diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4830 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 436,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0022935779816513763,
+      "grad_norm": 0.12869106233119965,
+      "learning_rate": 0.0,
+      "loss": 0.1978,
+      "memory/device_reserved (GiB)": 50.77,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 1,
+      "tokens_per_second_per_gpu": 354.96
+    },
+    {
+      "epoch": 0.0045871559633027525,
+      "grad_norm": 0.15667210519313812,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 0.2353,
+      "memory/device_reserved (GiB)": 50.77,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 2,
+      "tokens_per_second_per_gpu": 406.37
+    },
+    {
+      "epoch": 0.006880733944954129,
+      "grad_norm": 0.2217973917722702,
+      "learning_rate": 9.523809523809523e-06,
+      "loss": 0.2243,
+      "memory/device_reserved (GiB)": 50.87,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 3,
+      "tokens_per_second_per_gpu": 371.18
+    },
+    {
+      "epoch": 0.009174311926605505,
+      "grad_norm": 0.15948686003684998,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.2392,
+      "memory/device_reserved (GiB)": 50.87,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 4,
+      "tokens_per_second_per_gpu": 414.48
+    },
+    {
+      "epoch": 0.011467889908256881,
+      "grad_norm": 0.153566375374794,
+      "learning_rate": 1.9047619047619046e-05,
+      "loss": 0.2182,
+      "memory/device_reserved (GiB)": 50.87,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 5,
+      "tokens_per_second_per_gpu": 369.22
+    },
+    {
+      "epoch": 0.013761467889908258,
+      "grad_norm": 0.1521972268819809,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.2112,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 49.04,
+      "memory/max_allocated (GiB)": 49.04,
+      "step": 6,
+      "tokens_per_second_per_gpu": 429.31
+    },
+    {
+      "epoch": 0.016055045871559634,
+      "grad_norm": 0.168710395693779,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.226,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 7,
+      "tokens_per_second_per_gpu": 417.78
+    },
+    {
+      "epoch": 0.01834862385321101,
+      "grad_norm": 0.13864850997924805,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.1884,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 8,
+      "tokens_per_second_per_gpu": 439.56
+    },
+    {
+      "epoch": 0.020642201834862386,
+      "grad_norm": 0.15227903425693512,
+      "learning_rate": 3.809523809523809e-05,
+      "loss": 0.1996,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 9,
+      "tokens_per_second_per_gpu": 411.33
+    },
+    {
+      "epoch": 0.022935779816513763,
+      "grad_norm": 0.13421630859375,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.1599,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 10,
+      "tokens_per_second_per_gpu": 496.3
+    },
+    {
+      "epoch": 0.02522935779816514,
+      "grad_norm": 0.14955134689807892,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 0.1735,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 11,
+      "tokens_per_second_per_gpu": 372.95
+    },
+    {
+      "epoch": 0.027522935779816515,
+      "grad_norm": 0.1432778388261795,
+      "learning_rate": 5.2380952380952384e-05,
+      "loss": 0.1515,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 12,
+      "tokens_per_second_per_gpu": 398.65
+    },
+    {
+      "epoch": 0.02981651376146789,
+      "grad_norm": 0.14163611829280853,
+      "learning_rate": 5.714285714285714e-05,
+      "loss": 0.1517,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 13,
+      "tokens_per_second_per_gpu": 440.5
+    },
+    {
+      "epoch": 0.03211009174311927,
+      "grad_norm": 0.15477906167507172,
+      "learning_rate": 6.19047619047619e-05,
+      "loss": 0.1444,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 14,
+      "tokens_per_second_per_gpu": 385.32
+    },
+    {
+      "epoch": 0.034403669724770644,
+      "grad_norm": 0.1055532768368721,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.1292,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 15,
+      "tokens_per_second_per_gpu": 453.02
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 0.10180933028459549,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 0.1208,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 16,
+      "tokens_per_second_per_gpu": 474.27
+    },
+    {
+      "epoch": 0.0389908256880734,
+      "grad_norm": 0.07999677956104279,
+      "learning_rate": 7.619047619047618e-05,
+      "loss": 0.132,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 17,
+      "tokens_per_second_per_gpu": 382.05
+    },
+    {
+      "epoch": 0.04128440366972477,
+      "grad_norm": 0.09194924682378769,
+      "learning_rate": 8.095238095238096e-05,
+      "loss": 0.1067,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 18,
+      "tokens_per_second_per_gpu": 398.61
+    },
+    {
+      "epoch": 0.04357798165137615,
+      "grad_norm": 0.0931428000330925,
+      "learning_rate": 8.571428571428571e-05,
+      "loss": 0.1088,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 19,
+      "tokens_per_second_per_gpu": 447.07
+    },
+    {
+      "epoch": 0.045871559633027525,
+      "grad_norm": 0.06202042102813721,
+      "learning_rate": 9.047619047619048e-05,
+      "loss": 0.0962,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 20,
+      "tokens_per_second_per_gpu": 382.57
+    },
+    {
+      "epoch": 0.0481651376146789,
+      "grad_norm": 0.04220607504248619,
+      "learning_rate": 9.523809523809524e-05,
+      "loss": 0.0963,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 21,
+      "tokens_per_second_per_gpu": 423.29
+    },
+    {
+      "epoch": 0.05045871559633028,
+      "grad_norm": 0.050066106021404266,
+      "learning_rate": 0.0001,
+      "loss": 0.1032,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 22,
+      "tokens_per_second_per_gpu": 381.35
+    },
+    {
+      "epoch": 0.052752293577981654,
+      "grad_norm": 0.0557384118437767,
+      "learning_rate": 9.999856734543933e-05,
+      "loss": 0.1025,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 23,
+      "tokens_per_second_per_gpu": 393.62
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 0.04612402245402336,
+      "learning_rate": 9.999426946385727e-05,
+      "loss": 0.0985,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 24,
+      "tokens_per_second_per_gpu": 515.46
+    },
+    {
+      "epoch": 0.05733944954128441,
+      "grad_norm": 0.09721734374761581,
+      "learning_rate": 9.998710660154898e-05,
+      "loss": 0.1062,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 25,
+      "tokens_per_second_per_gpu": 398.15
+    },
+    {
+      "epoch": 0.05963302752293578,
+      "grad_norm": 0.036745935678482056,
+      "learning_rate": 9.997707916899079e-05,
+      "loss": 0.1045,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 26,
+      "tokens_per_second_per_gpu": 422.42
+    },
+    {
+      "epoch": 0.06192660550458716,
+      "grad_norm": 0.04298936203122139,
+      "learning_rate": 9.996418774081658e-05,
+      "loss": 0.0923,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 27,
+      "tokens_per_second_per_gpu": 440.87
+    },
+    {
+      "epoch": 0.06422018348623854,
+      "grad_norm": 0.033536747097969055,
+      "learning_rate": 9.994843305578486e-05,
+      "loss": 0.096,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 28,
+      "tokens_per_second_per_gpu": 370.28
+    },
+    {
+      "epoch": 0.06651376146788991,
+      "grad_norm": 0.03256046772003174,
+      "learning_rate": 9.99298160167365e-05,
+      "loss": 0.0832,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 29,
+      "tokens_per_second_per_gpu": 357.19
+    },
+    {
+      "epoch": 0.06880733944954129,
+      "grad_norm": 0.042709868401288986,
+      "learning_rate": 9.990833769054293e-05,
+      "loss": 0.086,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 30,
+      "tokens_per_second_per_gpu": 441.89
+    },
+    {
+      "epoch": 0.07110091743119266,
+      "grad_norm": 0.04347776621580124,
+      "learning_rate": 9.988399930804504e-05,
+      "loss": 0.1,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 48.77,
+      "memory/max_allocated (GiB)": 48.77,
+      "step": 31,
+      "tokens_per_second_per_gpu": 348.66
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 0.030414681881666183,
+      "learning_rate": 9.985680226398261e-05,
+      "loss": 0.0811,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 32,
+      "tokens_per_second_per_gpu": 435.28
+    },
+    {
+      "epoch": 0.07568807339449542,
+      "grad_norm": 0.034023743122816086,
+      "learning_rate": 9.98267481169144e-05,
+      "loss": 0.0743,
+      "memory/device_reserved (GiB)": 50.93,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 33,
+      "tokens_per_second_per_gpu": 482.51
+    },
+    {
+      "epoch": 0.0779816513761468,
+      "grad_norm": 0.03136487305164337,
+      "learning_rate": 9.979383858912885e-05,
+      "loss": 0.0739,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.08,
+      "memory/max_allocated (GiB)": 49.08,
+      "step": 34,
+      "tokens_per_second_per_gpu": 496.59
+    },
+    {
+      "epoch": 0.08027522935779817,
+      "grad_norm": 0.028108298778533936,
+      "learning_rate": 9.975807556654537e-05,
+      "loss": 0.077,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 35,
+      "tokens_per_second_per_gpu": 349.1
+    },
+    {
+      "epoch": 0.08256880733944955,
+      "grad_norm": 0.028020795434713364,
+      "learning_rate": 9.971946109860626e-05,
+      "loss": 0.0775,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 36,
+      "tokens_per_second_per_gpu": 351.02
+    },
+    {
+      "epoch": 0.08486238532110092,
+      "grad_norm": 0.028756650164723396,
+      "learning_rate": 9.967799739815925e-05,
+      "loss": 0.0788,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 37,
+      "tokens_per_second_per_gpu": 534.52
+    },
+    {
+      "epoch": 0.0871559633027523,
+      "grad_norm": 0.02806459739804268,
+      "learning_rate": 9.963368684133072e-05,
+      "loss": 0.0809,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 38,
+      "tokens_per_second_per_gpu": 367.94
+    },
+    {
+      "epoch": 0.08944954128440367,
+      "grad_norm": 0.02387731708586216,
+      "learning_rate": 9.958653196738954e-05,
+      "loss": 0.0642,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.04,
+      "memory/max_allocated (GiB)": 49.04,
+      "step": 39,
+      "tokens_per_second_per_gpu": 466.74
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 0.027889851480722427,
+      "learning_rate": 9.953653547860151e-05,
+      "loss": 0.0904,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 40,
+      "tokens_per_second_per_gpu": 371.51
+    },
+    {
+      "epoch": 0.09403669724770643,
+      "grad_norm": 0.031659577041864395,
+      "learning_rate": 9.948370024007454e-05,
+      "loss": 0.081,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 41,
+      "tokens_per_second_per_gpu": 479.04
+    },
+    {
+      "epoch": 0.0963302752293578,
+      "grad_norm": 0.03186093270778656,
+      "learning_rate": 9.942802927959443e-05,
+      "loss": 0.0881,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 42,
+      "tokens_per_second_per_gpu": 364.73
+    },
+    {
+      "epoch": 0.09862385321100918,
+      "grad_norm": 0.0313677079975605,
+      "learning_rate": 9.936952578745142e-05,
+      "loss": 0.0808,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 43,
+      "tokens_per_second_per_gpu": 418.0
+    },
+    {
+      "epoch": 0.10091743119266056,
+      "grad_norm": 0.0264989472925663,
+      "learning_rate": 9.93081931162573e-05,
+      "loss": 0.0664,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 44,
+      "tokens_per_second_per_gpu": 439.24
+    },
+    {
+      "epoch": 0.10321100917431193,
+      "grad_norm": 0.026272334158420563,
+      "learning_rate": 9.92440347807533e-05,
+      "loss": 0.0683,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.04,
+      "memory/max_allocated (GiB)": 49.04,
+      "step": 45,
+      "tokens_per_second_per_gpu": 482.81
+    },
+    {
+      "epoch": 0.10550458715596331,
+      "grad_norm": 0.029066840186715126,
+      "learning_rate": 9.91770544576087e-05,
+      "loss": 0.0737,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 46,
+      "tokens_per_second_per_gpu": 389.87
+    },
+    {
+      "epoch": 0.10779816513761468,
+      "grad_norm": 0.024542706087231636,
+      "learning_rate": 9.910725598521013e-05,
+      "loss": 0.0737,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 47,
+      "tokens_per_second_per_gpu": 473.12
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 0.042941153049468994,
+      "learning_rate": 9.90346433634416e-05,
+      "loss": 0.0951,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 48,
+      "tokens_per_second_per_gpu": 325.12
+    },
+    {
+      "epoch": 0.11238532110091744,
+      "grad_norm": 0.029044413939118385,
+      "learning_rate": 9.89592207534552e-05,
+      "loss": 0.0745,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.73,
+      "memory/max_allocated (GiB)": 48.73,
+      "step": 49,
+      "tokens_per_second_per_gpu": 315.62
+    },
+    {
+      "epoch": 0.11467889908256881,
+      "grad_norm": 0.028920788317918777,
+      "learning_rate": 9.888099247743283e-05,
+      "loss": 0.0818,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 50,
+      "tokens_per_second_per_gpu": 441.3
+    },
+    {
+      "epoch": 0.11697247706422019,
+      "grad_norm": 0.026095205917954445,
+      "learning_rate": 9.879996301833833e-05,
+      "loss": 0.0688,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 51,
+      "tokens_per_second_per_gpu": 386.22
+    },
+    {
+      "epoch": 0.11926605504587157,
+      "grad_norm": 0.024823926389217377,
+      "learning_rate": 9.871613701966067e-05,
+      "loss": 0.0701,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 52,
+      "tokens_per_second_per_gpu": 511.32
+    },
+    {
+      "epoch": 0.12155963302752294,
+      "grad_norm": 0.036093298345804214,
+      "learning_rate": 9.862951928514782e-05,
+      "loss": 0.0823,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 53,
+      "tokens_per_second_per_gpu": 323.2
+    },
+    {
+      "epoch": 0.12385321100917432,
+      "grad_norm": 0.03257686272263527,
+      "learning_rate": 9.854011477853146e-05,
+      "loss": 0.0769,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.04,
+      "memory/max_allocated (GiB)": 49.04,
+      "step": 54,
+      "tokens_per_second_per_gpu": 447.62
+    },
+    {
+      "epoch": 0.12614678899082568,
+      "grad_norm": 0.03413158655166626,
+      "learning_rate": 9.844792862324258e-05,
+      "loss": 0.0728,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 55,
+      "tokens_per_second_per_gpu": 451.05
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.02947932481765747,
+      "learning_rate": 9.835296610211779e-05,
+      "loss": 0.0713,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 56,
+      "tokens_per_second_per_gpu": 457.44
+    },
+    {
+      "epoch": 0.13073394495412843,
+      "grad_norm": 0.0220651775598526,
+      "learning_rate": 9.825523265709666e-05,
+      "loss": 0.0607,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 57,
+      "tokens_per_second_per_gpu": 456.49
+    },
+    {
+      "epoch": 0.13302752293577982,
+      "grad_norm": 0.026394842192530632,
+      "learning_rate": 9.815473388890983e-05,
+      "loss": 0.0716,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 58,
+      "tokens_per_second_per_gpu": 393.95
+    },
+    {
+      "epoch": 0.1353211009174312,
+      "grad_norm": 0.027936838567256927,
+      "learning_rate": 9.805147555675805e-05,
+      "loss": 0.0738,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 59,
+      "tokens_per_second_per_gpu": 464.83
+    },
+    {
+      "epoch": 0.13761467889908258,
+      "grad_norm": 0.023982539772987366,
+      "learning_rate": 9.794546357798208e-05,
+      "loss": 0.0608,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 60,
+      "tokens_per_second_per_gpu": 450.66
+    },
+    {
+      "epoch": 0.13990825688073394,
+      "grad_norm": 0.027479754760861397,
+      "learning_rate": 9.783670402772379e-05,
+      "loss": 0.0672,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 61,
+      "tokens_per_second_per_gpu": 455.94
+    },
+    {
+      "epoch": 0.14220183486238533,
+      "grad_norm": 0.02617599070072174,
+      "learning_rate": 9.772520313857775e-05,
+      "loss": 0.0804,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 62,
+      "tokens_per_second_per_gpu": 394.85
+    },
+    {
+      "epoch": 0.1444954128440367,
+      "grad_norm": 0.030884992331266403,
+      "learning_rate": 9.761096730023432e-05,
+      "loss": 0.0768,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 63,
+      "tokens_per_second_per_gpu": 446.63
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 0.027579287067055702,
+      "learning_rate": 9.749400305911322e-05,
+      "loss": 0.0659,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 64,
+      "tokens_per_second_per_gpu": 484.34
+    },
+    {
+      "epoch": 0.14908256880733944,
+      "grad_norm": 0.030303625389933586,
+      "learning_rate": 9.737431711798864e-05,
+      "loss": 0.0645,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 65,
+      "tokens_per_second_per_gpu": 437.07
+    },
+    {
+      "epoch": 0.15137614678899083,
+      "grad_norm": 0.027446158230304718,
+      "learning_rate": 9.725191633560491e-05,
+      "loss": 0.08,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 66,
+      "tokens_per_second_per_gpu": 411.5
+    },
+    {
+      "epoch": 0.1536697247706422,
+      "grad_norm": 0.03177177160978317,
+      "learning_rate": 9.712680772628364e-05,
+      "loss": 0.0801,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 67,
+      "tokens_per_second_per_gpu": 429.18
+    },
+    {
+      "epoch": 0.1559633027522936,
+      "grad_norm": 0.0288909412920475,
+      "learning_rate": 9.69989984595216e-05,
+      "loss": 0.0707,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.04,
+      "memory/max_allocated (GiB)": 49.04,
+      "step": 68,
+      "tokens_per_second_per_gpu": 408.55
+    },
+    {
+      "epoch": 0.15825688073394495,
+      "grad_norm": 0.02751251310110092,
+      "learning_rate": 9.686849585957994e-05,
+      "loss": 0.0736,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 69,
+      "tokens_per_second_per_gpu": 420.0
+    },
+    {
+      "epoch": 0.16055045871559634,
+      "grad_norm": 0.023428168147802353,
+      "learning_rate": 9.673530740506447e-05,
+      "loss": 0.0648,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 70,
+      "tokens_per_second_per_gpu": 512.59
+    },
+    {
+      "epoch": 0.1628440366972477,
+      "grad_norm": 0.031534772366285324,
+      "learning_rate": 9.659944072849707e-05,
+      "loss": 0.0818,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 71,
+      "tokens_per_second_per_gpu": 456.9
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 0.027208171784877777,
+      "learning_rate": 9.646090361587827e-05,
+      "loss": 0.0709,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 72,
+      "tokens_per_second_per_gpu": 378.48
+    },
+    {
+      "epoch": 0.16743119266055045,
+      "grad_norm": 0.02961639314889908,
+      "learning_rate": 9.631970400624113e-05,
+      "loss": 0.0764,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 73,
+      "tokens_per_second_per_gpu": 316.38
+    },
+    {
+      "epoch": 0.16972477064220184,
+      "grad_norm": 0.027367761358618736,
+      "learning_rate": 9.617584999119625e-05,
+      "loss": 0.0672,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 74,
+      "tokens_per_second_per_gpu": 402.44
+    },
+    {
+      "epoch": 0.1720183486238532,
+      "grad_norm": 0.030167503282427788,
+      "learning_rate": 9.602934981446803e-05,
+      "loss": 0.0743,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 75,
+      "tokens_per_second_per_gpu": 531.29
+    },
+    {
+      "epoch": 0.1743119266055046,
+      "grad_norm": 0.0387263149023056,
+      "learning_rate": 9.588021187142235e-05,
+      "loss": 0.083,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 76,
+      "tokens_per_second_per_gpu": 424.59
+    },
+    {
+      "epoch": 0.17660550458715596,
+      "grad_norm": 0.027617793530225754,
+      "learning_rate": 9.572844470858537e-05,
+      "loss": 0.0769,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 77,
+      "tokens_per_second_per_gpu": 461.9
+    },
+    {
+      "epoch": 0.17889908256880735,
+      "grad_norm": 0.029771512374281883,
+      "learning_rate": 9.557405702315381e-05,
+      "loss": 0.0658,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 78,
+      "tokens_per_second_per_gpu": 475.77
+    },
+    {
+      "epoch": 0.1811926605504587,
+      "grad_norm": 0.029358675703406334,
+      "learning_rate": 9.541705766249655e-05,
+      "loss": 0.066,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 79,
+      "tokens_per_second_per_gpu": 489.33
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 0.023111771792173386,
+      "learning_rate": 9.525745562364756e-05,
+      "loss": 0.066,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 80,
+      "tokens_per_second_per_gpu": 382.84
+    },
+    {
+      "epoch": 0.18577981651376146,
+      "grad_norm": 0.029448291286826134,
+      "learning_rate": 9.509526005279044e-05,
+      "loss": 0.0608,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 81,
+      "tokens_per_second_per_gpu": 415.81
+    },
+    {
+      "epoch": 0.18807339449541285,
+      "grad_norm": 0.02794116735458374,
+      "learning_rate": 9.493048024473412e-05,
+      "loss": 0.0736,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 82,
+      "tokens_per_second_per_gpu": 400.02
+    },
+    {
+      "epoch": 0.19036697247706422,
+      "grad_norm": 0.04534873738884926,
+      "learning_rate": 9.476312564238034e-05,
+      "loss": 0.0673,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 83,
+      "tokens_per_second_per_gpu": 369.1
+    },
+    {
+      "epoch": 0.1926605504587156,
+      "grad_norm": 0.026540853083133698,
+      "learning_rate": 9.459320583618252e-05,
+      "loss": 0.0558,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.04,
+      "memory/max_allocated (GiB)": 49.04,
+      "step": 84,
+      "tokens_per_second_per_gpu": 611.61
+    },
+    {
+      "epoch": 0.19495412844036697,
+      "grad_norm": 0.03129403293132782,
+      "learning_rate": 9.442073056359604e-05,
+      "loss": 0.0741,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 85,
+      "tokens_per_second_per_gpu": 492.16
+    },
+    {
+      "epoch": 0.19724770642201836,
+      "grad_norm": 0.027526071295142174,
+      "learning_rate": 9.424570970852034e-05,
+      "loss": 0.0733,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 86,
+      "tokens_per_second_per_gpu": 427.76
+    },
+    {
+      "epoch": 0.19954128440366972,
+      "grad_norm": 0.025468798354268074,
+      "learning_rate": 9.406815330073244e-05,
+      "loss": 0.0613,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 87,
+      "tokens_per_second_per_gpu": 462.82
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 0.029043635353446007,
+      "learning_rate": 9.388807151531229e-05,
+      "loss": 0.0758,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 88,
+      "tokens_per_second_per_gpu": 353.91
+    },
+    {
+      "epoch": 0.20412844036697247,
+      "grad_norm": 0.03196391835808754,
+      "learning_rate": 9.37054746720595e-05,
+      "loss": 0.0678,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 89,
+      "tokens_per_second_per_gpu": 411.71
+    },
+    {
+      "epoch": 0.20642201834862386,
+      "grad_norm": 0.033272091299295425,
+      "learning_rate": 9.352037323490208e-05,
+      "loss": 0.0722,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 90,
+      "tokens_per_second_per_gpu": 398.81
+    },
+    {
+      "epoch": 0.20871559633027523,
+      "grad_norm": 0.03096090629696846,
+      "learning_rate": 9.333277781129678e-05,
+      "loss": 0.0809,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 91,
+      "tokens_per_second_per_gpu": 393.81
+    },
+    {
+      "epoch": 0.21100917431192662,
+      "grad_norm": 0.026267440989613533,
+      "learning_rate": 9.314269915162114e-05,
+      "loss": 0.0604,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 92,
+      "tokens_per_second_per_gpu": 453.78
+    },
+    {
+      "epoch": 0.21330275229357798,
+      "grad_norm": 0.02608361840248108,
+      "learning_rate": 9.295014814855753e-05,
+      "loss": 0.0663,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 93,
+      "tokens_per_second_per_gpu": 430.47
+    },
+    {
+      "epoch": 0.21559633027522937,
+      "grad_norm": 0.024829065427184105,
+      "learning_rate": 9.275513583646884e-05,
+      "loss": 0.0598,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 94,
+      "tokens_per_second_per_gpu": 384.01
+    },
+    {
+      "epoch": 0.21788990825688073,
+      "grad_norm": 0.03385532647371292,
+      "learning_rate": 9.255767339076622e-05,
+      "loss": 0.0719,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 95,
+      "tokens_per_second_per_gpu": 440.35
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 0.029608217999339104,
+      "learning_rate": 9.23577721272686e-05,
+      "loss": 0.094,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.04,
+      "memory/max_allocated (GiB)": 49.04,
+      "step": 96,
+      "tokens_per_second_per_gpu": 485.56
+    },
+    {
+      "epoch": 0.22247706422018348,
+      "grad_norm": 0.02693762816488743,
+      "learning_rate": 9.215544350155422e-05,
+      "loss": 0.0755,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 97,
+      "tokens_per_second_per_gpu": 432.16
+    },
+    {
+      "epoch": 0.22477064220183487,
+      "grad_norm": 0.02771424688398838,
+      "learning_rate": 9.195069910830427e-05,
+      "loss": 0.0692,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 98,
+      "tokens_per_second_per_gpu": 412.93
+    },
+    {
+      "epoch": 0.22706422018348624,
+      "grad_norm": 0.02276022732257843,
+      "learning_rate": 9.174355068063828e-05,
+      "loss": 0.0637,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 99,
+      "tokens_per_second_per_gpu": 418.24
+    },
+    {
+      "epoch": 0.22935779816513763,
+      "grad_norm": 0.026155246421694756,
+      "learning_rate": 9.15340100894418e-05,
+      "loss": 0.0698,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 100,
+      "tokens_per_second_per_gpu": 403.6
+    },
+    {
+      "epoch": 0.231651376146789,
+      "grad_norm": 0.022778436541557312,
+      "learning_rate": 9.132208934268622e-05,
+      "loss": 0.0654,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 101,
+      "tokens_per_second_per_gpu": 491.32
+    },
+    {
+      "epoch": 0.23394495412844038,
+      "grad_norm": 0.04701945558190346,
+      "learning_rate": 9.110780058474052e-05,
+      "loss": 0.0741,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 102,
+      "tokens_per_second_per_gpu": 444.03
+    },
+    {
+      "epoch": 0.23623853211009174,
+      "grad_norm": 0.030211661010980606,
+      "learning_rate": 9.08911560956753e-05,
+      "loss": 0.0789,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 103,
+      "tokens_per_second_per_gpu": 514.87
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 0.026159459725022316,
+      "learning_rate": 9.067216829055922e-05,
+      "loss": 0.0637,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 104,
+      "tokens_per_second_per_gpu": 446.47
+    },
+    {
+      "epoch": 0.2408256880733945,
+      "grad_norm": 0.02918146923184395,
+      "learning_rate": 9.045084971874738e-05,
+      "loss": 0.0727,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 105,
+      "tokens_per_second_per_gpu": 425.37
+    },
+    {
+      "epoch": 0.24311926605504589,
+      "grad_norm": 0.03170175105333328,
+      "learning_rate": 9.022721306316222e-05,
+      "loss": 0.0857,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 106,
+      "tokens_per_second_per_gpu": 301.79
+    },
+    {
+      "epoch": 0.24541284403669725,
+      "grad_norm": 0.032674651592969894,
+      "learning_rate": 9.000127113956674e-05,
+      "loss": 0.0795,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.77,
+      "memory/max_allocated (GiB)": 48.77,
+      "step": 107,
+      "tokens_per_second_per_gpu": 338.41
+    },
+    {
+      "epoch": 0.24770642201834864,
+      "grad_norm": 0.026492780074477196,
+      "learning_rate": 8.977303689583e-05,
+      "loss": 0.0775,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 108,
+      "tokens_per_second_per_gpu": 383.35
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0290480125695467,
+      "learning_rate": 8.954252341118523e-05,
+      "loss": 0.076,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 109,
+      "tokens_per_second_per_gpu": 382.78
+    },
+    {
+      "epoch": 0.25229357798165136,
+      "grad_norm": 0.030473977327346802,
+      "learning_rate": 8.930974389548023e-05,
+      "loss": 0.0761,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 110,
+      "tokens_per_second_per_gpu": 476.56
+    },
+    {
+      "epoch": 0.2545871559633027,
+      "grad_norm": 0.02930077351629734,
+      "learning_rate": 8.90747116884204e-05,
+      "loss": 0.0691,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 111,
+      "tokens_per_second_per_gpu": 441.2
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 0.02884151227772236,
+      "learning_rate": 8.883744025880428e-05,
+      "loss": 0.0806,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 112,
+      "tokens_per_second_per_gpu": 406.96
+    },
+    {
+      "epoch": 0.2591743119266055,
+      "grad_norm": 0.02618175558745861,
+      "learning_rate": 8.859794320375168e-05,
+      "loss": 0.0677,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 113,
+      "tokens_per_second_per_gpu": 430.04
+    },
+    {
+      "epoch": 0.26146788990825687,
+      "grad_norm": 0.026963548734784126,
+      "learning_rate": 8.835623424792452e-05,
+      "loss": 0.0694,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 114,
+      "tokens_per_second_per_gpu": 351.9
+    },
+    {
+      "epoch": 0.26376146788990823,
+      "grad_norm": 0.021544624119997025,
+      "learning_rate": 8.811232724274035e-05,
+      "loss": 0.0613,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 115,
+      "tokens_per_second_per_gpu": 480.22
+    },
+    {
+      "epoch": 0.26605504587155965,
+      "grad_norm": 0.03840009495615959,
+      "learning_rate": 8.786623616557847e-05,
+      "loss": 0.0723,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 116,
+      "tokens_per_second_per_gpu": 433.18
+    },
+    {
+      "epoch": 0.268348623853211,
+      "grad_norm": 0.022571468725800514,
+      "learning_rate": 8.761797511897906e-05,
+      "loss": 0.065,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 117,
+      "tokens_per_second_per_gpu": 421.92
+    },
+    {
+      "epoch": 0.2706422018348624,
+      "grad_norm": 0.02688576467335224,
+      "learning_rate": 8.736755832983497e-05,
+      "loss": 0.0772,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 118,
+      "tokens_per_second_per_gpu": 354.3
+    },
+    {
+      "epoch": 0.27293577981651373,
+      "grad_norm": 0.025858785957098007,
+      "learning_rate": 8.711500014857634e-05,
+      "loss": 0.0745,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 119,
+      "tokens_per_second_per_gpu": 365.46
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 0.02718079835176468,
+      "learning_rate": 8.686031504834843e-05,
+      "loss": 0.0759,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 120,
+      "tokens_per_second_per_gpu": 426.06
+    },
+    {
+      "epoch": 0.2775229357798165,
+      "grad_norm": 0.028197383508086205,
+      "learning_rate": 8.660351762418203e-05,
+      "loss": 0.0753,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 121,
+      "tokens_per_second_per_gpu": 483.89
+    },
+    {
+      "epoch": 0.2798165137614679,
+      "grad_norm": 0.02615584433078766,
+      "learning_rate": 8.634462259215719e-05,
+      "loss": 0.0692,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 122,
+      "tokens_per_second_per_gpu": 347.59
+    },
+    {
+      "epoch": 0.28211009174311924,
+      "grad_norm": 0.028645118698477745,
+      "learning_rate": 8.608364478855983e-05,
+      "loss": 0.0784,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 123,
+      "tokens_per_second_per_gpu": 472.02
+    },
+    {
+      "epoch": 0.28440366972477066,
+      "grad_norm": 0.03761473670601845,
+      "learning_rate": 8.58205991690316e-05,
+      "loss": 0.0663,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 124,
+      "tokens_per_second_per_gpu": 439.34
+    },
+    {
+      "epoch": 0.286697247706422,
+      "grad_norm": 0.024080324918031693,
+      "learning_rate": 8.555550080771273e-05,
+      "loss": 0.0685,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 125,
+      "tokens_per_second_per_gpu": 413.4
+    },
+    {
+      "epoch": 0.2889908256880734,
+      "grad_norm": 0.03224342688918114,
+      "learning_rate": 8.528836489637828e-05,
+      "loss": 0.0813,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.77,
+      "memory/max_allocated (GiB)": 48.77,
+      "step": 126,
+      "tokens_per_second_per_gpu": 299.66
+    },
+    {
+      "epoch": 0.29128440366972475,
+      "grad_norm": 0.02632022649049759,
+      "learning_rate": 8.501920674356754e-05,
+      "loss": 0.0649,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 127,
+      "tokens_per_second_per_gpu": 424.46
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.025439690798521042,
+      "learning_rate": 8.47480417737067e-05,
+      "loss": 0.0692,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 128,
+      "tokens_per_second_per_gpu": 443.94
+    },
+    {
+      "epoch": 0.2958715596330275,
+      "grad_norm": 0.028366245329380035,
+      "learning_rate": 8.447488552622498e-05,
+      "loss": 0.0743,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 129,
+      "tokens_per_second_per_gpu": 392.47
+    },
+    {
+      "epoch": 0.2981651376146789,
+      "grad_norm": 0.028246046975255013,
+      "learning_rate": 8.419975365466415e-05,
+      "loss": 0.0693,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.77,
+      "memory/max_allocated (GiB)": 48.77,
+      "step": 130,
+      "tokens_per_second_per_gpu": 385.79
+    },
+    {
+      "epoch": 0.30045871559633025,
+      "grad_norm": 0.029451027512550354,
+      "learning_rate": 8.392266192578143e-05,
+      "loss": 0.0731,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 131,
+      "tokens_per_second_per_gpu": 401.98
+    },
+    {
+      "epoch": 0.30275229357798167,
+      "grad_norm": 0.03156789019703865,
+      "learning_rate": 8.364362621864595e-05,
+      "loss": 0.0733,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 132,
+      "tokens_per_second_per_gpu": 406.2
+    },
+    {
+      "epoch": 0.30504587155963303,
+      "grad_norm": 0.0247171763330698,
+      "learning_rate": 8.336266252372889e-05,
+      "loss": 0.0723,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 133,
+      "tokens_per_second_per_gpu": 467.27
+    },
+    {
+      "epoch": 0.3073394495412844,
+      "grad_norm": 0.024775700643658638,
+      "learning_rate": 8.307978694198699e-05,
+      "loss": 0.0644,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 134,
+      "tokens_per_second_per_gpu": 377.14
+    },
+    {
+      "epoch": 0.30963302752293576,
+      "grad_norm": 0.025003118440508842,
+      "learning_rate": 8.279501568393994e-05,
+      "loss": 0.0684,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 135,
+      "tokens_per_second_per_gpu": 368.87
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 0.028482772409915924,
+      "learning_rate": 8.250836506874142e-05,
+      "loss": 0.0705,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 136,
+      "tokens_per_second_per_gpu": 439.79
+    },
+    {
+      "epoch": 0.31422018348623854,
+      "grad_norm": 0.02605322189629078,
+      "learning_rate": 8.221985152324385e-05,
+      "loss": 0.0638,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.04,
+      "memory/max_allocated (GiB)": 49.04,
+      "step": 137,
+      "tokens_per_second_per_gpu": 438.9
+    },
+    {
+      "epoch": 0.3165137614678899,
+      "grad_norm": 0.030314577743411064,
+      "learning_rate": 8.192949158105713e-05,
+      "loss": 0.0682,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 138,
+      "tokens_per_second_per_gpu": 355.22
+    },
+    {
+      "epoch": 0.31880733944954126,
+      "grad_norm": 0.02862844057381153,
+      "learning_rate": 8.163730188160105e-05,
+      "loss": 0.0764,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 139,
+      "tokens_per_second_per_gpu": 430.33
+    },
+    {
+      "epoch": 0.3211009174311927,
+      "grad_norm": 0.030885115265846252,
+      "learning_rate": 8.134329916915184e-05,
+      "loss": 0.0774,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 140,
+      "tokens_per_second_per_gpu": 369.87
+    },
+    {
+      "epoch": 0.32339449541284404,
+      "grad_norm": 0.025037452578544617,
+      "learning_rate": 8.104750029188257e-05,
+      "loss": 0.0695,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 141,
+      "tokens_per_second_per_gpu": 538.21
+    },
+    {
+      "epoch": 0.3256880733944954,
+      "grad_norm": 0.02607853338122368,
+      "learning_rate": 8.074992220089769e-05,
+      "loss": 0.066,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 142,
+      "tokens_per_second_per_gpu": 443.91
+    },
+    {
+      "epoch": 0.32798165137614677,
+      "grad_norm": 0.028251491487026215,
+      "learning_rate": 8.045058194926153e-05,
+      "loss": 0.0691,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 143,
+      "tokens_per_second_per_gpu": 403.07
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 0.02848455123603344,
+      "learning_rate": 8.014949669102117e-05,
+      "loss": 0.0712,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 144,
+      "tokens_per_second_per_gpu": 421.87
+    },
+    {
+      "epoch": 0.33256880733944955,
+      "grad_norm": 0.027499854564666748,
+      "learning_rate": 7.984668368022335e-05,
+      "loss": 0.071,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 145,
+      "tokens_per_second_per_gpu": 310.07
+    },
+    {
+      "epoch": 0.3348623853211009,
+      "grad_norm": 0.05668507516384125,
+      "learning_rate": 7.954216026992571e-05,
+      "loss": 0.072,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 146,
+      "tokens_per_second_per_gpu": 430.94
+    },
+    {
+      "epoch": 0.33715596330275227,
+      "grad_norm": 0.023797793313860893,
+      "learning_rate": 7.923594391120236e-05,
+      "loss": 0.0724,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 147,
+      "tokens_per_second_per_gpu": 506.38
+    },
+    {
+      "epoch": 0.3394495412844037,
+      "grad_norm": 0.03140917047858238,
+      "learning_rate": 7.892805215214381e-05,
+      "loss": 0.0707,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 148,
+      "tokens_per_second_per_gpu": 392.49
+    },
+    {
+      "epoch": 0.34174311926605505,
+      "grad_norm": 0.023651011288166046,
+      "learning_rate": 7.861850263685134e-05,
+      "loss": 0.0675,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 149,
+      "tokens_per_second_per_gpu": 468.39
+    },
+    {
+      "epoch": 0.3440366972477064,
+      "grad_norm": 0.028501421213150024,
+      "learning_rate": 7.830731310442599e-05,
+      "loss": 0.0677,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 150,
+      "tokens_per_second_per_gpu": 377.79
+    },
+    {
+      "epoch": 0.3463302752293578,
+      "grad_norm": 0.028334010392427444,
+      "learning_rate": 7.799450138795185e-05,
+      "loss": 0.0749,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 151,
+      "tokens_per_second_per_gpu": 370.82
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.029713135212659836,
+      "learning_rate": 7.768008541347423e-05,
+      "loss": 0.066,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 152,
+      "tokens_per_second_per_gpu": 403.75
+    },
+    {
+      "epoch": 0.35091743119266056,
+      "grad_norm": 0.030461538583040237,
+      "learning_rate": 7.73640831989723e-05,
+      "loss": 0.0667,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 153,
+      "tokens_per_second_per_gpu": 473.97
+    },
+    {
+      "epoch": 0.3532110091743119,
+      "grad_norm": 0.02694588340818882,
+      "learning_rate": 7.704651285332663e-05,
+      "loss": 0.0642,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 154,
+      "tokens_per_second_per_gpu": 421.0
+    },
+    {
+      "epoch": 0.3555045871559633,
+      "grad_norm": 0.025780972093343735,
+      "learning_rate": 7.672739257528134e-05,
+      "loss": 0.0727,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 155,
+      "tokens_per_second_per_gpu": 507.84
+    },
+    {
+      "epoch": 0.3577981651376147,
+      "grad_norm": 0.027480922639369965,
+      "learning_rate": 7.640674065240136e-05,
+      "loss": 0.078,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 156,
+      "tokens_per_second_per_gpu": 334.0
+    },
+    {
+      "epoch": 0.36009174311926606,
+      "grad_norm": 0.032992683351039886,
+      "learning_rate": 7.608457546002424e-05,
+      "loss": 0.0728,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 157,
+      "tokens_per_second_per_gpu": 315.95
+    },
+    {
+      "epoch": 0.3623853211009174,
+      "grad_norm": 0.029259737581014633,
+      "learning_rate": 7.576091546020725e-05,
+      "loss": 0.0721,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 158,
+      "tokens_per_second_per_gpu": 390.2
+    },
+    {
+      "epoch": 0.3646788990825688,
+      "grad_norm": 0.027205413207411766,
+      "learning_rate": 7.543577920066944e-05,
+      "loss": 0.0726,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 159,
+      "tokens_per_second_per_gpu": 459.84
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 0.028103800490498543,
+      "learning_rate": 7.510918531372857e-05,
+      "loss": 0.0723,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 160,
+      "tokens_per_second_per_gpu": 362.42
+    },
+    {
+      "epoch": 0.36926605504587157,
+      "grad_norm": 0.025422796607017517,
+      "learning_rate": 7.478115251523352e-05,
+      "loss": 0.0651,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 161,
+      "tokens_per_second_per_gpu": 409.68
+    },
+    {
+      "epoch": 0.37155963302752293,
+      "grad_norm": 0.0247375275939703,
+      "learning_rate": 7.445169960349167e-05,
+      "loss": 0.0648,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 162,
+      "tokens_per_second_per_gpu": 443.52
+    },
+    {
+      "epoch": 0.3738532110091743,
+      "grad_norm": 0.024430420249700546,
+      "learning_rate": 7.412084545819168e-05,
+      "loss": 0.0654,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 163,
+      "tokens_per_second_per_gpu": 439.86
+    },
+    {
+      "epoch": 0.3761467889908257,
+      "grad_norm": 0.02779349498450756,
+      "learning_rate": 7.378860903932159e-05,
+      "loss": 0.07,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 164,
+      "tokens_per_second_per_gpu": 387.88
+    },
+    {
+      "epoch": 0.37844036697247707,
+      "grad_norm": 0.028585737571120262,
+      "learning_rate": 7.34550093860822e-05,
+      "loss": 0.0794,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 165,
+      "tokens_per_second_per_gpu": 469.25
+    },
+    {
+      "epoch": 0.38073394495412843,
+      "grad_norm": 0.028040310367941856,
+      "learning_rate": 7.31200656157961e-05,
+      "loss": 0.0702,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 166,
+      "tokens_per_second_per_gpu": 340.82
+    },
+    {
+      "epoch": 0.3830275229357798,
+      "grad_norm": 0.030313577502965927,
+      "learning_rate": 7.278379692281208e-05,
+      "loss": 0.0694,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 167,
+      "tokens_per_second_per_gpu": 414.21
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 0.032695479691028595,
+      "learning_rate": 7.244622257740523e-05,
+      "loss": 0.0658,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 168,
+      "tokens_per_second_per_gpu": 435.84
+    },
+    {
+      "epoch": 0.3876146788990826,
+      "grad_norm": 0.02221628651022911,
+      "learning_rate": 7.210736192467256e-05,
+      "loss": 0.0596,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 169,
+      "tokens_per_second_per_gpu": 451.04
+    },
+    {
+      "epoch": 0.38990825688073394,
+      "grad_norm": 0.02417284995317459,
+      "learning_rate": 7.176723438342446e-05,
+      "loss": 0.0714,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 170,
+      "tokens_per_second_per_gpu": 444.02
+    },
+    {
+      "epoch": 0.3922018348623853,
+      "grad_norm": 0.027553344145417213,
+      "learning_rate": 7.142585944507185e-05,
+      "loss": 0.0613,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 171,
+      "tokens_per_second_per_gpu": 436.33
+    },
+    {
+      "epoch": 0.3944954128440367,
+      "grad_norm": 0.028384285047650337,
+      "learning_rate": 7.10832566725092e-05,
+      "loss": 0.0634,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 172,
+      "tokens_per_second_per_gpu": 389.66
+    },
+    {
+      "epoch": 0.3967889908256881,
+      "grad_norm": 0.024850716814398766,
+      "learning_rate": 7.073944569899354e-05,
+      "loss": 0.0717,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 173,
+      "tokens_per_second_per_gpu": 475.5
+    },
+    {
+      "epoch": 0.39908256880733944,
+      "grad_norm": 0.025330083444714546,
+      "learning_rate": 7.039444622701922e-05,
+      "loss": 0.0724,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 174,
+      "tokens_per_second_per_gpu": 383.81
+    },
+    {
+      "epoch": 0.4013761467889908,
+      "grad_norm": 0.025969544425606728,
+      "learning_rate": 7.00482780271889e-05,
+      "loss": 0.0712,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 175,
+      "tokens_per_second_per_gpu": 385.6
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 0.02731173112988472,
+      "learning_rate": 6.97009609370806e-05,
+      "loss": 0.0678,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 176,
+      "tokens_per_second_per_gpu": 430.61
+    },
+    {
+      "epoch": 0.4059633027522936,
+      "grad_norm": 0.028133299201726913,
+      "learning_rate": 6.935251486011087e-05,
+      "loss": 0.061,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 177,
+      "tokens_per_second_per_gpu": 379.64
+    },
+    {
+      "epoch": 0.40825688073394495,
+      "grad_norm": 0.02273411862552166,
+      "learning_rate": 6.900295976439413e-05,
+      "loss": 0.0604,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 178,
+      "tokens_per_second_per_gpu": 393.24
+    },
+    {
+      "epoch": 0.4105504587155963,
+      "grad_norm": 0.025121403858065605,
+      "learning_rate": 6.865231568159846e-05,
+      "loss": 0.0697,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 179,
+      "tokens_per_second_per_gpu": 453.6
+    },
+    {
+      "epoch": 0.41284403669724773,
+      "grad_norm": 0.029893774539232254,
+      "learning_rate": 6.830060270579768e-05,
+      "loss": 0.0743,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.77,
+      "memory/max_allocated (GiB)": 48.77,
+      "step": 180,
+      "tokens_per_second_per_gpu": 402.38
+    },
+    {
+      "epoch": 0.4151376146788991,
+      "grad_norm": 0.026196127757430077,
+      "learning_rate": 6.794784099231972e-05,
+      "loss": 0.0653,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 181,
+      "tokens_per_second_per_gpu": 369.19
+    },
+    {
+      "epoch": 0.41743119266055045,
+      "grad_norm": 0.03042738139629364,
+      "learning_rate": 6.759405075659166e-05,
+      "loss": 0.0654,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 182,
+      "tokens_per_second_per_gpu": 389.74
+    },
+    {
+      "epoch": 0.4197247706422018,
+      "grad_norm": 0.02454569563269615,
+      "learning_rate": 6.723925227298132e-05,
+      "loss": 0.0648,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 183,
+      "tokens_per_second_per_gpu": 383.9
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.03029336780309677,
+      "learning_rate": 6.688346587363533e-05,
+      "loss": 0.0711,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 184,
+      "tokens_per_second_per_gpu": 436.44
+    },
+    {
+      "epoch": 0.4243119266055046,
+      "grad_norm": 0.02716301940381527,
+      "learning_rate": 6.652671194731396e-05,
+      "loss": 0.0638,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 185,
+      "tokens_per_second_per_gpu": 405.73
+    },
+    {
+      "epoch": 0.42660550458715596,
+      "grad_norm": 0.030476156622171402,
+      "learning_rate": 6.616901093822283e-05,
+      "loss": 0.0742,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 186,
+      "tokens_per_second_per_gpu": 417.15
+    },
+    {
+      "epoch": 0.4288990825688073,
+      "grad_norm": 0.024246055632829666,
+      "learning_rate": 6.58103833448412e-05,
+      "loss": 0.0606,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 187,
+      "tokens_per_second_per_gpu": 418.65
+    },
+    {
+      "epoch": 0.43119266055045874,
+      "grad_norm": 0.025659549981355667,
+      "learning_rate": 6.545084971874738e-05,
+      "loss": 0.0643,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 188,
+      "tokens_per_second_per_gpu": 524.48
+    },
+    {
+      "epoch": 0.4334862385321101,
+      "grad_norm": 0.02851368486881256,
+      "learning_rate": 6.509043066344092e-05,
+      "loss": 0.0728,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 189,
+      "tokens_per_second_per_gpu": 470.95
+    },
+    {
+      "epoch": 0.43577981651376146,
+      "grad_norm": 0.03035641275346279,
+      "learning_rate": 6.472914683316195e-05,
+      "loss": 0.0797,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.04,
+      "memory/max_allocated (GiB)": 49.04,
+      "step": 190,
+      "tokens_per_second_per_gpu": 409.73
+    },
+    {
+      "epoch": 0.4380733944954128,
+      "grad_norm": 0.026916082948446274,
+      "learning_rate": 6.436701893170756e-05,
+      "loss": 0.06,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 191,
+      "tokens_per_second_per_gpu": 424.58
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 0.035412922501564026,
+      "learning_rate": 6.400406771124536e-05,
+      "loss": 0.0699,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 192,
+      "tokens_per_second_per_gpu": 372.44
+    },
+    {
+      "epoch": 0.4426605504587156,
+      "grad_norm": 0.02869465760886669,
+      "learning_rate": 6.364031397112416e-05,
+      "loss": 0.0709,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 193,
+      "tokens_per_second_per_gpu": 411.07
+    },
+    {
+      "epoch": 0.44495412844036697,
+      "grad_norm": 0.02998914197087288,
+      "learning_rate": 6.327577855668216e-05,
+      "loss": 0.0693,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 194,
+      "tokens_per_second_per_gpu": 473.81
+    },
+    {
+      "epoch": 0.44724770642201833,
+      "grad_norm": 0.029111091047525406,
+      "learning_rate": 6.291048235805234e-05,
+      "loss": 0.0789,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 195,
+      "tokens_per_second_per_gpu": 393.48
+    },
+    {
+      "epoch": 0.44954128440366975,
+      "grad_norm": 0.028819169849157333,
+      "learning_rate": 6.254444630896529e-05,
+      "loss": 0.0738,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 196,
+      "tokens_per_second_per_gpu": 339.21
+    },
+    {
+      "epoch": 0.4518348623853211,
+      "grad_norm": 0.027091829106211662,
+      "learning_rate": 6.21776913855496e-05,
+      "loss": 0.0606,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 197,
+      "tokens_per_second_per_gpu": 490.05
+    },
+    {
+      "epoch": 0.4541284403669725,
+      "grad_norm": 0.023907724767923355,
+      "learning_rate": 6.181023860512984e-05,
+      "loss": 0.0664,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 198,
+      "tokens_per_second_per_gpu": 437.98
+    },
+    {
+      "epoch": 0.45642201834862384,
+      "grad_norm": 0.026607749983668327,
+      "learning_rate": 6.144210902502207e-05,
+      "loss": 0.0686,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 199,
+      "tokens_per_second_per_gpu": 518.9
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 0.028734847903251648,
+      "learning_rate": 6.107332374132715e-05,
+      "loss": 0.0709,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 200,
+      "tokens_per_second_per_gpu": 448.6
+    },
+    {
+      "epoch": 0.4610091743119266,
+      "grad_norm": 0.027956590056419373,
+      "learning_rate": 6.0703903887721837e-05,
+      "loss": 0.0645,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 201,
+      "tokens_per_second_per_gpu": 450.75
+    },
+    {
+      "epoch": 0.463302752293578,
+      "grad_norm": 0.02955472283065319,
+      "learning_rate": 6.0333870634247645e-05,
+      "loss": 0.0749,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 202,
+      "tokens_per_second_per_gpu": 366.38
+    },
+    {
+      "epoch": 0.46559633027522934,
+      "grad_norm": 0.033545345067977905,
+      "learning_rate": 5.9963245186097725e-05,
+      "loss": 0.0714,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 203,
+      "tokens_per_second_per_gpu": 409.9
+    },
+    {
+      "epoch": 0.46788990825688076,
+      "grad_norm": 0.027358222752809525,
+      "learning_rate": 5.95920487824016e-05,
+      "loss": 0.0632,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 204,
+      "tokens_per_second_per_gpu": 409.18
+    },
+    {
+      "epoch": 0.4701834862385321,
+      "grad_norm": 0.026303566992282867,
+      "learning_rate": 5.922030269500809e-05,
+      "loss": 0.0621,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.77,
+      "memory/max_allocated (GiB)": 48.77,
+      "step": 205,
+      "tokens_per_second_per_gpu": 344.97
+    },
+    {
+      "epoch": 0.4724770642201835,
+      "grad_norm": 0.023472387343645096,
+      "learning_rate": 5.8848028227266325e-05,
+      "loss": 0.0642,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 206,
+      "tokens_per_second_per_gpu": 458.3
+    },
+    {
+      "epoch": 0.47477064220183485,
+      "grad_norm": 0.02930634468793869,
+      "learning_rate": 5.847524671280484e-05,
+      "loss": 0.07,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 207,
+      "tokens_per_second_per_gpu": 386.88
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 0.02035793662071228,
+      "learning_rate": 5.810197951430911e-05,
+      "loss": 0.0558,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 208,
+      "tokens_per_second_per_gpu": 479.37
+    },
+    {
+      "epoch": 0.4793577981651376,
+      "grad_norm": 0.027948010712862015,
+      "learning_rate": 5.772824802229733e-05,
+      "loss": 0.07,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.77,
+      "memory/max_allocated (GiB)": 48.77,
+      "step": 209,
+      "tokens_per_second_per_gpu": 352.97
+    },
+    {
+      "epoch": 0.481651376146789,
+      "grad_norm": 0.027743425220251083,
+      "learning_rate": 5.735407365389453e-05,
+      "loss": 0.0686,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 210,
+      "tokens_per_second_per_gpu": 419.65
+    },
+    {
+      "epoch": 0.48394495412844035,
+      "grad_norm": 0.03574339672923088,
+      "learning_rate": 5.697947785160532e-05,
+      "loss": 0.0593,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 211,
+      "tokens_per_second_per_gpu": 391.99
+    },
+    {
+      "epoch": 0.48623853211009177,
+      "grad_norm": 0.03303733468055725,
+      "learning_rate": 5.660448208208513e-05,
+      "loss": 0.0615,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 212,
+      "tokens_per_second_per_gpu": 420.47
+    },
+    {
+      "epoch": 0.48853211009174313,
+      "grad_norm": 0.030316850170493126,
+      "learning_rate": 5.622910783490988e-05,
+      "loss": 0.0745,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.77,
+      "memory/max_allocated (GiB)": 48.77,
+      "step": 213,
+      "tokens_per_second_per_gpu": 379.16
+    },
+    {
+      "epoch": 0.4908256880733945,
+      "grad_norm": 0.031506236642599106,
+      "learning_rate": 5.585337662134471e-05,
+      "loss": 0.0724,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 214,
+      "tokens_per_second_per_gpu": 376.6
+    },
+    {
+      "epoch": 0.49311926605504586,
+      "grad_norm": 0.025807412341237068,
+      "learning_rate": 5.5477309973111046e-05,
+      "loss": 0.0628,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 215,
+      "tokens_per_second_per_gpu": 386.77
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.02294624038040638,
+      "learning_rate": 5.510092944115286e-05,
+      "loss": 0.0629,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 216,
+      "tokens_per_second_per_gpu": 473.64
+    },
+    {
+      "epoch": 0.49770642201834864,
+      "grad_norm": 0.027048619464039803,
+      "learning_rate": 5.472425659440157e-05,
+      "loss": 0.0675,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 217,
+      "tokens_per_second_per_gpu": 374.21
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.026564767584204674,
+      "learning_rate": 5.4347313018540056e-05,
+      "loss": 0.0697,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 218,
+      "tokens_per_second_per_gpu": 442.12
+    },
+    {
+      "epoch": 0.5022935779816514,
+      "grad_norm": 0.03516434505581856,
+      "learning_rate": 5.397012031476562e-05,
+      "loss": 0.082,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.04,
+      "memory/max_allocated (GiB)": 49.04,
+      "step": 219,
+      "tokens_per_second_per_gpu": 380.84
+    },
+    {
+      "epoch": 0.5045871559633027,
+      "grad_norm": 0.021558105945587158,
+      "learning_rate": 5.359270009855216e-05,
+      "loss": 0.0585,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 220,
+      "tokens_per_second_per_gpu": 509.31
+    },
+    {
+      "epoch": 0.5068807339449541,
+      "grad_norm": 0.024724913761019707,
+      "learning_rate": 5.321507399841148e-05,
+      "loss": 0.0632,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 221,
+      "tokens_per_second_per_gpu": 438.7
+    },
+    {
+      "epoch": 0.5091743119266054,
+      "grad_norm": 0.02698579616844654,
+      "learning_rate": 5.2837263654653715e-05,
+      "loss": 0.0715,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 222,
+      "tokens_per_second_per_gpu": 337.92
+    },
+    {
+      "epoch": 0.5114678899082569,
+      "grad_norm": 0.03043169341981411,
+      "learning_rate": 5.2459290718147344e-05,
+      "loss": 0.0755,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 223,
+      "tokens_per_second_per_gpu": 485.96
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 0.026405537500977516,
+      "learning_rate": 5.2081176849078464e-05,
+      "loss": 0.0641,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 224,
+      "tokens_per_second_per_gpu": 434.97
+    },
+    {
+      "epoch": 0.5160550458715596,
+      "grad_norm": 0.024269182235002518,
+      "learning_rate": 5.170294371570939e-05,
+      "loss": 0.0666,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 225,
+      "tokens_per_second_per_gpu": 399.27
+    },
+    {
+      "epoch": 0.518348623853211,
+      "grad_norm": 0.03496242314577103,
+      "learning_rate": 5.132461299313709e-05,
+      "loss": 0.073,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.04,
+      "memory/max_allocated (GiB)": 49.04,
+      "step": 226,
+      "tokens_per_second_per_gpu": 422.84
+    },
+    {
+      "epoch": 0.5206422018348624,
+      "grad_norm": 0.029179584234952927,
+      "learning_rate": 5.094620636205095e-05,
+      "loss": 0.0697,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.04,
+      "memory/max_allocated (GiB)": 49.04,
+      "step": 227,
+      "tokens_per_second_per_gpu": 357.38
+    },
+    {
+      "epoch": 0.5229357798165137,
+      "grad_norm": 0.027006233111023903,
+      "learning_rate": 5.056774550749043e-05,
+      "loss": 0.0614,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 228,
+      "tokens_per_second_per_gpu": 316.93
+    },
+    {
+      "epoch": 0.5252293577981652,
+      "grad_norm": 0.028260482475161552,
+      "learning_rate": 5.018925211760227e-05,
+      "loss": 0.0634,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 229,
+      "tokens_per_second_per_gpu": 417.85
+    },
+    {
+      "epoch": 0.5275229357798165,
+      "grad_norm": 0.025130394846200943,
+      "learning_rate": 4.981074788239773e-05,
+      "loss": 0.0588,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 230,
+      "tokens_per_second_per_gpu": 413.46
+    },
+    {
+      "epoch": 0.5298165137614679,
+      "grad_norm": 0.025551561266183853,
+      "learning_rate": 4.943225449250958e-05,
+      "loss": 0.0688,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 231,
+      "tokens_per_second_per_gpu": 445.27
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 0.028664810582995415,
+      "learning_rate": 4.9053793637949067e-05,
+      "loss": 0.0689,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 232,
+      "tokens_per_second_per_gpu": 395.88
+    },
+    {
+      "epoch": 0.5344036697247706,
+      "grad_norm": 0.02686873823404312,
+      "learning_rate": 4.8675387006862914e-05,
+      "loss": 0.0656,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 233,
+      "tokens_per_second_per_gpu": 544.1
+    },
+    {
+      "epoch": 0.536697247706422,
+      "grad_norm": 0.03144492581486702,
+      "learning_rate": 4.829705628429061e-05,
+      "loss": 0.0795,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 234,
+      "tokens_per_second_per_gpu": 356.41
+    },
+    {
+      "epoch": 0.5389908256880734,
+      "grad_norm": 0.02188139036297798,
+      "learning_rate": 4.7918823150921555e-05,
+      "loss": 0.0611,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 235,
+      "tokens_per_second_per_gpu": 368.54
+    },
+    {
+      "epoch": 0.5412844036697247,
+      "grad_norm": 0.02784140035510063,
+      "learning_rate": 4.754070928185266e-05,
+      "loss": 0.0604,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 236,
+      "tokens_per_second_per_gpu": 445.84
+    },
+    {
+      "epoch": 0.5435779816513762,
+      "grad_norm": 0.02372545376420021,
+      "learning_rate": 4.7162736345346303e-05,
+      "loss": 0.0604,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 237,
+      "tokens_per_second_per_gpu": 467.85
+    },
+    {
+      "epoch": 0.5458715596330275,
+      "grad_norm": 0.03274843469262123,
+      "learning_rate": 4.6784926001588544e-05,
+      "loss": 0.0817,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 238,
+      "tokens_per_second_per_gpu": 438.54
+    },
+    {
+      "epoch": 0.5481651376146789,
+      "grad_norm": 0.02551015093922615,
+      "learning_rate": 4.640729990144784e-05,
+      "loss": 0.0631,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 239,
+      "tokens_per_second_per_gpu": 486.44
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 0.04315930977463722,
+      "learning_rate": 4.6029879685234395e-05,
+      "loss": 0.0661,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.04,
+      "memory/max_allocated (GiB)": 49.04,
+      "step": 240,
+      "tokens_per_second_per_gpu": 450.6
+    },
+    {
+      "epoch": 0.5527522935779816,
+      "grad_norm": 0.024066558107733727,
+      "learning_rate": 4.565268698145997e-05,
+      "loss": 0.0612,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 241,
+      "tokens_per_second_per_gpu": 462.93
+    },
+    {
+      "epoch": 0.555045871559633,
+      "grad_norm": 0.026846949011087418,
+      "learning_rate": 4.527574340559844e-05,
+      "loss": 0.0754,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 242,
+      "tokens_per_second_per_gpu": 392.01
+    },
+    {
+      "epoch": 0.5573394495412844,
+      "grad_norm": 0.02346811629831791,
+      "learning_rate": 4.4899070558847154e-05,
+      "loss": 0.0675,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 243,
+      "tokens_per_second_per_gpu": 468.19
+    },
+    {
+      "epoch": 0.5596330275229358,
+      "grad_norm": 0.02288683131337166,
+      "learning_rate": 4.452269002688897e-05,
+      "loss": 0.064,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 244,
+      "tokens_per_second_per_gpu": 306.21
+    },
+    {
+      "epoch": 0.5619266055045872,
+      "grad_norm": 0.0288680586963892,
+      "learning_rate": 4.4146623378655296e-05,
+      "loss": 0.0677,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.73,
+      "memory/max_allocated (GiB)": 48.73,
+      "step": 245,
+      "tokens_per_second_per_gpu": 325.4
+    },
+    {
+      "epoch": 0.5642201834862385,
+      "grad_norm": 0.02450747601687908,
+      "learning_rate": 4.3770892165090126e-05,
+      "loss": 0.0638,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 246,
+      "tokens_per_second_per_gpu": 401.5
+    },
+    {
+      "epoch": 0.5665137614678899,
+      "grad_norm": 0.028074199333786964,
+      "learning_rate": 4.3395517917914895e-05,
+      "loss": 0.0615,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 247,
+      "tokens_per_second_per_gpu": 537.03
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 0.02514073997735977,
+      "learning_rate": 4.3020522148394676e-05,
+      "loss": 0.0669,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 248,
+      "tokens_per_second_per_gpu": 409.93
+    },
+    {
+      "epoch": 0.5711009174311926,
+      "grad_norm": 0.029449012130498886,
+      "learning_rate": 4.2645926346105484e-05,
+      "loss": 0.0711,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 249,
+      "tokens_per_second_per_gpu": 344.6
+    },
+    {
+      "epoch": 0.573394495412844,
+      "grad_norm": 0.024152036756277084,
+      "learning_rate": 4.22717519777027e-05,
+      "loss": 0.0652,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 250,
+      "tokens_per_second_per_gpu": 417.48
+    },
+    {
+      "epoch": 0.5756880733944955,
+      "grad_norm": 0.02781221643090248,
+      "learning_rate": 4.189802048569089e-05,
+      "loss": 0.0598,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 251,
+      "tokens_per_second_per_gpu": 477.01
+    },
+    {
+      "epoch": 0.5779816513761468,
+      "grad_norm": 0.02137266844511032,
+      "learning_rate": 4.1524753287195165e-05,
+      "loss": 0.0584,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 252,
+      "tokens_per_second_per_gpu": 475.28
+    },
+    {
+      "epoch": 0.5802752293577982,
+      "grad_norm": 0.03145367652177811,
+      "learning_rate": 4.1151971772733686e-05,
+      "loss": 0.0742,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 253,
+      "tokens_per_second_per_gpu": 416.81
+    },
+    {
+      "epoch": 0.5825688073394495,
+      "grad_norm": 0.026259735226631165,
+      "learning_rate": 4.07796973049919e-05,
+      "loss": 0.0704,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 254,
+      "tokens_per_second_per_gpu": 432.14
+    },
+    {
+      "epoch": 0.5848623853211009,
+      "grad_norm": 0.029704980552196503,
+      "learning_rate": 4.04079512175984e-05,
+      "loss": 0.0751,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 255,
+      "tokens_per_second_per_gpu": 368.81
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 0.037060242146253586,
+      "learning_rate": 4.003675481390228e-05,
+      "loss": 0.081,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 256,
+      "tokens_per_second_per_gpu": 400.19
+    },
+    {
+      "epoch": 0.5894495412844036,
+      "grad_norm": 0.027513017877936363,
+      "learning_rate": 3.966612936575235e-05,
+      "loss": 0.0597,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 257,
+      "tokens_per_second_per_gpu": 381.21
+    },
+    {
+      "epoch": 0.591743119266055,
+      "grad_norm": 0.037167515605688095,
+      "learning_rate": 3.929609611227817e-05,
+      "loss": 0.0639,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 258,
+      "tokens_per_second_per_gpu": 357.36
+    },
+    {
+      "epoch": 0.5940366972477065,
+      "grad_norm": 0.0229306872934103,
+      "learning_rate": 3.8926676258672866e-05,
+      "loss": 0.0626,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 259,
+      "tokens_per_second_per_gpu": 387.68
+    },
+    {
+      "epoch": 0.5963302752293578,
+      "grad_norm": 0.027137834578752518,
+      "learning_rate": 3.855789097497794e-05,
+      "loss": 0.0711,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 260,
+      "tokens_per_second_per_gpu": 377.29
+    },
+    {
+      "epoch": 0.5986238532110092,
+      "grad_norm": 0.027339540421962738,
+      "learning_rate": 3.818976139487017e-05,
+      "loss": 0.0644,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 261,
+      "tokens_per_second_per_gpu": 476.61
+    },
+    {
+      "epoch": 0.6009174311926605,
+      "grad_norm": 0.02739766612648964,
+      "learning_rate": 3.7822308614450406e-05,
+      "loss": 0.0711,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 262,
+      "tokens_per_second_per_gpu": 426.96
+    },
+    {
+      "epoch": 0.6032110091743119,
+      "grad_norm": 0.02805398218333721,
+      "learning_rate": 3.745555369103471e-05,
+      "loss": 0.0669,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 263,
+      "tokens_per_second_per_gpu": 363.5
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 0.03466130048036575,
+      "learning_rate": 3.708951764194767e-05,
+      "loss": 0.0771,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 264,
+      "tokens_per_second_per_gpu": 383.75
+    },
+    {
+      "epoch": 0.6077981651376146,
+      "grad_norm": 0.02684733085334301,
+      "learning_rate": 3.6724221443317855e-05,
+      "loss": 0.0613,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 265,
+      "tokens_per_second_per_gpu": 545.2
+    },
+    {
+      "epoch": 0.6100917431192661,
+      "grad_norm": 0.025042880326509476,
+      "learning_rate": 3.635968602887585e-05,
+      "loss": 0.0706,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 266,
+      "tokens_per_second_per_gpu": 420.58
+    },
+    {
+      "epoch": 0.6123853211009175,
+      "grad_norm": 0.02610246278345585,
+      "learning_rate": 3.599593228875465e-05,
+      "loss": 0.0749,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 267,
+      "tokens_per_second_per_gpu": 443.84
+    },
+    {
+      "epoch": 0.6146788990825688,
+      "grad_norm": 0.02343624271452427,
+      "learning_rate": 3.563298106829244e-05,
+      "loss": 0.0676,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 268,
+      "tokens_per_second_per_gpu": 409.89
+    },
+    {
+      "epoch": 0.6169724770642202,
+      "grad_norm": 0.02438695915043354,
+      "learning_rate": 3.527085316683805e-05,
+      "loss": 0.0648,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 269,
+      "tokens_per_second_per_gpu": 475.38
+    },
+    {
+      "epoch": 0.6192660550458715,
+      "grad_norm": 0.02070113644003868,
+      "learning_rate": 3.490956933655909e-05,
+      "loss": 0.0605,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 270,
+      "tokens_per_second_per_gpu": 341.3
+    },
+    {
+      "epoch": 0.6215596330275229,
+      "grad_norm": 0.03797437623143196,
+      "learning_rate": 3.4549150281252636e-05,
+      "loss": 0.0674,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 271,
+      "tokens_per_second_per_gpu": 391.97
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.02536945417523384,
+      "learning_rate": 3.41896166551588e-05,
+      "loss": 0.0649,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 272,
+      "tokens_per_second_per_gpu": 461.4
+    },
+    {
+      "epoch": 0.6261467889908257,
+      "grad_norm": 0.032918062061071396,
+      "learning_rate": 3.383098906177719e-05,
+      "loss": 0.0769,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 273,
+      "tokens_per_second_per_gpu": 495.12
+    },
+    {
+      "epoch": 0.6284403669724771,
+      "grad_norm": 0.03230955824255943,
+      "learning_rate": 3.347328805268605e-05,
+      "loss": 0.0687,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 274,
+      "tokens_per_second_per_gpu": 355.59
+    },
+    {
+      "epoch": 0.6307339449541285,
+      "grad_norm": 0.045344047248363495,
+      "learning_rate": 3.3116534126364685e-05,
+      "loss": 0.0748,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 275,
+      "tokens_per_second_per_gpu": 339.05
+    },
+    {
+      "epoch": 0.6330275229357798,
+      "grad_norm": 0.021811284124851227,
+      "learning_rate": 3.2760747727018694e-05,
+      "loss": 0.0646,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 276,
+      "tokens_per_second_per_gpu": 334.53
+    },
+    {
+      "epoch": 0.6353211009174312,
+      "grad_norm": 0.02648971416056156,
+      "learning_rate": 3.240594924340835e-05,
+      "loss": 0.068,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 277,
+      "tokens_per_second_per_gpu": 375.76
+    },
+    {
+      "epoch": 0.6376146788990825,
+      "grad_norm": 0.022893795743584633,
+      "learning_rate": 3.205215900768029e-05,
+      "loss": 0.0627,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 278,
+      "tokens_per_second_per_gpu": 412.8
+    },
+    {
+      "epoch": 0.6399082568807339,
+      "grad_norm": 0.027191977947950363,
+      "learning_rate": 3.169939729420233e-05,
+      "loss": 0.0632,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 279,
+      "tokens_per_second_per_gpu": 408.91
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 0.023182721808552742,
+      "learning_rate": 3.1347684318401536e-05,
+      "loss": 0.0631,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 280,
+      "tokens_per_second_per_gpu": 434.99
+    },
+    {
+      "epoch": 0.6444954128440367,
+      "grad_norm": 0.03368153050541878,
+      "learning_rate": 3.099704023560587e-05,
+      "loss": 0.0762,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 281,
+      "tokens_per_second_per_gpu": 393.93
+    },
+    {
+      "epoch": 0.6467889908256881,
+      "grad_norm": 0.023287048563361168,
+      "learning_rate": 3.0647485139889145e-05,
+      "loss": 0.0629,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 282,
+      "tokens_per_second_per_gpu": 335.84
+    },
+    {
+      "epoch": 0.6490825688073395,
+      "grad_norm": 0.027626749128103256,
+      "learning_rate": 3.0299039062919416e-05,
+      "loss": 0.0631,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 283,
+      "tokens_per_second_per_gpu": 446.72
+    },
+    {
+      "epoch": 0.6513761467889908,
+      "grad_norm": 0.02671007066965103,
+      "learning_rate": 2.995172197281113e-05,
+      "loss": 0.0684,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 284,
+      "tokens_per_second_per_gpu": 419.33
+    },
+    {
+      "epoch": 0.6536697247706422,
+      "grad_norm": 0.026775743812322617,
+      "learning_rate": 2.96055537729808e-05,
+      "loss": 0.063,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 285,
+      "tokens_per_second_per_gpu": 456.33
+    },
+    {
+      "epoch": 0.6559633027522935,
+      "grad_norm": 0.024690093472599983,
+      "learning_rate": 2.926055430100647e-05,
+      "loss": 0.0601,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 286,
+      "tokens_per_second_per_gpu": 363.46
+    },
+    {
+      "epoch": 0.658256880733945,
+      "grad_norm": 0.021927161142230034,
+      "learning_rate": 2.8916743327490803e-05,
+      "loss": 0.0598,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 287,
+      "tokens_per_second_per_gpu": 395.18
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 0.029110578820109367,
+      "learning_rate": 2.8574140554928175e-05,
+      "loss": 0.0732,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 288,
+      "tokens_per_second_per_gpu": 395.82
+    },
+    {
+      "epoch": 0.6628440366972477,
+      "grad_norm": 0.025474051013588905,
+      "learning_rate": 2.8232765616575563e-05,
+      "loss": 0.0674,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 289,
+      "tokens_per_second_per_gpu": 435.52
+    },
+    {
+      "epoch": 0.6651376146788991,
+      "grad_norm": 0.02178235538303852,
+      "learning_rate": 2.789263807532746e-05,
+      "loss": 0.0616,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 290,
+      "tokens_per_second_per_gpu": 442.39
+    },
+    {
+      "epoch": 0.6674311926605505,
+      "grad_norm": 0.023412682116031647,
+      "learning_rate": 2.7553777422594774e-05,
+      "loss": 0.0673,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 291,
+      "tokens_per_second_per_gpu": 412.59
+    },
+    {
+      "epoch": 0.6697247706422018,
+      "grad_norm": 0.023469222709536552,
+      "learning_rate": 2.721620307718793e-05,
+      "loss": 0.0682,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.77,
+      "memory/max_allocated (GiB)": 48.77,
+      "step": 292,
+      "tokens_per_second_per_gpu": 276.54
+    },
+    {
+      "epoch": 0.6720183486238532,
+      "grad_norm": 0.03131282329559326,
+      "learning_rate": 2.687993438420392e-05,
+      "loss": 0.0647,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 293,
+      "tokens_per_second_per_gpu": 392.4
+    },
+    {
+      "epoch": 0.6743119266055045,
+      "grad_norm": 0.02991569973528385,
+      "learning_rate": 2.65449906139178e-05,
+      "loss": 0.0681,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 294,
+      "tokens_per_second_per_gpu": 377.34
+    },
+    {
+      "epoch": 0.676605504587156,
+      "grad_norm": 0.02651585452258587,
+      "learning_rate": 2.6211390960678413e-05,
+      "loss": 0.0802,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.77,
+      "memory/max_allocated (GiB)": 48.77,
+      "step": 295,
+      "tokens_per_second_per_gpu": 358.61
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 0.022964881733059883,
+      "learning_rate": 2.5879154541808337e-05,
+      "loss": 0.0643,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 296,
+      "tokens_per_second_per_gpu": 484.52
+    },
+    {
+      "epoch": 0.6811926605504587,
+      "grad_norm": 0.028967639431357384,
+      "learning_rate": 2.554830039650834e-05,
+      "loss": 0.0632,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 297,
+      "tokens_per_second_per_gpu": 440.4
+    },
+    {
+      "epoch": 0.6834862385321101,
+      "grad_norm": 0.02948296256363392,
+      "learning_rate": 2.5218847484766495e-05,
+      "loss": 0.0752,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 298,
+      "tokens_per_second_per_gpu": 288.6
+    },
+    {
+      "epoch": 0.6857798165137615,
+      "grad_norm": 0.03220253810286522,
+      "learning_rate": 2.4890814686271448e-05,
+      "loss": 0.0634,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 299,
+      "tokens_per_second_per_gpu": 447.81
+    },
+    {
+      "epoch": 0.6880733944954128,
+      "grad_norm": 0.028979238122701645,
+      "learning_rate": 2.456422079933056e-05,
+      "loss": 0.0689,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 300,
+      "tokens_per_second_per_gpu": 458.5
+    },
+    {
+      "epoch": 0.6903669724770642,
+      "grad_norm": 0.024549167603254318,
+      "learning_rate": 2.4239084539792745e-05,
+      "loss": 0.0593,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 301,
+      "tokens_per_second_per_gpu": 419.65
+    },
+    {
+      "epoch": 0.6926605504587156,
+      "grad_norm": 0.02671237848699093,
+      "learning_rate": 2.391542453997578e-05,
+      "loss": 0.0657,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 302,
+      "tokens_per_second_per_gpu": 368.27
+    },
+    {
+      "epoch": 0.694954128440367,
+      "grad_norm": 0.03672722727060318,
+      "learning_rate": 2.3593259347598657e-05,
+      "loss": 0.0535,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 303,
+      "tokens_per_second_per_gpu": 474.85
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 0.03666655346751213,
+      "learning_rate": 2.3272607424718675e-05,
+      "loss": 0.0646,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 304,
+      "tokens_per_second_per_gpu": 393.88
+    },
+    {
+      "epoch": 0.6995412844036697,
+      "grad_norm": 0.025117024779319763,
+      "learning_rate": 2.29534871466734e-05,
+      "loss": 0.0699,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.04,
+      "memory/max_allocated (GiB)": 49.04,
+      "step": 305,
+      "tokens_per_second_per_gpu": 449.5
+    },
+    {
+      "epoch": 0.7018348623853211,
+      "grad_norm": 0.035403817892074585,
+      "learning_rate": 2.2635916801027706e-05,
+      "loss": 0.0769,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 306,
+      "tokens_per_second_per_gpu": 420.33
+    },
+    {
+      "epoch": 0.7041284403669725,
+      "grad_norm": 0.026707297191023827,
+      "learning_rate": 2.2319914586525777e-05,
+      "loss": 0.0633,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 307,
+      "tokens_per_second_per_gpu": 451.77
+    },
+    {
+      "epoch": 0.7064220183486238,
+      "grad_norm": 0.02504413016140461,
+      "learning_rate": 2.2005498612048155e-05,
+      "loss": 0.0597,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 308,
+      "tokens_per_second_per_gpu": 357.06
+    },
+    {
+      "epoch": 0.7087155963302753,
+      "grad_norm": 0.02307130955159664,
+      "learning_rate": 2.1692686895574005e-05,
+      "loss": 0.064,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 309,
+      "tokens_per_second_per_gpu": 474.84
+    },
+    {
+      "epoch": 0.7110091743119266,
+      "grad_norm": 0.026173440739512444,
+      "learning_rate": 2.1381497363148673e-05,
+      "loss": 0.063,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 310,
+      "tokens_per_second_per_gpu": 403.88
+    },
+    {
+      "epoch": 0.713302752293578,
+      "grad_norm": 0.027350088581442833,
+      "learning_rate": 2.1071947847856222e-05,
+      "loss": 0.0674,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 311,
+      "tokens_per_second_per_gpu": 409.62
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 0.02530243620276451,
+      "learning_rate": 2.0764056088797645e-05,
+      "loss": 0.063,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 312,
+      "tokens_per_second_per_gpu": 385.83
+    },
+    {
+      "epoch": 0.7178899082568807,
+      "grad_norm": 0.028018414974212646,
+      "learning_rate": 2.045783973007429e-05,
+      "loss": 0.0634,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 313,
+      "tokens_per_second_per_gpu": 395.65
+    },
+    {
+      "epoch": 0.7201834862385321,
+      "grad_norm": 0.02613895572721958,
+      "learning_rate": 2.0153316319776662e-05,
+      "loss": 0.0653,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 314,
+      "tokens_per_second_per_gpu": 357.1
+    },
+    {
+      "epoch": 0.7224770642201835,
+      "grad_norm": 0.026048416271805763,
+      "learning_rate": 1.985050330897883e-05,
+      "loss": 0.0644,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 315,
+      "tokens_per_second_per_gpu": 395.0
+    },
+    {
+      "epoch": 0.7247706422018348,
+      "grad_norm": 0.030031291767954826,
+      "learning_rate": 1.954941805073848e-05,
+      "loss": 0.078,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 316,
+      "tokens_per_second_per_gpu": 372.79
+    },
+    {
+      "epoch": 0.7270642201834863,
+      "grad_norm": 0.029979195445775986,
+      "learning_rate": 1.9250077799102322e-05,
+      "loss": 0.0651,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 317,
+      "tokens_per_second_per_gpu": 438.54
+    },
+    {
+      "epoch": 0.7293577981651376,
+      "grad_norm": 0.025628041476011276,
+      "learning_rate": 1.8952499708117432e-05,
+      "loss": 0.0669,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 318,
+      "tokens_per_second_per_gpu": 474.63
+    },
+    {
+      "epoch": 0.731651376146789,
+      "grad_norm": 0.024868648499250412,
+      "learning_rate": 1.8656700830848174e-05,
+      "loss": 0.0656,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 319,
+      "tokens_per_second_per_gpu": 445.15
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.024810567498207092,
+      "learning_rate": 1.8362698118398967e-05,
+      "loss": 0.064,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 320,
+      "tokens_per_second_per_gpu": 383.63
+    },
+    {
+      "epoch": 0.7362385321100917,
+      "grad_norm": 0.02743346616625786,
+      "learning_rate": 1.8070508418942876e-05,
+      "loss": 0.0758,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 321,
+      "tokens_per_second_per_gpu": 386.17
+    },
+    {
+      "epoch": 0.7385321100917431,
+      "grad_norm": 0.028884073719382286,
+      "learning_rate": 1.7780148476756147e-05,
+      "loss": 0.0675,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 322,
+      "tokens_per_second_per_gpu": 498.58
+    },
+    {
+      "epoch": 0.7408256880733946,
+      "grad_norm": 0.028301537036895752,
+      "learning_rate": 1.7491634931258587e-05,
+      "loss": 0.0734,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 323,
+      "tokens_per_second_per_gpu": 392.47
+    },
+    {
+      "epoch": 0.7431192660550459,
+      "grad_norm": 0.02405114285647869,
+      "learning_rate": 1.7204984316060063e-05,
+      "loss": 0.0538,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 324,
+      "tokens_per_second_per_gpu": 409.24
+    },
+    {
+      "epoch": 0.7454128440366973,
+      "grad_norm": 0.029399245977401733,
+      "learning_rate": 1.6920213058013022e-05,
+      "loss": 0.0693,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.08,
+      "memory/max_allocated (GiB)": 49.08,
+      "step": 325,
+      "tokens_per_second_per_gpu": 461.77
+    },
+    {
+      "epoch": 0.7477064220183486,
+      "grad_norm": 0.02802177332341671,
+      "learning_rate": 1.6637337476271124e-05,
+      "loss": 0.0647,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 326,
+      "tokens_per_second_per_gpu": 389.28
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.024391207844018936,
+      "learning_rate": 1.6356373781354058e-05,
+      "loss": 0.066,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 327,
+      "tokens_per_second_per_gpu": 376.51
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 0.02589585818350315,
+      "learning_rate": 1.6077338074218596e-05,
+      "loss": 0.0676,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 328,
+      "tokens_per_second_per_gpu": 422.1
+    },
+    {
+      "epoch": 0.7545871559633027,
+      "grad_norm": 0.022877002134919167,
+      "learning_rate": 1.580024634533587e-05,
+      "loss": 0.0653,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 329,
+      "tokens_per_second_per_gpu": 440.68
+    },
+    {
+      "epoch": 0.7568807339449541,
+      "grad_norm": 0.029319310560822487,
+      "learning_rate": 1.5525114473775014e-05,
+      "loss": 0.0871,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 330,
+      "tokens_per_second_per_gpu": 435.63
+    },
+    {
+      "epoch": 0.7591743119266054,
+      "grad_norm": 0.03219328075647354,
+      "learning_rate": 1.5251958226293306e-05,
+      "loss": 0.0801,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 331,
+      "tokens_per_second_per_gpu": 363.82
+    },
+    {
+      "epoch": 0.7614678899082569,
+      "grad_norm": 0.024657782167196274,
+      "learning_rate": 1.4980793256432474e-05,
+      "loss": 0.0622,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 332,
+      "tokens_per_second_per_gpu": 342.94
+    },
+    {
+      "epoch": 0.7637614678899083,
+      "grad_norm": 0.03142733871936798,
+      "learning_rate": 1.4711635103621719e-05,
+      "loss": 0.0681,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 333,
+      "tokens_per_second_per_gpu": 404.61
+    },
+    {
+      "epoch": 0.7660550458715596,
+      "grad_norm": 0.026000676676630974,
+      "learning_rate": 1.4444499192287275e-05,
+      "loss": 0.065,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 334,
+      "tokens_per_second_per_gpu": 367.91
+    },
+    {
+      "epoch": 0.768348623853211,
+      "grad_norm": 0.03227536380290985,
+      "learning_rate": 1.4179400830968415e-05,
+      "loss": 0.0767,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 335,
+      "tokens_per_second_per_gpu": 314.19
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 0.025221284478902817,
+      "learning_rate": 1.3916355211440164e-05,
+      "loss": 0.0645,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 336,
+      "tokens_per_second_per_gpu": 362.08
+    },
+    {
+      "epoch": 0.7729357798165137,
+      "grad_norm": 0.030213654041290283,
+      "learning_rate": 1.3655377407842812e-05,
+      "loss": 0.066,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 337,
+      "tokens_per_second_per_gpu": 466.08
+    },
+    {
+      "epoch": 0.7752293577981652,
+      "grad_norm": 0.026164716109633446,
+      "learning_rate": 1.3396482375817975e-05,
+      "loss": 0.0656,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 338,
+      "tokens_per_second_per_gpu": 458.34
+    },
+    {
+      "epoch": 0.7775229357798165,
+      "grad_norm": 0.0265730619430542,
+      "learning_rate": 1.3139684951651588e-05,
+      "loss": 0.0636,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 339,
+      "tokens_per_second_per_gpu": 399.93
+    },
+    {
+      "epoch": 0.7798165137614679,
+      "grad_norm": 0.026285763829946518,
+      "learning_rate": 1.2884999851423673e-05,
+      "loss": 0.0682,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 340,
+      "tokens_per_second_per_gpu": 421.4
+    },
+    {
+      "epoch": 0.7821100917431193,
+      "grad_norm": 0.023802319541573524,
+      "learning_rate": 1.2632441670165056e-05,
+      "loss": 0.0641,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.08,
+      "memory/max_allocated (GiB)": 49.08,
+      "step": 341,
+      "tokens_per_second_per_gpu": 439.55
+    },
+    {
+      "epoch": 0.7844036697247706,
+      "grad_norm": 0.024973031133413315,
+      "learning_rate": 1.2382024881020937e-05,
+      "loss": 0.0615,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 342,
+      "tokens_per_second_per_gpu": 492.26
+    },
+    {
+      "epoch": 0.786697247706422,
+      "grad_norm": 0.029818380251526833,
+      "learning_rate": 1.213376383442153e-05,
+      "loss": 0.0746,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 343,
+      "tokens_per_second_per_gpu": 394.15
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 0.028851691633462906,
+      "learning_rate": 1.188767275725966e-05,
+      "loss": 0.0744,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 344,
+      "tokens_per_second_per_gpu": 422.33
+    },
+    {
+      "epoch": 0.7912844036697247,
+      "grad_norm": 0.03523954004049301,
+      "learning_rate": 1.164376575207547e-05,
+      "loss": 0.077,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 345,
+      "tokens_per_second_per_gpu": 286.3
+    },
+    {
+      "epoch": 0.7935779816513762,
+      "grad_norm": 0.023627813905477524,
+      "learning_rate": 1.140205679624834e-05,
+      "loss": 0.0641,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 346,
+      "tokens_per_second_per_gpu": 351.44
+    },
+    {
+      "epoch": 0.7958715596330275,
+      "grad_norm": 0.026164906099438667,
+      "learning_rate": 1.1162559741195733e-05,
+      "loss": 0.0658,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 347,
+      "tokens_per_second_per_gpu": 389.38
+    },
+    {
+      "epoch": 0.7981651376146789,
+      "grad_norm": 0.023336883634328842,
+      "learning_rate": 1.092528831157959e-05,
+      "loss": 0.062,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 348,
+      "tokens_per_second_per_gpu": 472.15
+    },
+    {
+      "epoch": 0.8004587155963303,
+      "grad_norm": 0.02306864783167839,
+      "learning_rate": 1.0690256104519764e-05,
+      "loss": 0.0629,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 349,
+      "tokens_per_second_per_gpu": 422.0
+    },
+    {
+      "epoch": 0.8027522935779816,
+      "grad_norm": 0.026163572445511818,
+      "learning_rate": 1.0457476588814774e-05,
+      "loss": 0.0667,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.77,
+      "memory/max_allocated (GiB)": 48.77,
+      "step": 350,
+      "tokens_per_second_per_gpu": 389.59
+    },
+    {
+      "epoch": 0.805045871559633,
+      "grad_norm": 0.024867909029126167,
+      "learning_rate": 1.0226963104170002e-05,
+      "loss": 0.0674,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 351,
+      "tokens_per_second_per_gpu": 429.91
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 0.023188138380646706,
+      "learning_rate": 9.998728860433276e-06,
+      "loss": 0.0645,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 352,
+      "tokens_per_second_per_gpu": 388.46
+    },
+    {
+      "epoch": 0.8096330275229358,
+      "grad_norm": 0.03035775013267994,
+      "learning_rate": 9.772786936837785e-06,
+      "loss": 0.0707,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 353,
+      "tokens_per_second_per_gpu": 397.77
+    },
+    {
+      "epoch": 0.8119266055045872,
+      "grad_norm": 0.04821021109819412,
+      "learning_rate": 9.549150281252633e-06,
+      "loss": 0.0646,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 354,
+      "tokens_per_second_per_gpu": 445.4
+    },
+    {
+      "epoch": 0.8142201834862385,
+      "grad_norm": 0.030557144433259964,
+      "learning_rate": 9.327831709440792e-06,
+      "loss": 0.0659,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 355,
+      "tokens_per_second_per_gpu": 382.05
+    },
+    {
+      "epoch": 0.8165137614678899,
+      "grad_norm": 0.02662436105310917,
+      "learning_rate": 9.108843904324715e-06,
+      "loss": 0.0626,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 356,
+      "tokens_per_second_per_gpu": 412.62
+    },
+    {
+      "epoch": 0.8188073394495413,
+      "grad_norm": 0.027914568781852722,
+      "learning_rate": 8.8921994152595e-06,
+      "loss": 0.0681,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 357,
+      "tokens_per_second_per_gpu": 297.66
+    },
+    {
+      "epoch": 0.8211009174311926,
+      "grad_norm": 0.027242561802268028,
+      "learning_rate": 8.677910657313782e-06,
+      "loss": 0.067,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 358,
+      "tokens_per_second_per_gpu": 457.91
+    },
+    {
+      "epoch": 0.823394495412844,
+      "grad_norm": 0.030475802719593048,
+      "learning_rate": 8.465989910558209e-06,
+      "loss": 0.0689,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 359,
+      "tokens_per_second_per_gpu": 368.2
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.028360676020383835,
+      "learning_rate": 8.256449319361748e-06,
+      "loss": 0.0687,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 360,
+      "tokens_per_second_per_gpu": 384.64
+    },
+    {
+      "epoch": 0.8279816513761468,
+      "grad_norm": 0.031053343787789345,
+      "learning_rate": 8.049300891695744e-06,
+      "loss": 0.0754,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 361,
+      "tokens_per_second_per_gpu": 320.92
+    },
+    {
+      "epoch": 0.8302752293577982,
+      "grad_norm": 0.030271202325820923,
+      "learning_rate": 7.844556498445788e-06,
+      "loss": 0.072,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 362,
+      "tokens_per_second_per_gpu": 437.12
+    },
+    {
+      "epoch": 0.8325688073394495,
+      "grad_norm": 0.027202172204852104,
+      "learning_rate": 7.642227872731417e-06,
+      "loss": 0.0696,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 363,
+      "tokens_per_second_per_gpu": 332.31
+    },
+    {
+      "epoch": 0.8348623853211009,
+      "grad_norm": 0.02677847445011139,
+      "learning_rate": 7.4423266092337855e-06,
+      "loss": 0.0703,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 364,
+      "tokens_per_second_per_gpu": 364.39
+    },
+    {
+      "epoch": 0.8371559633027523,
+      "grad_norm": 0.0259072408080101,
+      "learning_rate": 7.244864163531162e-06,
+      "loss": 0.0678,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.77,
+      "memory/max_allocated (GiB)": 48.77,
+      "step": 365,
+      "tokens_per_second_per_gpu": 367.02
+    },
+    {
+      "epoch": 0.8394495412844036,
+      "grad_norm": 0.02673807553946972,
+      "learning_rate": 7.049851851442468e-06,
+      "loss": 0.0661,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 366,
+      "tokens_per_second_per_gpu": 475.35
+    },
+    {
+      "epoch": 0.841743119266055,
+      "grad_norm": 0.027974814176559448,
+      "learning_rate": 6.857300848378856e-06,
+      "loss": 0.0747,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 367,
+      "tokens_per_second_per_gpu": 409.23
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 0.022259563207626343,
+      "learning_rate": 6.667222188703226e-06,
+      "loss": 0.064,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 368,
+      "tokens_per_second_per_gpu": 440.59
+    },
+    {
+      "epoch": 0.8463302752293578,
+      "grad_norm": 0.02939799055457115,
+      "learning_rate": 6.479626765097918e-06,
+      "loss": 0.0693,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 369,
+      "tokens_per_second_per_gpu": 455.83
+    },
+    {
+      "epoch": 0.8486238532110092,
+      "grad_norm": 0.029195845127105713,
+      "learning_rate": 6.294525327940515e-06,
+      "loss": 0.0711,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 370,
+      "tokens_per_second_per_gpu": 394.89
+    },
+    {
+      "epoch": 0.8509174311926605,
+      "grad_norm": 0.0236493106931448,
+      "learning_rate": 6.111928484687723e-06,
+      "loss": 0.0643,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 371,
+      "tokens_per_second_per_gpu": 408.68
+    },
+    {
+      "epoch": 0.8532110091743119,
+      "grad_norm": 0.02727104350924492,
+      "learning_rate": 5.931846699267557e-06,
+      "loss": 0.067,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 372,
+      "tokens_per_second_per_gpu": 509.27
+    },
+    {
+      "epoch": 0.8555045871559633,
+      "grad_norm": 0.034410908818244934,
+      "learning_rate": 5.7542902914796745e-06,
+      "loss": 0.0624,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 373,
+      "tokens_per_second_per_gpu": 556.2
+    },
+    {
+      "epoch": 0.8577981651376146,
+      "grad_norm": 0.0287538543343544,
+      "learning_rate": 5.579269436403967e-06,
+      "loss": 0.0651,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.08,
+      "memory/max_allocated (GiB)": 49.08,
+      "step": 374,
+      "tokens_per_second_per_gpu": 381.61
+    },
+    {
+      "epoch": 0.8600917431192661,
+      "grad_norm": 0.02870243228971958,
+      "learning_rate": 5.4067941638174806e-06,
+      "loss": 0.0731,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 375,
+      "tokens_per_second_per_gpu": 361.13
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.026416806504130363,
+      "learning_rate": 5.2368743576196536e-06,
+      "loss": 0.064,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 376,
+      "tokens_per_second_per_gpu": 326.97
+    },
+    {
+      "epoch": 0.8646788990825688,
+      "grad_norm": 0.023003704845905304,
+      "learning_rate": 5.0695197552659e-06,
+      "loss": 0.0625,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 377,
+      "tokens_per_second_per_gpu": 438.76
+    },
+    {
+      "epoch": 0.8669724770642202,
+      "grad_norm": 0.037476420402526855,
+      "learning_rate": 4.9047399472095746e-06,
+      "loss": 0.0697,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 378,
+      "tokens_per_second_per_gpu": 345.81
+    },
+    {
+      "epoch": 0.8692660550458715,
+      "grad_norm": 0.02971925400197506,
+      "learning_rate": 4.742544376352443e-06,
+      "loss": 0.0663,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 379,
+      "tokens_per_second_per_gpu": 436.62
+    },
+    {
+      "epoch": 0.8715596330275229,
+      "grad_norm": 0.023713113740086555,
+      "learning_rate": 4.582942337503465e-06,
+      "loss": 0.0602,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 380,
+      "tokens_per_second_per_gpu": 448.46
+    },
+    {
+      "epoch": 0.8738532110091743,
+      "grad_norm": 0.02941006049513817,
+      "learning_rate": 4.425942976846187e-06,
+      "loss": 0.0725,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 381,
+      "tokens_per_second_per_gpu": 329.17
+    },
+    {
+      "epoch": 0.8761467889908257,
+      "grad_norm": 0.028299743309617043,
+      "learning_rate": 4.271555291414636e-06,
+      "loss": 0.072,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 382,
+      "tokens_per_second_per_gpu": 340.56
+    },
+    {
+      "epoch": 0.8784403669724771,
+      "grad_norm": 0.03180241584777832,
+      "learning_rate": 4.119788128577667e-06,
+      "loss": 0.0766,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 383,
+      "tokens_per_second_per_gpu": 446.49
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.026926379650831223,
+      "learning_rate": 3.9706501855319765e-06,
+      "loss": 0.0683,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 384,
+      "tokens_per_second_per_gpu": 440.01
+    },
+    {
+      "epoch": 0.8830275229357798,
+      "grad_norm": 0.03347824513912201,
+      "learning_rate": 3.824150008803767e-06,
+      "loss": 0.0751,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 385,
+      "tokens_per_second_per_gpu": 343.0
+    },
+    {
+      "epoch": 0.8853211009174312,
+      "grad_norm": 0.030953101813793182,
+      "learning_rate": 3.680295993758881e-06,
+      "loss": 0.0689,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 386,
+      "tokens_per_second_per_gpu": 393.96
+    },
+    {
+      "epoch": 0.8876146788990825,
+      "grad_norm": 0.032475098967552185,
+      "learning_rate": 3.539096384121743e-06,
+      "loss": 0.0828,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 387,
+      "tokens_per_second_per_gpu": 378.9
+    },
+    {
+      "epoch": 0.8899082568807339,
+      "grad_norm": 0.02490062825381756,
+      "learning_rate": 3.40055927150294e-06,
+      "loss": 0.0623,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 388,
+      "tokens_per_second_per_gpu": 408.26
+    },
+    {
+      "epoch": 0.8922018348623854,
+      "grad_norm": 0.02600006014108658,
+      "learning_rate": 3.2646925949355312e-06,
+      "loss": 0.0658,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 389,
+      "tokens_per_second_per_gpu": 446.25
+    },
+    {
+      "epoch": 0.8944954128440367,
+      "grad_norm": 0.024244820699095726,
+      "learning_rate": 3.1315041404200663e-06,
+      "loss": 0.0655,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 390,
+      "tokens_per_second_per_gpu": 420.39
+    },
+    {
+      "epoch": 0.8967889908256881,
+      "grad_norm": 0.0253219585865736,
+      "learning_rate": 3.00100154047841e-06,
+      "loss": 0.0674,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 391,
+      "tokens_per_second_per_gpu": 463.53
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.027757421135902405,
+      "learning_rate": 2.8731922737163685e-06,
+      "loss": 0.0681,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 392,
+      "tokens_per_second_per_gpu": 472.68
+    },
+    {
+      "epoch": 0.9013761467889908,
+      "grad_norm": 0.02381259575486183,
+      "learning_rate": 2.7480836643950956e-06,
+      "loss": 0.0596,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.08,
+      "memory/max_allocated (GiB)": 49.08,
+      "step": 393,
+      "tokens_per_second_per_gpu": 452.37
+    },
+    {
+      "epoch": 0.9036697247706422,
+      "grad_norm": 0.024906722828745842,
+      "learning_rate": 2.6256828820113766e-06,
+      "loss": 0.0669,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 394,
+      "tokens_per_second_per_gpu": 327.71
+    },
+    {
+      "epoch": 0.9059633027522935,
+      "grad_norm": 0.025515113025903702,
+      "learning_rate": 2.5059969408867843e-06,
+      "loss": 0.0636,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 395,
+      "tokens_per_second_per_gpu": 409.3
+    },
+    {
+      "epoch": 0.908256880733945,
+      "grad_norm": 0.026188403367996216,
+      "learning_rate": 2.3890326997656975e-06,
+      "loss": 0.0688,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 396,
+      "tokens_per_second_per_gpu": 371.81
+    },
+    {
+      "epoch": 0.9105504587155964,
+      "grad_norm": 0.027840575203299522,
+      "learning_rate": 2.274796861422246e-06,
+      "loss": 0.0737,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 397,
+      "tokens_per_second_per_gpu": 447.44
+    },
+    {
+      "epoch": 0.9128440366972477,
+      "grad_norm": 0.0268483255058527,
+      "learning_rate": 2.163295972276219e-06,
+      "loss": 0.0583,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 398,
+      "tokens_per_second_per_gpu": 383.02
+    },
+    {
+      "epoch": 0.9151376146788991,
+      "grad_norm": 0.027824856340885162,
+      "learning_rate": 2.054536422017922e-06,
+      "loss": 0.0767,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 399,
+      "tokens_per_second_per_gpu": 331.96
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 0.024313461035490036,
+      "learning_rate": 1.9485244432419667e-06,
+      "loss": 0.0694,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 400,
+      "tokens_per_second_per_gpu": 371.17
+    },
+    {
+      "epoch": 0.9197247706422018,
+      "grad_norm": 0.02038564346730709,
+      "learning_rate": 1.8452661110901715e-06,
+      "loss": 0.0563,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 401,
+      "tokens_per_second_per_gpu": 474.9
+    },
+    {
+      "epoch": 0.9220183486238532,
+      "grad_norm": 0.030249858275055885,
+      "learning_rate": 1.7447673429033362e-06,
+      "loss": 0.0685,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 402,
+      "tokens_per_second_per_gpu": 324.48
+    },
+    {
+      "epoch": 0.9243119266055045,
+      "grad_norm": 0.027523530647158623,
+      "learning_rate": 1.6470338978822108e-06,
+      "loss": 0.0666,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 403,
+      "tokens_per_second_per_gpu": 405.82
+    },
+    {
+      "epoch": 0.926605504587156,
+      "grad_norm": 0.026385333389043808,
+      "learning_rate": 1.5520713767574246e-06,
+      "loss": 0.0768,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 404,
+      "tokens_per_second_per_gpu": 380.7
+    },
+    {
+      "epoch": 0.9288990825688074,
+      "grad_norm": 0.02548050880432129,
+      "learning_rate": 1.4598852214685488e-06,
+      "loss": 0.0649,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 405,
+      "tokens_per_second_per_gpu": 421.77
+    },
+    {
+      "epoch": 0.9311926605504587,
+      "grad_norm": 0.0276033915579319,
+      "learning_rate": 1.3704807148521903e-06,
+      "loss": 0.0722,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 406,
+      "tokens_per_second_per_gpu": 391.2
+    },
+    {
+      "epoch": 0.9334862385321101,
+      "grad_norm": 0.025824090465903282,
+      "learning_rate": 1.2838629803393342e-06,
+      "loss": 0.0658,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 407,
+      "tokens_per_second_per_gpu": 363.71
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 0.032180044800043106,
+      "learning_rate": 1.2000369816616674e-06,
+      "loss": 0.0677,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 408,
+      "tokens_per_second_per_gpu": 490.46
+    },
+    {
+      "epoch": 0.9380733944954128,
+      "grad_norm": 0.03195993974804878,
+      "learning_rate": 1.119007522567167e-06,
+      "loss": 0.08,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 409,
+      "tokens_per_second_per_gpu": 443.35
+    },
+    {
+      "epoch": 0.9403669724770642,
+      "grad_norm": 0.024462653324007988,
+      "learning_rate": 1.0407792465447986e-06,
+      "loss": 0.0589,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 410,
+      "tokens_per_second_per_gpu": 511.33
+    },
+    {
+      "epoch": 0.9426605504587156,
+      "grad_norm": 0.02783488854765892,
+      "learning_rate": 9.653566365584176e-07,
+      "loss": 0.0705,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 411,
+      "tokens_per_second_per_gpu": 407.62
+    },
+    {
+      "epoch": 0.944954128440367,
+      "grad_norm": 0.03449428081512451,
+      "learning_rate": 8.927440147898702e-07,
+      "loss": 0.0801,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 412,
+      "tokens_per_second_per_gpu": 306.83
+    },
+    {
+      "epoch": 0.9472477064220184,
+      "grad_norm": 0.027761735022068024,
+      "learning_rate": 8.229455423913013e-07,
+      "loss": 0.0749,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.77,
+      "memory/max_allocated (GiB)": 48.77,
+      "step": 413,
+      "tokens_per_second_per_gpu": 327.64
+    },
+    {
+      "epoch": 0.9495412844036697,
+      "grad_norm": 0.029755057767033577,
+      "learning_rate": 7.559652192467126e-07,
+      "loss": 0.0778,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 414,
+      "tokens_per_second_per_gpu": 384.79
+    },
+    {
+      "epoch": 0.9518348623853211,
+      "grad_norm": 0.028378870338201523,
+      "learning_rate": 6.918068837427128e-07,
+      "loss": 0.0672,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 415,
+      "tokens_per_second_per_gpu": 406.2
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 0.02773345075547695,
+      "learning_rate": 6.304742125485874e-07,
+      "loss": 0.06,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 416,
+      "tokens_per_second_per_gpu": 387.18
+    },
+    {
+      "epoch": 0.9564220183486238,
+      "grad_norm": 0.0268245879560709,
+      "learning_rate": 5.719707204055735e-07,
+      "loss": 0.0621,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 417,
+      "tokens_per_second_per_gpu": 411.81
+    },
+    {
+      "epoch": 0.9587155963302753,
+      "grad_norm": 0.033236313611269,
+      "learning_rate": 5.162997599254704e-07,
+      "loss": 0.0578,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.08,
+      "memory/max_allocated (GiB)": 49.08,
+      "step": 418,
+      "tokens_per_second_per_gpu": 471.32
+    },
+    {
+      "epoch": 0.9610091743119266,
+      "grad_norm": 0.022961758077144623,
+      "learning_rate": 4.634645213984934e-07,
+      "loss": 0.0643,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 419,
+      "tokens_per_second_per_gpu": 436.57
+    },
+    {
+      "epoch": 0.963302752293578,
+      "grad_norm": 0.028307458385825157,
+      "learning_rate": 4.134680326104645e-07,
+      "loss": 0.0691,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 420,
+      "tokens_per_second_per_gpu": 492.18
+    },
+    {
+      "epoch": 0.9655963302752294,
+      "grad_norm": 0.026976363733410835,
+      "learning_rate": 3.663131586692792e-07,
+      "loss": 0.0655,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 421,
+      "tokens_per_second_per_gpu": 327.53
+    },
+    {
+      "epoch": 0.9678899082568807,
+      "grad_norm": 0.024504244327545166,
+      "learning_rate": 3.2200260184075406e-07,
+      "loss": 0.0658,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 422,
+      "tokens_per_second_per_gpu": 419.83
+    },
+    {
+      "epoch": 0.9701834862385321,
+      "grad_norm": 0.023533035069704056,
+      "learning_rate": 2.805389013937454e-07,
+      "loss": 0.0556,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 423,
+      "tokens_per_second_per_gpu": 401.57
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 0.022774042561650276,
+      "learning_rate": 2.419244334546267e-07,
+      "loss": 0.0581,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.73,
+      "memory/max_allocated (GiB)": 48.73,
+      "step": 424,
+      "tokens_per_second_per_gpu": 329.45
+    },
+    {
+      "epoch": 0.9747706422018348,
+      "grad_norm": 0.03273961320519447,
+      "learning_rate": 2.061614108711474e-07,
+      "loss": 0.0824,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 425,
+      "tokens_per_second_per_gpu": 373.34
+    },
+    {
+      "epoch": 0.9770642201834863,
+      "grad_norm": 0.02143704518675804,
+      "learning_rate": 1.732518830856067e-07,
+      "loss": 0.0588,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 426,
+      "tokens_per_second_per_gpu": 431.87
+    },
+    {
+      "epoch": 0.9793577981651376,
+      "grad_norm": 0.026173925027251244,
+      "learning_rate": 1.431977360173975e-07,
+      "loss": 0.0678,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 427,
+      "tokens_per_second_per_gpu": 439.15
+    },
+    {
+      "epoch": 0.981651376146789,
+      "grad_norm": 0.026415711268782616,
+      "learning_rate": 1.16000691954965e-07,
+      "loss": 0.067,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.0,
+      "memory/max_allocated (GiB)": 49.0,
+      "step": 428,
+      "tokens_per_second_per_gpu": 480.36
+    },
+    {
+      "epoch": 0.9839449541284404,
+      "grad_norm": 0.025120330974459648,
+      "learning_rate": 9.1662309457069e-08,
+      "loss": 0.0644,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.97,
+      "memory/max_allocated (GiB)": 48.97,
+      "step": 429,
+      "tokens_per_second_per_gpu": 414.88
+    },
+    {
+      "epoch": 0.9862385321100917,
+      "grad_norm": 0.023429665714502335,
+      "learning_rate": 7.018398326350539e-08,
+      "loss": 0.0645,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.85,
+      "memory/max_allocated (GiB)": 48.85,
+      "step": 430,
+      "tokens_per_second_per_gpu": 330.04
+    },
+    {
+      "epoch": 0.9885321100917431,
+      "grad_norm": 0.03130911663174629,
+      "learning_rate": 5.15669442151423e-08,
+      "loss": 0.0723,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.81,
+      "memory/max_allocated (GiB)": 48.81,
+      "step": 431,
+      "tokens_per_second_per_gpu": 269.97
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 0.026494460180401802,
+      "learning_rate": 3.581225918342646e-08,
+      "loss": 0.0685,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.93,
+      "memory/max_allocated (GiB)": 48.93,
+      "step": 432,
+      "tokens_per_second_per_gpu": 387.46
+    },
+    {
+      "epoch": 0.9931192660550459,
+      "grad_norm": 0.032140735536813736,
+      "learning_rate": 2.292083100920994e-08,
+      "loss": 0.0631,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 433,
+      "tokens_per_second_per_gpu": 427.4
+    },
+    {
+      "epoch": 0.9954128440366973,
+      "grad_norm": 0.025799578055739403,
+      "learning_rate": 1.2893398451024886e-08,
+      "loss": 0.0695,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 49.04,
+      "memory/max_allocated (GiB)": 49.04,
+      "step": 434,
+      "tokens_per_second_per_gpu": 461.05
+    },
+    {
+      "epoch": 0.9977064220183486,
+      "grad_norm": 0.031855881214141846,
+      "learning_rate": 5.730536142745102e-09,
+      "loss": 0.0818,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 435,
+      "tokens_per_second_per_gpu": 415.34
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.026343608275055885,
+      "learning_rate": 1.432654560679092e-09,
+      "loss": 0.0674,
+      "memory/device_reserved (GiB)": 50.97,
+      "memory/max_active (GiB)": 48.89,
+      "memory/max_allocated (GiB)": 48.89,
+      "step": 436,
+      "tokens_per_second_per_gpu": 342.89
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 436,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 60,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.337198826144924e+18,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}