": 185800
+    },
+    {
+      "epoch": 1.181562789988178,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.230769230769231e-07,
+      "loss": 1.6183,
+      "step": 185900
+    },
+    {
+      "epoch": 1.1821983805153369,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.179487179487179e-07,
+      "loss": 1.6314,
+      "step": 186000
+    },
+    {
+      "epoch": 1.1828339710424955,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.128205128205129e-07,
+      "loss": 1.6269,
+      "step": 186100
+    },
+    {
+      "epoch": 1.1834695615696544,
+      "grad_norm": 1.2890625,
+      "learning_rate": 7.076923076923077e-07,
+      "loss": 1.6146,
+      "step": 186200
+    },
+    {
+      "epoch": 1.184105152096813,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.025641025641026e-07,
+      "loss": 1.6253,
+      "step": 186300
+    },
+    {
+      "epoch": 1.184740742623972,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.974358974358974e-07,
+      "loss": 1.6206,
+      "step": 186400
+    },
+    {
+      "epoch": 1.1853763331511307,
+      "grad_norm": 1.3125,
+      "learning_rate": 6.923076923076924e-07,
+      "loss": 1.6378,
+      "step": 186500
+    },
+    {
+      "epoch": 1.1860119236782896,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.871794871794872e-07,
+      "loss": 1.6195,
+      "step": 186600
+    },
+    {
+      "epoch": 1.1866475142054482,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.820512820512821e-07,
+      "loss": 1.6105,
+      "step": 186700
+    },
+    {
+      "epoch": 1.1872831047326071,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.769230769230769e-07,
+      "loss": 1.634,
+      "step": 186800
+    },
+    {
+      "epoch": 1.1879186952597658,
+      "grad_norm": 0.86328125,
+      "learning_rate": 6.717948717948719e-07,
+      "loss": 1.6297,
+      "step": 186900
+    },
+    {
+      "epoch": 1.1885542857869247,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 1.6124,
+      "step": 187000
+    },
+    {
+      "epoch": 1.1891898763140833,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.615384615384616e-07,
+      "loss": 1.6296,
+      "step": 187100
+    },
+    {
+      "epoch": 1.1898254668412422,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.564102564102565e-07,
+      "loss": 1.6216,
+      "step": 187200
+    },
+    {
+      "epoch": 1.190461057368401,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.512820512820513e-07,
+      "loss": 1.6228,
+      "step": 187300
+    },
+    {
+      "epoch": 1.1910966478955598,
+      "grad_norm": 1.2890625,
+      "learning_rate": 6.461538461538462e-07,
+      "loss": 1.6217,
+      "step": 187400
+    },
+    {
+      "epoch": 1.1917322384227185,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 1.6235,
+      "step": 187500
+    },
+    {
+      "epoch": 1.1923678289498774,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.35897435897436e-07,
+      "loss": 1.6218,
+      "step": 187600
+    },
+    {
+      "epoch": 1.193003419477036,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.307692307692308e-07,
+      "loss": 1.6405,
+      "step": 187700
+    },
+    {
+      "epoch": 1.193639010004195,
+      "grad_norm": 1.2890625,
+      "learning_rate": 6.256410256410257e-07,
+      "loss": 1.6174,
+      "step": 187800
+    },
+    {
+      "epoch": 1.1942746005313536,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.205128205128206e-07,
+      "loss": 1.6174,
+      "step": 187900
+    },
+    {
+      "epoch": 1.1949101910585125,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.153846153846155e-07,
+      "loss": 1.6336,
+      "step": 188000
+    },
+    {
+      "epoch": 1.1955457815856712,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.102564102564103e-07,
+      "loss": 1.6303,
+      "step": 188100
+    },
+    {
+      "epoch": 1.19618137211283,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.051282051282052e-07,
+      "loss": 1.629,
+      "step": 188200
+    },
+    {
+      "epoch": 1.1968169626399887,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 1.6121,
+      "step": 188300
+    },
+    {
+      "epoch": 1.1974525531671476,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.948717948717949e-07,
+      "loss": 1.6173,
+      "step": 188400
+    },
+    {
+      "epoch": 1.1980881436943065,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.897435897435898e-07,
+      "loss": 1.6073,
+      "step": 188500
+    },
+    {
+      "epoch": 1.1987237342214652,
+      "grad_norm": 0.85546875,
+      "learning_rate": 5.846153846153847e-07,
+      "loss": 1.6165,
+      "step": 188600
+    },
+    {
+      "epoch": 1.1993593247486238,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.794871794871796e-07,
+      "loss": 1.6199,
+      "step": 188700
+    },
+    {
+      "epoch": 1.1999949152757827,
+      "grad_norm": 0.8046875,
+      "learning_rate": 5.743589743589744e-07,
+      "loss": 1.6232,
+      "step": 188800
+    },
+    {
+      "epoch": 1.2006305058029416,
+      "grad_norm": 0.86328125,
+      "learning_rate": 5.692307692307693e-07,
+      "loss": 1.6187,
+      "step": 188900
+    },
+    {
+      "epoch": 1.2012660963301003,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.641025641025642e-07,
+      "loss": 1.6207,
+      "step": 189000
+    },
+    {
+      "epoch": 1.201901686857259,
+      "grad_norm": 0.8125,
+      "learning_rate": 5.589743589743591e-07,
+      "loss": 1.613,
+      "step": 189100
+    },
+    {
+      "epoch": 1.2025372773844178,
+      "grad_norm": 0.89453125,
+      "learning_rate": 5.53846153846154e-07,
+      "loss": 1.6118,
+      "step": 189200
+    },
+    {
+      "epoch": 1.2031728679115767,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.487179487179488e-07,
+      "loss": 1.6395,
+      "step": 189300
+    },
+    {
+      "epoch": 1.2038084584387354,
+      "grad_norm": 0.82421875,
+      "learning_rate": 5.435897435897437e-07,
+      "loss": 1.6283,
+      "step": 189400
+    },
+    {
+      "epoch": 1.2044440489658943,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.384615384615386e-07,
+      "loss": 1.639,
+      "step": 189500
+    },
+    {
+      "epoch": 1.205079639493053,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.333333333333335e-07,
+      "loss": 1.6224,
+      "step": 189600
+    },
+    {
+      "epoch": 1.2057152300202119,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.282051282051282e-07,
+      "loss": 1.6401,
+      "step": 189700
+    },
+    {
+      "epoch": 1.2063508205473705,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.230769230769231e-07,
+      "loss": 1.6279,
+      "step": 189800
+    },
+    {
+      "epoch": 1.2069864110745294,
+      "grad_norm": 1.125,
+      "learning_rate": 5.179487179487179e-07,
+      "loss": 1.6324,
+      "step": 189900
+    },
+    {
+      "epoch": 1.207622001601688,
+      "grad_norm": 0.76171875,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 1.6146,
+      "step": 190000
+    },
+    {
+      "epoch": 1.208257592128847,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.076923076923077e-07,
+      "loss": 1.6171,
+      "step": 190100
+    },
+    {
+      "epoch": 1.2088931826560056,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.025641025641026e-07,
+      "loss": 1.6087,
+      "step": 190200
+    },
+    {
+      "epoch": 1.2095287731831645,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.974358974358974e-07,
+      "loss": 1.6004,
+      "step": 190300
+    },
+    {
+      "epoch": 1.2101643637103232,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.923076923076923e-07,
+      "loss": 1.6331,
+      "step": 190400
+    },
+    {
+      "epoch": 1.210799954237482,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.871794871794872e-07,
+      "loss": 1.6412,
+      "step": 190500
+    },
+    {
+      "epoch": 1.2114355447646408,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.820512820512821e-07,
+      "loss": 1.6085,
+      "step": 190600
+    },
+    {
+      "epoch": 1.2120711352917997,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.76923076923077e-07,
+      "loss": 1.6188,
+      "step": 190700
+    },
+    {
+      "epoch": 1.2127067258189583,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.7179487179487187e-07,
+      "loss": 1.6264,
+      "step": 190800
+    },
+    {
+      "epoch": 1.2133423163461172,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.666666666666667e-07,
+      "loss": 1.6233,
+      "step": 190900
+    },
+    {
+      "epoch": 1.213977906873276,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.615384615384616e-07,
+      "loss": 1.6342,
+      "step": 191000
+    },
+    {
+      "epoch": 1.2146134974004348,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.5641025641025646e-07,
+      "loss": 1.6317,
+      "step": 191100
+    },
+    {
+      "epoch": 1.2152490879275935,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.5128205128205136e-07,
+      "loss": 1.6204,
+      "step": 191200
+    },
+    {
+      "epoch": 1.2158846784547523,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.4615384615384615e-07,
+      "loss": 1.6211,
+      "step": 191300
+    },
+    {
+      "epoch": 1.216520268981911,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.4102564102564105e-07,
+      "loss": 1.6256,
+      "step": 191400
+    },
+    {
+      "epoch": 1.21715585950907,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.358974358974359e-07,
+      "loss": 1.6148,
+      "step": 191500
+    },
+    {
+      "epoch": 1.2177914500362286,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.307692307692308e-07,
+      "loss": 1.6202,
+      "step": 191600
+    },
+    {
+      "epoch": 1.2184270405633875,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.2564102564102564e-07,
+      "loss": 1.627,
+      "step": 191700
+    },
+    {
+      "epoch": 1.2190626310905461,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.2051282051282054e-07,
+      "loss": 1.6217,
+      "step": 191800
+    },
+    {
+      "epoch": 1.219698221617705,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.153846153846154e-07,
+      "loss": 1.6366,
+      "step": 191900
+    },
+    {
+      "epoch": 1.220333812144864,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.102564102564103e-07,
+      "loss": 1.6388,
+      "step": 192000
+    },
+    {
+      "epoch": 1.2209694026720226,
+      "grad_norm": 1.3984375,
+      "learning_rate": 4.051282051282052e-07,
+      "loss": 1.6081,
+      "step": 192100
+    },
+    {
+      "epoch": 1.2216049931991813,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 1.6327,
+      "step": 192200
+    },
+    {
+      "epoch": 1.2222405837263401,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.9487179487179493e-07,
+      "loss": 1.6144,
+      "step": 192300
+    },
+    {
+      "epoch": 1.222876174253499,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.897435897435898e-07,
+      "loss": 1.6078,
+      "step": 192400
+    },
+    {
+      "epoch": 1.2235117647806577,
+      "grad_norm": 1.1171875,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 1.635,
+      "step": 192500
+    },
+    {
+      "epoch": 1.2241473553078164,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.794871794871795e-07,
+      "loss": 1.6114,
+      "step": 192600
+    },
+    {
+      "epoch": 1.2247829458349753,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.743589743589744e-07,
+      "loss": 1.629,
+      "step": 192700
+    },
+    {
+      "epoch": 1.2254185363621342,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.6923076923076927e-07,
+      "loss": 1.6034,
+      "step": 192800
+    },
+    {
+      "epoch": 1.2260541268892928,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.641025641025641e-07,
+      "loss": 1.6224,
+      "step": 192900
+    },
+    {
+      "epoch": 1.2266897174164517,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.5897435897435896e-07,
+      "loss": 1.6244,
+      "step": 193000
+    },
+    {
+      "epoch": 1.2273253079436104,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.5384615384615386e-07,
+      "loss": 1.6263,
+      "step": 193100
+    },
+    {
+      "epoch": 1.2279608984707693,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.487179487179487e-07,
+      "loss": 1.6212,
+      "step": 193200
+    },
+    {
+      "epoch": 1.228596488997928,
+      "grad_norm": 1.1328125,
+      "learning_rate": 3.435897435897436e-07,
+      "loss": 1.6289,
+      "step": 193300
+    },
+    {
+      "epoch": 1.2292320795250868,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.3846153846153845e-07,
+      "loss": 1.6209,
+      "step": 193400
+    },
+    {
+      "epoch": 1.2298676700522455,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 1.6283,
+      "step": 193500
+    },
+    {
+      "epoch": 1.2305032605794044,
+      "grad_norm": 0.875,
+      "learning_rate": 3.2820512820512825e-07,
+      "loss": 1.6371,
+      "step": 193600
+    },
+    {
+      "epoch": 1.231138851106563,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.230769230769231e-07,
+      "loss": 1.6415,
+      "step": 193700
+    },
+    {
+      "epoch": 1.231774441633722,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.17948717948718e-07,
+      "loss": 1.6124,
+      "step": 193800
+    },
+    {
+      "epoch": 1.2324100321608806,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.1282051282051284e-07,
+      "loss": 1.6265,
+      "step": 193900
+    },
+    {
+      "epoch": 1.2330456226880395,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.0769230769230774e-07,
+      "loss": 1.6312,
+      "step": 194000
+    },
+    {
+      "epoch": 1.2336812132151982,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.025641025641026e-07,
+      "loss": 1.6242,
+      "step": 194100
+    },
+    {
+      "epoch": 1.234316803742357,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.9743589743589744e-07,
+      "loss": 1.6233,
+      "step": 194200
+    },
+    {
+      "epoch": 1.2349523942695158,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.9230769230769234e-07,
+      "loss": 1.6164,
+      "step": 194300
+    },
+    {
+      "epoch": 1.2355879847966746,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.871794871794872e-07,
+      "loss": 1.6302,
+      "step": 194400
+    },
+    {
+      "epoch": 1.2362235753238333,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.820512820512821e-07,
+      "loss": 1.6237,
+      "step": 194500
+    },
+    {
+      "epoch": 1.2368591658509922,
+      "grad_norm": 1.1328125,
+      "learning_rate": 2.76923076923077e-07,
+      "loss": 1.6299,
+      "step": 194600
+    },
+    {
+      "epoch": 1.2374947563781509,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.717948717948718e-07,
+      "loss": 1.6201,
+      "step": 194700
+    },
+    {
+      "epoch": 1.2381303469053098,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.666666666666667e-07,
+      "loss": 1.6312,
+      "step": 194800
+    },
+    {
+      "epoch": 1.2387659374324684,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.6153846153846157e-07,
+      "loss": 1.616,
+      "step": 194900
+    },
+    {
+      "epoch": 1.2394015279596273,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.564102564102564e-07,
+      "loss": 1.6243,
+      "step": 195000
+    },
+    {
+      "epoch": 1.240037118486786,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.512820512820513e-07,
+      "loss": 1.6187,
+      "step": 195100
+    },
+    {
+      "epoch": 1.2406727090139449,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.4615384615384616e-07,
+      "loss": 1.6108,
+      "step": 195200
+    },
+    {
+      "epoch": 1.2413082995411036,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.4102564102564106e-07,
+      "loss": 1.6293,
+      "step": 195300
+    },
+    {
+      "epoch": 1.2419438900682624,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.3589743589743593e-07,
+      "loss": 1.624,
+      "step": 195400
+    },
+    {
+      "epoch": 1.2425794805954211,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.307692307692308e-07,
+      "loss": 1.6333,
+      "step": 195500
+    },
+    {
+      "epoch": 1.24321507112258,
+      "grad_norm": 0.875,
+      "learning_rate": 2.2564102564102568e-07,
+      "loss": 1.6186,
+      "step": 195600
+    },
+    {
+      "epoch": 1.2438506616497387,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.2051282051282053e-07,
+      "loss": 1.6238,
+      "step": 195700
+    },
+    {
+      "epoch": 1.2444862521768976,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.153846153846154e-07,
+      "loss": 1.6251,
+      "step": 195800
+    },
+    {
+      "epoch": 1.2451218427040565,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.1025641025641027e-07,
+      "loss": 1.6266,
+      "step": 195900
+    },
+    {
+      "epoch": 1.2457574332312151,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.0512820512820514e-07,
+      "loss": 1.629,
+      "step": 196000
+    },
+    {
+      "epoch": 1.2463930237583738,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 1.6201,
+      "step": 196100
+    },
+    {
+      "epoch": 1.2470286142855327,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.948717948717949e-07,
+      "loss": 1.6353,
+      "step": 196200
+    },
+    {
+      "epoch": 1.2476642048126916,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.8974358974358976e-07,
+      "loss": 1.6252,
+      "step": 196300
+    },
+    {
+      "epoch": 1.2482997953398502,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.8461538461538464e-07,
+      "loss": 1.6206,
+      "step": 196400
+    },
+    {
+      "epoch": 1.248935385867009,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.7948717948717948e-07,
+      "loss": 1.628,
+      "step": 196500
+    },
+    {
+      "epoch": 1.2495709763941678,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.7435897435897435e-07,
+      "loss": 1.6288,
+      "step": 196600
+    },
+    {
+      "epoch": 1.2502065669213267,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.6923076923076923e-07,
+      "loss": 1.634,
+      "step": 196700
+    },
+    {
+      "epoch": 1.2508421574484854,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.6410256410256413e-07,
+      "loss": 1.6227,
+      "step": 196800
+    },
+    {
+      "epoch": 1.251477747975644,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.58974358974359e-07,
+      "loss": 1.6262,
+      "step": 196900
+    },
+    {
+      "epoch": 1.252113338502803,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.5384615384615387e-07,
+      "loss": 1.6395,
+      "step": 197000
+    },
+    {
+      "epoch": 1.2527489290299618,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.4871794871794872e-07,
+      "loss": 1.6084,
+      "step": 197100
+    },
+    {
+      "epoch": 1.2533845195571205,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.435897435897436e-07,
+      "loss": 1.6321,
+      "step": 197200
+    },
+    {
+      "epoch": 1.2540201100842794,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.384615384615385e-07,
+      "loss": 1.6181,
+      "step": 197300
+    },
+    {
+      "epoch": 1.254655700611438,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3333333333333336e-07,
+      "loss": 1.627,
+      "step": 197400
+    },
+    {
+      "epoch": 1.255291291138597,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 1.6061,
+      "step": 197500
+    },
+    {
+      "epoch": 1.2559268816657556,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.2307692307692308e-07,
+      "loss": 1.629,
+      "step": 197600
+    },
+    {
+      "epoch": 1.2565624721929145,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.1794871794871797e-07,
+      "loss": 1.6315,
+      "step": 197700
+    },
+    {
+      "epoch": 1.2571980627200732,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.1282051282051284e-07,
+      "loss": 1.6194,
+      "step": 197800
+    },
+    {
+      "epoch": 1.257833653247232,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.076923076923077e-07,
+      "loss": 1.6216,
+      "step": 197900
+    },
+    {
+      "epoch": 1.2584692437743907,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.0256410256410257e-07,
+      "loss": 1.6253,
+      "step": 198000
+    },
+    {
+      "epoch": 1.2591048343015496,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.743589743589745e-08,
+      "loss": 1.6114,
+      "step": 198100
+    },
+    {
+      "epoch": 1.2597404248287083,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.230769230769232e-08,
+      "loss": 1.6078,
+      "step": 198200
+    },
+    {
+      "epoch": 1.2603760153558672,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.717948717948718e-08,
+      "loss": 1.6376,
+      "step": 198300
+    },
+    {
+      "epoch": 1.2610116058830259,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.205128205128206e-08,
+      "loss": 1.6233,
+      "step": 198400
+    },
+    {
+      "epoch": 1.2616471964101847,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.692307692307694e-08,
+      "loss": 1.6106,
+      "step": 198500
+    },
+    {
+      "epoch": 1.2622827869373434,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.17948717948718e-08,
+      "loss": 1.6388,
+      "step": 198600
+    },
+    {
+      "epoch": 1.2629183774645023,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.666666666666668e-08,
+      "loss": 1.6184,
+      "step": 198700
+    },
+    {
+      "epoch": 1.263553967991661,
+      "grad_norm": 1.2734375,
+      "learning_rate": 6.153846153846154e-08,
+      "loss": 1.616,
+      "step": 198800
+    },
+    {
+      "epoch": 1.2641895585188199,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.641025641025642e-08,
+      "loss": 1.6208,
+      "step": 198900
+    },
+    {
+      "epoch": 1.2648251490459785,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.1282051282051286e-08,
+      "loss": 1.606,
+      "step": 199000
+    },
+    {
+      "epoch": 1.2654607395731374,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.615384615384616e-08,
+      "loss": 1.6127,
+      "step": 199100
+    },
+    {
+      "epoch": 1.266096330100296,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.102564102564103e-08,
+      "loss": 1.6287,
+      "step": 199200
+    },
+    {
+      "epoch": 1.266731920627455,
+      "grad_norm": 1.2421875,
+      "learning_rate": 3.58974358974359e-08,
+      "loss": 1.6168,
+      "step": 199300
+    },
+    {
+      "epoch": 1.2673675111546139,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.076923076923077e-08,
+      "loss": 1.6341,
+      "step": 199400
+    },
+    {
+      "epoch": 1.2680031016817725,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.5641025641025643e-08,
+      "loss": 1.6117,
+      "step": 199500
+    },
+    {
+      "epoch": 1.2686386922089312,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.0512820512820516e-08,
+      "loss": 1.6181,
+      "step": 199600
+    },
+    {
+      "epoch": 1.26927428273609,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.5384615384615385e-08,
+      "loss": 1.624,
+      "step": 199700
+    },
+    {
+      "epoch": 1.269909873263249,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.0256410256410258e-08,
+      "loss": 1.6318,
+      "step": 199800
+    },
+    {
+      "epoch": 1.2705454637904077,
+      "grad_norm": 0.953125,
+      "learning_rate": 5.128205128205129e-09,
+      "loss": 1.6126,
+      "step": 199900
+    },
+    {
+      "epoch": 1.2711810543175663,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0,
+      "loss": 1.6342,
+      "step": 200000
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 200000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 10000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.369174972818627e+19,
+  "train_batch_size": 3,
+  "trial_name": null,
+  "trial_params": null
+}