diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,19886 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2836,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003527259099226208,
+      "grad_norm": 9.0625,
+      "learning_rate": 0.0,
+      "loss": 1.5425,
+      "step": 1
+    },
+    {
+      "epoch": 0.0007054518198452416,
+      "grad_norm": 6.5625,
+      "learning_rate": 1.4084507042253522e-07,
+      "loss": 1.3258,
+      "step": 2
+    },
+    {
+      "epoch": 0.0010581777297678622,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.8169014084507043e-07,
+      "loss": 1.4303,
+      "step": 3
+    },
+    {
+      "epoch": 0.0014109036396904831,
+      "grad_norm": 7.46875,
+      "learning_rate": 4.225352112676057e-07,
+      "loss": 1.4012,
+      "step": 4
+    },
+    {
+      "epoch": 0.0017636295496131038,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.633802816901409e-07,
+      "loss": 1.4349,
+      "step": 5
+    },
+    {
+      "epoch": 0.0021163554595357245,
+      "grad_norm": 7.5,
+      "learning_rate": 7.042253521126762e-07,
+      "loss": 1.5161,
+      "step": 6
+    },
+    {
+      "epoch": 0.002469081369458345,
+      "grad_norm": 7.5,
+      "learning_rate": 8.450704225352114e-07,
+      "loss": 1.5384,
+      "step": 7
+    },
+    {
+      "epoch": 0.0028218072793809662,
+      "grad_norm": 7.1875,
+      "learning_rate": 9.859154929577465e-07,
+      "loss": 1.4748,
+      "step": 8
+    },
+    {
+      "epoch": 0.003174533189303587,
+      "grad_norm": 7.875,
+      "learning_rate": 1.1267605633802817e-06,
+      "loss": 1.4806,
+      "step": 9
+    },
+    {
+      "epoch": 0.0035272590992262076,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.267605633802817e-06,
+      "loss": 1.3946,
+      "step": 10
+    },
+    {
+      "epoch": 0.0038799850091488283,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.4084507042253523e-06,
+      "loss": 1.3772,
+      "step": 11
+    },
+    {
+      "epoch": 0.004232710919071449,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.5492957746478873e-06,
+      "loss": 1.6112,
+      "step": 12
+    },
+    {
+      "epoch": 0.00458543682899407,
+      "grad_norm": 6.21875,
+      "learning_rate": 1.6901408450704227e-06,
+      "loss": 1.4249,
+      "step": 13
+    },
+    {
+      "epoch": 0.00493816273891669,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.8309859154929579e-06,
+      "loss": 1.4448,
+      "step": 14
+    },
+    {
+      "epoch": 0.005290888648839311,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.971830985915493e-06,
+      "loss": 1.3779,
+      "step": 15
+    },
+    {
+      "epoch": 0.0056436145587619325,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.1126760563380285e-06,
+      "loss": 1.4188,
+      "step": 16
+    },
+    {
+      "epoch": 0.005996340468684553,
+      "grad_norm": 5.5625,
+      "learning_rate": 2.2535211267605635e-06,
+      "loss": 1.3251,
+      "step": 17
+    },
+    {
+      "epoch": 0.006349066378607174,
+      "grad_norm": 5.4375,
+      "learning_rate": 2.3943661971830984e-06,
+      "loss": 1.4313,
+      "step": 18
+    },
+    {
+      "epoch": 0.0067017922885297945,
+      "grad_norm": 5.21875,
+      "learning_rate": 2.535211267605634e-06,
+      "loss": 1.3798,
+      "step": 19
+    },
+    {
+      "epoch": 0.007054518198452415,
+      "grad_norm": 4.59375,
+      "learning_rate": 2.676056338028169e-06,
+      "loss": 1.4177,
+      "step": 20
+    },
+    {
+      "epoch": 0.007407244108375036,
+      "grad_norm": 4.71875,
+      "learning_rate": 2.8169014084507046e-06,
+      "loss": 1.4336,
+      "step": 21
+    },
+    {
+      "epoch": 0.0077599700182976565,
+      "grad_norm": 4.21875,
+      "learning_rate": 2.9577464788732396e-06,
+      "loss": 1.4378,
+      "step": 22
+    },
+    {
+      "epoch": 0.008112695928220278,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.0985915492957746e-06,
+      "loss": 1.4078,
+      "step": 23
+    },
+    {
+      "epoch": 0.008465421838142898,
+      "grad_norm": 4.1875,
+      "learning_rate": 3.2394366197183104e-06,
+      "loss": 1.405,
+      "step": 24
+    },
+    {
+      "epoch": 0.00881814774806552,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.3802816901408454e-06,
+      "loss": 1.3539,
+      "step": 25
+    },
+    {
+      "epoch": 0.00917087365798814,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.5211267605633804e-06,
+      "loss": 1.3613,
+      "step": 26
+    },
+    {
+      "epoch": 0.00952359956791076,
+      "grad_norm": 3.453125,
+      "learning_rate": 3.6619718309859158e-06,
+      "loss": 1.474,
+      "step": 27
+    },
+    {
+      "epoch": 0.00987632547783338,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.8028169014084508e-06,
+      "loss": 1.2334,
+      "step": 28
+    },
+    {
+      "epoch": 0.010229051387756002,
+      "grad_norm": 3.203125,
+      "learning_rate": 3.943661971830986e-06,
+      "loss": 1.4626,
+      "step": 29
+    },
+    {
+      "epoch": 0.010581777297678622,
+      "grad_norm": 2.75,
+      "learning_rate": 4.0845070422535216e-06,
+      "loss": 1.3544,
+      "step": 30
+    },
+    {
+      "epoch": 0.010934503207601243,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.225352112676057e-06,
+      "loss": 1.2513,
+      "step": 31
+    },
+    {
+      "epoch": 0.011287229117523865,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.3661971830985915e-06,
+      "loss": 1.3558,
+      "step": 32
+    },
+    {
+      "epoch": 0.011639955027446485,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.507042253521127e-06,
+      "loss": 1.4187,
+      "step": 33
+    },
+    {
+      "epoch": 0.011992680937369106,
+      "grad_norm": 1.8203125,
+      "learning_rate": 4.647887323943662e-06,
+      "loss": 1.263,
+      "step": 34
+    },
+    {
+      "epoch": 0.012345406847291726,
+      "grad_norm": 1.8046875,
+      "learning_rate": 4.788732394366197e-06,
+      "loss": 1.288,
+      "step": 35
+    },
+    {
+      "epoch": 0.012698132757214348,
+      "grad_norm": 1.7421875,
+      "learning_rate": 4.929577464788733e-06,
+      "loss": 1.3913,
+      "step": 36
+    },
+    {
+      "epoch": 0.013050858667136967,
+      "grad_norm": 1.4375,
+      "learning_rate": 5.070422535211268e-06,
+      "loss": 1.2193,
+      "step": 37
+    },
+    {
+      "epoch": 0.013403584577059589,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.211267605633803e-06,
+      "loss": 1.2279,
+      "step": 38
+    },
+    {
+      "epoch": 0.013756310486982209,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.352112676056338e-06,
+      "loss": 1.2713,
+      "step": 39
+    },
+    {
+      "epoch": 0.01410903639690483,
+      "grad_norm": 1.40625,
+      "learning_rate": 5.492957746478874e-06,
+      "loss": 1.2482,
+      "step": 40
+    },
+    {
+      "epoch": 0.01446176230682745,
+      "grad_norm": 1.25,
+      "learning_rate": 5.633802816901409e-06,
+      "loss": 1.2335,
+      "step": 41
+    },
+    {
+      "epoch": 0.014814488216750072,
+      "grad_norm": 1.28125,
+      "learning_rate": 5.774647887323944e-06,
+      "loss": 1.2969,
+      "step": 42
+    },
+    {
+      "epoch": 0.015167214126672693,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.915492957746479e-06,
+      "loss": 1.2158,
+      "step": 43
+    },
+    {
+      "epoch": 0.015519940036595313,
+      "grad_norm": 1.3046875,
+      "learning_rate": 6.056338028169015e-06,
+      "loss": 1.2714,
+      "step": 44
+    },
+    {
+      "epoch": 0.015872665946517935,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.197183098591549e-06,
+      "loss": 1.2197,
+      "step": 45
+    },
+    {
+      "epoch": 0.016225391856440556,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.3380281690140855e-06,
+      "loss": 1.2832,
+      "step": 46
+    },
+    {
+      "epoch": 0.016578117766363174,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.478873239436621e-06,
+      "loss": 1.128,
+      "step": 47
+    },
+    {
+      "epoch": 0.016930843676285796,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.619718309859155e-06,
+      "loss": 1.2838,
+      "step": 48
+    },
+    {
+      "epoch": 0.017283569586208417,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.760563380281691e-06,
+      "loss": 1.3424,
+      "step": 49
+    },
+    {
+      "epoch": 0.01763629549613104,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.901408450704225e-06,
+      "loss": 1.235,
+      "step": 50
+    },
+    {
+      "epoch": 0.017989021406053657,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.042253521126761e-06,
+      "loss": 1.1872,
+      "step": 51
+    },
+    {
+      "epoch": 0.01834174731597628,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.183098591549297e-06,
+      "loss": 1.1716,
+      "step": 52
+    },
+    {
+      "epoch": 0.0186944732258989,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.3239436619718316e-06,
+      "loss": 1.2438,
+      "step": 53
+    },
+    {
+      "epoch": 0.01904719913582152,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.464788732394367e-06,
+      "loss": 1.2019,
+      "step": 54
+    },
+    {
+      "epoch": 0.019399925045744143,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.6056338028169015e-06,
+      "loss": 1.2172,
+      "step": 55
+    },
+    {
+      "epoch": 0.01975265095566676,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.746478873239436e-06,
+      "loss": 1.2326,
+      "step": 56
+    },
+    {
+      "epoch": 0.020105376865589383,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.887323943661972e-06,
+      "loss": 1.1734,
+      "step": 57
+    },
+    {
+      "epoch": 0.020458102775512004,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.028169014084509e-06,
+      "loss": 1.1533,
+      "step": 58
+    },
+    {
+      "epoch": 0.020810828685434626,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.169014084507043e-06,
+      "loss": 1.1012,
+      "step": 59
+    },
+    {
+      "epoch": 0.021163554595357244,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.309859154929578e-06,
+      "loss": 1.0607,
+      "step": 60
+    },
+    {
+      "epoch": 0.021516280505279865,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.450704225352114e-06,
+      "loss": 1.2002,
+      "step": 61
+    },
+    {
+      "epoch": 0.021869006415202487,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.591549295774648e-06,
+      "loss": 1.1043,
+      "step": 62
+    },
+    {
+      "epoch": 0.02222173232512511,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.732394366197183e-06,
+      "loss": 1.1364,
+      "step": 63
+    },
+    {
+      "epoch": 0.02257445823504773,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.87323943661972e-06,
+      "loss": 1.1508,
+      "step": 64
+    },
+    {
+      "epoch": 0.022927184144970348,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.014084507042254e-06,
+      "loss": 1.1009,
+      "step": 65
+    },
+    {
+      "epoch": 0.02327991005489297,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.15492957746479e-06,
+      "loss": 1.2577,
+      "step": 66
+    },
+    {
+      "epoch": 0.02363263596481559,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.295774647887325e-06,
+      "loss": 1.1656,
+      "step": 67
+    },
+    {
+      "epoch": 0.023985361874738213,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.43661971830986e-06,
+      "loss": 1.1696,
+      "step": 68
+    },
+    {
+      "epoch": 0.02433808778466083,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.577464788732394e-06,
+      "loss": 1.1377,
+      "step": 69
+    },
+    {
+      "epoch": 0.024690813694583452,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.71830985915493e-06,
+      "loss": 1.1267,
+      "step": 70
+    },
+    {
+      "epoch": 0.025043539604506074,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.859154929577466e-06,
+      "loss": 1.0336,
+      "step": 71
+    },
+    {
+      "epoch": 0.025396265514428695,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1e-05,
+      "loss": 1.1695,
+      "step": 72
+    },
+    {
+      "epoch": 0.025748991424351313,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0140845070422535e-05,
+      "loss": 1.2087,
+      "step": 73
+    },
+    {
+      "epoch": 0.026101717334273935,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.0281690140845072e-05,
+      "loss": 1.1042,
+      "step": 74
+    },
+    {
+      "epoch": 0.026454443244196556,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0422535211267606e-05,
+      "loss": 1.1087,
+      "step": 75
+    },
+    {
+      "epoch": 0.026807169154119178,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.056338028169014e-05,
+      "loss": 1.0927,
+      "step": 76
+    },
+    {
+      "epoch": 0.0271598950640418,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.0704225352112675e-05,
+      "loss": 1.0465,
+      "step": 77
+    },
+    {
+      "epoch": 0.027512620973964418,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.0845070422535213e-05,
+      "loss": 1.1633,
+      "step": 78
+    },
+    {
+      "epoch": 0.02786534688388704,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.0985915492957748e-05,
+      "loss": 1.1555,
+      "step": 79
+    },
+    {
+      "epoch": 0.02821807279380966,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1126760563380284e-05,
+      "loss": 1.17,
+      "step": 80
+    },
+    {
+      "epoch": 0.028570798703732282,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1267605633802819e-05,
+      "loss": 1.131,
+      "step": 81
+    },
+    {
+      "epoch": 0.0289235246136549,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1408450704225353e-05,
+      "loss": 1.1565,
+      "step": 82
+    },
+    {
+      "epoch": 0.029276250523577522,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.1549295774647888e-05,
+      "loss": 1.1044,
+      "step": 83
+    },
+    {
+      "epoch": 0.029628976433500143,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1690140845070424e-05,
+      "loss": 1.1791,
+      "step": 84
+    },
+    {
+      "epoch": 0.029981702343422765,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.1830985915492958e-05,
+      "loss": 1.158,
+      "step": 85
+    },
+    {
+      "epoch": 0.030334428253345386,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1971830985915493e-05,
+      "loss": 1.0884,
+      "step": 86
+    },
+    {
+      "epoch": 0.030687154163268004,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.211267605633803e-05,
+      "loss": 1.1371,
+      "step": 87
+    },
+    {
+      "epoch": 0.031039880073190626,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2253521126760564e-05,
+      "loss": 1.066,
+      "step": 88
+    },
+    {
+      "epoch": 0.03139260598311325,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.2394366197183098e-05,
+      "loss": 1.1076,
+      "step": 89
+    },
+    {
+      "epoch": 0.03174533189303587,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.2535211267605636e-05,
+      "loss": 1.2457,
+      "step": 90
+    },
+    {
+      "epoch": 0.03209805780295849,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2676056338028171e-05,
+      "loss": 1.1313,
+      "step": 91
+    },
+    {
+      "epoch": 0.03245078371288111,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.2816901408450705e-05,
+      "loss": 1.1282,
+      "step": 92
+    },
+    {
+      "epoch": 0.03280350962280373,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.2957746478873242e-05,
+      "loss": 1.105,
+      "step": 93
+    },
+    {
+      "epoch": 0.03315623553272635,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3098591549295776e-05,
+      "loss": 1.1364,
+      "step": 94
+    },
+    {
+      "epoch": 0.03350896144264897,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.323943661971831e-05,
+      "loss": 1.1215,
+      "step": 95
+    },
+    {
+      "epoch": 0.03386168735257159,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3380281690140845e-05,
+      "loss": 1.0771,
+      "step": 96
+    },
+    {
+      "epoch": 0.03421441326249421,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.3521126760563382e-05,
+      "loss": 1.2036,
+      "step": 97
+    },
+    {
+      "epoch": 0.034567139172416834,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.3661971830985916e-05,
+      "loss": 0.9837,
+      "step": 98
+    },
+    {
+      "epoch": 0.034919865082339456,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.380281690140845e-05,
+      "loss": 1.1185,
+      "step": 99
+    },
+    {
+      "epoch": 0.03527259099226208,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.3943661971830987e-05,
+      "loss": 1.0752,
+      "step": 100
+    },
+    {
+      "epoch": 0.0356253169021847,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.4084507042253522e-05,
+      "loss": 1.0353,
+      "step": 101
+    },
+    {
+      "epoch": 0.035978042812107314,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.4225352112676058e-05,
+      "loss": 0.991,
+      "step": 102
+    },
+    {
+      "epoch": 0.036330768722029935,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4366197183098594e-05,
+      "loss": 1.0691,
+      "step": 103
+    },
+    {
+      "epoch": 0.03668349463195256,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4507042253521129e-05,
+      "loss": 1.031,
+      "step": 104
+    },
+    {
+      "epoch": 0.03703622054187518,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4647887323943663e-05,
+      "loss": 1.2358,
+      "step": 105
+    },
+    {
+      "epoch": 0.0373889464517978,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.47887323943662e-05,
+      "loss": 1.2409,
+      "step": 106
+    },
+    {
+      "epoch": 0.03774167236172042,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4929577464788734e-05,
+      "loss": 1.0894,
+      "step": 107
+    },
+    {
+      "epoch": 0.03809439827164304,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5070422535211269e-05,
+      "loss": 1.0806,
+      "step": 108
+    },
+    {
+      "epoch": 0.038447124181565664,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5211267605633803e-05,
+      "loss": 1.1366,
+      "step": 109
+    },
+    {
+      "epoch": 0.038799850091488286,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.535211267605634e-05,
+      "loss": 1.0159,
+      "step": 110
+    },
+    {
+      "epoch": 0.0391525760014109,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5492957746478872e-05,
+      "loss": 1.1443,
+      "step": 111
+    },
+    {
+      "epoch": 0.03950530191133352,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.563380281690141e-05,
+      "loss": 1.0168,
+      "step": 112
+    },
+    {
+      "epoch": 0.039858027821256144,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5774647887323945e-05,
+      "loss": 0.9803,
+      "step": 113
+    },
+    {
+      "epoch": 0.040210753731178765,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.591549295774648e-05,
+      "loss": 1.0476,
+      "step": 114
+    },
+    {
+      "epoch": 0.04056347964110139,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.6056338028169017e-05,
+      "loss": 1.1826,
+      "step": 115
+    },
+    {
+      "epoch": 0.04091620555102401,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.619718309859155e-05,
+      "loss": 1.208,
+      "step": 116
+    },
+    {
+      "epoch": 0.04126893146094663,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.6338028169014086e-05,
+      "loss": 1.0559,
+      "step": 117
+    },
+    {
+      "epoch": 0.04162165737086925,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.6478873239436623e-05,
+      "loss": 1.1268,
+      "step": 118
+    },
+    {
+      "epoch": 0.04197438328079187,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.6619718309859155e-05,
+      "loss": 1.1456,
+      "step": 119
+    },
+    {
+      "epoch": 0.04232710919071449,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.676056338028169e-05,
+      "loss": 1.0405,
+      "step": 120
+    },
+    {
+      "epoch": 0.04267983510063711,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6901408450704228e-05,
+      "loss": 1.1343,
+      "step": 121
+    },
+    {
+      "epoch": 0.04303256101055973,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.704225352112676e-05,
+      "loss": 1.0845,
+      "step": 122
+    },
+    {
+      "epoch": 0.04338528692048235,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7183098591549297e-05,
+      "loss": 1.0866,
+      "step": 123
+    },
+    {
+      "epoch": 0.043738012830404974,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.732394366197183e-05,
+      "loss": 1.0625,
+      "step": 124
+    },
+    {
+      "epoch": 0.044090738740327595,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.7464788732394366e-05,
+      "loss": 1.0321,
+      "step": 125
+    },
+    {
+      "epoch": 0.04444346465025022,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.7605633802816902e-05,
+      "loss": 1.2201,
+      "step": 126
+    },
+    {
+      "epoch": 0.04479619056017284,
+      "grad_norm": 0.75,
+      "learning_rate": 1.774647887323944e-05,
+      "loss": 0.9987,
+      "step": 127
+    },
+    {
+      "epoch": 0.04514891647009546,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.7887323943661975e-05,
+      "loss": 0.9943,
+      "step": 128
+    },
+    {
+      "epoch": 0.045501642380018074,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.8028169014084508e-05,
+      "loss": 0.995,
+      "step": 129
+    },
+    {
+      "epoch": 0.045854368289940696,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.8169014084507044e-05,
+      "loss": 1.07,
+      "step": 130
+    },
+    {
+      "epoch": 0.04620709419986332,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.830985915492958e-05,
+      "loss": 1.1025,
+      "step": 131
+    },
+    {
+      "epoch": 0.04655982010978594,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.8450704225352113e-05,
+      "loss": 1.0071,
+      "step": 132
+    },
+    {
+      "epoch": 0.04691254601970856,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.859154929577465e-05,
+      "loss": 0.9805,
+      "step": 133
+    },
+    {
+      "epoch": 0.04726527192963118,
+      "grad_norm": 4.875,
+      "learning_rate": 1.8732394366197186e-05,
+      "loss": 1.0181,
+      "step": 134
+    },
+    {
+      "epoch": 0.047617997839553804,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.887323943661972e-05,
+      "loss": 1.0583,
+      "step": 135
+    },
+    {
+      "epoch": 0.047970723749476425,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9014084507042255e-05,
+      "loss": 1.081,
+      "step": 136
+    },
+    {
+      "epoch": 0.04832344965939904,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9154929577464788e-05,
+      "loss": 1.0621,
+      "step": 137
+    },
+    {
+      "epoch": 0.04867617556932166,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9295774647887327e-05,
+      "loss": 0.9868,
+      "step": 138
+    },
+    {
+      "epoch": 0.04902890147924428,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.943661971830986e-05,
+      "loss": 1.1053,
+      "step": 139
+    },
+    {
+      "epoch": 0.049381627389166904,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9577464788732396e-05,
+      "loss": 1.0616,
+      "step": 140
+    },
+    {
+      "epoch": 0.049734353299089526,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.9718309859154933e-05,
+      "loss": 1.0527,
+      "step": 141
+    },
+    {
+      "epoch": 0.05008707920901215,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9859154929577465e-05,
+      "loss": 1.1594,
+      "step": 142
+    },
+    {
+      "epoch": 0.05043980511893477,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2e-05,
+      "loss": 1.0565,
+      "step": 143
+    },
+    {
+      "epoch": 0.05079253102885739,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9999993200538522e-05,
+      "loss": 1.1768,
+      "step": 144
+    },
+    {
+      "epoch": 0.05114525693878001,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.999997280216333e-05,
+      "loss": 1.1254,
+      "step": 145
+    },
+    {
+      "epoch": 0.05149798284870263,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9999938804902167e-05,
+      "loss": 1.1456,
+      "step": 146
+    },
+    {
+      "epoch": 0.05185070875862525,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9999891208801262e-05,
+      "loss": 1.1024,
+      "step": 147
+    },
+    {
+      "epoch": 0.05220343466854787,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9999830013925347e-05,
+      "loss": 1.0523,
+      "step": 148
+    },
+    {
+      "epoch": 0.05255616057847049,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.999975522035763e-05,
+      "loss": 1.0825,
+      "step": 149
+    },
+    {
+      "epoch": 0.05290888648839311,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.999966682819983e-05,
+      "loss": 1.028,
+      "step": 150
+    },
+    {
+      "epoch": 0.053261612398315734,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.999956483757215e-05,
+      "loss": 1.0204,
+      "step": 151
+    },
+    {
+      "epoch": 0.053614338308238356,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.999944924861329e-05,
+      "loss": 1.0192,
+      "step": 152
+    },
+    {
+      "epoch": 0.05396706421816098,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9999320061480424e-05,
+      "loss": 1.1087,
+      "step": 153
+    },
+    {
+      "epoch": 0.0543197901280836,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.999917727634925e-05,
+      "loss": 1.0401,
+      "step": 154
+    },
+    {
+      "epoch": 0.054672516038006214,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.999902089341393e-05,
+      "loss": 1.0994,
+      "step": 155
+    },
+    {
+      "epoch": 0.055025241947928835,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9998850912887132e-05,
+      "loss": 1.085,
+      "step": 156
+    },
+    {
+      "epoch": 0.05537796785785146,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9998667335000008e-05,
+      "loss": 1.072,
+      "step": 157
+    },
+    {
+      "epoch": 0.05573069376777408,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.999847016000221e-05,
+      "loss": 1.0694,
+      "step": 158
+    },
+    {
+      "epoch": 0.0560834196776967,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9998259388161866e-05,
+      "loss": 1.0011,
+      "step": 159
+    },
+    {
+      "epoch": 0.05643614558761932,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.999803501976561e-05,
+      "loss": 1.0635,
+      "step": 160
+    },
+    {
+      "epoch": 0.05678887149754194,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.999779705511856e-05,
+      "loss": 1.1201,
+      "step": 161
+    },
+    {
+      "epoch": 0.057141597407464564,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9997545494544316e-05,
+      "loss": 1.0502,
+      "step": 162
+    },
+    {
+      "epoch": 0.057494323317387186,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9997280338384974e-05,
+      "loss": 1.087,
+      "step": 163
+    },
+    {
+      "epoch": 0.0578470492273098,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9997001587001125e-05,
+      "loss": 1.0675,
+      "step": 164
+    },
+    {
+      "epoch": 0.05819977513723242,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9996709240771835e-05,
+      "loss": 1.1547,
+      "step": 165
+    },
+    {
+      "epoch": 0.058552501047155044,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.999640330009466e-05,
+      "loss": 1.1197,
+      "step": 166
+    },
+    {
+      "epoch": 0.058905226957077665,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.9996083765385656e-05,
+      "loss": 1.0619,
+      "step": 167
+    },
+    {
+      "epoch": 0.05925795286700029,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9995750637079347e-05,
+      "loss": 1.0949,
+      "step": 168
+    },
+    {
+      "epoch": 0.05961067877692291,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.999540391562876e-05,
+      "loss": 1.0608,
+      "step": 169
+    },
+    {
+      "epoch": 0.05996340468684553,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.999504360150539e-05,
+      "loss": 1.0499,
+      "step": 170
+    },
+    {
+      "epoch": 0.06031613059676815,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.999466969519923e-05,
+      "loss": 1.0834,
+      "step": 171
+    },
+    {
+      "epoch": 0.06066885650669077,
+      "grad_norm": 0.75,
+      "learning_rate": 1.999428219721875e-05,
+      "loss": 1.1442,
+      "step": 172
+    },
+    {
+      "epoch": 0.06102158241661339,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.999388110809091e-05,
+      "loss": 1.0727,
+      "step": 173
+    },
+    {
+      "epoch": 0.06137430832653601,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9993466428361144e-05,
+      "loss": 1.1547,
+      "step": 174
+    },
+    {
+      "epoch": 0.06172703423645863,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9993038158593374e-05,
+      "loss": 1.0928,
+      "step": 175
+    },
+    {
+      "epoch": 0.06207976014638125,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.999259629937e-05,
+      "loss": 1.0616,
+      "step": 176
+    },
+    {
+      "epoch": 0.062432486056303874,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9992140851291904e-05,
+      "loss": 1.1913,
+      "step": 177
+    },
+    {
+      "epoch": 0.0627852119662265,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9991671814978443e-05,
+      "loss": 1.1597,
+      "step": 178
+    },
+    {
+      "epoch": 0.06313793787614912,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.9991189191067457e-05,
+      "loss": 1.1163,
+      "step": 179
+    },
+    {
+      "epoch": 0.06349066378607174,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.9990692980215264e-05,
+      "loss": 1.1293,
+      "step": 180
+    },
+    {
+      "epoch": 0.06384338969599436,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.999018318309666e-05,
+      "loss": 1.0169,
+      "step": 181
+    },
+    {
+      "epoch": 0.06419611560591698,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.998965980040491e-05,
+      "loss": 1.0775,
+      "step": 182
+    },
+    {
+      "epoch": 0.0645488415158396,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9989122832851754e-05,
+      "loss": 1.1141,
+      "step": 183
+    },
+    {
+      "epoch": 0.06490156742576222,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.998857228116742e-05,
+      "loss": 0.976,
+      "step": 184
+    },
+    {
+      "epoch": 0.06525429333568485,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9988008146100594e-05,
+      "loss": 1.0116,
+      "step": 185
+    },
+    {
+      "epoch": 0.06560701924560745,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9987430428418436e-05,
+      "loss": 1.2115,
+      "step": 186
+    },
+    {
+      "epoch": 0.06595974515553008,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.9986839128906586e-05,
+      "loss": 1.0558,
+      "step": 187
+    },
+    {
+      "epoch": 0.0663124710654527,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9986234248369144e-05,
+      "loss": 1.0774,
+      "step": 188
+    },
+    {
+      "epoch": 0.06666519697537532,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9985615787628682e-05,
+      "loss": 1.0395,
+      "step": 189
+    },
+    {
+      "epoch": 0.06701792288529794,
+      "grad_norm": 0.75,
+      "learning_rate": 1.998498374752624e-05,
+      "loss": 0.975,
+      "step": 190
+    },
+    {
+      "epoch": 0.06737064879522056,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9984338128921327e-05,
+      "loss": 1.0451,
+      "step": 191
+    },
+    {
+      "epoch": 0.06772337470514318,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.998367893269191e-05,
+      "loss": 1.0771,
+      "step": 192
+    },
+    {
+      "epoch": 0.0680761006150658,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9983006159734433e-05,
+      "loss": 1.1368,
+      "step": 193
+    },
+    {
+      "epoch": 0.06842882652498843,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9982319810963786e-05,
+      "loss": 1.0246,
+      "step": 194
+    },
+    {
+      "epoch": 0.06878155243491105,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9981619887313335e-05,
+      "loss": 1.1786,
+      "step": 195
+    },
+    {
+      "epoch": 0.06913427834483367,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.99809063897349e-05,
+      "loss": 1.0365,
+      "step": 196
+    },
+    {
+      "epoch": 0.06948700425475629,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9980179319198757e-05,
+      "loss": 1.1482,
+      "step": 197
+    },
+    {
+      "epoch": 0.06983973016467891,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9979438676693647e-05,
+      "loss": 1.1349,
+      "step": 198
+    },
+    {
+      "epoch": 0.07019245607460153,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9978684463226766e-05,
+      "loss": 1.1789,
+      "step": 199
+    },
+    {
+      "epoch": 0.07054518198452416,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.997791667982376e-05,
+      "loss": 1.1127,
+      "step": 200
+    },
+    {
+      "epoch": 0.07089790789444678,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.997713532752873e-05,
+      "loss": 1.1251,
+      "step": 201
+    },
+    {
+      "epoch": 0.0712506338043694,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9976340407404233e-05,
+      "loss": 1.1256,
+      "step": 202
+    },
+    {
+      "epoch": 0.07160335971429202,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9975531920531277e-05,
+      "loss": 1.0255,
+      "step": 203
+    },
+    {
+      "epoch": 0.07195608562421463,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9974709868009313e-05,
+      "loss": 1.0415,
+      "step": 204
+    },
+    {
+      "epoch": 0.07230881153413725,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.997387425095625e-05,
+      "loss": 1.0463,
+      "step": 205
+    },
+    {
+      "epoch": 0.07266153744405987,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.997302507050843e-05,
+      "loss": 1.0703,
+      "step": 206
+    },
+    {
+      "epoch": 0.07301426335398249,
+      "grad_norm": 0.75,
+      "learning_rate": 1.997216232782065e-05,
+      "loss": 1.1451,
+      "step": 207
+    },
+    {
+      "epoch": 0.07336698926390511,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.997128602406615e-05,
+      "loss": 1.0317,
+      "step": 208
+    },
+    {
+      "epoch": 0.07371971517382774,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9970396160436602e-05,
+      "loss": 1.1593,
+      "step": 209
+    },
+    {
+      "epoch": 0.07407244108375036,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9969492738142133e-05,
+      "loss": 0.9953,
+      "step": 210
+    },
+    {
+      "epoch": 0.07442516699367298,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.996857575841129e-05,
+      "loss": 0.9734,
+      "step": 211
+    },
+    {
+      "epoch": 0.0747778929035956,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9967645222491074e-05,
+      "loss": 1.0819,
+      "step": 212
+    },
+    {
+      "epoch": 0.07513061881351822,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9966701131646914e-05,
+      "loss": 1.0168,
+      "step": 213
+    },
+    {
+      "epoch": 0.07548334472344084,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.996574348716267e-05,
+      "loss": 1.1088,
+      "step": 214
+    },
+    {
+      "epoch": 0.07583607063336346,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9964772290340633e-05,
+      "loss": 1.0637,
+      "step": 215
+    },
+    {
+      "epoch": 0.07618879654328609,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.996378754250153e-05,
+      "loss": 1.1932,
+      "step": 216
+    },
+    {
+      "epoch": 0.07654152245320871,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.996278924498451e-05,
+      "loss": 1.0963,
+      "step": 217
+    },
+    {
+      "epoch": 0.07689424836313133,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9961777399147152e-05,
+      "loss": 1.0332,
+      "step": 218
+    },
+    {
+      "epoch": 0.07724697427305395,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9960752006365455e-05,
+      "loss": 0.9238,
+      "step": 219
+    },
+    {
+      "epoch": 0.07759970018297657,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9959713068033842e-05,
+      "loss": 1.1263,
+      "step": 220
+    },
+    {
+      "epoch": 0.07795242609289918,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.995866058556516e-05,
+      "loss": 1.0471,
+      "step": 221
+    },
+    {
+      "epoch": 0.0783051520028218,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9957594560390666e-05,
+      "loss": 1.1127,
+      "step": 222
+    },
+    {
+      "epoch": 0.07865787791274442,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.995651499396005e-05,
+      "loss": 1.0057,
+      "step": 223
+    },
+    {
+      "epoch": 0.07901060382266704,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.99554218877414e-05,
+      "loss": 1.0171,
+      "step": 224
+    },
+    {
+      "epoch": 0.07936332973258967,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.995431524322122e-05,
+      "loss": 1.2039,
+      "step": 225
+    },
+    {
+      "epoch": 0.07971605564251229,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9953195061904427e-05,
+      "loss": 1.0996,
+      "step": 226
+    },
+    {
+      "epoch": 0.08006878155243491,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9952061345314355e-05,
+      "loss": 1.0396,
+      "step": 227
+    },
+    {
+      "epoch": 0.08042150746235753,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9950914094992733e-05,
+      "loss": 1.0478,
+      "step": 228
+    },
+    {
+      "epoch": 0.08077423337228015,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9949753312499692e-05,
+      "loss": 0.9604,
+      "step": 229
+    },
+    {
+      "epoch": 0.08112695928220277,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9948578999413775e-05,
+      "loss": 1.0457,
+      "step": 230
+    },
+    {
+      "epoch": 0.0814796851921254,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9947391157331925e-05,
+      "loss": 0.9789,
+      "step": 231
+    },
+    {
+      "epoch": 0.08183241110204802,
+      "grad_norm": 14.25,
+      "learning_rate": 1.9946189787869477e-05,
+      "loss": 1.0924,
+      "step": 232
+    },
+    {
+      "epoch": 0.08218513701197064,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9944974892660158e-05,
+      "loss": 1.0606,
+      "step": 233
+    },
+    {
+      "epoch": 0.08253786292189326,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9943746473356105e-05,
+      "loss": 1.0675,
+      "step": 234
+    },
+    {
+      "epoch": 0.08289058883181588,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9942504531627827e-05,
+      "loss": 1.0987,
+      "step": 235
+    },
+    {
+      "epoch": 0.0832433147417385,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.994124906916424e-05,
+      "loss": 1.0105,
+      "step": 236
+    },
+    {
+      "epoch": 0.08359604065166112,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9939980087672627e-05,
+      "loss": 1.0717,
+      "step": 237
+    },
+    {
+      "epoch": 0.08394876656158375,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9938697588878675e-05,
+      "loss": 0.9982,
+      "step": 238
+    },
+    {
+      "epoch": 0.08430149247150635,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.993740157452644e-05,
+      "loss": 1.0735,
+      "step": 239
+    },
+    {
+      "epoch": 0.08465421838142898,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9936092046378363e-05,
+      "loss": 1.0616,
+      "step": 240
+    },
+    {
+      "epoch": 0.0850069442913516,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.993476900621526e-05,
+      "loss": 1.0205,
+      "step": 241
+    },
+    {
+      "epoch": 0.08535967020127422,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9933432455836326e-05,
+      "loss": 1.0564,
+      "step": 242
+    },
+    {
+      "epoch": 0.08571239611119684,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9932082397059124e-05,
+      "loss": 1.0474,
+      "step": 243
+    },
+    {
+      "epoch": 0.08606512202111946,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9930718831719586e-05,
+      "loss": 1.0562,
+      "step": 244
+    },
+    {
+      "epoch": 0.08641784793104208,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9929341761672017e-05,
+      "loss": 1.0923,
+      "step": 245
+    },
+    {
+      "epoch": 0.0867705738409647,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9927951188789084e-05,
+      "loss": 1.0468,
+      "step": 246
+    },
+    {
+      "epoch": 0.08712329975088733,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9926547114961814e-05,
+      "loss": 1.083,
+      "step": 247
+    },
+    {
+      "epoch": 0.08747602566080995,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9925129542099598e-05,
+      "loss": 1.039,
+      "step": 248
+    },
+    {
+      "epoch": 0.08782875157073257,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9923698472130183e-05,
+      "loss": 1.1202,
+      "step": 249
+    },
+    {
+      "epoch": 0.08818147748065519,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.992225390699967e-05,
+      "loss": 1.0373,
+      "step": 250
+    },
+    {
+      "epoch": 0.08853420339057781,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9920795848672512e-05,
+      "loss": 1.1197,
+      "step": 251
+    },
+    {
+      "epoch": 0.08888692930050043,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9919324299131508e-05,
+      "loss": 0.9593,
+      "step": 252
+    },
+    {
+      "epoch": 0.08923965521042306,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.991783926037781e-05,
+      "loss": 1.0468,
+      "step": 253
+    },
+    {
+      "epoch": 0.08959238112034568,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.9916340734430912e-05,
+      "loss": 1.0166,
+      "step": 254
+    },
+    {
+      "epoch": 0.0899451070302683,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.991482872332864e-05,
+      "loss": 1.0669,
+      "step": 255
+    },
+    {
+      "epoch": 0.09029783294019092,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9913303229127182e-05,
+      "loss": 1.0255,
+      "step": 256
+    },
+    {
+      "epoch": 0.09065055885011353,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9911764253901032e-05,
+      "loss": 0.9739,
+      "step": 257
+    },
+    {
+      "epoch": 0.09100328476003615,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.991021179974303e-05,
+      "loss": 1.1084,
+      "step": 258
+    },
+    {
+      "epoch": 0.09135601066995877,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9908645868764358e-05,
+      "loss": 0.9585,
+      "step": 259
+    },
+    {
+      "epoch": 0.09170873657988139,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9907066463094503e-05,
+      "loss": 1.0385,
+      "step": 260
+    },
+    {
+      "epoch": 0.09206146248980401,
+      "grad_norm": 6.625,
+      "learning_rate": 1.990547358488129e-05,
+      "loss": 1.0099,
+      "step": 261
+    },
+    {
+      "epoch": 0.09241418839972664,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9903867236290863e-05,
+      "loss": 1.0178,
+      "step": 262
+    },
+    {
+      "epoch": 0.09276691430964926,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.990224741950768e-05,
+      "loss": 1.1679,
+      "step": 263
+    },
+    {
+      "epoch": 0.09311964021957188,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.990061413673452e-05,
+      "loss": 1.1432,
+      "step": 264
+    },
+    {
+      "epoch": 0.0934723661294945,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9898967390192473e-05,
+      "loss": 1.0084,
+      "step": 265
+    },
+    {
+      "epoch": 0.09382509203941712,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9897307182120936e-05,
+      "loss": 1.042,
+      "step": 266
+    },
+    {
+      "epoch": 0.09417781794933974,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.989563351477761e-05,
+      "loss": 1.0418,
+      "step": 267
+    },
+    {
+      "epoch": 0.09453054385926236,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9893946390438504e-05,
+      "loss": 1.0524,
+      "step": 268
+    },
+    {
+      "epoch": 0.09488326976918499,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9892245811397924e-05,
+      "loss": 1.0597,
+      "step": 269
+    },
+    {
+      "epoch": 0.09523599567910761,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9890531779968484e-05,
+      "loss": 1.1716,
+      "step": 270
+    },
+    {
+      "epoch": 0.09558872158903023,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.9888804298481067e-05,
+      "loss": 1.0704,
+      "step": 271
+    },
+    {
+      "epoch": 0.09594144749895285,
+      "grad_norm": 0.75,
+      "learning_rate": 1.988706336928487e-05,
+      "loss": 1.0433,
+      "step": 272
+    },
+    {
+      "epoch": 0.09629417340887547,
+      "grad_norm": 0.75,
+      "learning_rate": 1.988530899474737e-05,
+      "loss": 1.0244,
+      "step": 273
+    },
+    {
+      "epoch": 0.09664689931879808,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9883541177254325e-05,
+      "loss": 1.108,
+      "step": 274
+    },
+    {
+      "epoch": 0.0969996252287207,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.988175991920978e-05,
+      "loss": 1.0747,
+      "step": 275
+    },
+    {
+      "epoch": 0.09735235113864332,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.9879965223036047e-05,
+      "loss": 1.0846,
+      "step": 276
+    },
+    {
+      "epoch": 0.09770507704856594,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9878157091173726e-05,
+      "loss": 1.047,
+      "step": 277
+    },
+    {
+      "epoch": 0.09805780295848857,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.987633552608168e-05,
+      "loss": 1.1664,
+      "step": 278
+    },
+    {
+      "epoch": 0.09841052886841119,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.987450053023704e-05,
+      "loss": 1.059,
+      "step": 279
+    },
+    {
+      "epoch": 0.09876325477833381,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9872652106135203e-05,
+      "loss": 0.8738,
+      "step": 280
+    },
+    {
+      "epoch": 0.09911598068825643,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9870790256289827e-05,
+      "loss": 1.0737,
+      "step": 281
+    },
+    {
+      "epoch": 0.09946870659817905,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9868914983232827e-05,
+      "loss": 1.0943,
+      "step": 282
+    },
+    {
+      "epoch": 0.09982143250810167,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9867026289514373e-05,
+      "loss": 0.9617,
+      "step": 283
+    },
+    {
+      "epoch": 0.1001741584180243,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9865124177702887e-05,
+      "loss": 1.1207,
+      "step": 284
+    },
+    {
+      "epoch": 0.10052688432794692,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9863208650385028e-05,
+      "loss": 1.0375,
+      "step": 285
+    },
+    {
+      "epoch": 0.10087961023786954,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.986127971016572e-05,
+      "loss": 1.0948,
+      "step": 286
+    },
+    {
+      "epoch": 0.10123233614779216,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.9859337359668102e-05,
+      "loss": 1.0312,
+      "step": 287
+    },
+    {
+      "epoch": 0.10158506205771478,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.985738160153357e-05,
+      "loss": 1.082,
+      "step": 288
+    },
+    {
+      "epoch": 0.1019377879676374,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.985541243842174e-05,
+      "loss": 1.0602,
+      "step": 289
+    },
+    {
+      "epoch": 0.10229051387756002,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9853429873010456e-05,
+      "loss": 1.0984,
+      "step": 290
+    },
+    {
+      "epoch": 0.10264323978748265,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9851433907995804e-05,
+      "loss": 1.0885,
+      "step": 291
+    },
+    {
+      "epoch": 0.10299596569740525,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.984942454609208e-05,
+      "loss": 1.0057,
+      "step": 292
+    },
+    {
+      "epoch": 0.10334869160732787,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9847401790031792e-05,
+      "loss": 1.0389,
+      "step": 293
+    },
+    {
+      "epoch": 0.1037014175172505,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.9845365642565678e-05,
+      "loss": 1.0149,
+      "step": 294
+    },
+    {
+      "epoch": 0.10405414342717312,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9843316106462677e-05,
+      "loss": 1.1332,
+      "step": 295
+    },
+    {
+      "epoch": 0.10440686933709574,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.984125318450994e-05,
+      "loss": 1.1017,
+      "step": 296
+    },
+    {
+      "epoch": 0.10475959524701836,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.983917687951281e-05,
+      "loss": 0.9944,
+      "step": 297
+    },
+    {
+      "epoch": 0.10511232115694098,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9837087194294845e-05,
+      "loss": 1.0053,
+      "step": 298
+    },
+    {
+      "epoch": 0.1054650470668636,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9834984131697796e-05,
+      "loss": 1.082,
+      "step": 299
+    },
+    {
+      "epoch": 0.10581777297678623,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9832867694581587e-05,
+      "loss": 0.9901,
+      "step": 300
+    },
+    {
+      "epoch": 0.10617049888670885,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.983073788582436e-05,
+      "loss": 1.0909,
+      "step": 301
+    },
+    {
+      "epoch": 0.10652322479663147,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.982859470832242e-05,
+      "loss": 0.9731,
+      "step": 302
+    },
+    {
+      "epoch": 0.10687595070655409,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9826438164990256e-05,
+      "loss": 0.9999,
+      "step": 303
+    },
+    {
+      "epoch": 0.10722867661647671,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9824268258760534e-05,
+      "loss": 1.0692,
+      "step": 304
+    },
+    {
+      "epoch": 0.10758140252639933,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9822084992584098e-05,
+      "loss": 1.1204,
+      "step": 305
+    },
+    {
+      "epoch": 0.10793412843632195,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9819888369429944e-05,
+      "loss": 1.0517,
+      "step": 306
+    },
+    {
+      "epoch": 0.10828685434624458,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9817678392285256e-05,
+      "loss": 0.9667,
+      "step": 307
+    },
+    {
+      "epoch": 0.1086395802561672,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.981545506415536e-05,
+      "loss": 0.9873,
+      "step": 308
+    },
+    {
+      "epoch": 0.10899230616608982,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9813218388063738e-05,
+      "loss": 1.0446,
+      "step": 309
+    },
+    {
+      "epoch": 0.10934503207601243,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9810968367052034e-05,
+      "loss": 1.1857,
+      "step": 310
+    },
+    {
+      "epoch": 0.10969775798593505,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.9808705004180032e-05,
+      "loss": 1.0878,
+      "step": 311
+    },
+    {
+      "epoch": 0.11005048389585767,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9806428302525665e-05,
+      "loss": 0.9908,
+      "step": 312
+    },
+    {
+      "epoch": 0.11040320980578029,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9804138265184995e-05,
+      "loss": 0.967,
+      "step": 313
+    },
+    {
+      "epoch": 0.11075593571570291,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9801834895272233e-05,
+      "loss": 0.8441,
+      "step": 314
+    },
+    {
+      "epoch": 0.11110866162562553,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9799518195919712e-05,
+      "loss": 1.1039,
+      "step": 315
+    },
+    {
+      "epoch": 0.11146138753554816,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9797188170277893e-05,
+      "loss": 1.0203,
+      "step": 316
+    },
+    {
+      "epoch": 0.11181411344547078,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.979484482151536e-05,
+      "loss": 1.0761,
+      "step": 317
+    },
+    {
+      "epoch": 0.1121668393553934,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9792488152818812e-05,
+      "loss": 1.1215,
+      "step": 318
+    },
+    {
+      "epoch": 0.11251956526531602,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.979011816739307e-05,
+      "loss": 1.0715,
+      "step": 319
+    },
+    {
+      "epoch": 0.11287229117523864,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9787734868461058e-05,
+      "loss": 1.0365,
+      "step": 320
+    },
+    {
+      "epoch": 0.11322501708516126,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9785338259263804e-05,
+      "loss": 0.8883,
+      "step": 321
+    },
+    {
+      "epoch": 0.11357774299508389,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9782928343060436e-05,
+      "loss": 0.9438,
+      "step": 322
+    },
+    {
+      "epoch": 0.11393046890500651,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.9780505123128187e-05,
+      "loss": 1.117,
+      "step": 323
+    },
+    {
+      "epoch": 0.11428319481492913,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.977806860276237e-05,
+      "loss": 0.9446,
+      "step": 324
+    },
+    {
+      "epoch": 0.11463592072485175,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.977561878527639e-05,
+      "loss": 1.0111,
+      "step": 325
+    },
+    {
+      "epoch": 0.11498864663477437,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9773155674001737e-05,
+      "loss": 0.9562,
+      "step": 326
+    },
+    {
+      "epoch": 0.11534137254469698,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9770679272287976e-05,
+      "loss": 1.0872,
+      "step": 327
+    },
+    {
+      "epoch": 0.1156940984546196,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.976818958350275e-05,
+      "loss": 0.998,
+      "step": 328
+    },
+    {
+      "epoch": 0.11604682436454222,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.976568661103176e-05,
+      "loss": 1.0092,
+      "step": 329
+    },
+    {
+      "epoch": 0.11639955027446484,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.976317035827879e-05,
+      "loss": 1.0568,
+      "step": 330
+    },
+    {
+      "epoch": 0.11675227618438747,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9760640828665662e-05,
+      "loss": 0.9924,
+      "step": 331
+    },
+    {
+      "epoch": 0.11710500209431009,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.975809802563227e-05,
+      "loss": 1.0173,
+      "step": 332
+    },
+    {
+      "epoch": 0.11745772800423271,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9755541952636553e-05,
+      "loss": 0.9518,
+      "step": 333
+    },
+    {
+      "epoch": 0.11781045391415533,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9752972613154488e-05,
+      "loss": 1.0331,
+      "step": 334
+    },
+    {
+      "epoch": 0.11816317982407795,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.975039001068011e-05,
+      "loss": 0.9417,
+      "step": 335
+    },
+    {
+      "epoch": 0.11851590573400057,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.974779414872547e-05,
+      "loss": 1.0188,
+      "step": 336
+    },
+    {
+      "epoch": 0.1188686316439232,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9745185030820667e-05,
+      "loss": 1.0776,
+      "step": 337
+    },
+    {
+      "epoch": 0.11922135755384582,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.9742562660513815e-05,
+      "loss": 1.0231,
+      "step": 338
+    },
+    {
+      "epoch": 0.11957408346376844,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9739927041371065e-05,
+      "loss": 0.9083,
+      "step": 339
+    },
+    {
+      "epoch": 0.11992680937369106,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.9737278176976565e-05,
+      "loss": 0.9707,
+      "step": 340
+    },
+    {
+      "epoch": 0.12027953528361368,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.973461607093249e-05,
+      "loss": 1.0512,
+      "step": 341
+    },
+    {
+      "epoch": 0.1206322611935363,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.973194072685902e-05,
+      "loss": 1.0327,
+      "step": 342
+    },
+    {
+      "epoch": 0.12098498710345892,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9729252148394327e-05,
+      "loss": 1.1822,
+      "step": 343
+    },
+    {
+      "epoch": 0.12133771301338155,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.97265503391946e-05,
+      "loss": 0.991,
+      "step": 344
+    },
+    {
+      "epoch": 0.12169043892330415,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9723835302933993e-05,
+      "loss": 1.0498,
+      "step": 345
+    },
+    {
+      "epoch": 0.12204316483322677,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9721107043304677e-05,
+      "loss": 1.0392,
+      "step": 346
+    },
+    {
+      "epoch": 0.1223958907431494,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9718365564016785e-05,
+      "loss": 1.1071,
+      "step": 347
+    },
+    {
+      "epoch": 0.12274861665307202,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.971561086879843e-05,
+      "loss": 1.0742,
+      "step": 348
+    },
+    {
+      "epoch": 0.12310134256299464,
+      "grad_norm": 5.28125,
+      "learning_rate": 1.9712842961395707e-05,
+      "loss": 0.9824,
+      "step": 349
+    },
+    {
+      "epoch": 0.12345406847291726,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.971006184557267e-05,
+      "loss": 1.0962,
+      "step": 350
+    },
+    {
+      "epoch": 0.12380679438283988,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9707267525111337e-05,
+      "loss": 1.0011,
+      "step": 351
+    },
+    {
+      "epoch": 0.1241595202927625,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9704460003811684e-05,
+      "loss": 1.0353,
+      "step": 352
+    },
+    {
+      "epoch": 0.12451224620268513,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9701639285491633e-05,
+      "loss": 0.9377,
+      "step": 353
+    },
+    {
+      "epoch": 0.12486497211260775,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.9698805373987065e-05,
+      "loss": 1.1055,
+      "step": 354
+    },
+    {
+      "epoch": 0.12521769802253035,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9695958273151785e-05,
+      "loss": 1.0456,
+      "step": 355
+    },
+    {
+      "epoch": 0.125570423932453,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.969309798685755e-05,
+      "loss": 1.0591,
+      "step": 356
+    },
+    {
+      "epoch": 0.1259231498423756,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.969022451899404e-05,
+      "loss": 0.9976,
+      "step": 357
+    },
+    {
+      "epoch": 0.12627587575229823,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.968733787346886e-05,
+      "loss": 1.0997,
+      "step": 358
+    },
+    {
+      "epoch": 0.12662860166222084,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.968443805420754e-05,
+      "loss": 1.1155,
+      "step": 359
+    },
+    {
+      "epoch": 0.12698132757214348,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.968152506515352e-05,
+      "loss": 0.8887,
+      "step": 360
+    },
+    {
+      "epoch": 0.12733405348206608,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9678598910268154e-05,
+      "loss": 1.1393,
+      "step": 361
+    },
+    {
+      "epoch": 0.12768677939198872,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.967565959353069e-05,
+      "loss": 1.0491,
+      "step": 362
+    },
+    {
+      "epoch": 0.12803950530191133,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.967270711893829e-05,
+      "loss": 0.9819,
+      "step": 363
+    },
+    {
+      "epoch": 0.12839223121183396,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9669741490506e-05,
+      "loss": 0.9905,
+      "step": 364
+    },
+    {
+      "epoch": 0.12874495712175657,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.966676271226675e-05,
+      "loss": 0.9891,
+      "step": 365
+    },
+    {
+      "epoch": 0.1290976830316792,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9663770788271366e-05,
+      "loss": 1.0977,
+      "step": 366
+    },
+    {
+      "epoch": 0.1294504089416018,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9660765722588538e-05,
+      "loss": 1.0905,
+      "step": 367
+    },
+    {
+      "epoch": 0.12980313485152445,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.965774751930483e-05,
+      "loss": 0.9685,
+      "step": 368
+    },
+    {
+      "epoch": 0.13015586076144706,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9654716182524674e-05,
+      "loss": 0.9819,
+      "step": 369
+    },
+    {
+      "epoch": 0.1305085866713697,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9651671716370364e-05,
+      "loss": 0.9523,
+      "step": 370
+    },
+    {
+      "epoch": 0.1308613125812923,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9648614124982044e-05,
+      "loss": 1.033,
+      "step": 371
+    },
+    {
+      "epoch": 0.1312140384912149,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.9645543412517708e-05,
+      "loss": 1.0601,
+      "step": 372
+    },
+    {
+      "epoch": 0.13156676440113754,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.96424595831532e-05,
+      "loss": 0.9516,
+      "step": 373
+    },
+    {
+      "epoch": 0.13191949031106015,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.963936264108219e-05,
+      "loss": 0.9791,
+      "step": 374
+    },
+    {
+      "epoch": 0.13227221622098279,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.9636252590516184e-05,
+      "loss": 0.9771,
+      "step": 375
+    },
+    {
+      "epoch": 0.1326249421309054,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.963312943568452e-05,
+      "loss": 0.9594,
+      "step": 376
+    },
+    {
+      "epoch": 0.13297766804082803,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9629993180834356e-05,
+      "loss": 1.0027,
+      "step": 377
+    },
+    {
+      "epoch": 0.13333039395075064,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9626843830230654e-05,
+      "loss": 1.1103,
+      "step": 378
+    },
+    {
+      "epoch": 0.13368311986067327,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9623681388156193e-05,
+      "loss": 1.08,
+      "step": 379
+    },
+    {
+      "epoch": 0.13403584577059588,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9620505858911554e-05,
+      "loss": 1.0662,
+      "step": 380
+    },
+    {
+      "epoch": 0.13438857168051851,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9617317246815114e-05,
+      "loss": 0.9297,
+      "step": 381
+    },
+    {
+      "epoch": 0.13474129759044112,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9614115556203047e-05,
+      "loss": 0.9221,
+      "step": 382
+    },
+    {
+      "epoch": 0.13509402350036376,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.96109007914293e-05,
+      "loss": 1.0903,
+      "step": 383
+    },
+    {
+      "epoch": 0.13544674941028637,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.9607672956865608e-05,
+      "loss": 0.9736,
+      "step": 384
+    },
+    {
+      "epoch": 0.135799475320209,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9604432056901487e-05,
+      "loss": 0.9991,
+      "step": 385
+    },
+    {
+      "epoch": 0.1361522012301316,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.9601178095944202e-05,
+      "loss": 1.1147,
+      "step": 386
+    },
+    {
+      "epoch": 0.13650492714005424,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9597911078418796e-05,
+      "loss": 0.9653,
+      "step": 387
+    },
+    {
+      "epoch": 0.13685765304997685,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9594631008768053e-05,
+      "loss": 1.0561,
+      "step": 388
+    },
+    {
+      "epoch": 0.13721037895989946,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9591337891452525e-05,
+      "loss": 0.9653,
+      "step": 389
+    },
+    {
+      "epoch": 0.1375631048698221,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9588031730950488e-05,
+      "loss": 0.9729,
+      "step": 390
+    },
+    {
+      "epoch": 0.1379158307797447,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.958471253175797e-05,
+      "loss": 1.0324,
+      "step": 391
+    },
+    {
+      "epoch": 0.13826855668966734,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.958138029838872e-05,
+      "loss": 0.9398,
+      "step": 392
+    },
+    {
+      "epoch": 0.13862128259958995,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9578035035374214e-05,
+      "loss": 1.0411,
+      "step": 393
+    },
+    {
+      "epoch": 0.13897400850951258,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.957467674726366e-05,
+      "loss": 1.0712,
+      "step": 394
+    },
+    {
+      "epoch": 0.1393267344194352,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.957130543862396e-05,
+      "loss": 1.0629,
+      "step": 395
+    },
+    {
+      "epoch": 0.13967946032935782,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9567921114039732e-05,
+      "loss": 1.0637,
+      "step": 396
+    },
+    {
+      "epoch": 0.14003218623928043,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9564523778113297e-05,
+      "loss": 1.134,
+      "step": 397
+    },
+    {
+      "epoch": 0.14038491214920307,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.956111343546466e-05,
+      "loss": 1.1563,
+      "step": 398
+    },
+    {
+      "epoch": 0.14073763805912567,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.9557690090731517e-05,
+      "loss": 0.9929,
+      "step": 399
+    },
+    {
+      "epoch": 0.1410903639690483,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.955425374856926e-05,
+      "loss": 0.9438,
+      "step": 400
+    },
+    {
+      "epoch": 0.14144308987897092,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9550804413650935e-05,
+      "loss": 0.9654,
+      "step": 401
+    },
+    {
+      "epoch": 0.14179581578889355,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.954734209066727e-05,
+      "loss": 0.9959,
+      "step": 402
+    },
+    {
+      "epoch": 0.14214854169881616,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.954386678432665e-05,
+      "loss": 1.0529,
+      "step": 403
+    },
+    {
+      "epoch": 0.1425012676087388,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9540378499355118e-05,
+      "loss": 1.0475,
+      "step": 404
+    },
+    {
+      "epoch": 0.1428539935186614,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9536877240496364e-05,
+      "loss": 1.0033,
+      "step": 405
+    },
+    {
+      "epoch": 0.14320671942858404,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9533363012511724e-05,
+      "loss": 1.1604,
+      "step": 406
+    },
+    {
+      "epoch": 0.14355944533850665,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9529835820180166e-05,
+      "loss": 1.0209,
+      "step": 407
+    },
+    {
+      "epoch": 0.14391217124842925,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.95262956682983e-05,
+      "loss": 1.0508,
+      "step": 408
+    },
+    {
+      "epoch": 0.1442648971583519,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9522742561680344e-05,
+      "loss": 1.0153,
+      "step": 409
+    },
+    {
+      "epoch": 0.1446176230682745,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9519176505158146e-05,
+      "loss": 0.9623,
+      "step": 410
+    },
+    {
+      "epoch": 0.14497034897819713,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9515597503581153e-05,
+      "loss": 0.9678,
+      "step": 411
+    },
+    {
+      "epoch": 0.14532307488811974,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9512005561816424e-05,
+      "loss": 1.0877,
+      "step": 412
+    },
+    {
+      "epoch": 0.14567580079804238,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9508400684748615e-05,
+      "loss": 1.0385,
+      "step": 413
+    },
+    {
+      "epoch": 0.14602852670796498,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9504782877279966e-05,
+      "loss": 1.0309,
+      "step": 414
+    },
+    {
+      "epoch": 0.14638125261788762,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.950115214433031e-05,
+      "loss": 1.1131,
+      "step": 415
+    },
+    {
+      "epoch": 0.14673397852781023,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9497508490837053e-05,
+      "loss": 1.0346,
+      "step": 416
+    },
+    {
+      "epoch": 0.14708670443773286,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9493851921755164e-05,
+      "loss": 1.0352,
+      "step": 417
+    },
+    {
+      "epoch": 0.14743943034765547,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9490182442057194e-05,
+      "loss": 1.0801,
+      "step": 418
+    },
+    {
+      "epoch": 0.1477921562575781,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.948650005673323e-05,
+      "loss": 0.9961,
+      "step": 419
+    },
+    {
+      "epoch": 0.1481448821675007,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.948280477079093e-05,
+      "loss": 1.0846,
+      "step": 420
+    },
+    {
+      "epoch": 0.14849760807742335,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9479096589255472e-05,
+      "loss": 1.059,
+      "step": 421
+    },
+    {
+      "epoch": 0.14885033398734596,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9475375517169595e-05,
+      "loss": 0.9374,
+      "step": 422
+    },
+    {
+      "epoch": 0.1492030598972686,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.947164155959355e-05,
+      "loss": 1.0119,
+      "step": 423
+    },
+    {
+      "epoch": 0.1495557858071912,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9467894721605122e-05,
+      "loss": 1.1006,
+      "step": 424
+    },
+    {
+      "epoch": 0.1499085117171138,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.94641350082996e-05,
+      "loss": 1.026,
+      "step": 425
+    },
+    {
+      "epoch": 0.15026123762703644,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9460362424789795e-05,
+      "loss": 1.0153,
+      "step": 426
+    },
+    {
+      "epoch": 0.15061396353695905,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.945657697620601e-05,
+      "loss": 1.0136,
+      "step": 427
+    },
+    {
+      "epoch": 0.15096668944688169,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9452778667696053e-05,
+      "loss": 1.043,
+      "step": 428
+    },
+    {
+      "epoch": 0.1513194153568043,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9448967504425213e-05,
+      "loss": 1.1106,
+      "step": 429
+    },
+    {
+      "epoch": 0.15167214126672693,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9445143491576254e-05,
+      "loss": 1.0602,
+      "step": 430
+    },
+    {
+      "epoch": 0.15202486717664954,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.944130663434943e-05,
+      "loss": 0.9094,
+      "step": 431
+    },
+    {
+      "epoch": 0.15237759308657217,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.9437456937962454e-05,
+      "loss": 1.0497,
+      "step": 432
+    },
+    {
+      "epoch": 0.15273031899649478,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9433594407650493e-05,
+      "loss": 1.0618,
+      "step": 433
+    },
+    {
+      "epoch": 0.15308304490641741,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9429719048666173e-05,
+      "loss": 0.9667,
+      "step": 434
+    },
+    {
+      "epoch": 0.15343577081634002,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9425830866279567e-05,
+      "loss": 1.0697,
+      "step": 435
+    },
+    {
+      "epoch": 0.15378849672626266,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9421929865778187e-05,
+      "loss": 0.9001,
+      "step": 436
+    },
+    {
+      "epoch": 0.15414122263618527,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.9418016052466965e-05,
+      "loss": 1.0421,
+      "step": 437
+    },
+    {
+      "epoch": 0.1544939485461079,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9414089431668276e-05,
+      "loss": 1.0498,
+      "step": 438
+    },
+    {
+      "epoch": 0.1548466744560305,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9410150008721894e-05,
+      "loss": 0.9289,
+      "step": 439
+    },
+    {
+      "epoch": 0.15519940036595314,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9406197788985013e-05,
+      "loss": 1.0805,
+      "step": 440
+    },
+    {
+      "epoch": 0.15555212627587575,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9402232777832225e-05,
+      "loss": 0.9332,
+      "step": 441
+    },
+    {
+      "epoch": 0.15590485218579836,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.9398254980655516e-05,
+      "loss": 0.9848,
+      "step": 442
+    },
+    {
+      "epoch": 0.156257578095721,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9394264402864265e-05,
+      "loss": 0.9871,
+      "step": 443
+    },
+    {
+      "epoch": 0.1566103040056436,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.939026104988523e-05,
+      "loss": 1.0788,
+      "step": 444
+    },
+    {
+      "epoch": 0.15696302991556624,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9386244927162536e-05,
+      "loss": 0.9189,
+      "step": 445
+    },
+    {
+      "epoch": 0.15731575582548885,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.938221604015768e-05,
+      "loss": 1.0967,
+      "step": 446
+    },
+    {
+      "epoch": 0.15766848173541148,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.937817439434951e-05,
+      "loss": 0.9432,
+      "step": 447
+    },
+    {
+      "epoch": 0.1580212076453341,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.9374119995234233e-05,
+      "loss": 1.0247,
+      "step": 448
+    },
+    {
+      "epoch": 0.15837393355525672,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9370052848325392e-05,
+      "loss": 0.9372,
+      "step": 449
+    },
+    {
+      "epoch": 0.15872665946517933,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9365972959153872e-05,
+      "loss": 1.0433,
+      "step": 450
+    },
+    {
+      "epoch": 0.15907938537510197,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.936188033326788e-05,
+      "loss": 0.9767,
+      "step": 451
+    },
+    {
+      "epoch": 0.15943211128502457,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.935777497623295e-05,
+      "loss": 1.0064,
+      "step": 452
+    },
+    {
+      "epoch": 0.1597848371949472,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.9353656893631923e-05,
+      "loss": 1.0221,
+      "step": 453
+    },
+    {
+      "epoch": 0.16013756310486982,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9349526091064952e-05,
+      "loss": 0.9193,
+      "step": 454
+    },
+    {
+      "epoch": 0.16049028901479245,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9345382574149473e-05,
+      "loss": 1.04,
+      "step": 455
+    },
+    {
+      "epoch": 0.16084301492471506,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9341226348520236e-05,
+      "loss": 1.0571,
+      "step": 456
+    },
+    {
+      "epoch": 0.1611957408346377,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.933705741982925e-05,
+      "loss": 0.9267,
+      "step": 457
+    },
+    {
+      "epoch": 0.1615484667445603,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.933287579374581e-05,
+      "loss": 0.9115,
+      "step": 458
+    },
+    {
+      "epoch": 0.16190119265448294,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9328681475956485e-05,
+      "loss": 1.0593,
+      "step": 459
+    },
+    {
+      "epoch": 0.16225391856440555,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9324474472165086e-05,
+      "loss": 1.0331,
+      "step": 460
+    },
+    {
+      "epoch": 0.16260664447432815,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.932025478809269e-05,
+      "loss": 1.0494,
+      "step": 461
+    },
+    {
+      "epoch": 0.1629593703842508,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9316022429477613e-05,
+      "loss": 0.9802,
+      "step": 462
+    },
+    {
+      "epoch": 0.1633120962941734,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9311777402075407e-05,
+      "loss": 0.9827,
+      "step": 463
+    },
+    {
+      "epoch": 0.16366482220409603,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.930751971165885e-05,
+      "loss": 1.0222,
+      "step": 464
+    },
+    {
+      "epoch": 0.16401754811401864,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9303249364017944e-05,
+      "loss": 0.8986,
+      "step": 465
+    },
+    {
+      "epoch": 0.16437027402394128,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.92989663649599e-05,
+      "loss": 0.9733,
+      "step": 466
+    },
+    {
+      "epoch": 0.16472299993386388,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.929467072030914e-05,
+      "loss": 0.9479,
+      "step": 467
+    },
+    {
+      "epoch": 0.16507572584378652,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.929036243590727e-05,
+      "loss": 1.0159,
+      "step": 468
+    },
+    {
+      "epoch": 0.16542845175370913,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9286041517613103e-05,
+      "loss": 1.013,
+      "step": 469
+    },
+    {
+      "epoch": 0.16578117766363176,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.9281707971302612e-05,
+      "loss": 0.8978,
+      "step": 470
+    },
+    {
+      "epoch": 0.16613390357355437,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.927736180286896e-05,
+      "loss": 1.0748,
+      "step": 471
+    },
+    {
+      "epoch": 0.166486629483477,
+      "grad_norm": 0.75,
+      "learning_rate": 1.927300301822247e-05,
+      "loss": 0.9961,
+      "step": 472
+    },
+    {
+      "epoch": 0.1668393553933996,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.926863162329061e-05,
+      "loss": 0.977,
+      "step": 473
+    },
+    {
+      "epoch": 0.16719208130332225,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9264247624018018e-05,
+      "loss": 0.9626,
+      "step": 474
+    },
+    {
+      "epoch": 0.16754480721324486,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.925985102636645e-05,
+      "loss": 0.9971,
+      "step": 475
+    },
+    {
+      "epoch": 0.1678975331231675,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9255441836314813e-05,
+      "loss": 0.9868,
+      "step": 476
+    },
+    {
+      "epoch": 0.1682502590330901,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9251020059859128e-05,
+      "loss": 0.9759,
+      "step": 477
+    },
+    {
+      "epoch": 0.1686029849430127,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9246585703012534e-05,
+      "loss": 0.9281,
+      "step": 478
+    },
+    {
+      "epoch": 0.16895571085293534,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.924213877180528e-05,
+      "loss": 0.9877,
+      "step": 479
+    },
+    {
+      "epoch": 0.16930843676285795,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9237679272284714e-05,
+      "loss": 0.975,
+      "step": 480
+    },
+    {
+      "epoch": 0.16966116267278059,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.923320721051527e-05,
+      "loss": 1.0737,
+      "step": 481
+    },
+    {
+      "epoch": 0.1700138885827032,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9228722592578478e-05,
+      "loss": 0.9716,
+      "step": 482
+    },
+    {
+      "epoch": 0.17036661449262583,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9224225424572927e-05,
+      "loss": 1.0771,
+      "step": 483
+    },
+    {
+      "epoch": 0.17071934040254844,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.921971571261429e-05,
+      "loss": 0.9219,
+      "step": 484
+    },
+    {
+      "epoch": 0.17107206631247107,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.9215193462835285e-05,
+      "loss": 0.9794,
+      "step": 485
+    },
+    {
+      "epoch": 0.17142479222239368,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9210658681385684e-05,
+      "loss": 1.0309,
+      "step": 486
+    },
+    {
+      "epoch": 0.17177751813231631,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9206111374432303e-05,
+      "loss": 0.9646,
+      "step": 487
+    },
+    {
+      "epoch": 0.17213024404223892,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9201551548158986e-05,
+      "loss": 1.0776,
+      "step": 488
+    },
+    {
+      "epoch": 0.17248296995216156,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.919697920876661e-05,
+      "loss": 1.1002,
+      "step": 489
+    },
+    {
+      "epoch": 0.17283569586208417,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9192394362473065e-05,
+      "loss": 1.0494,
+      "step": 490
+    },
+    {
+      "epoch": 0.1731884217720068,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9187797015513244e-05,
+      "loss": 0.9766,
+      "step": 491
+    },
+    {
+      "epoch": 0.1735411476819294,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9183187174139048e-05,
+      "loss": 0.9678,
+      "step": 492
+    },
+    {
+      "epoch": 0.17389387359185204,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.9178564844619362e-05,
+      "loss": 0.9624,
+      "step": 493
+    },
+    {
+      "epoch": 0.17424659950177465,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9173930033240055e-05,
+      "loss": 0.9656,
+      "step": 494
+    },
+    {
+      "epoch": 0.17459932541169726,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.9169282746303972e-05,
+      "loss": 1.0534,
+      "step": 495
+    },
+    {
+      "epoch": 0.1749520513216199,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9164622990130923e-05,
+      "loss": 1.0111,
+      "step": 496
+    },
+    {
+      "epoch": 0.1753047772315425,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.915995077105768e-05,
+      "loss": 1.0911,
+      "step": 497
+    },
+    {
+      "epoch": 0.17565750314146514,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.915526609543795e-05,
+      "loss": 1.0044,
+      "step": 498
+    },
+    {
+      "epoch": 0.17601022905138775,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9150568969642388e-05,
+      "loss": 0.9679,
+      "step": 499
+    },
+    {
+      "epoch": 0.17636295496131038,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9145859400058584e-05,
+      "loss": 0.9376,
+      "step": 500
+    },
+    {
+      "epoch": 0.176715680871233,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9141137393091043e-05,
+      "loss": 0.9308,
+      "step": 501
+    },
+    {
+      "epoch": 0.17706840678115562,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9136402955161183e-05,
+      "loss": 1.0058,
+      "step": 502
+    },
+    {
+      "epoch": 0.17742113269107823,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9131656092707337e-05,
+      "loss": 0.8984,
+      "step": 503
+    },
+    {
+      "epoch": 0.17777385860100087,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9126896812184718e-05,
+      "loss": 1.0147,
+      "step": 504
+    },
+    {
+      "epoch": 0.17812658451092347,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.912212512006544e-05,
+      "loss": 1.0494,
+      "step": 505
+    },
+    {
+      "epoch": 0.1784793104208461,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.911734102283849e-05,
+      "loss": 0.9816,
+      "step": 506
+    },
+    {
+      "epoch": 0.17883203633076872,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9112544527009725e-05,
+      "loss": 1.008,
+      "step": 507
+    },
+    {
+      "epoch": 0.17918476224069135,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9107735639101862e-05,
+      "loss": 0.9767,
+      "step": 508
+    },
+    {
+      "epoch": 0.17953748815061396,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9102914365654465e-05,
+      "loss": 0.9995,
+      "step": 509
+    },
+    {
+      "epoch": 0.1798902140605366,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9098080713223955e-05,
+      "loss": 1.0055,
+      "step": 510
+    },
+    {
+      "epoch": 0.1802429399704592,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9093234688383583e-05,
+      "loss": 1.0065,
+      "step": 511
+    },
+    {
+      "epoch": 0.18059566588038184,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9088376297723407e-05,
+      "loss": 0.9292,
+      "step": 512
+    },
+    {
+      "epoch": 0.18094839179030445,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.908350554785032e-05,
+      "loss": 0.8879,
+      "step": 513
+    },
+    {
+      "epoch": 0.18130111770022705,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9078622445388024e-05,
+      "loss": 0.9558,
+      "step": 514
+    },
+    {
+      "epoch": 0.1816538436101497,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9073726996977004e-05,
+      "loss": 1.043,
+      "step": 515
+    },
+    {
+      "epoch": 0.1820065695200723,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9068819209274548e-05,
+      "loss": 1.0153,
+      "step": 516
+    },
+    {
+      "epoch": 0.18235929542999493,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9063899088954715e-05,
+      "loss": 0.9683,
+      "step": 517
+    },
+    {
+      "epoch": 0.18271202133991754,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.905896664270834e-05,
+      "loss": 0.9974,
+      "step": 518
+    },
+    {
+      "epoch": 0.18306474724984018,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.905402187724302e-05,
+      "loss": 1.0921,
+      "step": 519
+    },
+    {
+      "epoch": 0.18341747315976278,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9049064799283107e-05,
+      "loss": 1.0459,
+      "step": 520
+    },
+    {
+      "epoch": 0.18377019906968542,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.904409541556968e-05,
+      "loss": 1.056,
+      "step": 521
+    },
+    {
+      "epoch": 0.18412292497960803,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.903911373286058e-05,
+      "loss": 0.9706,
+      "step": 522
+    },
+    {
+      "epoch": 0.18447565088953066,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9034119757930353e-05,
+      "loss": 1.0055,
+      "step": 523
+    },
+    {
+      "epoch": 0.18482837679945327,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.9029113497570266e-05,
+      "loss": 1.0271,
+      "step": 524
+    },
+    {
+      "epoch": 0.1851811027093759,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9024094958588294e-05,
+      "loss": 1.0446,
+      "step": 525
+    },
+    {
+      "epoch": 0.1855338286192985,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9019064147809113e-05,
+      "loss": 1.0482,
+      "step": 526
+    },
+    {
+      "epoch": 0.18588655452922115,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.901402107207408e-05,
+      "loss": 1.059,
+      "step": 527
+    },
+    {
+      "epoch": 0.18623928043914376,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9008965738241236e-05,
+      "loss": 1.0344,
+      "step": 528
+    },
+    {
+      "epoch": 0.1865920063490664,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9003898153185293e-05,
+      "loss": 0.9621,
+      "step": 529
+    },
+    {
+      "epoch": 0.186944732258989,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.8998818323797617e-05,
+      "loss": 0.9139,
+      "step": 530
+    },
+    {
+      "epoch": 0.1872974581689116,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.899372625698623e-05,
+      "loss": 0.9809,
+      "step": 531
+    },
+    {
+      "epoch": 0.18765018407883424,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8988621959675794e-05,
+      "loss": 0.9624,
+      "step": 532
+    },
+    {
+      "epoch": 0.18800290998875685,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.898350543880761e-05,
+      "loss": 1.0383,
+      "step": 533
+    },
+    {
+      "epoch": 0.18835563589867949,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8978376701339583e-05,
+      "loss": 0.9665,
+      "step": 534
+    },
+    {
+      "epoch": 0.1887083618086021,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8973235754246257e-05,
+      "loss": 0.9442,
+      "step": 535
+    },
+    {
+      "epoch": 0.18906108771852473,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8968082604518754e-05,
+      "loss": 1.0764,
+      "step": 536
+    },
+    {
+      "epoch": 0.18941381362844734,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.896291725916481e-05,
+      "loss": 1.0245,
+      "step": 537
+    },
+    {
+      "epoch": 0.18976653953836997,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.895773972520874e-05,
+      "loss": 1.0089,
+      "step": 538
+    },
+    {
+      "epoch": 0.19011926544829258,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.8952550009691424e-05,
+      "loss": 0.9042,
+      "step": 539
+    },
+    {
+      "epoch": 0.19047199135821521,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.8947348119670325e-05,
+      "loss": 0.9374,
+      "step": 540
+    },
+    {
+      "epoch": 0.19082471726813782,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8942134062219448e-05,
+      "loss": 0.9895,
+      "step": 541
+    },
+    {
+      "epoch": 0.19117744317806046,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.893690784442935e-05,
+      "loss": 1.0096,
+      "step": 542
+    },
+    {
+      "epoch": 0.19153016908798307,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.893166947340713e-05,
+      "loss": 1.0552,
+      "step": 543
+    },
+    {
+      "epoch": 0.1918828949979057,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.89264189562764e-05,
+      "loss": 1.0943,
+      "step": 544
+    },
+    {
+      "epoch": 0.1922356209078283,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.8921156300177303e-05,
+      "loss": 1.074,
+      "step": 545
+    },
+    {
+      "epoch": 0.19258834681775094,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.8915881512266483e-05,
+      "loss": 0.988,
+      "step": 546
+    },
+    {
+      "epoch": 0.19294107272767355,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8910594599717082e-05,
+      "loss": 0.9767,
+      "step": 547
+    },
+    {
+      "epoch": 0.19329379863759616,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8905295569718734e-05,
+      "loss": 0.8705,
+      "step": 548
+    },
+    {
+      "epoch": 0.1936465245475188,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8899984429477554e-05,
+      "loss": 1.0319,
+      "step": 549
+    },
+    {
+      "epoch": 0.1939992504574414,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8894661186216108e-05,
+      "loss": 0.9913,
+      "step": 550
+    },
+    {
+      "epoch": 0.19435197636736404,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.8889325847173444e-05,
+      "loss": 1.0296,
+      "step": 551
+    },
+    {
+      "epoch": 0.19470470227728665,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.8883978419605042e-05,
+      "loss": 0.9197,
+      "step": 552
+    },
+    {
+      "epoch": 0.19505742818720928,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.8878618910782834e-05,
+      "loss": 0.9475,
+      "step": 553
+    },
+    {
+      "epoch": 0.1954101540971319,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8873247327995168e-05,
+      "loss": 1.0234,
+      "step": 554
+    },
+    {
+      "epoch": 0.19576288000705452,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.8867863678546823e-05,
+      "loss": 0.9979,
+      "step": 555
+    },
+    {
+      "epoch": 0.19611560591697713,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.886246796975898e-05,
+      "loss": 0.9285,
+      "step": 556
+    },
+    {
+      "epoch": 0.19646833182689977,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8857060208969225e-05,
+      "loss": 1.0024,
+      "step": 557
+    },
+    {
+      "epoch": 0.19682105773682237,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.8851640403531524e-05,
+      "loss": 0.8489,
+      "step": 558
+    },
+    {
+      "epoch": 0.197173783646745,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8846208560816236e-05,
+      "loss": 1.035,
+      "step": 559
+    },
+    {
+      "epoch": 0.19752650955666762,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.8840764688210074e-05,
+      "loss": 1.0814,
+      "step": 560
+    },
+    {
+      "epoch": 0.19787923546659025,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.8835308793116126e-05,
+      "loss": 0.9326,
+      "step": 561
+    },
+    {
+      "epoch": 0.19823196137651286,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.882984088295382e-05,
+      "loss": 1.0715,
+      "step": 562
+    },
+    {
+      "epoch": 0.1985846872864355,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.882436096515892e-05,
+      "loss": 1.0664,
+      "step": 563
+    },
+    {
+      "epoch": 0.1989374131963581,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.881886904718353e-05,
+      "loss": 1.0361,
+      "step": 564
+    },
+    {
+      "epoch": 0.19929013910628074,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.8813365136496062e-05,
+      "loss": 1.0147,
+      "step": 565
+    },
+    {
+      "epoch": 0.19964286501620335,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.8807849240581247e-05,
+      "loss": 0.941,
+      "step": 566
+    },
+    {
+      "epoch": 0.19999559092612595,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.8802321366940103e-05,
+      "loss": 0.9849,
+      "step": 567
+    },
+    {
+      "epoch": 0.2003483168360486,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.879678152308995e-05,
+      "loss": 0.9806,
+      "step": 568
+    },
+    {
+      "epoch": 0.2007010427459712,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.8791229716564374e-05,
+      "loss": 1.1239,
+      "step": 569
+    },
+    {
+      "epoch": 0.20105376865589383,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.8785665954913238e-05,
+      "loss": 1.0748,
+      "step": 570
+    },
+    {
+      "epoch": 0.20140649456581644,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.878009024570265e-05,
+      "loss": 0.955,
+      "step": 571
+    },
+    {
+      "epoch": 0.20175922047573908,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.8774502596514984e-05,
+      "loss": 0.9723,
+      "step": 572
+    },
+    {
+      "epoch": 0.20211194638566168,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.8768903014948838e-05,
+      "loss": 0.9836,
+      "step": 573
+    },
+    {
+      "epoch": 0.20246467229558432,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.8763291508619035e-05,
+      "loss": 0.9589,
+      "step": 574
+    },
+    {
+      "epoch": 0.20281739820550693,
+      "grad_norm": 0.75,
+      "learning_rate": 1.8757668085156628e-05,
+      "loss": 0.9172,
+      "step": 575
+    },
+    {
+      "epoch": 0.20317012411542956,
+      "grad_norm": 0.75,
+      "learning_rate": 1.8752032752208856e-05,
+      "loss": 0.9668,
+      "step": 576
+    },
+    {
+      "epoch": 0.20352285002535217,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.8746385517439173e-05,
+      "loss": 0.9867,
+      "step": 577
+    },
+    {
+      "epoch": 0.2038755759352748,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.874072638852721e-05,
+      "loss": 0.8649,
+      "step": 578
+    },
+    {
+      "epoch": 0.2042283018451974,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.873505537316877e-05,
+      "loss": 0.8687,
+      "step": 579
+    },
+    {
+      "epoch": 0.20458102775512005,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.8729372479075825e-05,
+      "loss": 0.9594,
+      "step": 580
+    },
+    {
+      "epoch": 0.20493375366504266,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.8723677713976494e-05,
+      "loss": 1.04,
+      "step": 581
+    },
+    {
+      "epoch": 0.2052864795749653,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.871797108561505e-05,
+      "loss": 1.0669,
+      "step": 582
+    },
+    {
+      "epoch": 0.2056392054848879,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.8712252601751893e-05,
+      "loss": 1.0372,
+      "step": 583
+    },
+    {
+      "epoch": 0.2059919313948105,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.8706522270163545e-05,
+      "loss": 1.0683,
+      "step": 584
+    },
+    {
+      "epoch": 0.20634465730473314,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8700780098642636e-05,
+      "loss": 1.0125,
+      "step": 585
+    },
+    {
+      "epoch": 0.20669738321465575,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.86950260949979e-05,
+      "loss": 0.9962,
+      "step": 586
+    },
+    {
+      "epoch": 0.20705010912457839,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.868926026705417e-05,
+      "loss": 0.9606,
+      "step": 587
+    },
+    {
+      "epoch": 0.207402835034501,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.868348262265234e-05,
+      "loss": 0.9519,
+      "step": 588
+    },
+    {
+      "epoch": 0.20775556094442363,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.8677693169649393e-05,
+      "loss": 1.0144,
+      "step": 589
+    },
+    {
+      "epoch": 0.20810828685434624,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.867189191591836e-05,
+      "loss": 0.9148,
+      "step": 590
+    },
+    {
+      "epoch": 0.20846101276426887,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.866607886934832e-05,
+      "loss": 0.9645,
+      "step": 591
+    },
+    {
+      "epoch": 0.20881373867419148,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 1.0131,
+      "step": 592
+    },
+    {
+      "epoch": 0.20916646458411411,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.865441742932771e-05,
+      "loss": 0.9361,
+      "step": 593
+    },
+    {
+      "epoch": 0.20951919049403672,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.864856905173545e-05,
+      "loss": 0.9302,
+      "step": 594
+    },
+    {
+      "epoch": 0.20987191640395936,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.8642708913020756e-05,
+      "loss": 1.0698,
+      "step": 595
+    },
+    {
+      "epoch": 0.21022464231388197,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.8636837021152803e-05,
+      "loss": 0.9186,
+      "step": 596
+    },
+    {
+      "epoch": 0.2105773682238046,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.8630953384116722e-05,
+      "loss": 1.0223,
+      "step": 597
+    },
+    {
+      "epoch": 0.2109300941337272,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.8625058009913627e-05,
+      "loss": 0.9891,
+      "step": 598
+    },
+    {
+      "epoch": 0.21128282004364984,
+      "grad_norm": 0.75,
+      "learning_rate": 1.86191509065606e-05,
+      "loss": 1.0292,
+      "step": 599
+    },
+    {
+      "epoch": 0.21163554595357245,
+      "grad_norm": 0.75,
+      "learning_rate": 1.861323208209065e-05,
+      "loss": 0.969,
+      "step": 600
+    },
+    {
+      "epoch": 0.21198827186349506,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.8607301544552752e-05,
+      "loss": 0.9548,
+      "step": 601
+    },
+    {
+      "epoch": 0.2123409977734177,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.8601359302011795e-05,
+      "loss": 0.879,
+      "step": 602
+    },
+    {
+      "epoch": 0.2126937236833403,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.859540536254859e-05,
+      "loss": 1.0281,
+      "step": 603
+    },
+    {
+      "epoch": 0.21304644959326294,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.858943973425985e-05,
+      "loss": 1.0357,
+      "step": 604
+    },
+    {
+      "epoch": 0.21339917550318555,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.8583462425258192e-05,
+      "loss": 0.9963,
+      "step": 605
+    },
+    {
+      "epoch": 0.21375190141310818,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.857747344367211e-05,
+      "loss": 0.974,
+      "step": 606
+    },
+    {
+      "epoch": 0.2141046273230308,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.8571472797645973e-05,
+      "loss": 1.1248,
+      "step": 607
+    },
+    {
+      "epoch": 0.21445735323295342,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.8565460495340017e-05,
+      "loss": 1.0289,
+      "step": 608
+    },
+    {
+      "epoch": 0.21481007914287603,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.8559436544930318e-05,
+      "loss": 0.9794,
+      "step": 609
+    },
+    {
+      "epoch": 0.21516280505279867,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.8553400954608806e-05,
+      "loss": 0.9116,
+      "step": 610
+    },
+    {
+      "epoch": 0.21551553096272127,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.8547353732583234e-05,
+      "loss": 1.0435,
+      "step": 611
+    },
+    {
+      "epoch": 0.2158682568726439,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.854129488707717e-05,
+      "loss": 0.9652,
+      "step": 612
+    },
+    {
+      "epoch": 0.21622098278256652,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.8535224426329992e-05,
+      "loss": 1.0545,
+      "step": 613
+    },
+    {
+      "epoch": 0.21657370869248915,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8529142358596877e-05,
+      "loss": 0.8997,
+      "step": 614
+    },
+    {
+      "epoch": 0.21692643460241176,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.852304869214877e-05,
+      "loss": 0.9896,
+      "step": 615
+    },
+    {
+      "epoch": 0.2172791605123344,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.8516943435272417e-05,
+      "loss": 1.0055,
+      "step": 616
+    },
+    {
+      "epoch": 0.217631886422257,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.85108265962703e-05,
+      "loss": 0.9077,
+      "step": 617
+    },
+    {
+      "epoch": 0.21798461233217964,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.8504698183460662e-05,
+      "loss": 0.9351,
+      "step": 618
+    },
+    {
+      "epoch": 0.21833733824210225,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.8498558205177487e-05,
+      "loss": 0.9411,
+      "step": 619
+    },
+    {
+      "epoch": 0.21869006415202485,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.849240666977048e-05,
+      "loss": 0.9504,
+      "step": 620
+    },
+    {
+      "epoch": 0.2190427900619475,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.848624358560507e-05,
+      "loss": 1.0742,
+      "step": 621
+    },
+    {
+      "epoch": 0.2193955159718701,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.8480068961062388e-05,
+      "loss": 1.0484,
+      "step": 622
+    },
+    {
+      "epoch": 0.21974824188179273,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.847388280453926e-05,
+      "loss": 1.0187,
+      "step": 623
+    },
+    {
+      "epoch": 0.22010096779171534,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.8467685124448186e-05,
+      "loss": 1.0029,
+      "step": 624
+    },
+    {
+      "epoch": 0.22045369370163798,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.8461475929217353e-05,
+      "loss": 0.9887,
+      "step": 625
+    },
+    {
+      "epoch": 0.22080641961156058,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.8455255227290588e-05,
+      "loss": 0.9413,
+      "step": 626
+    },
+    {
+      "epoch": 0.22115914552148322,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.844902302712738e-05,
+      "loss": 1.0943,
+      "step": 627
+    },
+    {
+      "epoch": 0.22151187143140583,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.844277933720285e-05,
+      "loss": 1.0302,
+      "step": 628
+    },
+    {
+      "epoch": 0.22186459734132846,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.8436524166007742e-05,
+      "loss": 0.8825,
+      "step": 629
+    },
+    {
+      "epoch": 0.22221732325125107,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.8430257522048416e-05,
+      "loss": 1.0155,
+      "step": 630
+    },
+    {
+      "epoch": 0.2225700491611737,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.8423979413846838e-05,
+      "loss": 1.0011,
+      "step": 631
+    },
+    {
+      "epoch": 0.2229227750710963,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.8417689849940552e-05,
+      "loss": 1.0321,
+      "step": 632
+    },
+    {
+      "epoch": 0.22327550098101895,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.841138883888269e-05,
+      "loss": 1.0117,
+      "step": 633
+    },
+    {
+      "epoch": 0.22362822689094156,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.8405076389241948e-05,
+      "loss": 0.9083,
+      "step": 634
+    },
+    {
+      "epoch": 0.2239809528008642,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.839875250960258e-05,
+      "loss": 0.9817,
+      "step": 635
+    },
+    {
+      "epoch": 0.2243336787107868,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8392417208564377e-05,
+      "loss": 1.0552,
+      "step": 636
+    },
+    {
+      "epoch": 0.2246864046207094,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.8386070494742672e-05,
+      "loss": 1.0184,
+      "step": 637
+    },
+    {
+      "epoch": 0.22503913053063204,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8379712376768304e-05,
+      "loss": 1.0443,
+      "step": 638
+    },
+    {
+      "epoch": 0.22539185644055465,
+      "grad_norm": 0.75,
+      "learning_rate": 1.837334286328764e-05,
+      "loss": 0.9986,
+      "step": 639
+    },
+    {
+      "epoch": 0.22574458235047729,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8366961962962517e-05,
+      "loss": 0.9091,
+      "step": 640
+    },
+    {
+      "epoch": 0.2260973082603999,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.8360569684470286e-05,
+      "loss": 0.9964,
+      "step": 641
+    },
+    {
+      "epoch": 0.22645003417032253,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.835416603650375e-05,
+      "loss": 0.9967,
+      "step": 642
+    },
+    {
+      "epoch": 0.22680276008024514,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.8347751027771185e-05,
+      "loss": 1.0296,
+      "step": 643
+    },
+    {
+      "epoch": 0.22715548599016777,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.8341324666996307e-05,
+      "loss": 1.0251,
+      "step": 644
+    },
+    {
+      "epoch": 0.22750821190009038,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.8334886962918274e-05,
+      "loss": 1.0689,
+      "step": 645
+    },
+    {
+      "epoch": 0.22786093781001301,
+      "grad_norm": 0.75,
+      "learning_rate": 1.832843792429168e-05,
+      "loss": 0.985,
+      "step": 646
+    },
+    {
+      "epoch": 0.22821366371993562,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8321977559886514e-05,
+      "loss": 0.9986,
+      "step": 647
+    },
+    {
+      "epoch": 0.22856638962985826,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.8315505878488174e-05,
+      "loss": 0.9553,
+      "step": 648
+    },
+    {
+      "epoch": 0.22891911553978087,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.830902288889746e-05,
+      "loss": 0.9804,
+      "step": 649
+    },
+    {
+      "epoch": 0.2292718414497035,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.8302528599930526e-05,
+      "loss": 1.0425,
+      "step": 650
+    },
+    {
+      "epoch": 0.2296245673596261,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8296023020418917e-05,
+      "loss": 1.0213,
+      "step": 651
+    },
+    {
+      "epoch": 0.22997729326954874,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.8289506159209517e-05,
+      "loss": 0.958,
+      "step": 652
+    },
+    {
+      "epoch": 0.23033001917947135,
+      "grad_norm": 0.75,
+      "learning_rate": 1.8282978025164553e-05,
+      "loss": 1.0171,
+      "step": 653
+    },
+    {
+      "epoch": 0.23068274508939396,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.8276438627161586e-05,
+      "loss": 0.9414,
+      "step": 654
+    },
+    {
+      "epoch": 0.2310354709993166,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.826988797409349e-05,
+      "loss": 1.0632,
+      "step": 655
+    },
+    {
+      "epoch": 0.2313881969092392,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8263326074868456e-05,
+      "loss": 1.1421,
+      "step": 656
+    },
+    {
+      "epoch": 0.23174092281916184,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.8256752938409953e-05,
+      "loss": 1.0069,
+      "step": 657
+    },
+    {
+      "epoch": 0.23209364872908445,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.8250168573656736e-05,
+      "loss": 0.9596,
+      "step": 658
+    },
+    {
+      "epoch": 0.23244637463900708,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.824357298956284e-05,
+      "loss": 0.9765,
+      "step": 659
+    },
+    {
+      "epoch": 0.2327991005489297,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.823696619509754e-05,
+      "loss": 0.9405,
+      "step": 660
+    },
+    {
+      "epoch": 0.23315182645885232,
+      "grad_norm": 0.75,
+      "learning_rate": 1.8230348199245377e-05,
+      "loss": 0.9792,
+      "step": 661
+    },
+    {
+      "epoch": 0.23350455236877493,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.8223719011006104e-05,
+      "loss": 0.9061,
+      "step": 662
+    },
+    {
+      "epoch": 0.23385727827869757,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.82170786393947e-05,
+      "loss": 0.9786,
+      "step": 663
+    },
+    {
+      "epoch": 0.23421000418862017,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.8210427093441363e-05,
+      "loss": 0.9569,
+      "step": 664
+    },
+    {
+      "epoch": 0.2345627300985428,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8203764382191476e-05,
+      "loss": 0.9626,
+      "step": 665
+    },
+    {
+      "epoch": 0.23491545600846542,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.8197090514705606e-05,
+      "loss": 0.9125,
+      "step": 666
+    },
+    {
+      "epoch": 0.23526818191838805,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.8190405500059502e-05,
+      "loss": 0.9638,
+      "step": 667
+    },
+    {
+      "epoch": 0.23562090782831066,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.8183709347344052e-05,
+      "loss": 0.9243,
+      "step": 668
+    },
+    {
+      "epoch": 0.2359736337382333,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.8177002065665312e-05,
+      "loss": 0.9834,
+      "step": 669
+    },
+    {
+      "epoch": 0.2363263596481559,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.8170283664144458e-05,
+      "loss": 0.9332,
+      "step": 670
+    },
+    {
+      "epoch": 0.23667908555807854,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8163554151917796e-05,
+      "loss": 0.8725,
+      "step": 671
+    },
+    {
+      "epoch": 0.23703181146800115,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8156813538136736e-05,
+      "loss": 0.9775,
+      "step": 672
+    },
+    {
+      "epoch": 0.23738453737792375,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.8150061831967786e-05,
+      "loss": 0.9808,
+      "step": 673
+    },
+    {
+      "epoch": 0.2377372632878464,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.814329904259254e-05,
+      "loss": 0.983,
+      "step": 674
+    },
+    {
+      "epoch": 0.238089989197769,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.8136525179207662e-05,
+      "loss": 0.9581,
+      "step": 675
+    },
+    {
+      "epoch": 0.23844271510769163,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.8129740251024878e-05,
+      "loss": 0.9902,
+      "step": 676
+    },
+    {
+      "epoch": 0.23879544101761424,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.812294426727096e-05,
+      "loss": 0.9617,
+      "step": 677
+    },
+    {
+      "epoch": 0.23914816692753688,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.8116137237187713e-05,
+      "loss": 0.9068,
+      "step": 678
+    },
+    {
+      "epoch": 0.23950089283745948,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.8109319170031966e-05,
+      "loss": 0.8448,
+      "step": 679
+    },
+    {
+      "epoch": 0.23985361874738212,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.8102490075075553e-05,
+      "loss": 0.9889,
+      "step": 680
+    },
+    {
+      "epoch": 0.24020634465730473,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.809564996160531e-05,
+      "loss": 1.0287,
+      "step": 681
+    },
+    {
+      "epoch": 0.24055907056722736,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.8088798838923054e-05,
+      "loss": 0.9922,
+      "step": 682
+    },
+    {
+      "epoch": 0.24091179647714997,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.8081936716345574e-05,
+      "loss": 0.9978,
+      "step": 683
+    },
+    {
+      "epoch": 0.2412645223870726,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.8075063603204618e-05,
+      "loss": 0.9694,
+      "step": 684
+    },
+    {
+      "epoch": 0.2416172482969952,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.806817950884688e-05,
+      "loss": 0.9066,
+      "step": 685
+    },
+    {
+      "epoch": 0.24196997420691785,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.8061284442633986e-05,
+      "loss": 0.9329,
+      "step": 686
+    },
+    {
+      "epoch": 0.24232270011684046,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.805437841394248e-05,
+      "loss": 0.9187,
+      "step": 687
+    },
+    {
+      "epoch": 0.2426754260267631,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.8047461432163826e-05,
+      "loss": 0.9524,
+      "step": 688
+    },
+    {
+      "epoch": 0.2430281519366857,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.804053350670437e-05,
+      "loss": 1.1422,
+      "step": 689
+    },
+    {
+      "epoch": 0.2433808778466083,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.803359464698534e-05,
+      "loss": 0.9033,
+      "step": 690
+    },
+    {
+      "epoch": 0.24373360375653094,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.8026644862442843e-05,
+      "loss": 1.0562,
+      "step": 691
+    },
+    {
+      "epoch": 0.24408632966645355,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.8019684162527834e-05,
+      "loss": 0.9112,
+      "step": 692
+    },
+    {
+      "epoch": 0.24443905557637619,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8012712556706117e-05,
+      "loss": 0.9331,
+      "step": 693
+    },
+    {
+      "epoch": 0.2447917814862988,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.8005730054458328e-05,
+      "loss": 0.9761,
+      "step": 694
+    },
+    {
+      "epoch": 0.24514450739622143,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.7998736665279914e-05,
+      "loss": 1.0185,
+      "step": 695
+    },
+    {
+      "epoch": 0.24549723330614404,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.7991732398681132e-05,
+      "loss": 0.9192,
+      "step": 696
+    },
+    {
+      "epoch": 0.24584995921606667,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.798471726418703e-05,
+      "loss": 0.9506,
+      "step": 697
+    },
+    {
+      "epoch": 0.24620268512598928,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.7977691271337438e-05,
+      "loss": 1.044,
+      "step": 698
+    },
+    {
+      "epoch": 0.24655541103591191,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.7970654429686944e-05,
+      "loss": 0.9968,
+      "step": 699
+    },
+    {
+      "epoch": 0.24690813694583452,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.79636067488049e-05,
+      "loss": 1.0569,
+      "step": 700
+    },
+    {
+      "epoch": 0.24726086285575716,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7956548238275387e-05,
+      "loss": 0.9074,
+      "step": 701
+    },
+    {
+      "epoch": 0.24761358876567977,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.7949478907697228e-05,
+      "loss": 0.9825,
+      "step": 702
+    },
+    {
+      "epoch": 0.2479663146756024,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.7942398766683945e-05,
+      "loss": 0.925,
+      "step": 703
+    },
+    {
+      "epoch": 0.248319040585525,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.7935307824863767e-05,
+      "loss": 1.0548,
+      "step": 704
+    },
+    {
+      "epoch": 0.24867176649544764,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.7928206091879614e-05,
+      "loss": 0.9776,
+      "step": 705
+    },
+    {
+      "epoch": 0.24902449240537025,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.7921093577389076e-05,
+      "loss": 1.0272,
+      "step": 706
+    },
+    {
+      "epoch": 0.24937721831529286,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.791397029106441e-05,
+      "loss": 0.9806,
+      "step": 707
+    },
+    {
+      "epoch": 0.2497299442252155,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.790683624259251e-05,
+      "loss": 0.9718,
+      "step": 708
+    },
+    {
+      "epoch": 0.25008267013513813,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.789969144167492e-05,
+      "loss": 0.9514,
+      "step": 709
+    },
+    {
+      "epoch": 0.2504353960450607,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.7892535898027804e-05,
+      "loss": 0.9675,
+      "step": 710
+    },
+    {
+      "epoch": 0.25078812195498335,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.7885369621381922e-05,
+      "loss": 1.0766,
+      "step": 711
+    },
+    {
+      "epoch": 0.251140847864906,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.7878192621482638e-05,
+      "loss": 0.9015,
+      "step": 712
+    },
+    {
+      "epoch": 0.2514935737748286,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.787100490808991e-05,
+      "loss": 0.9856,
+      "step": 713
+    },
+    {
+      "epoch": 0.2518462996847512,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.7863806490978243e-05,
+      "loss": 0.9837,
+      "step": 714
+    },
+    {
+      "epoch": 0.25219902559467383,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.7856597379936713e-05,
+      "loss": 1.0258,
+      "step": 715
+    },
+    {
+      "epoch": 0.25255175150459647,
+      "grad_norm": 5.25,
+      "learning_rate": 1.7849377584768934e-05,
+      "loss": 0.9867,
+      "step": 716
+    },
+    {
+      "epoch": 0.2529044774145191,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.7842147115293048e-05,
+      "loss": 1.0021,
+      "step": 717
+    },
+    {
+      "epoch": 0.2532572033244417,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.783490598134172e-05,
+      "loss": 0.9249,
+      "step": 718
+    },
+    {
+      "epoch": 0.2536099292343643,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.782765419276211e-05,
+      "loss": 0.9893,
+      "step": 719
+    },
+    {
+      "epoch": 0.25396265514428695,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.7820391759415868e-05,
+      "loss": 0.979,
+      "step": 720
+    },
+    {
+      "epoch": 0.2543153810542096,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.7813118691179122e-05,
+      "loss": 1.0259,
+      "step": 721
+    },
+    {
+      "epoch": 0.25466810696413217,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.7805834997942456e-05,
+      "loss": 1.0119,
+      "step": 722
+    },
+    {
+      "epoch": 0.2550208328740548,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.779854068961092e-05,
+      "loss": 0.9594,
+      "step": 723
+    },
+    {
+      "epoch": 0.25537355878397744,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.7791235776103976e-05,
+      "loss": 0.9852,
+      "step": 724
+    },
+    {
+      "epoch": 0.2557262846939,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.7783920267355527e-05,
+      "loss": 0.8684,
+      "step": 725
+    },
+    {
+      "epoch": 0.25607901060382265,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.7776594173313877e-05,
+      "loss": 0.9039,
+      "step": 726
+    },
+    {
+      "epoch": 0.2564317365137453,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.776925750394172e-05,
+      "loss": 0.9715,
+      "step": 727
+    },
+    {
+      "epoch": 0.2567844624236679,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.7761910269216138e-05,
+      "loss": 0.9087,
+      "step": 728
+    },
+    {
+      "epoch": 0.2571371883335905,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.775455247912858e-05,
+      "loss": 0.9786,
+      "step": 729
+    },
+    {
+      "epoch": 0.25748991424351314,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.774718414368485e-05,
+      "loss": 0.8384,
+      "step": 730
+    },
+    {
+      "epoch": 0.2578426401534358,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.7739805272905087e-05,
+      "loss": 0.9543,
+      "step": 731
+    },
+    {
+      "epoch": 0.2581953660633584,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.773241587682376e-05,
+      "loss": 1.0125,
+      "step": 732
+    },
+    {
+      "epoch": 0.258548091973281,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.7725015965489656e-05,
+      "loss": 1.085,
+      "step": 733
+    },
+    {
+      "epoch": 0.2589008178832036,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.7717605548965856e-05,
+      "loss": 0.9544,
+      "step": 734
+    },
+    {
+      "epoch": 0.25925354379312626,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.7710184637329722e-05,
+      "loss": 0.9293,
+      "step": 735
+    },
+    {
+      "epoch": 0.2596062697030489,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.7702753240672907e-05,
+      "loss": 1.027,
+      "step": 736
+    },
+    {
+      "epoch": 0.2599589956129715,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.76953113691013e-05,
+      "loss": 0.9761,
+      "step": 737
+    },
+    {
+      "epoch": 0.2603117215228941,
+      "grad_norm": 0.75,
+      "learning_rate": 1.768785903273505e-05,
+      "loss": 0.9638,
+      "step": 738
+    },
+    {
+      "epoch": 0.26066444743281675,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.7680396241708526e-05,
+      "loss": 0.874,
+      "step": 739
+    },
+    {
+      "epoch": 0.2610171733427394,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.7672923006170324e-05,
+      "loss": 0.9396,
+      "step": 740
+    },
+    {
+      "epoch": 0.26136989925266196,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.7665439336283243e-05,
+      "loss": 0.9031,
+      "step": 741
+    },
+    {
+      "epoch": 0.2617226251625846,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.7657945242224256e-05,
+      "loss": 0.896,
+      "step": 742
+    },
+    {
+      "epoch": 0.26207535107250723,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.765044073418454e-05,
+      "loss": 0.9423,
+      "step": 743
+    },
+    {
+      "epoch": 0.2624280769824298,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.7642925822369406e-05,
+      "loss": 0.9348,
+      "step": 744
+    },
+    {
+      "epoch": 0.26278080289235245,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.763540051699833e-05,
+      "loss": 0.9675,
+      "step": 745
+    },
+    {
+      "epoch": 0.2631335288022751,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.7627864828304916e-05,
+      "loss": 0.9826,
+      "step": 746
+    },
+    {
+      "epoch": 0.2634862547121977,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.7620318766536883e-05,
+      "loss": 0.9548,
+      "step": 747
+    },
+    {
+      "epoch": 0.2638389806221203,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.7612762341956072e-05,
+      "loss": 0.951,
+      "step": 748
+    },
+    {
+      "epoch": 0.26419170653204294,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.76051955648384e-05,
+      "loss": 1.0383,
+      "step": 749
+    },
+    {
+      "epoch": 0.26454443244196557,
+      "grad_norm": 0.75,
+      "learning_rate": 1.759761844547387e-05,
+      "loss": 0.9519,
+      "step": 750
+    },
+    {
+      "epoch": 0.2648971583518882,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.7590030994166553e-05,
+      "loss": 0.9753,
+      "step": 751
+    },
+    {
+      "epoch": 0.2652498842618108,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.7582433221234557e-05,
+      "loss": 0.9519,
+      "step": 752
+    },
+    {
+      "epoch": 0.2656026101717334,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.757482513701004e-05,
+      "loss": 0.9694,
+      "step": 753
+    },
+    {
+      "epoch": 0.26595533608165606,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.756720675183918e-05,
+      "loss": 0.8695,
+      "step": 754
+    },
+    {
+      "epoch": 0.2663080619915787,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.7559578076082156e-05,
+      "loss": 0.964,
+      "step": 755
+    },
+    {
+      "epoch": 0.2666607879015013,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.7551939120113145e-05,
+      "loss": 0.9434,
+      "step": 756
+    },
+    {
+      "epoch": 0.2670135138114239,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.7544289894320307e-05,
+      "loss": 1.1606,
+      "step": 757
+    },
+    {
+      "epoch": 0.26736623972134654,
+      "grad_norm": 5.5,
+      "learning_rate": 1.753663040910576e-05,
+      "loss": 0.9369,
+      "step": 758
+    },
+    {
+      "epoch": 0.2677189656312692,
+      "grad_norm": 0.75,
+      "learning_rate": 1.7528960674885588e-05,
+      "loss": 1.0033,
+      "step": 759
+    },
+    {
+      "epoch": 0.26807169154119176,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.7521280702089797e-05,
+      "loss": 0.9316,
+      "step": 760
+    },
+    {
+      "epoch": 0.2684244174511144,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.751359050116232e-05,
+      "loss": 0.9938,
+      "step": 761
+    },
+    {
+      "epoch": 0.26877714336103703,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.750589008256101e-05,
+      "loss": 1.0566,
+      "step": 762
+    },
+    {
+      "epoch": 0.2691298692709596,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.74981794567576e-05,
+      "loss": 0.9871,
+      "step": 763
+    },
+    {
+      "epoch": 0.26948259518088225,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.7490458634237718e-05,
+      "loss": 1.0069,
+      "step": 764
+    },
+    {
+      "epoch": 0.2698353210908049,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.7482727625500846e-05,
+      "loss": 1.0184,
+      "step": 765
+    },
+    {
+      "epoch": 0.2701880470007275,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.747498644106032e-05,
+      "loss": 0.9934,
+      "step": 766
+    },
+    {
+      "epoch": 0.2705407729106501,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.7467235091443326e-05,
+      "loss": 0.9992,
+      "step": 767
+    },
+    {
+      "epoch": 0.27089349882057273,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.745947358719086e-05,
+      "loss": 0.9859,
+      "step": 768
+    },
+    {
+      "epoch": 0.27124622473049537,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.745170193885773e-05,
+      "loss": 0.9323,
+      "step": 769
+    },
+    {
+      "epoch": 0.271598950640418,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.7443920157012538e-05,
+      "loss": 0.9447,
+      "step": 770
+    },
+    {
+      "epoch": 0.2719516765503406,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.7436128252237678e-05,
+      "loss": 0.9307,
+      "step": 771
+    },
+    {
+      "epoch": 0.2723044024602632,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.7428326235129294e-05,
+      "loss": 0.9239,
+      "step": 772
+    },
+    {
+      "epoch": 0.27265712837018585,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.7420514116297294e-05,
+      "loss": 0.9506,
+      "step": 773
+    },
+    {
+      "epoch": 0.2730098542801085,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.7412691906365317e-05,
+      "loss": 0.9475,
+      "step": 774
+    },
+    {
+      "epoch": 0.27336258019003107,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.740485961597072e-05,
+      "loss": 1.0328,
+      "step": 775
+    },
+    {
+      "epoch": 0.2737153060999537,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.7397017255764583e-05,
+      "loss": 1.0248,
+      "step": 776
+    },
+    {
+      "epoch": 0.27406803200987634,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.7389164836411663e-05,
+      "loss": 0.9763,
+      "step": 777
+    },
+    {
+      "epoch": 0.2744207579197989,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.7381302368590417e-05,
+      "loss": 0.9405,
+      "step": 778
+    },
+    {
+      "epoch": 0.27477348382972155,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.737342986299294e-05,
+      "loss": 1.0008,
+      "step": 779
+    },
+    {
+      "epoch": 0.2751262097396442,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.7365547330325e-05,
+      "loss": 0.9694,
+      "step": 780
+    },
+    {
+      "epoch": 0.2754789356495668,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.735765478130599e-05,
+      "loss": 0.905,
+      "step": 781
+    },
+    {
+      "epoch": 0.2758316615594894,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.734975222666893e-05,
+      "loss": 1.076,
+      "step": 782
+    },
+    {
+      "epoch": 0.27618438746941204,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.734183967716044e-05,
+      "loss": 0.9141,
+      "step": 783
+    },
+    {
+      "epoch": 0.2765371133793347,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7333917143540734e-05,
+      "loss": 0.9778,
+      "step": 784
+    },
+    {
+      "epoch": 0.2768898392892573,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7325984636583606e-05,
+      "loss": 0.9578,
+      "step": 785
+    },
+    {
+      "epoch": 0.2772425651991799,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.7318042167076413e-05,
+      "loss": 0.9661,
+      "step": 786
+    },
+    {
+      "epoch": 0.2775952911091025,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.731008974582005e-05,
+      "loss": 1.0347,
+      "step": 787
+    },
+    {
+      "epoch": 0.27794801701902516,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.7302127383628966e-05,
+      "loss": 1.0427,
+      "step": 788
+    },
+    {
+      "epoch": 0.2783007429289478,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.729415509133111e-05,
+      "loss": 0.9827,
+      "step": 789
+    },
+    {
+      "epoch": 0.2786534688388704,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.7286172879767935e-05,
+      "loss": 1.0259,
+      "step": 790
+    },
+    {
+      "epoch": 0.279006194748793,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.7278180759794397e-05,
+      "loss": 1.0111,
+      "step": 791
+    },
+    {
+      "epoch": 0.27935892065871565,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.7270178742278915e-05,
+      "loss": 0.9789,
+      "step": 792
+    },
+    {
+      "epoch": 0.2797116465686383,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.7262166838103373e-05,
+      "loss": 0.8851,
+      "step": 793
+    },
+    {
+      "epoch": 0.28006437247856086,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.7254145058163097e-05,
+      "loss": 1.0636,
+      "step": 794
+    },
+    {
+      "epoch": 0.2804170983884835,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.7246113413366843e-05,
+      "loss": 1.0061,
+      "step": 795
+    },
+    {
+      "epoch": 0.28076982429840613,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.7238071914636782e-05,
+      "loss": 0.9642,
+      "step": 796
+    },
+    {
+      "epoch": 0.2811225502083287,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.723002057290849e-05,
+      "loss": 0.9053,
+      "step": 797
+    },
+    {
+      "epoch": 0.28147527611825135,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.722195939913092e-05,
+      "loss": 0.9396,
+      "step": 798
+    },
+    {
+      "epoch": 0.281828002028174,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.72138884042664e-05,
+      "loss": 1.0239,
+      "step": 799
+    },
+    {
+      "epoch": 0.2821807279380966,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.720580759929062e-05,
+      "loss": 1.0068,
+      "step": 800
+    },
+    {
+      "epoch": 0.2825334538480192,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.7197716995192596e-05,
+      "loss": 0.9746,
+      "step": 801
+    },
+    {
+      "epoch": 0.28288617975794184,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.7189616602974687e-05,
+      "loss": 0.9921,
+      "step": 802
+    },
+    {
+      "epoch": 0.28323890566786447,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.7181506433652545e-05,
+      "loss": 1.0413,
+      "step": 803
+    },
+    {
+      "epoch": 0.2835916315777871,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.7173386498255133e-05,
+      "loss": 0.9603,
+      "step": 804
+    },
+    {
+      "epoch": 0.2839443574877097,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.7165256807824686e-05,
+      "loss": 1.021,
+      "step": 805
+    },
+    {
+      "epoch": 0.2842970833976323,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.7157117373416713e-05,
+      "loss": 0.9143,
+      "step": 806
+    },
+    {
+      "epoch": 0.28464980930755496,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.7148968206099956e-05,
+      "loss": 0.9631,
+      "step": 807
+    },
+    {
+      "epoch": 0.2850025352174776,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.7140809316956413e-05,
+      "loss": 1.06,
+      "step": 808
+    },
+    {
+      "epoch": 0.2853552611274002,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.71326407170813e-05,
+      "loss": 1.0486,
+      "step": 809
+    },
+    {
+      "epoch": 0.2857079870373228,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.7124462417583023e-05,
+      "loss": 1.0422,
+      "step": 810
+    },
+    {
+      "epoch": 0.28606071294724544,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.711627442958319e-05,
+      "loss": 0.9214,
+      "step": 811
+    },
+    {
+      "epoch": 0.2864134388571681,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.7108076764216585e-05,
+      "loss": 0.9525,
+      "step": 812
+    },
+    {
+      "epoch": 0.28676616476709066,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.7099869432631152e-05,
+      "loss": 0.9101,
+      "step": 813
+    },
+    {
+      "epoch": 0.2871188906770133,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.7091652445987976e-05,
+      "loss": 0.9787,
+      "step": 814
+    },
+    {
+      "epoch": 0.28747161658693593,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.7083425815461273e-05,
+      "loss": 0.951,
+      "step": 815
+    },
+    {
+      "epoch": 0.2878243424968585,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.7075189552238375e-05,
+      "loss": 0.9916,
+      "step": 816
+    },
+    {
+      "epoch": 0.28817706840678114,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.706694366751971e-05,
+      "loss": 1.1035,
+      "step": 817
+    },
+    {
+      "epoch": 0.2885297943167038,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.7058688172518802e-05,
+      "loss": 0.8761,
+      "step": 818
+    },
+    {
+      "epoch": 0.2888825202266264,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.705042307846223e-05,
+      "loss": 0.9569,
+      "step": 819
+    },
+    {
+      "epoch": 0.289235246136549,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.7042148396589624e-05,
+      "loss": 0.8767,
+      "step": 820
+    },
+    {
+      "epoch": 0.28958797204647163,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.703386413815367e-05,
+      "loss": 0.9863,
+      "step": 821
+    },
+    {
+      "epoch": 0.28994069795639427,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.702557031442007e-05,
+      "loss": 0.8866,
+      "step": 822
+    },
+    {
+      "epoch": 0.2902934238663169,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.701726693666752e-05,
+      "loss": 0.9081,
+      "step": 823
+    },
+    {
+      "epoch": 0.2906461497762395,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.7008954016187725e-05,
+      "loss": 0.8925,
+      "step": 824
+    },
+    {
+      "epoch": 0.2909988756861621,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.7000631564285363e-05,
+      "loss": 1.0457,
+      "step": 825
+    },
+    {
+      "epoch": 0.29135160159608475,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6992299592278065e-05,
+      "loss": 0.8859,
+      "step": 826
+    },
+    {
+      "epoch": 0.2917043275060074,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.6983958111496428e-05,
+      "loss": 1.0515,
+      "step": 827
+    },
+    {
+      "epoch": 0.29205705341592997,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.6975607133283958e-05,
+      "loss": 0.9198,
+      "step": 828
+    },
+    {
+      "epoch": 0.2924097793258526,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6967246668997086e-05,
+      "loss": 0.9083,
+      "step": 829
+    },
+    {
+      "epoch": 0.29276250523577524,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.6958876730005146e-05,
+      "loss": 0.9058,
+      "step": 830
+    },
+    {
+      "epoch": 0.2931152311456978,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.6950497327690357e-05,
+      "loss": 0.9866,
+      "step": 831
+    },
+    {
+      "epoch": 0.29346795705562045,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.6942108473447796e-05,
+      "loss": 1.0966,
+      "step": 832
+    },
+    {
+      "epoch": 0.2938206829655431,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.6933710178685406e-05,
+      "loss": 1.0164,
+      "step": 833
+    },
+    {
+      "epoch": 0.2941734088754657,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.692530245482396e-05,
+      "loss": 1.0489,
+      "step": 834
+    },
+    {
+      "epoch": 0.2945261347853883,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.6916885313297063e-05,
+      "loss": 1.0835,
+      "step": 835
+    },
+    {
+      "epoch": 0.29487886069531094,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.6908458765551116e-05,
+      "loss": 0.8496,
+      "step": 836
+    },
+    {
+      "epoch": 0.2952315866052336,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.6900022823045317e-05,
+      "loss": 1.0073,
+      "step": 837
+    },
+    {
+      "epoch": 0.2955843125151562,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.689157749725164e-05,
+      "loss": 0.9269,
+      "step": 838
+    },
+    {
+      "epoch": 0.2959370384250788,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.6883122799654814e-05,
+      "loss": 0.9681,
+      "step": 839
+    },
+    {
+      "epoch": 0.2962897643350014,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.6874658741752327e-05,
+      "loss": 0.9633,
+      "step": 840
+    },
+    {
+      "epoch": 0.29664249024492406,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.686618533505438e-05,
+      "loss": 0.9027,
+      "step": 841
+    },
+    {
+      "epoch": 0.2969952161548467,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.685770259108389e-05,
+      "loss": 0.9897,
+      "step": 842
+    },
+    {
+      "epoch": 0.2973479420647693,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.684921052137648e-05,
+      "loss": 1.0995,
+      "step": 843
+    },
+    {
+      "epoch": 0.2977006679746919,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.684070913748045e-05,
+      "loss": 0.8927,
+      "step": 844
+    },
+    {
+      "epoch": 0.29805339388461455,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6832198450956766e-05,
+      "loss": 1.0478,
+      "step": 845
+    },
+    {
+      "epoch": 0.2984061197945372,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6823678473379046e-05,
+      "loss": 0.9626,
+      "step": 846
+    },
+    {
+      "epoch": 0.29875884570445976,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.6815149216333538e-05,
+      "loss": 1.0428,
+      "step": 847
+    },
+    {
+      "epoch": 0.2991115716143824,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.6806610691419117e-05,
+      "loss": 1.028,
+      "step": 848
+    },
+    {
+      "epoch": 0.29946429752430503,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.6798062910247254e-05,
+      "loss": 0.9671,
+      "step": 849
+    },
+    {
+      "epoch": 0.2998170234342276,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.6789505884442015e-05,
+      "loss": 0.9878,
+      "step": 850
+    },
+    {
+      "epoch": 0.30016974934415025,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.678093962564003e-05,
+      "loss": 0.9375,
+      "step": 851
+    },
+    {
+      "epoch": 0.3005224752540729,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.6772364145490486e-05,
+      "loss": 0.8304,
+      "step": 852
+    },
+    {
+      "epoch": 0.3008752011639955,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.6763779455655122e-05,
+      "loss": 0.9883,
+      "step": 853
+    },
+    {
+      "epoch": 0.3012279270739181,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.6755185567808185e-05,
+      "loss": 0.9941,
+      "step": 854
+    },
+    {
+      "epoch": 0.30158065298384074,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.6746582493636433e-05,
+      "loss": 0.9761,
+      "step": 855
+    },
+    {
+      "epoch": 0.30193337889376337,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.6737970244839125e-05,
+      "loss": 0.7738,
+      "step": 856
+    },
+    {
+      "epoch": 0.302286104803686,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.672934883312799e-05,
+      "loss": 0.9891,
+      "step": 857
+    },
+    {
+      "epoch": 0.3026388307136086,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.6720718270227224e-05,
+      "loss": 0.9568,
+      "step": 858
+    },
+    {
+      "epoch": 0.3029915566235312,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.6712078567873458e-05,
+      "loss": 0.9114,
+      "step": 859
+    },
+    {
+      "epoch": 0.30334428253345386,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.6703429737815756e-05,
+      "loss": 0.9453,
+      "step": 860
+    },
+    {
+      "epoch": 0.3036970084433765,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.66947717918156e-05,
+      "loss": 1.0401,
+      "step": 861
+    },
+    {
+      "epoch": 0.3040497343532991,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.6686104741646863e-05,
+      "loss": 1.0635,
+      "step": 862
+    },
+    {
+      "epoch": 0.3044024602632217,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.6677428599095796e-05,
+      "loss": 1.0077,
+      "step": 863
+    },
+    {
+      "epoch": 0.30475518617314434,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.6668743375961025e-05,
+      "loss": 1.0388,
+      "step": 864
+    },
+    {
+      "epoch": 0.305107912083067,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.6660049084053512e-05,
+      "loss": 1.0174,
+      "step": 865
+    },
+    {
+      "epoch": 0.30546063799298956,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.6651345735196555e-05,
+      "loss": 0.9669,
+      "step": 866
+    },
+    {
+      "epoch": 0.3058133639029122,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.6642633341225782e-05,
+      "loss": 0.9906,
+      "step": 867
+    },
+    {
+      "epoch": 0.30616608981283483,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.66339119139891e-05,
+      "loss": 1.031,
+      "step": 868
+    },
+    {
+      "epoch": 0.3065188157227574,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.6625181465346717e-05,
+      "loss": 0.9352,
+      "step": 869
+    },
+    {
+      "epoch": 0.30687154163268004,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.66164420071711e-05,
+      "loss": 0.9691,
+      "step": 870
+    },
+    {
+      "epoch": 0.3072242675426027,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.660769355134697e-05,
+      "loss": 0.9338,
+      "step": 871
+    },
+    {
+      "epoch": 0.3075769934525253,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.6598936109771285e-05,
+      "loss": 0.938,
+      "step": 872
+    },
+    {
+      "epoch": 0.3079297193624479,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.6590169694353227e-05,
+      "loss": 0.9928,
+      "step": 873
+    },
+    {
+      "epoch": 0.30828244527237053,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.6581394317014173e-05,
+      "loss": 1.0752,
+      "step": 874
+    },
+    {
+      "epoch": 0.30863517118229317,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.6572609989687687e-05,
+      "loss": 1.0013,
+      "step": 875
+    },
+    {
+      "epoch": 0.3089878970922158,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.6563816724319516e-05,
+      "loss": 0.8965,
+      "step": 876
+    },
+    {
+      "epoch": 0.3093406230021384,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.6555014532867548e-05,
+      "loss": 0.9742,
+      "step": 877
+    },
+    {
+      "epoch": 0.309693348912061,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.654620342730182e-05,
+      "loss": 0.9272,
+      "step": 878
+    },
+    {
+      "epoch": 0.31004607482198365,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.653738341960448e-05,
+      "loss": 0.9977,
+      "step": 879
+    },
+    {
+      "epoch": 0.3103988007319063,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.65285545217698e-05,
+      "loss": 0.9728,
+      "step": 880
+    },
+    {
+      "epoch": 0.31075152664182887,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.6519716745804112e-05,
+      "loss": 0.9562,
+      "step": 881
+    },
+    {
+      "epoch": 0.3111042525517515,
+      "grad_norm": 0.75,
+      "learning_rate": 1.651087010372586e-05,
+      "loss": 0.9845,
+      "step": 882
+    },
+    {
+      "epoch": 0.31145697846167414,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.650201460756551e-05,
+      "loss": 0.9441,
+      "step": 883
+    },
+    {
+      "epoch": 0.3118097043715967,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.6493150269365584e-05,
+      "loss": 1.0183,
+      "step": 884
+    },
+    {
+      "epoch": 0.31216243028151935,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.6484277101180635e-05,
+      "loss": 0.9086,
+      "step": 885
+    },
+    {
+      "epoch": 0.312515156191442,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.6475395115077213e-05,
+      "loss": 0.9194,
+      "step": 886
+    },
+    {
+      "epoch": 0.3128678821013646,
+      "grad_norm": 0.75,
+      "learning_rate": 1.6466504323133857e-05,
+      "loss": 0.9439,
+      "step": 887
+    },
+    {
+      "epoch": 0.3132206080112872,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6457604737441094e-05,
+      "loss": 0.8764,
+      "step": 888
+    },
+    {
+      "epoch": 0.31357333392120984,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.64486963701014e-05,
+      "loss": 0.9481,
+      "step": 889
+    },
+    {
+      "epoch": 0.3139260598311325,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.643977923322919e-05,
+      "loss": 1.0339,
+      "step": 890
+    },
+    {
+      "epoch": 0.3142787857410551,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.6430853338950815e-05,
+      "loss": 1.067,
+      "step": 891
+    },
+    {
+      "epoch": 0.3146315116509777,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.6421918699404532e-05,
+      "loss": 0.8997,
+      "step": 892
+    },
+    {
+      "epoch": 0.3149842375609003,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.6412975326740485e-05,
+      "loss": 0.9487,
+      "step": 893
+    },
+    {
+      "epoch": 0.31533696347082296,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.6404023233120693e-05,
+      "loss": 0.8344,
+      "step": 894
+    },
+    {
+      "epoch": 0.3156896893807456,
+      "grad_norm": 0.75,
+      "learning_rate": 1.6395062430719043e-05,
+      "loss": 1.0349,
+      "step": 895
+    },
+    {
+      "epoch": 0.3160424152906682,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.638609293172127e-05,
+      "loss": 1.0463,
+      "step": 896
+    },
+    {
+      "epoch": 0.3163951412005908,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6377114748324913e-05,
+      "loss": 0.9539,
+      "step": 897
+    },
+    {
+      "epoch": 0.31674786711051345,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6368127892739337e-05,
+      "loss": 0.882,
+      "step": 898
+    },
+    {
+      "epoch": 0.3171005930204361,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.63591323771857e-05,
+      "loss": 0.9788,
+      "step": 899
+    },
+    {
+      "epoch": 0.31745331893035866,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.6350128213896942e-05,
+      "loss": 0.9328,
+      "step": 900
+    },
+    {
+      "epoch": 0.3178060448402813,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.634111541511774e-05,
+      "loss": 0.9488,
+      "step": 901
+    },
+    {
+      "epoch": 0.31815877075020393,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6332093993104545e-05,
+      "loss": 0.9039,
+      "step": 902
+    },
+    {
+      "epoch": 0.3185114966601265,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.632306396012551e-05,
+      "loss": 0.9784,
+      "step": 903
+    },
+    {
+      "epoch": 0.31886422257004915,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.6314025328460507e-05,
+      "loss": 0.9473,
+      "step": 904
+    },
+    {
+      "epoch": 0.3192169484799718,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.6304978110401106e-05,
+      "loss": 0.9145,
+      "step": 905
+    },
+    {
+      "epoch": 0.3195696743898944,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.6295922318250544e-05,
+      "loss": 0.9167,
+      "step": 906
+    },
+    {
+      "epoch": 0.319922400299817,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.628685796432373e-05,
+      "loss": 0.8607,
+      "step": 907
+    },
+    {
+      "epoch": 0.32027512620973964,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.6277785060947203e-05,
+      "loss": 0.9668,
+      "step": 908
+    },
+    {
+      "epoch": 0.32062785211966227,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.6268703620459136e-05,
+      "loss": 0.904,
+      "step": 909
+    },
+    {
+      "epoch": 0.3209805780295849,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.6259613655209314e-05,
+      "loss": 1.0186,
+      "step": 910
+    },
+    {
+      "epoch": 0.3213333039395075,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.6250515177559106e-05,
+      "loss": 0.9815,
+      "step": 911
+    },
+    {
+      "epoch": 0.3216860298494301,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.624140819988146e-05,
+      "loss": 1.0358,
+      "step": 912
+    },
+    {
+      "epoch": 0.32203875575935276,
+      "grad_norm": 4.8125,
+      "learning_rate": 1.623229273456089e-05,
+      "loss": 0.9553,
+      "step": 913
+    },
+    {
+      "epoch": 0.3223914816692754,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.622316879399344e-05,
+      "loss": 0.9444,
+      "step": 914
+    },
+    {
+      "epoch": 0.322744207579198,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.62140363905867e-05,
+      "loss": 1.0303,
+      "step": 915
+    },
+    {
+      "epoch": 0.3230969334891206,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.6204895536759738e-05,
+      "loss": 0.8918,
+      "step": 916
+    },
+    {
+      "epoch": 0.32344965939904324,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.6195746244943142e-05,
+      "loss": 0.9438,
+      "step": 917
+    },
+    {
+      "epoch": 0.3238023853089659,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.6186588527578963e-05,
+      "loss": 0.93,
+      "step": 918
+    },
+    {
+      "epoch": 0.32415511121888846,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.61774223971207e-05,
+      "loss": 0.9825,
+      "step": 919
+    },
+    {
+      "epoch": 0.3245078371288111,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.616824786603332e-05,
+      "loss": 0.9265,
+      "step": 920
+    },
+    {
+      "epoch": 0.32486056303873373,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.6159064946793185e-05,
+      "loss": 0.8819,
+      "step": 921
+    },
+    {
+      "epoch": 0.3252132889486563,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.6149873651888075e-05,
+      "loss": 0.9396,
+      "step": 922
+    },
+    {
+      "epoch": 0.32556601485857894,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.614067399381717e-05,
+      "loss": 0.9203,
+      "step": 923
+    },
+    {
+      "epoch": 0.3259187407685016,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6131465985091007e-05,
+      "loss": 0.9863,
+      "step": 924
+    },
+    {
+      "epoch": 0.3262714666784242,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.612224963823149e-05,
+      "loss": 0.9572,
+      "step": 925
+    },
+    {
+      "epoch": 0.3266241925883468,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.6113024965771854e-05,
+      "loss": 0.9344,
+      "step": 926
+    },
+    {
+      "epoch": 0.32697691849826943,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.6103791980256663e-05,
+      "loss": 1.0024,
+      "step": 927
+    },
+    {
+      "epoch": 0.32732964440819207,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.6094550694241783e-05,
+      "loss": 0.9684,
+      "step": 928
+    },
+    {
+      "epoch": 0.3276823703181147,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.608530112029437e-05,
+      "loss": 1.0828,
+      "step": 929
+    },
+    {
+      "epoch": 0.3280350962280373,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.607604327099284e-05,
+      "loss": 1.0742,
+      "step": 930
+    },
+    {
+      "epoch": 0.3283878221379599,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.606677715892688e-05,
+      "loss": 0.993,
+      "step": 931
+    },
+    {
+      "epoch": 0.32874054804788255,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.6057502796697403e-05,
+      "loss": 0.9256,
+      "step": 932
+    },
+    {
+      "epoch": 0.3290932739578052,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.6048220196916537e-05,
+      "loss": 1.1202,
+      "step": 933
+    },
+    {
+      "epoch": 0.32944599986772777,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.603892937220762e-05,
+      "loss": 0.9715,
+      "step": 934
+    },
+    {
+      "epoch": 0.3297987257776504,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.602963033520518e-05,
+      "loss": 0.9632,
+      "step": 935
+    },
+    {
+      "epoch": 0.33015145168757304,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.60203230985549e-05,
+      "loss": 1.0691,
+      "step": 936
+    },
+    {
+      "epoch": 0.3305041775974956,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.601100767491361e-05,
+      "loss": 0.8851,
+      "step": 937
+    },
+    {
+      "epoch": 0.33085690350741825,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.6001684076949304e-05,
+      "loss": 0.9352,
+      "step": 938
+    },
+    {
+      "epoch": 0.3312096294173409,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5992352317341055e-05,
+      "loss": 0.9713,
+      "step": 939
+    },
+    {
+      "epoch": 0.3315623553272635,
+      "grad_norm": 0.75,
+      "learning_rate": 1.598301240877905e-05,
+      "loss": 0.9449,
+      "step": 940
+    },
+    {
+      "epoch": 0.3319150812371861,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5973664363964573e-05,
+      "loss": 0.895,
+      "step": 941
+    },
+    {
+      "epoch": 0.33226780714710874,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.5964308195609944e-05,
+      "loss": 0.9448,
+      "step": 942
+    },
+    {
+      "epoch": 0.3326205330570314,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5954943916438552e-05,
+      "loss": 1.0261,
+      "step": 943
+    },
+    {
+      "epoch": 0.332973258966954,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.594557153918481e-05,
+      "loss": 0.848,
+      "step": 944
+    },
+    {
+      "epoch": 0.3333259848768766,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.593619107659413e-05,
+      "loss": 0.8739,
+      "step": 945
+    },
+    {
+      "epoch": 0.3336787107867992,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.5926802541422938e-05,
+      "loss": 1.0085,
+      "step": 946
+    },
+    {
+      "epoch": 0.33403143669672186,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5917405946438635e-05,
+      "loss": 0.9469,
+      "step": 947
+    },
+    {
+      "epoch": 0.3343841626066445,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5908001304419572e-05,
+      "loss": 1.0331,
+      "step": 948
+    },
+    {
+      "epoch": 0.3347368885165671,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.589858862815505e-05,
+      "loss": 0.9215,
+      "step": 949
+    },
+    {
+      "epoch": 0.3350896144264897,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.58891679304453e-05,
+      "loss": 0.9211,
+      "step": 950
+    },
+    {
+      "epoch": 0.33544234033641235,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.5879739224101448e-05,
+      "loss": 1.024,
+      "step": 951
+    },
+    {
+      "epoch": 0.335795066246335,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.5870302521945526e-05,
+      "loss": 0.9666,
+      "step": 952
+    },
+    {
+      "epoch": 0.33614779215625756,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.5860857836810427e-05,
+      "loss": 0.9932,
+      "step": 953
+    },
+    {
+      "epoch": 0.3365005180661802,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5851405181539912e-05,
+      "loss": 0.9506,
+      "step": 954
+    },
+    {
+      "epoch": 0.33685324397610283,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.584194456898857e-05,
+      "loss": 0.9384,
+      "step": 955
+    },
+    {
+      "epoch": 0.3372059698860254,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5832476012021812e-05,
+      "loss": 0.9602,
+      "step": 956
+    },
+    {
+      "epoch": 0.33755869579594805,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.5822999523515863e-05,
+      "loss": 0.9397,
+      "step": 957
+    },
+    {
+      "epoch": 0.3379114217058707,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.5813515116357724e-05,
+      "loss": 0.9791,
+      "step": 958
+    },
+    {
+      "epoch": 0.3382641476157933,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.5804022803445164e-05,
+      "loss": 0.9716,
+      "step": 959
+    },
+    {
+      "epoch": 0.3386168735257159,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.579452259768671e-05,
+      "loss": 0.8937,
+      "step": 960
+    },
+    {
+      "epoch": 0.33896959943563854,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.578501451200162e-05,
+      "loss": 0.965,
+      "step": 961
+    },
+    {
+      "epoch": 0.33932232534556117,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5775498559319856e-05,
+      "loss": 0.9708,
+      "step": 962
+    },
+    {
+      "epoch": 0.3396750512554838,
+      "grad_norm": 0.875,
+      "learning_rate": 1.5765974752582102e-05,
+      "loss": 0.9519,
+      "step": 963
+    },
+    {
+      "epoch": 0.3400277771654064,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.57564431047397e-05,
+      "loss": 1.0427,
+      "step": 964
+    },
+    {
+      "epoch": 0.340380503075329,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5746903628754672e-05,
+      "loss": 0.9388,
+      "step": 965
+    },
+    {
+      "epoch": 0.34073322898525166,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5737356337599675e-05,
+      "loss": 0.9667,
+      "step": 966
+    },
+    {
+      "epoch": 0.3410859548951743,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.5727801244257995e-05,
+      "loss": 0.8992,
+      "step": 967
+    },
+    {
+      "epoch": 0.3414386808050969,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.5718238361723525e-05,
+      "loss": 1.0136,
+      "step": 968
+    },
+    {
+      "epoch": 0.3417914067150195,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.570866770300077e-05,
+      "loss": 0.9609,
+      "step": 969
+    },
+    {
+      "epoch": 0.34214413262494214,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.5699089281104777e-05,
+      "loss": 0.9235,
+      "step": 970
+    },
+    {
+      "epoch": 0.3424968585348648,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5689503109061185e-05,
+      "loss": 0.9644,
+      "step": 971
+    },
+    {
+      "epoch": 0.34284958444478736,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5679909199906142e-05,
+      "loss": 1.0411,
+      "step": 972
+    },
+    {
+      "epoch": 0.34320231035471,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.567030756668634e-05,
+      "loss": 0.9608,
+      "step": 973
+    },
+    {
+      "epoch": 0.34355503626463263,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5660698222458962e-05,
+      "loss": 0.9288,
+      "step": 974
+    },
+    {
+      "epoch": 0.3439077621745552,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.565108118029168e-05,
+      "loss": 0.9534,
+      "step": 975
+    },
+    {
+      "epoch": 0.34426048808447784,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.564145645326264e-05,
+      "loss": 0.9146,
+      "step": 976
+    },
+    {
+      "epoch": 0.3446132139944005,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.563182405446043e-05,
+      "loss": 0.9765,
+      "step": 977
+    },
+    {
+      "epoch": 0.3449659399043231,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.562218399698408e-05,
+      "loss": 0.8996,
+      "step": 978
+    },
+    {
+      "epoch": 0.3453186658142457,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5612536293943023e-05,
+      "loss": 0.998,
+      "step": 979
+    },
+    {
+      "epoch": 0.34567139172416833,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.56028809584571e-05,
+      "loss": 1.0394,
+      "step": 980
+    },
+    {
+      "epoch": 0.34602411763409097,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.559321800365653e-05,
+      "loss": 1.0428,
+      "step": 981
+    },
+    {
+      "epoch": 0.3463768435440136,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.5583547442681884e-05,
+      "loss": 0.9042,
+      "step": 982
+    },
+    {
+      "epoch": 0.3467295694539362,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5573869288684087e-05,
+      "loss": 0.9842,
+      "step": 983
+    },
+    {
+      "epoch": 0.3470822953638588,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.556418355482439e-05,
+      "loss": 1.0134,
+      "step": 984
+    },
+    {
+      "epoch": 0.34743502127378145,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.5554490254274345e-05,
+      "loss": 0.8548,
+      "step": 985
+    },
+    {
+      "epoch": 0.3477877471837041,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.554478940021579e-05,
+      "loss": 0.9506,
+      "step": 986
+    },
+    {
+      "epoch": 0.34814047309362667,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.553508100584085e-05,
+      "loss": 0.8797,
+      "step": 987
+    },
+    {
+      "epoch": 0.3484931990035493,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.552536508435189e-05,
+      "loss": 0.8801,
+      "step": 988
+    },
+    {
+      "epoch": 0.34884592491347194,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.5515641648961526e-05,
+      "loss": 0.9017,
+      "step": 989
+    },
+    {
+      "epoch": 0.3491986508233945,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.5505910712892572e-05,
+      "loss": 0.9192,
+      "step": 990
+    },
+    {
+      "epoch": 0.34955137673331715,
+      "grad_norm": 7.0,
+      "learning_rate": 1.5496172289378062e-05,
+      "loss": 1.0291,
+      "step": 991
+    },
+    {
+      "epoch": 0.3499041026432398,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5486426391661193e-05,
+      "loss": 0.9444,
+      "step": 992
+    },
+    {
+      "epoch": 0.3502568285531624,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5476673032995345e-05,
+      "loss": 0.9555,
+      "step": 993
+    },
+    {
+      "epoch": 0.350609554463085,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5466912226644037e-05,
+      "loss": 0.9221,
+      "step": 994
+    },
+    {
+      "epoch": 0.35096228037300764,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5457143985880905e-05,
+      "loss": 0.9754,
+      "step": 995
+    },
+    {
+      "epoch": 0.3513150062829303,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5447368323989712e-05,
+      "loss": 1.0154,
+      "step": 996
+    },
+    {
+      "epoch": 0.3516677321928529,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.5437585254264297e-05,
+      "loss": 0.9278,
+      "step": 997
+    },
+    {
+      "epoch": 0.3520204581027755,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.542779479000859e-05,
+      "loss": 0.9478,
+      "step": 998
+    },
+    {
+      "epoch": 0.3523731840126981,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.5417996944536566e-05,
+      "loss": 0.9792,
+      "step": 999
+    },
+    {
+      "epoch": 0.35272590992262076,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.5408191731172238e-05,
+      "loss": 1.0488,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3530786358325434,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.5398379163249636e-05,
+      "loss": 1.0821,
+      "step": 1001
+    },
+    {
+      "epoch": 0.353431361742466,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.5388559254112806e-05,
+      "loss": 1.0407,
+      "step": 1002
+    },
+    {
+      "epoch": 0.3537840876523886,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5378732017115757e-05,
+      "loss": 0.9287,
+      "step": 1003
+    },
+    {
+      "epoch": 0.35413681356231125,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.5368897465622478e-05,
+      "loss": 0.8905,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3544895394722339,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.5359055613006892e-05,
+      "loss": 0.9352,
+      "step": 1005
+    },
+    {
+      "epoch": 0.35484226538215646,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.534920647265287e-05,
+      "loss": 1.0162,
+      "step": 1006
+    },
+    {
+      "epoch": 0.3551949912920791,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.5339350057954178e-05,
+      "loss": 0.9454,
+      "step": 1007
+    },
+    {
+      "epoch": 0.35554771720200173,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5329486382314475e-05,
+      "loss": 1.0025,
+      "step": 1008
+    },
+    {
+      "epoch": 0.3559004431119243,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.53196154591473e-05,
+      "loss": 0.9713,
+      "step": 1009
+    },
+    {
+      "epoch": 0.35625316902184695,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5309737301876045e-05,
+      "loss": 0.9465,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3566058949317696,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.529985192393394e-05,
+      "loss": 1.0401,
+      "step": 1011
+    },
+    {
+      "epoch": 0.3569586208416922,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.5289959338764036e-05,
+      "loss": 0.9799,
+      "step": 1012
+    },
+    {
+      "epoch": 0.3573113467516148,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.5280059559819177e-05,
+      "loss": 0.9251,
+      "step": 1013
+    },
+    {
+      "epoch": 0.35766407266153744,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.5270152600562006e-05,
+      "loss": 0.8728,
+      "step": 1014
+    },
+    {
+      "epoch": 0.35801679857146007,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5260238474464913e-05,
+      "loss": 1.041,
+      "step": 1015
+    },
+    {
+      "epoch": 0.3583695244813827,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5250317195010044e-05,
+      "loss": 0.926,
+      "step": 1016
+    },
+    {
+      "epoch": 0.3587222503913053,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.5240388775689274e-05,
+      "loss": 1.1001,
+      "step": 1017
+    },
+    {
+      "epoch": 0.3590749763012279,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.523045323000418e-05,
+      "loss": 0.9967,
+      "step": 1018
+    },
+    {
+      "epoch": 0.35942770221115056,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.522051057146603e-05,
+      "loss": 0.8976,
+      "step": 1019
+    },
+    {
+      "epoch": 0.3597804281210732,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.5210560813595776e-05,
+      "loss": 0.9577,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3601331540309958,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.5200603969924012e-05,
+      "loss": 0.9428,
+      "step": 1021
+    },
+    {
+      "epoch": 0.3604858799409184,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.5190640053990982e-05,
+      "loss": 1.0208,
+      "step": 1022
+    },
+    {
+      "epoch": 0.36083860585084104,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.5180669079346527e-05,
+      "loss": 0.9196,
+      "step": 1023
+    },
+    {
+      "epoch": 0.3611913317607637,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5170691059550105e-05,
+      "loss": 0.9222,
+      "step": 1024
+    },
+    {
+      "epoch": 0.36154405767068626,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5160706008170744e-05,
+      "loss": 0.9481,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3618967835806089,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.5150713938787044e-05,
+      "loss": 0.9821,
+      "step": 1026
+    },
+    {
+      "epoch": 0.36224950949053153,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.5140714864987143e-05,
+      "loss": 1.0289,
+      "step": 1027
+    },
+    {
+      "epoch": 0.3626022354004541,
+      "grad_norm": 5.6875,
+      "learning_rate": 1.5130708800368697e-05,
+      "loss": 0.9201,
+      "step": 1028
+    },
+    {
+      "epoch": 0.36295496131037674,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.512069575853888e-05,
+      "loss": 0.9145,
+      "step": 1029
+    },
+    {
+      "epoch": 0.3633076872202994,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.5110675753114357e-05,
+      "loss": 0.9416,
+      "step": 1030
+    },
+    {
+      "epoch": 0.363660413130222,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.510064879772125e-05,
+      "loss": 0.8707,
+      "step": 1031
+    },
+    {
+      "epoch": 0.3640131390401446,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5090614905995135e-05,
+      "loss": 0.9588,
+      "step": 1032
+    },
+    {
+      "epoch": 0.36436586495006723,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.5080574091581031e-05,
+      "loss": 0.8527,
+      "step": 1033
+    },
+    {
+      "epoch": 0.36471859085998987,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5070526368133358e-05,
+      "loss": 0.9853,
+      "step": 1034
+    },
+    {
+      "epoch": 0.3650713167699125,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.5060471749315943e-05,
+      "loss": 0.9505,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3654240426798351,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.5050410248801986e-05,
+      "loss": 1.0374,
+      "step": 1036
+    },
+    {
+      "epoch": 0.3657767685897577,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5040341880274038e-05,
+      "loss": 0.8629,
+      "step": 1037
+    },
+    {
+      "epoch": 0.36612949449968035,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.5030266657423998e-05,
+      "loss": 0.991,
+      "step": 1038
+    },
+    {
+      "epoch": 0.366482220409603,
+      "grad_norm": 0.875,
+      "learning_rate": 1.5020184593953082e-05,
+      "loss": 0.8893,
+      "step": 1039
+    },
+    {
+      "epoch": 0.36683494631952557,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.5010095703571816e-05,
+      "loss": 0.9277,
+      "step": 1040
+    },
+    {
+      "epoch": 0.3671876722294482,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.9441,
+      "step": 1041
+    },
+    {
+      "epoch": 0.36754039813937084,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.4989897496966707e-05,
+      "loss": 0.9125,
+      "step": 1042
+    },
+    {
+      "epoch": 0.3678931240492934,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.4979788208210249e-05,
+      "loss": 0.9047,
+      "step": 1043
+    },
+    {
+      "epoch": 0.36824584995921605,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4969672147478172e-05,
+      "loss": 1.0238,
+      "step": 1044
+    },
+    {
+      "epoch": 0.3685985758691387,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.495954932852723e-05,
+      "loss": 0.9515,
+      "step": 1045
+    },
+    {
+      "epoch": 0.3689513017790613,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.494941976512336e-05,
+      "loss": 0.9831,
+      "step": 1046
+    },
+    {
+      "epoch": 0.3693040276889839,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.4939283471041686e-05,
+      "loss": 0.9375,
+      "step": 1047
+    },
+    {
+      "epoch": 0.36965675359890654,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.4929140460066474e-05,
+      "loss": 0.9888,
+      "step": 1048
+    },
+    {
+      "epoch": 0.3700094795088292,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.4918990745991122e-05,
+      "loss": 0.8899,
+      "step": 1049
+    },
+    {
+      "epoch": 0.3703622054187518,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4908834342618153e-05,
+      "loss": 0.966,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3707149313286744,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.489867126375918e-05,
+      "loss": 0.9499,
+      "step": 1051
+    },
+    {
+      "epoch": 0.371067657238597,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.4888501523234894e-05,
+      "loss": 1.1049,
+      "step": 1052
+    },
+    {
+      "epoch": 0.37142038314851966,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4878325134875046e-05,
+      "loss": 1.0351,
+      "step": 1053
+    },
+    {
+      "epoch": 0.3717731090584423,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4868142112518436e-05,
+      "loss": 1.0721,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3721258349683649,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4857952470012871e-05,
+      "loss": 0.9423,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3724785608782875,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4847756221215167e-05,
+      "loss": 0.9888,
+      "step": 1056
+    },
+    {
+      "epoch": 0.37283128678821015,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4837553379991123e-05,
+      "loss": 1.013,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3731840126981328,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.4827343960215508e-05,
+      "loss": 0.8107,
+      "step": 1058
+    },
+    {
+      "epoch": 0.37353673860805536,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4817127975772034e-05,
+      "loss": 0.9987,
+      "step": 1059
+    },
+    {
+      "epoch": 0.373889464517978,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4806905440553336e-05,
+      "loss": 0.8656,
+      "step": 1060
+    },
+    {
+      "epoch": 0.37424219042790063,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.4796676368460963e-05,
+      "loss": 0.8996,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3745949163378232,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4786440773405351e-05,
+      "loss": 0.892,
+      "step": 1062
+    },
+    {
+      "epoch": 0.37494764224774585,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.4776198669305803e-05,
+      "loss": 0.8182,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3753003681576685,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.4765950070090484e-05,
+      "loss": 0.9942,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3756530940675911,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4755694989696383e-05,
+      "loss": 1.0073,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3760058199775137,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4745433442069304e-05,
+      "loss": 0.8099,
+      "step": 1066
+    },
+    {
+      "epoch": 0.37635854588743634,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4735165441163846e-05,
+      "loss": 0.9493,
+      "step": 1067
+    },
+    {
+      "epoch": 0.37671127179735897,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4724891000943384e-05,
+      "loss": 0.8733,
+      "step": 1068
+    },
+    {
+      "epoch": 0.3770639977072816,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4714610135380052e-05,
+      "loss": 0.9897,
+      "step": 1069
+    },
+    {
+      "epoch": 0.3774167236172042,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4704322858454718e-05,
+      "loss": 0.9094,
+      "step": 1070
+    },
+    {
+      "epoch": 0.3777694495271268,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4694029184156973e-05,
+      "loss": 0.9343,
+      "step": 1071
+    },
+    {
+      "epoch": 0.37812217543704946,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.4683729126485103e-05,
+      "loss": 0.9072,
+      "step": 1072
+    },
+    {
+      "epoch": 0.3784749013469721,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4673422699446078e-05,
+      "loss": 1.0008,
+      "step": 1073
+    },
+    {
+      "epoch": 0.3788276272568947,
+      "grad_norm": 0.75,
+      "learning_rate": 1.466310991705553e-05,
+      "loss": 0.9236,
+      "step": 1074
+    },
+    {
+      "epoch": 0.3791803531668173,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4652790793337729e-05,
+      "loss": 0.9591,
+      "step": 1075
+    },
+    {
+      "epoch": 0.37953307907673994,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4642465342325574e-05,
+      "loss": 0.9212,
+      "step": 1076
+    },
+    {
+      "epoch": 0.3798858049866626,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.4632133578060566e-05,
+      "loss": 1.0065,
+      "step": 1077
+    },
+    {
+      "epoch": 0.38023853089658516,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4621795514592794e-05,
+      "loss": 0.9198,
+      "step": 1078
+    },
+    {
+      "epoch": 0.3805912568065078,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.4611451165980905e-05,
+      "loss": 0.9755,
+      "step": 1079
+    },
+    {
+      "epoch": 0.38094398271643043,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.4601100546292103e-05,
+      "loss": 0.983,
+      "step": 1080
+    },
+    {
+      "epoch": 0.381296708626353,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.4590743669602118e-05,
+      "loss": 0.8859,
+      "step": 1081
+    },
+    {
+      "epoch": 0.38164943453627564,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.4580380549995183e-05,
+      "loss": 1.0132,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3820021604461983,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4570011201564029e-05,
+      "loss": 0.8678,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3823548863561209,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4559635638409848e-05,
+      "loss": 0.9027,
+      "step": 1084
+    },
+    {
+      "epoch": 0.3827076122660435,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4549253874642289e-05,
+      "loss": 0.9666,
+      "step": 1085
+    },
+    {
+      "epoch": 0.38306033817596613,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.4538865924379435e-05,
+      "loss": 0.9394,
+      "step": 1086
+    },
+    {
+      "epoch": 0.38341306408588877,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.4528471801747778e-05,
+      "loss": 1.0332,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3837657899958114,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.4518071520882206e-05,
+      "loss": 1.0082,
+      "step": 1088
+    },
+    {
+      "epoch": 0.384118515905734,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.450766509592598e-05,
+      "loss": 0.9077,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3844712418156566,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.4497252541030719e-05,
+      "loss": 1.0047,
+      "step": 1090
+    },
+    {
+      "epoch": 0.38482396772557925,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4486833870356374e-05,
+      "loss": 1.0549,
+      "step": 1091
+    },
+    {
+      "epoch": 0.3851766936355019,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4476409098071215e-05,
+      "loss": 1.0063,
+      "step": 1092
+    },
+    {
+      "epoch": 0.38552941954542447,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4465978238351812e-05,
+      "loss": 0.9053,
+      "step": 1093
+    },
+    {
+      "epoch": 0.3858821454553471,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.4455541305383007e-05,
+      "loss": 1.1241,
+      "step": 1094
+    },
+    {
+      "epoch": 0.38623487136526974,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.444509831335791e-05,
+      "loss": 0.8614,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3865875972751923,
+      "grad_norm": 13.25,
+      "learning_rate": 1.4434649276477859e-05,
+      "loss": 0.9902,
+      "step": 1096
+    },
+    {
+      "epoch": 0.38694032318511495,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4424194208952427e-05,
+      "loss": 0.9382,
+      "step": 1097
+    },
+    {
+      "epoch": 0.3872930490950376,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.441373312499937e-05,
+      "loss": 0.8969,
+      "step": 1098
+    },
+    {
+      "epoch": 0.3876457750049602,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4403266038844641e-05,
+      "loss": 0.9393,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3879985009148828,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4392792964722351e-05,
+      "loss": 0.9678,
+      "step": 1100
+    },
+    {
+      "epoch": 0.38835122682480544,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.438231391687475e-05,
+      "loss": 0.8819,
+      "step": 1101
+    },
+    {
+      "epoch": 0.3887039527347281,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4371828909552217e-05,
+      "loss": 0.9777,
+      "step": 1102
+    },
+    {
+      "epoch": 0.3890566786446507,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4361337957013227e-05,
+      "loss": 1.0016,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3894094045545733,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.435084107352435e-05,
+      "loss": 1.0313,
+      "step": 1104
+    },
+    {
+      "epoch": 0.3897621304644959,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.4340338273360219e-05,
+      "loss": 1.0087,
+      "step": 1105
+    },
+    {
+      "epoch": 0.39011485637441856,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4329829570803511e-05,
+      "loss": 0.9336,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3904675822843412,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4319314980144922e-05,
+      "loss": 0.958,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3908203081942638,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.4308794515683172e-05,
+      "loss": 0.9702,
+      "step": 1108
+    },
+    {
+      "epoch": 0.3911730341041864,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.4298268191724951e-05,
+      "loss": 0.9189,
+      "step": 1109
+    },
+    {
+      "epoch": 0.39152576001410905,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.4287736022584935e-05,
+      "loss": 0.9916,
+      "step": 1110
+    },
+    {
+      "epoch": 0.3918784859240317,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.4277198022585734e-05,
+      "loss": 0.9508,
+      "step": 1111
+    },
+    {
+      "epoch": 0.39223121183395426,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.4266654206057892e-05,
+      "loss": 1.0529,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3925839377438769,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4256104587339869e-05,
+      "loss": 0.7978,
+      "step": 1113
+    },
+    {
+      "epoch": 0.39293666365379953,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.4245549180778007e-05,
+      "loss": 1.0596,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3932893895637221,
+      "grad_norm": 0.75,
+      "learning_rate": 1.423498800072652e-05,
+      "loss": 1.0357,
+      "step": 1115
+    },
+    {
+      "epoch": 0.39364211547364475,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.422442106154748e-05,
+      "loss": 0.8834,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3939948413835674,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.4213848377610782e-05,
+      "loss": 1.0003,
+      "step": 1117
+    },
+    {
+      "epoch": 0.39434756729349,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.420326996329414e-05,
+      "loss": 0.9774,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3947002932034126,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.419268583298306e-05,
+      "loss": 0.9334,
+      "step": 1119
+    },
+    {
+      "epoch": 0.39505301911333524,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.4182096001070812e-05,
+      "loss": 0.8949,
+      "step": 1120
+    },
+    {
+      "epoch": 0.39540574502325787,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.4171500481958433e-05,
+      "loss": 1.0099,
+      "step": 1121
+    },
+    {
+      "epoch": 0.3957584709331805,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4160899290054689e-05,
+      "loss": 0.9626,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3961111968431031,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.4150292439776054e-05,
+      "loss": 1.013,
+      "step": 1123
+    },
+    {
+      "epoch": 0.3964639227530257,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.4139679945546705e-05,
+      "loss": 0.928,
+      "step": 1124
+    },
+    {
+      "epoch": 0.39681664866294836,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.4129061821798494e-05,
+      "loss": 0.8822,
+      "step": 1125
+    },
+    {
+      "epoch": 0.397169374572871,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4118438082970918e-05,
+      "loss": 0.9606,
+      "step": 1126
+    },
+    {
+      "epoch": 0.39752210048279357,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4107808743511124e-05,
+      "loss": 0.9015,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3978748263927162,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.4097173817873864e-05,
+      "loss": 0.9271,
+      "step": 1128
+    },
+    {
+      "epoch": 0.39822755230263884,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.40865333205215e-05,
+      "loss": 0.9601,
+      "step": 1129
+    },
+    {
+      "epoch": 0.3985802782125615,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4075887265923954e-05,
+      "loss": 0.9486,
+      "step": 1130
+    },
+    {
+      "epoch": 0.39893300412248406,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4065235668558712e-05,
+      "loss": 0.9704,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3992857300324067,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4054578542910808e-05,
+      "loss": 0.992,
+      "step": 1132
+    },
+    {
+      "epoch": 0.39963845594232933,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.4043915903472777e-05,
+      "loss": 0.9193,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3999911818522519,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.4033247764744669e-05,
+      "loss": 1.0611,
+      "step": 1134
+    },
+    {
+      "epoch": 0.40034390776217454,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4022574141233995e-05,
+      "loss": 0.9596,
+      "step": 1135
+    },
+    {
+      "epoch": 0.4006966336720972,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.4011895047455738e-05,
+      "loss": 0.9907,
+      "step": 1136
+    },
+    {
+      "epoch": 0.4010493595820198,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.4001210497932312e-05,
+      "loss": 1.0011,
+      "step": 1137
+    },
+    {
+      "epoch": 0.4014020854919424,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.3990520507193556e-05,
+      "loss": 0.8978,
+      "step": 1138
+    },
+    {
+      "epoch": 0.40175481140186503,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3979825089776708e-05,
+      "loss": 0.9687,
+      "step": 1139
+    },
+    {
+      "epoch": 0.40210753731178767,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.396912426022638e-05,
+      "loss": 0.9763,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4024602632217103,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.3958418033094549e-05,
+      "loss": 1.0174,
+      "step": 1141
+    },
+    {
+      "epoch": 0.4028129891316329,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.3947706422940532e-05,
+      "loss": 0.9983,
+      "step": 1142
+    },
+    {
+      "epoch": 0.4031657150415555,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.3936989444330965e-05,
+      "loss": 0.9424,
+      "step": 1143
+    },
+    {
+      "epoch": 0.40351844095147815,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3926267111839781e-05,
+      "loss": 0.9835,
+      "step": 1144
+    },
+    {
+      "epoch": 0.4038711668614008,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.3915539440048203e-05,
+      "loss": 0.9299,
+      "step": 1145
+    },
+    {
+      "epoch": 0.40422389277132337,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3904806443544706e-05,
+      "loss": 0.9727,
+      "step": 1146
+    },
+    {
+      "epoch": 0.404576618681246,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.3894068136925006e-05,
+      "loss": 1.0059,
+      "step": 1147
+    },
+    {
+      "epoch": 0.40492934459116864,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.3883324534792051e-05,
+      "loss": 0.9951,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4052820705010912,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.3872575651755976e-05,
+      "loss": 1.0304,
+      "step": 1149
+    },
+    {
+      "epoch": 0.40563479641101385,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.3861821502434107e-05,
+      "loss": 1.0094,
+      "step": 1150
+    },
+    {
+      "epoch": 0.4059875223209365,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.385106210145093e-05,
+      "loss": 0.9281,
+      "step": 1151
+    },
+    {
+      "epoch": 0.4063402482308591,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.384029746343807e-05,
+      "loss": 0.9449,
+      "step": 1152
+    },
+    {
+      "epoch": 0.4066929741407817,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.382952760303428e-05,
+      "loss": 0.9506,
+      "step": 1153
+    },
+    {
+      "epoch": 0.40704570005070434,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.38187525348854e-05,
+      "loss": 0.9309,
+      "step": 1154
+    },
+    {
+      "epoch": 0.407398425960627,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3807972273644367e-05,
+      "loss": 1.0301,
+      "step": 1155
+    },
+    {
+      "epoch": 0.4077511518705496,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.379718683397118e-05,
+      "loss": 0.9592,
+      "step": 1156
+    },
+    {
+      "epoch": 0.4081038777804722,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3786396230532871e-05,
+      "loss": 0.9846,
+      "step": 1157
+    },
+    {
+      "epoch": 0.4084566036903948,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3775600478003498e-05,
+      "loss": 1.0624,
+      "step": 1158
+    },
+    {
+      "epoch": 0.40880932960031746,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3764799591064124e-05,
+      "loss": 0.9096,
+      "step": 1159
+    },
+    {
+      "epoch": 0.4091620555102401,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.3753993584402787e-05,
+      "loss": 0.9179,
+      "step": 1160
+    },
+    {
+      "epoch": 0.4095147814201627,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.3743182472714498e-05,
+      "loss": 0.9313,
+      "step": 1161
+    },
+    {
+      "epoch": 0.4098675073300853,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.3732366270701203e-05,
+      "loss": 0.984,
+      "step": 1162
+    },
+    {
+      "epoch": 0.41022023324000795,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3721544993071774e-05,
+      "loss": 0.9786,
+      "step": 1163
+    },
+    {
+      "epoch": 0.4105729591499306,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3710718654541974e-05,
+      "loss": 0.9832,
+      "step": 1164
+    },
+    {
+      "epoch": 0.41092568505985316,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.3699887269834468e-05,
+      "loss": 0.9608,
+      "step": 1165
+    },
+    {
+      "epoch": 0.4112784109697758,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3689050853678767e-05,
+      "loss": 0.9775,
+      "step": 1166
+    },
+    {
+      "epoch": 0.41163113687969843,
+      "grad_norm": 0.875,
+      "learning_rate": 1.367820942081123e-05,
+      "loss": 0.9552,
+      "step": 1167
+    },
+    {
+      "epoch": 0.411983862789621,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.366736298597504e-05,
+      "loss": 0.9264,
+      "step": 1168
+    },
+    {
+      "epoch": 0.41233658869954365,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.3656511563920178e-05,
+      "loss": 0.9542,
+      "step": 1169
+    },
+    {
+      "epoch": 0.4126893146094663,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.364565516940341e-05,
+      "loss": 0.9519,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4130420405193889,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3634793817188268e-05,
+      "loss": 0.9149,
+      "step": 1171
+    },
+    {
+      "epoch": 0.4133947664293115,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3623927522045014e-05,
+      "loss": 1.0677,
+      "step": 1172
+    },
+    {
+      "epoch": 0.41374749233923414,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.3613056298750641e-05,
+      "loss": 0.9398,
+      "step": 1173
+    },
+    {
+      "epoch": 0.41410021824915677,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.3602180162088843e-05,
+      "loss": 0.9514,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4144529441590794,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.3591299126849994e-05,
+      "loss": 0.9908,
+      "step": 1175
+    },
+    {
+      "epoch": 0.414805670069002,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.3580413207831136e-05,
+      "loss": 0.9156,
+      "step": 1176
+    },
+    {
+      "epoch": 0.4151583959789246,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.3569522419835934e-05,
+      "loss": 0.9957,
+      "step": 1177
+    },
+    {
+      "epoch": 0.41551112188884726,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.3558626777674695e-05,
+      "loss": 0.9883,
+      "step": 1178
+    },
+    {
+      "epoch": 0.4158638477987699,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.3547726296164316e-05,
+      "loss": 0.9488,
+      "step": 1179
+    },
+    {
+      "epoch": 0.41621657370869247,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3536820990128281e-05,
+      "loss": 0.9368,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4165692996186151,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.352591087439663e-05,
+      "loss": 0.9806,
+      "step": 1181
+    },
+    {
+      "epoch": 0.41692202552853774,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.3514995963805946e-05,
+      "loss": 0.9369,
+      "step": 1182
+    },
+    {
+      "epoch": 0.4172747514384604,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3504076273199326e-05,
+      "loss": 1.031,
+      "step": 1183
+    },
+    {
+      "epoch": 0.41762747734838296,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3493151817426382e-05,
+      "loss": 1.0194,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4179802032583056,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.3482222611343193e-05,
+      "loss": 0.9728,
+      "step": 1185
+    },
+    {
+      "epoch": 0.41833292916822823,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.34712886698123e-05,
+      "loss": 1.0131,
+      "step": 1186
+    },
+    {
+      "epoch": 0.4186856550781508,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.3460350007702691e-05,
+      "loss": 0.889,
+      "step": 1187
+    },
+    {
+      "epoch": 0.41903838098807344,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3449406639889764e-05,
+      "loss": 1.0086,
+      "step": 1188
+    },
+    {
+      "epoch": 0.4193911068979961,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.3438458581255322e-05,
+      "loss": 0.8554,
+      "step": 1189
+    },
+    {
+      "epoch": 0.4197438328079187,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.3427505846687548e-05,
+      "loss": 1.0022,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4200965587178413,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.3416548451080975e-05,
+      "loss": 0.9666,
+      "step": 1191
+    },
+    {
+      "epoch": 0.42044928462776393,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3405586409336489e-05,
+      "loss": 0.8706,
+      "step": 1192
+    },
+    {
+      "epoch": 0.42080201053768657,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.3394619736361278e-05,
+      "loss": 0.9358,
+      "step": 1193
+    },
+    {
+      "epoch": 0.4211547364476092,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.3383648447068843e-05,
+      "loss": 1.062,
+      "step": 1194
+    },
+    {
+      "epoch": 0.4215074623575318,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.3372672556378951e-05,
+      "loss": 1.0354,
+      "step": 1195
+    },
+    {
+      "epoch": 0.4218601882674544,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3361692079217634e-05,
+      "loss": 0.9724,
+      "step": 1196
+    },
+    {
+      "epoch": 0.42221291417737705,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3350707030517159e-05,
+      "loss": 0.9341,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4225656400872997,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3339717425216005e-05,
+      "loss": 0.9251,
+      "step": 1198
+    },
+    {
+      "epoch": 0.42291836599722227,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.3328723278258857e-05,
+      "loss": 1.1241,
+      "step": 1199
+    },
+    {
+      "epoch": 0.4232710919071449,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.3317724604596565e-05,
+      "loss": 0.8429,
+      "step": 1200
+    },
+    {
+      "epoch": 0.42362381781706754,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.3306721419186143e-05,
+      "loss": 0.9911,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4239765437269901,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.3295713736990741e-05,
+      "loss": 0.9226,
+      "step": 1202
+    },
+    {
+      "epoch": 0.42432926963691275,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.3284701572979618e-05,
+      "loss": 0.9289,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4246819955468354,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.3273684942128131e-05,
+      "loss": 0.941,
+      "step": 1204
+    },
+    {
+      "epoch": 0.425034721456758,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.326266385941771e-05,
+      "loss": 0.9588,
+      "step": 1205
+    },
+    {
+      "epoch": 0.4253874473666806,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.3251638339835847e-05,
+      "loss": 1.0347,
+      "step": 1206
+    },
+    {
+      "epoch": 0.42574017327660324,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.3240608398376051e-05,
+      "loss": 0.927,
+      "step": 1207
+    },
+    {
+      "epoch": 0.4260928991865259,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.3229574050037863e-05,
+      "loss": 0.975,
+      "step": 1208
+    },
+    {
+      "epoch": 0.4264456250964485,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.321853530982681e-05,
+      "loss": 1.0639,
+      "step": 1209
+    },
+    {
+      "epoch": 0.4267983510063711,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.320749219275438e-05,
+      "loss": 1.0597,
+      "step": 1210
+    },
+    {
+      "epoch": 0.4271510769162937,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3196444713838028e-05,
+      "loss": 1.0353,
+      "step": 1211
+    },
+    {
+      "epoch": 0.42750380282621636,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3185392888101135e-05,
+      "loss": 0.8998,
+      "step": 1212
+    },
+    {
+      "epoch": 0.427856528736139,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.3174336730572997e-05,
+      "loss": 0.9903,
+      "step": 1213
+    },
+    {
+      "epoch": 0.4282092546460616,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.3163276256288793e-05,
+      "loss": 0.9469,
+      "step": 1214
+    },
+    {
+      "epoch": 0.4285619805559842,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.3152211480289581e-05,
+      "loss": 0.8972,
+      "step": 1215
+    },
+    {
+      "epoch": 0.42891470646590685,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.314114241762226e-05,
+      "loss": 0.8905,
+      "step": 1216
+    },
+    {
+      "epoch": 0.4292674323758295,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3130069083339563e-05,
+      "loss": 1.0237,
+      "step": 1217
+    },
+    {
+      "epoch": 0.42962015828575206,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3118991492500034e-05,
+      "loss": 0.9197,
+      "step": 1218
+    },
+    {
+      "epoch": 0.4299728841956747,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.3107909660168006e-05,
+      "loss": 0.9794,
+      "step": 1219
+    },
+    {
+      "epoch": 0.43032561010559733,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.3096823601413573e-05,
+      "loss": 0.9171,
+      "step": 1220
+    },
+    {
+      "epoch": 0.4306783360155199,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3085733331312585e-05,
+      "loss": 0.9633,
+      "step": 1221
+    },
+    {
+      "epoch": 0.43103106192544255,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.3074638864946612e-05,
+      "loss": 0.9813,
+      "step": 1222
+    },
+    {
+      "epoch": 0.4313837878353652,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.3063540217402934e-05,
+      "loss": 0.948,
+      "step": 1223
+    },
+    {
+      "epoch": 0.4317365137452878,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3052437403774514e-05,
+      "loss": 1.0296,
+      "step": 1224
+    },
+    {
+      "epoch": 0.4320892396552104,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.3041330439159987e-05,
+      "loss": 0.9071,
+      "step": 1225
+    },
+    {
+      "epoch": 0.43244196556513304,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.3030219338663628e-05,
+      "loss": 1.0379,
+      "step": 1226
+    },
+    {
+      "epoch": 0.43279469147505567,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3019104117395334e-05,
+      "loss": 0.9648,
+      "step": 1227
+    },
+    {
+      "epoch": 0.4331474173849783,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.300798479047061e-05,
+      "loss": 0.9531,
+      "step": 1228
+    },
+    {
+      "epoch": 0.4335001432949009,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.2996861373010543e-05,
+      "loss": 0.8299,
+      "step": 1229
+    },
+    {
+      "epoch": 0.4338528692048235,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2985733880141785e-05,
+      "loss": 0.9483,
+      "step": 1230
+    },
+    {
+      "epoch": 0.43420559511474616,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.2974602326996522e-05,
+      "loss": 0.8909,
+      "step": 1231
+    },
+    {
+      "epoch": 0.4345583210246688,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.2963466728712474e-05,
+      "loss": 0.9928,
+      "step": 1232
+    },
+    {
+      "epoch": 0.43491104693459137,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.2952327100432853e-05,
+      "loss": 0.9401,
+      "step": 1233
+    },
+    {
+      "epoch": 0.435263772844514,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2941183457306355e-05,
+      "loss": 0.9619,
+      "step": 1234
+    },
+    {
+      "epoch": 0.43561649875443664,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.2930035814487128e-05,
+      "loss": 1.0585,
+      "step": 1235
+    },
+    {
+      "epoch": 0.4359692246643593,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.2918884187134774e-05,
+      "loss": 0.9065,
+      "step": 1236
+    },
+    {
+      "epoch": 0.43632195057428186,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.2907728590414299e-05,
+      "loss": 0.7564,
+      "step": 1237
+    },
+    {
+      "epoch": 0.4366746764842045,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.2896569039496116e-05,
+      "loss": 0.8913,
+      "step": 1238
+    },
+    {
+      "epoch": 0.43702740239412713,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.2885405549556012e-05,
+      "loss": 1.0038,
+      "step": 1239
+    },
+    {
+      "epoch": 0.4373801283040497,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.287423813577513e-05,
+      "loss": 0.8793,
+      "step": 1240
+    },
+    {
+      "epoch": 0.43773285421397234,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.2863066813339953e-05,
+      "loss": 0.8376,
+      "step": 1241
+    },
+    {
+      "epoch": 0.438085580123895,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2851891597442272e-05,
+      "loss": 1.001,
+      "step": 1242
+    },
+    {
+      "epoch": 0.4384383060338176,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.284071250327918e-05,
+      "loss": 0.9524,
+      "step": 1243
+    },
+    {
+      "epoch": 0.4387910319437402,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.282952954605304e-05,
+      "loss": 0.8758,
+      "step": 1244
+    },
+    {
+      "epoch": 0.43914375785366283,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2818342740971468e-05,
+      "loss": 0.9796,
+      "step": 1245
+    },
+    {
+      "epoch": 0.43949648376358547,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2807152103247318e-05,
+      "loss": 0.8898,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4398492096735081,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.2795957648098645e-05,
+      "loss": 1.0303,
+      "step": 1247
+    },
+    {
+      "epoch": 0.4402019355834307,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.278475939074871e-05,
+      "loss": 0.9565,
+      "step": 1248
+    },
+    {
+      "epoch": 0.4405546614933533,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.2773557346425933e-05,
+      "loss": 0.879,
+      "step": 1249
+    },
+    {
+      "epoch": 0.44090738740327595,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.276235153036389e-05,
+      "loss": 0.9881,
+      "step": 1250
+    },
+    {
+      "epoch": 0.4412601133131986,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.2751141957801282e-05,
+      "loss": 0.9085,
+      "step": 1251
+    },
+    {
+      "epoch": 0.44161283922312117,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.2739928643981917e-05,
+      "loss": 0.8676,
+      "step": 1252
+    },
+    {
+      "epoch": 0.4419655651330438,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.2728711604154702e-05,
+      "loss": 0.9733,
+      "step": 1253
+    },
+    {
+      "epoch": 0.44231829104296644,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2717490853573596e-05,
+      "loss": 1.056,
+      "step": 1254
+    },
+    {
+      "epoch": 0.442671016952889,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2706266407497618e-05,
+      "loss": 1.002,
+      "step": 1255
+    },
+    {
+      "epoch": 0.44302374286281165,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.2695038281190798e-05,
+      "loss": 1.0,
+      "step": 1256
+    },
+    {
+      "epoch": 0.4433764687727343,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2683806489922183e-05,
+      "loss": 0.9206,
+      "step": 1257
+    },
+    {
+      "epoch": 0.4437291946826569,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.2672571048965798e-05,
+      "loss": 0.9984,
+      "step": 1258
+    },
+    {
+      "epoch": 0.4440819205925795,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2661331973600635e-05,
+      "loss": 0.9473,
+      "step": 1259
+    },
+    {
+      "epoch": 0.44443464650250214,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.2650089279110627e-05,
+      "loss": 0.9802,
+      "step": 1260
+    },
+    {
+      "epoch": 0.4447873724124248,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.263884298078462e-05,
+      "loss": 0.9056,
+      "step": 1261
+    },
+    {
+      "epoch": 0.4451400983223474,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.2627593093916377e-05,
+      "loss": 0.9846,
+      "step": 1262
+    },
+    {
+      "epoch": 0.44549282423227,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2616339633804527e-05,
+      "loss": 1.0284,
+      "step": 1263
+    },
+    {
+      "epoch": 0.4458455501421926,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2605082615752564e-05,
+      "loss": 0.8812,
+      "step": 1264
+    },
+    {
+      "epoch": 0.44619827605211526,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.2593822055068825e-05,
+      "loss": 0.9679,
+      "step": 1265
+    },
+    {
+      "epoch": 0.4465510019620379,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.2582557967066453e-05,
+      "loss": 0.9534,
+      "step": 1266
+    },
+    {
+      "epoch": 0.4469037278719605,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2571290367063402e-05,
+      "loss": 0.9731,
+      "step": 1267
+    },
+    {
+      "epoch": 0.4472564537818831,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2560019270382389e-05,
+      "loss": 0.9765,
+      "step": 1268
+    },
+    {
+      "epoch": 0.44760917969180575,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2548744692350892e-05,
+      "loss": 0.8918,
+      "step": 1269
+    },
+    {
+      "epoch": 0.4479619056017284,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2537466648301124e-05,
+      "loss": 0.8566,
+      "step": 1270
+    },
+    {
+      "epoch": 0.44831463151165096,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.252618515357001e-05,
+      "loss": 1.0022,
+      "step": 1271
+    },
+    {
+      "epoch": 0.4486673574215736,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.2514900223499167e-05,
+      "loss": 0.917,
+      "step": 1272
+    },
+    {
+      "epoch": 0.44902008333149623,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2503611873434886e-05,
+      "loss": 0.8482,
+      "step": 1273
+    },
+    {
+      "epoch": 0.4493728092414188,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.2492320118728109e-05,
+      "loss": 0.9041,
+      "step": 1274
+    },
+    {
+      "epoch": 0.44972553515134145,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2481024974734401e-05,
+      "loss": 0.9497,
+      "step": 1275
+    },
+    {
+      "epoch": 0.4500782610612641,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.2469726456813942e-05,
+      "loss": 0.9451,
+      "step": 1276
+    },
+    {
+      "epoch": 0.4504309869711867,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2458424580331505e-05,
+      "loss": 0.9745,
+      "step": 1277
+    },
+    {
+      "epoch": 0.4507837128811093,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.244711936065642e-05,
+      "loss": 0.9374,
+      "step": 1278
+    },
+    {
+      "epoch": 0.45113643879103194,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.2435810813162569e-05,
+      "loss": 0.9724,
+      "step": 1279
+    },
+    {
+      "epoch": 0.45148916470095457,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.2424498953228362e-05,
+      "loss": 0.9636,
+      "step": 1280
+    },
+    {
+      "epoch": 0.4518418906108772,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2413183796236703e-05,
+      "loss": 1.0748,
+      "step": 1281
+    },
+    {
+      "epoch": 0.4521946165207998,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.240186535757499e-05,
+      "loss": 0.9435,
+      "step": 1282
+    },
+    {
+      "epoch": 0.4525473424307224,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2390543652635088e-05,
+      "loss": 0.8927,
+      "step": 1283
+    },
+    {
+      "epoch": 0.45290006834064506,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2379218696813285e-05,
+      "loss": 0.9663,
+      "step": 1284
+    },
+    {
+      "epoch": 0.4532527942505677,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2367890505510305e-05,
+      "loss": 0.9374,
+      "step": 1285
+    },
+    {
+      "epoch": 0.45360552016049027,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.235655909413127e-05,
+      "loss": 0.9249,
+      "step": 1286
+    },
+    {
+      "epoch": 0.4539582460704129,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.234522447808568e-05,
+      "loss": 1.0577,
+      "step": 1287
+    },
+    {
+      "epoch": 0.45431097198033554,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.2333886672787391e-05,
+      "loss": 0.9653,
+      "step": 1288
+    },
+    {
+      "epoch": 0.4546636978902582,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.2322545693654595e-05,
+      "loss": 0.9736,
+      "step": 1289
+    },
+    {
+      "epoch": 0.45501642380018076,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.2311201556109803e-05,
+      "loss": 1.0039,
+      "step": 1290
+    },
+    {
+      "epoch": 0.4553691497101034,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.229985427557982e-05,
+      "loss": 0.9214,
+      "step": 1291
+    },
+    {
+      "epoch": 0.45572187562002603,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.2288503867495728e-05,
+      "loss": 0.9572,
+      "step": 1292
+    },
+    {
+      "epoch": 0.4560746015299486,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.2277150347292854e-05,
+      "loss": 0.9636,
+      "step": 1293
+    },
+    {
+      "epoch": 0.45642732743987124,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.2265793730410769e-05,
+      "loss": 0.9347,
+      "step": 1294
+    },
+    {
+      "epoch": 0.4567800533497939,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2254434032293245e-05,
+      "loss": 1.0389,
+      "step": 1295
+    },
+    {
+      "epoch": 0.4571327792597165,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.224307126838825e-05,
+      "loss": 0.9889,
+      "step": 1296
+    },
+    {
+      "epoch": 0.4574855051696391,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2231705454147918e-05,
+      "loss": 1.0228,
+      "step": 1297
+    },
+    {
+      "epoch": 0.45783823107956173,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2220336605028532e-05,
+      "loss": 0.9166,
+      "step": 1298
+    },
+    {
+      "epoch": 0.45819095698948437,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.22089647364905e-05,
+      "loss": 0.9297,
+      "step": 1299
+    },
+    {
+      "epoch": 0.458543682899407,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2197589863998343e-05,
+      "loss": 0.9669,
+      "step": 1300
+    },
+    {
+      "epoch": 0.4588964088093296,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2186212003020658e-05,
+      "loss": 1.0255,
+      "step": 1301
+    },
+    {
+      "epoch": 0.4592491347192522,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.217483116903012e-05,
+      "loss": 0.8066,
+      "step": 1302
+    },
+    {
+      "epoch": 0.45960186062917485,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2163447377503423e-05,
+      "loss": 0.8922,
+      "step": 1303
+    },
+    {
+      "epoch": 0.4599545865390975,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2152060643921308e-05,
+      "loss": 1.0217,
+      "step": 1304
+    },
+    {
+      "epoch": 0.46030731244902007,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.2140670983768504e-05,
+      "loss": 0.9604,
+      "step": 1305
+    },
+    {
+      "epoch": 0.4606600383589427,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.212927841253372e-05,
+      "loss": 0.955,
+      "step": 1306
+    },
+    {
+      "epoch": 0.46101276426886534,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2117882945709626e-05,
+      "loss": 0.9113,
+      "step": 1307
+    },
+    {
+      "epoch": 0.4613654901787879,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2106484598792832e-05,
+      "loss": 0.9148,
+      "step": 1308
+    },
+    {
+      "epoch": 0.46171821608871055,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2095083387283859e-05,
+      "loss": 0.8977,
+      "step": 1309
+    },
+    {
+      "epoch": 0.4620709419986332,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.208367932668713e-05,
+      "loss": 0.984,
+      "step": 1310
+    },
+    {
+      "epoch": 0.4624236679085558,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2072272432510942e-05,
+      "loss": 0.9244,
+      "step": 1311
+    },
+    {
+      "epoch": 0.4627763938184784,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.206086272026743e-05,
+      "loss": 0.9071,
+      "step": 1312
+    },
+    {
+      "epoch": 0.46312911972840104,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2049450205472585e-05,
+      "loss": 0.9831,
+      "step": 1313
+    },
+    {
+      "epoch": 0.4634818456383237,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.2038034903646194e-05,
+      "loss": 0.8433,
+      "step": 1314
+    },
+    {
+      "epoch": 0.4638345715482463,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.202661683031184e-05,
+      "loss": 0.9624,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4641872974581689,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.2015196000996872e-05,
+      "loss": 1.021,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4645400233680915,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.2003772431232385e-05,
+      "loss": 0.8463,
+      "step": 1317
+    },
+    {
+      "epoch": 0.46489274927801416,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.199234613655321e-05,
+      "loss": 0.9284,
+      "step": 1318
+    },
+    {
+      "epoch": 0.4652454751879368,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.1980917132497871e-05,
+      "loss": 0.8922,
+      "step": 1319
+    },
+    {
+      "epoch": 0.4655982010978594,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.1969485434608587e-05,
+      "loss": 1.0073,
+      "step": 1320
+    },
+    {
+      "epoch": 0.465950927007782,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.1958051058431232e-05,
+      "loss": 0.8869,
+      "step": 1321
+    },
+    {
+      "epoch": 0.46630365291770465,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.1946614019515326e-05,
+      "loss": 0.894,
+      "step": 1322
+    },
+    {
+      "epoch": 0.4666563788276273,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.1935174333414014e-05,
+      "loss": 0.9551,
+      "step": 1323
+    },
+    {
+      "epoch": 0.46700910473754986,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.1923732015684033e-05,
+      "loss": 0.9848,
+      "step": 1324
+    },
+    {
+      "epoch": 0.4673618306474725,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.1912287081885708e-05,
+      "loss": 0.8676,
+      "step": 1325
+    },
+    {
+      "epoch": 0.46771455655739513,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.1900839547582912e-05,
+      "loss": 0.9915,
+      "step": 1326
+    },
+    {
+      "epoch": 0.4680672824673177,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.1889389428343057e-05,
+      "loss": 0.8597,
+      "step": 1327
+    },
+    {
+      "epoch": 0.46842000837724035,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.1877936739737075e-05,
+      "loss": 0.8732,
+      "step": 1328
+    },
+    {
+      "epoch": 0.468772734287163,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.1866481497339388e-05,
+      "loss": 0.9691,
+      "step": 1329
+    },
+    {
+      "epoch": 0.4691254601970856,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.1855023716727896e-05,
+      "loss": 0.875,
+      "step": 1330
+    },
+    {
+      "epoch": 0.4694781861070082,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.1843563413483941e-05,
+      "loss": 0.9112,
+      "step": 1331
+    },
+    {
+      "epoch": 0.46983091201693084,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.1832100603192302e-05,
+      "loss": 0.945,
+      "step": 1332
+    },
+    {
+      "epoch": 0.47018363792685347,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.1820635301441168e-05,
+      "loss": 0.9587,
+      "step": 1333
+    },
+    {
+      "epoch": 0.4705363638367761,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.1809167523822115e-05,
+      "loss": 0.9014,
+      "step": 1334
+    },
+    {
+      "epoch": 0.4708890897466987,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.1797697285930084e-05,
+      "loss": 1.0895,
+      "step": 1335
+    },
+    {
+      "epoch": 0.4712418156566213,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.1786224603363364e-05,
+      "loss": 0.8954,
+      "step": 1336
+    },
+    {
+      "epoch": 0.47159454156654396,
+      "grad_norm": 0.75,
+      "learning_rate": 1.1774749491723565e-05,
+      "loss": 0.8571,
+      "step": 1337
+    },
+    {
+      "epoch": 0.4719472674764666,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.1763271966615606e-05,
+      "loss": 0.9533,
+      "step": 1338
+    },
+    {
+      "epoch": 0.47229999338638917,
+      "grad_norm": 0.75,
+      "learning_rate": 1.1751792043647685e-05,
+      "loss": 1.007,
+      "step": 1339
+    },
+    {
+      "epoch": 0.4726527192963118,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.174030973843126e-05,
+      "loss": 0.8724,
+      "step": 1340
+    },
+    {
+      "epoch": 0.47300544520623444,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.1728825066581029e-05,
+      "loss": 0.9727,
+      "step": 1341
+    },
+    {
+      "epoch": 0.4733581711161571,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.1717338043714907e-05,
+      "loss": 1.0812,
+      "step": 1342
+    },
+    {
+      "epoch": 0.47371089702607966,
+      "grad_norm": 0.75,
+      "learning_rate": 1.1705848685454011e-05,
+      "loss": 0.9875,
+      "step": 1343
+    },
+    {
+      "epoch": 0.4740636229360023,
+      "grad_norm": 0.75,
+      "learning_rate": 1.1694357007422633e-05,
+      "loss": 0.913,
+      "step": 1344
+    },
+    {
+      "epoch": 0.47441634884592493,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.168286302524821e-05,
+      "loss": 0.9691,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4747690747558475,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.1671366754561327e-05,
+      "loss": 1.0101,
+      "step": 1346
+    },
+    {
+      "epoch": 0.47512180066577014,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.1659868210995669e-05,
+      "loss": 0.9422,
+      "step": 1347
+    },
+    {
+      "epoch": 0.4754745265756928,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.1648367410188019e-05,
+      "loss": 0.9546,
+      "step": 1348
+    },
+    {
+      "epoch": 0.4758272524856154,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.1636864367778226e-05,
+      "loss": 0.8227,
+      "step": 1349
+    },
+    {
+      "epoch": 0.476179978395538,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.1625359099409193e-05,
+      "loss": 0.9485,
+      "step": 1350
+    },
+    {
+      "epoch": 0.47653270430546063,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.161385162072684e-05,
+      "loss": 1.0524,
+      "step": 1351
+    },
+    {
+      "epoch": 0.47688543021538327,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.1602341947380103e-05,
+      "loss": 0.8611,
+      "step": 1352
+    },
+    {
+      "epoch": 0.4772381561253059,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.1590830095020896e-05,
+      "loss": 0.8741,
+      "step": 1353
+    },
+    {
+      "epoch": 0.4775908820352285,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.1579316079304096e-05,
+      "loss": 0.8813,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4779436079451511,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.1567799915887532e-05,
+      "loss": 0.9865,
+      "step": 1355
+    },
+    {
+      "epoch": 0.47829633385507375,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.1556281620431936e-05,
+      "loss": 1.0178,
+      "step": 1356
+    },
+    {
+      "epoch": 0.4786490597649964,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.1544761208600957e-05,
+      "loss": 1.0196,
+      "step": 1357
+    },
+    {
+      "epoch": 0.47900178567491897,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.1533238696061112e-05,
+      "loss": 1.0035,
+      "step": 1358
+    },
+    {
+      "epoch": 0.4793545115848416,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.1521714098481775e-05,
+      "loss": 0.9407,
+      "step": 1359
+    },
+    {
+      "epoch": 0.47970723749476424,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.1510187431535158e-05,
+      "loss": 0.938,
+      "step": 1360
+    },
+    {
+      "epoch": 0.4800599634046868,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.149865871089629e-05,
+      "loss": 0.9797,
+      "step": 1361
+    },
+    {
+      "epoch": 0.48041268931460945,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.1487127952242985e-05,
+      "loss": 0.9853,
+      "step": 1362
+    },
+    {
+      "epoch": 0.4807654152245321,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.1475595171255836e-05,
+      "loss": 0.9656,
+      "step": 1363
+    },
+    {
+      "epoch": 0.4811181411344547,
+      "grad_norm": 0.75,
+      "learning_rate": 1.1464060383618182e-05,
+      "loss": 0.9878,
+      "step": 1364
+    },
+    {
+      "epoch": 0.4814708670443773,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.145252360501609e-05,
+      "loss": 1.0102,
+      "step": 1365
+    },
+    {
+      "epoch": 0.48182359295429994,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.1440984851138339e-05,
+      "loss": 0.926,
+      "step": 1366
+    },
+    {
+      "epoch": 0.4821763188642226,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.1429444137676391e-05,
+      "loss": 0.9496,
+      "step": 1367
+    },
+    {
+      "epoch": 0.4825290447741452,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.1417901480324372e-05,
+      "loss": 0.9839,
+      "step": 1368
+    },
+    {
+      "epoch": 0.4828817706840678,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.1406356894779052e-05,
+      "loss": 0.8476,
+      "step": 1369
+    },
+    {
+      "epoch": 0.4832344965939904,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.1394810396739828e-05,
+      "loss": 0.9086,
+      "step": 1370
+    },
+    {
+      "epoch": 0.48358722250391306,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.138326200190869e-05,
+      "loss": 0.962,
+      "step": 1371
+    },
+    {
+      "epoch": 0.4839399484138357,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.1371711725990213e-05,
+      "loss": 0.9364,
+      "step": 1372
+    },
+    {
+      "epoch": 0.4842926743237583,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.1360159584691529e-05,
+      "loss": 0.9965,
+      "step": 1373
+    },
+    {
+      "epoch": 0.4846454002336809,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.1348605593722302e-05,
+      "loss": 1.0156,
+      "step": 1374
+    },
+    {
+      "epoch": 0.48499812614360355,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.1337049768794719e-05,
+      "loss": 1.0384,
+      "step": 1375
+    },
+    {
+      "epoch": 0.4853508520535262,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.1325492125623458e-05,
+      "loss": 1.0063,
+      "step": 1376
+    },
+    {
+      "epoch": 0.48570357796344876,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.1313932679925665e-05,
+      "loss": 0.9363,
+      "step": 1377
+    },
+    {
+      "epoch": 0.4860563038733714,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.1302371447420944e-05,
+      "loss": 0.8944,
+      "step": 1378
+    },
+    {
+      "epoch": 0.48640902978329403,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.1290808443831324e-05,
+      "loss": 0.9115,
+      "step": 1379
+    },
+    {
+      "epoch": 0.4867617556932166,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.1279243684881247e-05,
+      "loss": 0.9259,
+      "step": 1380
+    },
+    {
+      "epoch": 0.48711448160313925,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.126767718629754e-05,
+      "loss": 0.9543,
+      "step": 1381
+    },
+    {
+      "epoch": 0.4874672075130619,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.1256108963809393e-05,
+      "loss": 0.8192,
+      "step": 1382
+    },
+    {
+      "epoch": 0.4878199334229845,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.124453903314834e-05,
+      "loss": 0.8974,
+      "step": 1383
+    },
+    {
+      "epoch": 0.4881726593329071,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.1232967410048247e-05,
+      "loss": 1.0014,
+      "step": 1384
+    },
+    {
+      "epoch": 0.48852538524282973,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.1221394110245271e-05,
+      "loss": 0.8814,
+      "step": 1385
+    },
+    {
+      "epoch": 0.48887811115275237,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.1209819149477857e-05,
+      "loss": 0.9736,
+      "step": 1386
+    },
+    {
+      "epoch": 0.489230837062675,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.1198242543486699e-05,
+      "loss": 0.8525,
+      "step": 1387
+    },
+    {
+      "epoch": 0.4895835629725976,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.1186664308014738e-05,
+      "loss": 0.9506,
+      "step": 1388
+    },
+    {
+      "epoch": 0.4899362888825202,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.1175084458807126e-05,
+      "loss": 1.0022,
+      "step": 1389
+    },
+    {
+      "epoch": 0.49028901479244286,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.1163503011611216e-05,
+      "loss": 0.9337,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4906417407023655,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.115191998217652e-05,
+      "loss": 0.9624,
+      "step": 1391
+    },
+    {
+      "epoch": 0.49099446661228807,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.1140335386254715e-05,
+      "loss": 0.9084,
+      "step": 1392
+    },
+    {
+      "epoch": 0.4913471925222107,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.1128749239599603e-05,
+      "loss": 0.8387,
+      "step": 1393
+    },
+    {
+      "epoch": 0.49169991843213334,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.1117161557967098e-05,
+      "loss": 0.9544,
+      "step": 1394
+    },
+    {
+      "epoch": 0.492052644342056,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.1105572357115197e-05,
+      "loss": 1.0167,
+      "step": 1395
+    },
+    {
+      "epoch": 0.49240537025197856,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.1093981652803965e-05,
+      "loss": 0.8697,
+      "step": 1396
+    },
+    {
+      "epoch": 0.4927580961619012,
+      "grad_norm": 0.75,
+      "learning_rate": 1.108238946079551e-05,
+      "loss": 0.8986,
+      "step": 1397
+    },
+    {
+      "epoch": 0.49311082207182383,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.1070795796853969e-05,
+      "loss": 0.9096,
+      "step": 1398
+    },
+    {
+      "epoch": 0.4934635479817464,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.105920067674547e-05,
+      "loss": 1.0279,
+      "step": 1399
+    },
+    {
+      "epoch": 0.49381627389166904,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.1047604116238133e-05,
+      "loss": 0.9369,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4941689998015917,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.1036006131102029e-05,
+      "loss": 0.9898,
+      "step": 1401
+    },
+    {
+      "epoch": 0.4945217257115143,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.1024406737109167e-05,
+      "loss": 1.0333,
+      "step": 1402
+    },
+    {
+      "epoch": 0.4948744516214369,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.1012805950033476e-05,
+      "loss": 0.9853,
+      "step": 1403
+    },
+    {
+      "epoch": 0.49522717753135953,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.100120378565078e-05,
+      "loss": 0.8156,
+      "step": 1404
+    },
+    {
+      "epoch": 0.49557990344128217,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.0989600259738766e-05,
+      "loss": 1.0421,
+      "step": 1405
+    },
+    {
+      "epoch": 0.4959326293512048,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.0977995388076983e-05,
+      "loss": 0.8805,
+      "step": 1406
+    },
+    {
+      "epoch": 0.4962853552611274,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.0966389186446806e-05,
+      "loss": 0.8886,
+      "step": 1407
+    },
+    {
+      "epoch": 0.49663808117105,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.0954781670631421e-05,
+      "loss": 0.9154,
+      "step": 1408
+    },
+    {
+      "epoch": 0.49699080708097265,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.0943172856415798e-05,
+      "loss": 0.9636,
+      "step": 1409
+    },
+    {
+      "epoch": 0.4973435329908953,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.0931562759586672e-05,
+      "loss": 0.9149,
+      "step": 1410
+    },
+    {
+      "epoch": 0.49769625890081787,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.0919951395932527e-05,
+      "loss": 0.9148,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4980489848107405,
+      "grad_norm": 0.75,
+      "learning_rate": 1.0908338781243565e-05,
+      "loss": 0.935,
+      "step": 1412
+    },
+    {
+      "epoch": 0.49840171072066314,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.0896724931311693e-05,
+      "loss": 0.976,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4987544366305857,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.0885109861930496e-05,
+      "loss": 1.0158,
+      "step": 1414
+    },
+    {
+      "epoch": 0.49910716254050835,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.0873493588895216e-05,
+      "loss": 1.0029,
+      "step": 1415
+    },
+    {
+      "epoch": 0.499459888450431,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.0861876128002738e-05,
+      "loss": 0.9039,
+      "step": 1416
+    },
+    {
+      "epoch": 0.4998126143603536,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.0850257495051548e-05,
+      "loss": 0.8763,
+      "step": 1417
+    },
+    {
+      "epoch": 0.5001653402702763,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.0838637705841743e-05,
+      "loss": 0.8521,
+      "step": 1418
+    },
+    {
+      "epoch": 0.5005180661801989,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.0827016776174984e-05,
+      "loss": 0.8657,
+      "step": 1419
+    },
+    {
+      "epoch": 0.5008707920901214,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.081539472185448e-05,
+      "loss": 1.0071,
+      "step": 1420
+    },
+    {
+      "epoch": 0.501223518000044,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.0803771558684977e-05,
+      "loss": 0.9205,
+      "step": 1421
+    },
+    {
+      "epoch": 0.5015762439099667,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.0792147302472723e-05,
+      "loss": 0.9717,
+      "step": 1422
+    },
+    {
+      "epoch": 0.5019289698198893,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.0780521969025457e-05,
+      "loss": 0.9519,
+      "step": 1423
+    },
+    {
+      "epoch": 0.502281695729812,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.0768895574152378e-05,
+      "loss": 0.9999,
+      "step": 1424
+    },
+    {
+      "epoch": 0.5026344216397346,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.0757268133664129e-05,
+      "loss": 1.0155,
+      "step": 1425
+    },
+    {
+      "epoch": 0.5029871475496572,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.0745639663372781e-05,
+      "loss": 0.9229,
+      "step": 1426
+    },
+    {
+      "epoch": 0.5033398734595799,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.0734010179091798e-05,
+      "loss": 1.0177,
+      "step": 1427
+    },
+    {
+      "epoch": 0.5036925993695024,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.0722379696636032e-05,
+      "loss": 1.012,
+      "step": 1428
+    },
+    {
+      "epoch": 0.504045325279425,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.0710748231821678e-05,
+      "loss": 0.9704,
+      "step": 1429
+    },
+    {
+      "epoch": 0.5043980511893477,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.069911580046628e-05,
+      "loss": 0.9466,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5047507770992703,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.068748241838869e-05,
+      "loss": 0.9554,
+      "step": 1431
+    },
+    {
+      "epoch": 0.5051035030091929,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.0675848101409058e-05,
+      "loss": 1.059,
+      "step": 1432
+    },
+    {
+      "epoch": 0.5054562289191156,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.0664212865348798e-05,
+      "loss": 1.012,
+      "step": 1433
+    },
+    {
+      "epoch": 0.5058089548290382,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.0652576726030579e-05,
+      "loss": 0.9217,
+      "step": 1434
+    },
+    {
+      "epoch": 0.5061616807389607,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.0640939699278299e-05,
+      "loss": 0.9729,
+      "step": 1435
+    },
+    {
+      "epoch": 0.5065144066488834,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.0629301800917058e-05,
+      "loss": 1.0311,
+      "step": 1436
+    },
+    {
+      "epoch": 0.506867132558806,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.061766304677315e-05,
+      "loss": 0.9429,
+      "step": 1437
+    },
+    {
+      "epoch": 0.5072198584687286,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.0606023452674022e-05,
+      "loss": 0.9353,
+      "step": 1438
+    },
+    {
+      "epoch": 0.5075725843786513,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.0594383034448266e-05,
+      "loss": 0.9159,
+      "step": 1439
+    },
+    {
+      "epoch": 0.5079253102885739,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.0582741807925602e-05,
+      "loss": 0.9395,
+      "step": 1440
+    },
+    {
+      "epoch": 0.5082780361984965,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.0571099788936845e-05,
+      "loss": 0.7592,
+      "step": 1441
+    },
+    {
+      "epoch": 0.5086307621084192,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.0559456993313886e-05,
+      "loss": 0.8932,
+      "step": 1442
+    },
+    {
+      "epoch": 0.5089834880183417,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.0547813436889668e-05,
+      "loss": 0.8716,
+      "step": 1443
+    },
+    {
+      "epoch": 0.5093362139282643,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.053616913549818e-05,
+      "loss": 0.9495,
+      "step": 1444
+    },
+    {
+      "epoch": 0.509688939838187,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.0524524104974414e-05,
+      "loss": 0.9725,
+      "step": 1445
+    },
+    {
+      "epoch": 0.5100416657481096,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.0512878361154361e-05,
+      "loss": 0.9654,
+      "step": 1446
+    },
+    {
+      "epoch": 0.5103943916580322,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0501231919874974e-05,
+      "loss": 0.8664,
+      "step": 1447
+    },
+    {
+      "epoch": 0.5107471175679549,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.0489584796974162e-05,
+      "loss": 0.9267,
+      "step": 1448
+    },
+    {
+      "epoch": 0.5110998434778775,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.0477937008290753e-05,
+      "loss": 1.0021,
+      "step": 1449
+    },
+    {
+      "epoch": 0.5114525693878,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.0466288569664491e-05,
+      "loss": 0.9478,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5118052952977227,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.0454639496935997e-05,
+      "loss": 0.9501,
+      "step": 1451
+    },
+    {
+      "epoch": 0.5121580212076453,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.0442989805946753e-05,
+      "loss": 0.9537,
+      "step": 1452
+    },
+    {
+      "epoch": 0.5125107471175679,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.0431339512539082e-05,
+      "loss": 0.902,
+      "step": 1453
+    },
+    {
+      "epoch": 0.5128634730274906,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.0419688632556133e-05,
+      "loss": 0.8314,
+      "step": 1454
+    },
+    {
+      "epoch": 0.5132161989374132,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.0408037181841845e-05,
+      "loss": 0.8123,
+      "step": 1455
+    },
+    {
+      "epoch": 0.5135689248473359,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.0396385176240939e-05,
+      "loss": 0.9625,
+      "step": 1456
+    },
+    {
+      "epoch": 0.5139216507572585,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.0384732631598886e-05,
+      "loss": 1.0427,
+      "step": 1457
+    },
+    {
+      "epoch": 0.514274376667181,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.037307956376189e-05,
+      "loss": 0.9776,
+      "step": 1458
+    },
+    {
+      "epoch": 0.5146271025771036,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.0361425988576868e-05,
+      "loss": 0.9203,
+      "step": 1459
+    },
+    {
+      "epoch": 0.5149798284870263,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.0349771921891432e-05,
+      "loss": 0.9591,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5153325543969489,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.0338117379553851e-05,
+      "loss": 0.9524,
+      "step": 1461
+    },
+    {
+      "epoch": 0.5156852803068716,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.0326462377413051e-05,
+      "loss": 1.0201,
+      "step": 1462
+    },
+    {
+      "epoch": 0.5160380062167942,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.031480693131858e-05,
+      "loss": 0.9743,
+      "step": 1463
+    },
+    {
+      "epoch": 0.5163907321267168,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.0303151057120586e-05,
+      "loss": 0.993,
+      "step": 1464
+    },
+    {
+      "epoch": 0.5167434580366395,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.0291494770669808e-05,
+      "loss": 1.0703,
+      "step": 1465
+    },
+    {
+      "epoch": 0.517096183946562,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.0279838087817535e-05,
+      "loss": 0.89,
+      "step": 1466
+    },
+    {
+      "epoch": 0.5174489098564846,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.0268181024415602e-05,
+      "loss": 0.9987,
+      "step": 1467
+    },
+    {
+      "epoch": 0.5178016357664073,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.0256523596316364e-05,
+      "loss": 0.922,
+      "step": 1468
+    },
+    {
+      "epoch": 0.5181543616763299,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.024486581937266e-05,
+      "loss": 0.983,
+      "step": 1469
+    },
+    {
+      "epoch": 0.5185070875862525,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.0233207709437817e-05,
+      "loss": 1.0169,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5188598134961752,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.0221549282365603e-05,
+      "loss": 0.9513,
+      "step": 1471
+    },
+    {
+      "epoch": 0.5192125394060978,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.0209890554010228e-05,
+      "loss": 1.0778,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5195652653160203,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.0198231540226307e-05,
+      "loss": 0.9281,
+      "step": 1473
+    },
+    {
+      "epoch": 0.519917991225943,
+      "grad_norm": 0.75,
+      "learning_rate": 1.0186572256868843e-05,
+      "loss": 0.9713,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5202707171358656,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.0174912719793202e-05,
+      "loss": 0.9384,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5206234430457882,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.01632529448551e-05,
+      "loss": 0.9982,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5209761689557109,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.0151592947910578e-05,
+      "loss": 0.9983,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5213288948656335,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.0139932744815973e-05,
+      "loss": 0.9534,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5216816207755561,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.0128272351427908e-05,
+      "loss": 1.0692,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5220343466854788,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.0116611783603258e-05,
+      "loss": 0.8409,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5223870725954013,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.0104951057199144e-05,
+      "loss": 0.9872,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5227397985053239,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.0093290188072895e-05,
+      "loss": 1.0343,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5230925244152466,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.0081629192082037e-05,
+      "loss": 1.0359,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5234452503251692,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.0069968085084272e-05,
+      "loss": 0.8757,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5237979762350918,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.0058306882937445e-05,
+      "loss": 1.0655,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5241507021450145,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.0046645601499539e-05,
+      "loss": 1.0408,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5245034280549371,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.0034984256628637e-05,
+      "loss": 0.9688,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5248561539648596,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.0023322864182916e-05,
+      "loss": 0.8983,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5252088798747823,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.001166144002061e-05,
+      "loss": 0.9272,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5255616057847049,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1e-05,
+      "loss": 1.0348,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5259143316946275,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.988338559979395e-06,
+      "loss": 0.9491,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5262670576045502,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.976677135817087e-06,
+      "loss": 0.9628,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5266197835144728,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.965015743371368e-06,
+      "loss": 1.0067,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5269725094243954,
+      "grad_norm": 0.75,
+      "learning_rate": 9.953354398500463e-06,
+      "loss": 0.9558,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5273252353343181,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.941693117062558e-06,
+      "loss": 0.877,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5276779612442406,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.930031914915733e-06,
+      "loss": 0.9475,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5280306871541632,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.918370807917963e-06,
+      "loss": 0.8945,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5283834130640859,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.906709811927109e-06,
+      "loss": 0.9697,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5287361389740085,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.895048942800856e-06,
+      "loss": 0.9419,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5290888648839311,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.883388216396745e-06,
+      "loss": 0.9718,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5294415907938538,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.871727648572097e-06,
+      "loss": 0.9449,
+      "step": 1501
+    },
+    {
+      "epoch": 0.5297943167037764,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.860067255184027e-06,
+      "loss": 1.0647,
+      "step": 1502
+    },
+    {
+      "epoch": 0.5301470426136989,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.848407052089424e-06,
+      "loss": 1.037,
+      "step": 1503
+    },
+    {
+      "epoch": 0.5304997685236216,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.836747055144905e-06,
+      "loss": 0.8971,
+      "step": 1504
+    },
+    {
+      "epoch": 0.5308524944335442,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.825087280206801e-06,
+      "loss": 0.9511,
+      "step": 1505
+    },
+    {
+      "epoch": 0.5312052203434668,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.813427743131163e-06,
+      "loss": 0.9765,
+      "step": 1506
+    },
+    {
+      "epoch": 0.5315579462533895,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.801768459773694e-06,
+      "loss": 0.8721,
+      "step": 1507
+    },
+    {
+      "epoch": 0.5319106721633121,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.790109445989775e-06,
+      "loss": 0.9903,
+      "step": 1508
+    },
+    {
+      "epoch": 0.5322633980732348,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.7784507176344e-06,
+      "loss": 0.9137,
+      "step": 1509
+    },
+    {
+      "epoch": 0.5326161239831574,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.766792290562188e-06,
+      "loss": 1.0142,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5329688498930799,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.755134180627342e-06,
+      "loss": 0.975,
+      "step": 1511
+    },
+    {
+      "epoch": 0.5333215758030025,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.74347640368364e-06,
+      "loss": 0.9745,
+      "step": 1512
+    },
+    {
+      "epoch": 0.5336743017129252,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.7318189755844e-06,
+      "loss": 0.8472,
+      "step": 1513
+    },
+    {
+      "epoch": 0.5340270276228478,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.720161912182468e-06,
+      "loss": 0.9948,
+      "step": 1514
+    },
+    {
+      "epoch": 0.5343797535327705,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.708505229330197e-06,
+      "loss": 1.024,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5347324794426931,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.696848942879416e-06,
+      "loss": 1.0359,
+      "step": 1516
+    },
+    {
+      "epoch": 0.5350852053526157,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.685193068681423e-06,
+      "loss": 0.911,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5354379312625384,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.67353762258695e-06,
+      "loss": 0.8336,
+      "step": 1518
+    },
+    {
+      "epoch": 0.5357906571724609,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.661882620446154e-06,
+      "loss": 0.866,
+      "step": 1519
+    },
+    {
+      "epoch": 0.5361433830823835,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.650228078108572e-06,
+      "loss": 0.9904,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5364961089923062,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.638574011423136e-06,
+      "loss": 0.9711,
+      "step": 1521
+    },
+    {
+      "epoch": 0.5368488349022288,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.626920436238115e-06,
+      "loss": 1.0234,
+      "step": 1522
+    },
+    {
+      "epoch": 0.5372015608121514,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.615267368401118e-06,
+      "loss": 0.9208,
+      "step": 1523
+    },
+    {
+      "epoch": 0.5375542867220741,
+      "grad_norm": 0.75,
+      "learning_rate": 9.603614823759064e-06,
+      "loss": 0.8495,
+      "step": 1524
+    },
+    {
+      "epoch": 0.5379070126319967,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.591962818158155e-06,
+      "loss": 0.883,
+      "step": 1525
+    },
+    {
+      "epoch": 0.5382597385419192,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.58031136744387e-06,
+      "loss": 0.9003,
+      "step": 1526
+    },
+    {
+      "epoch": 0.5386124644518419,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.568660487460918e-06,
+      "loss": 1.0209,
+      "step": 1527
+    },
+    {
+      "epoch": 0.5389651903617645,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.55701019405325e-06,
+      "loss": 0.9552,
+      "step": 1528
+    },
+    {
+      "epoch": 0.5393179162716871,
+      "grad_norm": 0.75,
+      "learning_rate": 9.545360503064007e-06,
+      "loss": 0.9092,
+      "step": 1529
+    },
+    {
+      "epoch": 0.5396706421816098,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.533711430335509e-06,
+      "loss": 0.9204,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5400233680915324,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.522062991709249e-06,
+      "loss": 0.931,
+      "step": 1531
+    },
+    {
+      "epoch": 0.540376094001455,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.510415203025844e-06,
+      "loss": 0.9924,
+      "step": 1532
+    },
+    {
+      "epoch": 0.5407288199113777,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.49876808012503e-06,
+      "loss": 0.955,
+      "step": 1533
+    },
+    {
+      "epoch": 0.5410815458213002,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.487121638845644e-06,
+      "loss": 0.9268,
+      "step": 1534
+    },
+    {
+      "epoch": 0.5414342717312228,
+      "grad_norm": 0.7578125,
+      "learning_rate": 9.475475895025586e-06,
+      "loss": 0.8977,
+      "step": 1535
+    },
+    {
+      "epoch": 0.5417869976411455,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.463830864501822e-06,
+      "loss": 0.9428,
+      "step": 1536
+    },
+    {
+      "epoch": 0.5421397235510681,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.452186563110337e-06,
+      "loss": 1.0106,
+      "step": 1537
+    },
+    {
+      "epoch": 0.5424924494609907,
+      "grad_norm": 1.0,
+      "learning_rate": 9.440543006686119e-06,
+      "loss": 1.028,
+      "step": 1538
+    },
+    {
+      "epoch": 0.5428451753709134,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.428900211063159e-06,
+      "loss": 0.9687,
+      "step": 1539
+    },
+    {
+      "epoch": 0.543197901280836,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.417258192074398e-06,
+      "loss": 0.9854,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5435506271907585,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.405616965551738e-06,
+      "loss": 0.9127,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5439033531006812,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.393976547325985e-06,
+      "loss": 0.9688,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5442560790106038,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.382336953226852e-06,
+      "loss": 0.9574,
+      "step": 1543
+    },
+    {
+      "epoch": 0.5446088049205264,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.370698199082946e-06,
+      "loss": 0.9884,
+      "step": 1544
+    },
+    {
+      "epoch": 0.5449615308304491,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.359060300721703e-06,
+      "loss": 0.9684,
+      "step": 1545
+    },
+    {
+      "epoch": 0.5453142567403717,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.347423273969423e-06,
+      "loss": 0.8314,
+      "step": 1546
+    },
+    {
+      "epoch": 0.5456669826502943,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.335787134651207e-06,
+      "loss": 0.8531,
+      "step": 1547
+    },
+    {
+      "epoch": 0.546019708560217,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.324151898590945e-06,
+      "loss": 0.926,
+      "step": 1548
+    },
+    {
+      "epoch": 0.5463724344701395,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.312517581611313e-06,
+      "loss": 1.0918,
+      "step": 1549
+    },
+    {
+      "epoch": 0.5467251603800621,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.300884199533722e-06,
+      "loss": 0.8886,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5470778862899848,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.289251768178324e-06,
+      "loss": 0.9986,
+      "step": 1551
+    },
+    {
+      "epoch": 0.5474306121999074,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.277620303363973e-06,
+      "loss": 0.8878,
+      "step": 1552
+    },
+    {
+      "epoch": 0.54778333810983,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.2659898209082e-06,
+      "loss": 0.9566,
+      "step": 1553
+    },
+    {
+      "epoch": 0.5481360640197527,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.254360336627222e-06,
+      "loss": 0.8896,
+      "step": 1554
+    },
+    {
+      "epoch": 0.5484887899296753,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.242731866335871e-06,
+      "loss": 0.9326,
+      "step": 1555
+    },
+    {
+      "epoch": 0.5488415158395978,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.231104425847625e-06,
+      "loss": 0.9227,
+      "step": 1556
+    },
+    {
+      "epoch": 0.5491942417495205,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.219478030974548e-06,
+      "loss": 1.1315,
+      "step": 1557
+    },
+    {
+      "epoch": 0.5495469676594431,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.207852697527279e-06,
+      "loss": 0.9445,
+      "step": 1558
+    },
+    {
+      "epoch": 0.5498996935693657,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.196228441315028e-06,
+      "loss": 0.9576,
+      "step": 1559
+    },
+    {
+      "epoch": 0.5502524194792884,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.184605278145524e-06,
+      "loss": 0.9295,
+      "step": 1560
+    },
+    {
+      "epoch": 0.550605145389211,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.172983223825021e-06,
+      "loss": 0.908,
+      "step": 1561
+    },
+    {
+      "epoch": 0.5509578712991337,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.16136229415826e-06,
+      "loss": 0.929,
+      "step": 1562
+    },
+    {
+      "epoch": 0.5513105972090563,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.149742504948454e-06,
+      "loss": 0.9397,
+      "step": 1563
+    },
+    {
+      "epoch": 0.5516633231189788,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.138123871997266e-06,
+      "loss": 1.0009,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5520160490289014,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.126506411104786e-06,
+      "loss": 0.9218,
+      "step": 1565
+    },
+    {
+      "epoch": 0.5523687749388241,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.114890138069507e-06,
+      "loss": 0.9398,
+      "step": 1566
+    },
+    {
+      "epoch": 0.5527215008487467,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.10327506868831e-06,
+      "loss": 0.8664,
+      "step": 1567
+    },
+    {
+      "epoch": 0.5530742267586694,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.091661218756438e-06,
+      "loss": 0.9916,
+      "step": 1568
+    },
+    {
+      "epoch": 0.553426952668592,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.080048604067476e-06,
+      "loss": 0.9345,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5537796785785146,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.068437240413333e-06,
+      "loss": 1.0067,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5541324044884373,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.056827143584206e-06,
+      "loss": 1.0133,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5544851303983598,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.045218329368584e-06,
+      "loss": 0.9664,
+      "step": 1572
+    },
+    {
+      "epoch": 0.5548378563082824,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.033610813553196e-06,
+      "loss": 0.7963,
+      "step": 1573
+    },
+    {
+      "epoch": 0.555190582218205,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.02200461192302e-06,
+      "loss": 0.8762,
+      "step": 1574
+    },
+    {
+      "epoch": 0.5555433081281277,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.010399740261239e-06,
+      "loss": 0.977,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5558960340380503,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.998796214349223e-06,
+      "loss": 1.0294,
+      "step": 1576
+    },
+    {
+      "epoch": 0.556248759947973,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.987194049966526e-06,
+      "loss": 0.8706,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5566014858578956,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.975593262890833e-06,
+      "loss": 1.0799,
+      "step": 1578
+    },
+    {
+      "epoch": 0.5569542117678181,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.963993868897975e-06,
+      "loss": 1.011,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5573069376777408,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.952395883761872e-06,
+      "loss": 0.9517,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5576596635876634,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.940799323254532e-06,
+      "loss": 0.8478,
+      "step": 1581
+    },
+    {
+      "epoch": 0.558012389497586,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.929204203146036e-06,
+      "loss": 0.8731,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5583651154075087,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.91761053920449e-06,
+      "loss": 1.0238,
+      "step": 1583
+    },
+    {
+      "epoch": 0.5587178413174313,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.906018347196039e-06,
+      "loss": 0.9701,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5590705672273539,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.894427642884808e-06,
+      "loss": 1.0005,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5594232931372766,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.882838442032903e-06,
+      "loss": 0.9581,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5597760190471991,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.871250760400399e-06,
+      "loss": 0.8692,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5601287449571217,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.85966461374529e-06,
+      "loss": 0.9237,
+      "step": 1588
+    },
+    {
+      "epoch": 0.5604814708670444,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.848080017823482e-06,
+      "loss": 0.9415,
+      "step": 1589
+    },
+    {
+      "epoch": 0.560834196776967,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.83649698838879e-06,
+      "loss": 0.8998,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5611869226868896,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.824915541192874e-06,
+      "loss": 0.976,
+      "step": 1591
+    },
+    {
+      "epoch": 0.5615396485968123,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.813335691985264e-06,
+      "loss": 0.9621,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5618923745067349,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.801757456513306e-06,
+      "loss": 0.9151,
+      "step": 1593
+    },
+    {
+      "epoch": 0.5622451004166574,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.790180850522147e-06,
+      "loss": 1.0357,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5625978263265801,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.77860588975473e-06,
+      "loss": 0.9004,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5629505522365027,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.767032589951753e-06,
+      "loss": 0.936,
+      "step": 1596
+    },
+    {
+      "epoch": 0.5633032781464253,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.755460966851661e-06,
+      "loss": 0.9143,
+      "step": 1597
+    },
+    {
+      "epoch": 0.563656004056348,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.743891036190612e-06,
+      "loss": 0.8832,
+      "step": 1598
+    },
+    {
+      "epoch": 0.5640087299662706,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.732322813702461e-06,
+      "loss": 1.0396,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5643614558761932,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.720756315118754e-06,
+      "loss": 0.964,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5647141817861159,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.709191556168675e-06,
+      "loss": 0.9411,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5650669076960384,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.697628552579058e-06,
+      "loss": 1.0131,
+      "step": 1602
+    },
+    {
+      "epoch": 0.565419633605961,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.68606732007434e-06,
+      "loss": 0.9803,
+      "step": 1603
+    },
+    {
+      "epoch": 0.5657723595158837,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.674507874376544e-06,
+      "loss": 0.9187,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5661250854258063,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.662950231205283e-06,
+      "loss": 0.992,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5664778113357289,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.651394406277698e-06,
+      "loss": 0.965,
+      "step": 1606
+    },
+    {
+      "epoch": 0.5668305372456516,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.639840415308475e-06,
+      "loss": 0.9357,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5671832631555742,
+      "grad_norm": 1.34375,
+      "learning_rate": 8.628288274009789e-06,
+      "loss": 0.9745,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5675359890654967,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.616737998091312e-06,
+      "loss": 0.8935,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5678887149754194,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.605189603260175e-06,
+      "loss": 0.9519,
+      "step": 1610
+    },
+    {
+      "epoch": 0.568241440885342,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.593643105220948e-06,
+      "loss": 0.8559,
+      "step": 1611
+    },
+    {
+      "epoch": 0.5685941667952646,
+      "grad_norm": 0.75,
+      "learning_rate": 8.582098519675633e-06,
+      "loss": 0.96,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5689468927051873,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.570555862323612e-06,
+      "loss": 0.9154,
+      "step": 1613
+    },
+    {
+      "epoch": 0.5692996186151099,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.559015148861663e-06,
+      "loss": 0.9712,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5696523445250326,
+      "grad_norm": 0.75,
+      "learning_rate": 8.547476394983913e-06,
+      "loss": 1.0153,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5700050704349552,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.535939616381823e-06,
+      "loss": 1.0243,
+      "step": 1616
+    },
+    {
+      "epoch": 0.5703577963448777,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.524404828744168e-06,
+      "loss": 1.1414,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5707105222548003,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.512872047757018e-06,
+      "loss": 1.0127,
+      "step": 1618
+    },
+    {
+      "epoch": 0.571063248164723,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.501341289103712e-06,
+      "loss": 0.9362,
+      "step": 1619
+    },
+    {
+      "epoch": 0.5714159740746456,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.489812568464843e-06,
+      "loss": 0.8861,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5717686999845683,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.47828590151823e-06,
+      "loss": 0.9423,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5721214258944909,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.466761303938891e-06,
+      "loss": 0.9274,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5724741518044135,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.455238791399046e-06,
+      "loss": 0.852,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5728268777143362,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.443718379568065e-06,
+      "loss": 0.8854,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5731796036242587,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.432200084112473e-06,
+      "loss": 0.9444,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5735323295341813,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.420683920695907e-06,
+      "loss": 0.8564,
+      "step": 1626
+    },
+    {
+      "epoch": 0.573885055444104,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.409169904979106e-06,
+      "loss": 1.0572,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5742377813540266,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.397658052619902e-06,
+      "loss": 0.898,
+      "step": 1628
+    },
+    {
+      "epoch": 0.5745905072639492,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.38614837927316e-06,
+      "loss": 0.9653,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5749432331738719,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.37464090059081e-06,
+      "loss": 0.9111,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5752959590837945,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.363135632221777e-06,
+      "loss": 0.9967,
+      "step": 1631
+    },
+    {
+      "epoch": 0.575648684993717,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.351632589811983e-06,
+      "loss": 0.9783,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5760014109036397,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.340131789004334e-06,
+      "loss": 1.0541,
+      "step": 1633
+    },
+    {
+      "epoch": 0.5763541368135623,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.328633245438675e-06,
+      "loss": 0.9626,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5767068627234849,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.317136974751791e-06,
+      "loss": 0.9807,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5770595886334076,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.305642992577372e-06,
+      "loss": 0.8238,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5774123145433302,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.294151314545988e-06,
+      "loss": 0.9791,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5777650404532528,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.282661956285096e-06,
+      "loss": 1.0026,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5781177663631755,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.271174933418973e-06,
+      "loss": 0.942,
+      "step": 1639
+    },
+    {
+      "epoch": 0.578470492273098,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.259690261568742e-06,
+      "loss": 1.0238,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5788232181830206,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.248207956352318e-06,
+      "loss": 0.9581,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5791759440929433,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.236728033384396e-06,
+      "loss": 0.8739,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5795286700028659,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.225250508276439e-06,
+      "loss": 0.8289,
+      "step": 1643
+    },
+    {
+      "epoch": 0.5798813959127885,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.213775396636643e-06,
+      "loss": 0.9348,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5802341218227112,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.202302714069918e-06,
+      "loss": 0.9059,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5805868477326338,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.19083247617789e-06,
+      "loss": 0.8206,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5809395736425563,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.179364698558832e-06,
+      "loss": 1.1363,
+      "step": 1647
+    },
+    {
+      "epoch": 0.581292299552479,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.167899396807701e-06,
+      "loss": 0.8424,
+      "step": 1648
+    },
+    {
+      "epoch": 0.5816450254624016,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.156436586516064e-06,
+      "loss": 0.9683,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5819977513723242,
+      "grad_norm": 0.75,
+      "learning_rate": 8.144976283272106e-06,
+      "loss": 0.9781,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5823504772822469,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.133518502660613e-06,
+      "loss": 0.99,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5827032031921695,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.122063260262926e-06,
+      "loss": 1.0254,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5830559291020921,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.110610571656946e-06,
+      "loss": 0.9104,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5834086550120148,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.099160452417095e-06,
+      "loss": 0.9073,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5837613809219373,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.087712918114294e-06,
+      "loss": 0.9514,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5841141068318599,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.076267984315969e-06,
+      "loss": 0.9358,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5844668327417826,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.064825666585988e-06,
+      "loss": 0.8972,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5848195586517052,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.053385980484676e-06,
+      "loss": 0.9414,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5851722845616278,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.041948941568771e-06,
+      "loss": 0.8753,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5855250104715505,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.030514565391416e-06,
+      "loss": 0.9454,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5858777363814731,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.019082867502132e-06,
+      "loss": 0.9969,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5862304622913956,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.007653863446792e-06,
+      "loss": 0.9297,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5865831882013183,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.996227568767617e-06,
+      "loss": 1.0092,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5869359141112409,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.984803999003131e-06,
+      "loss": 0.917,
+      "step": 1664
+    },
+    {
+      "epoch": 0.5872886400211635,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.973383169688163e-06,
+      "loss": 0.9517,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5876413659310862,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.961965096353808e-06,
+      "loss": 0.851,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5879940918410088,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.950549794527418e-06,
+      "loss": 1.0222,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5883468177509315,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.939137279732571e-06,
+      "loss": 0.8983,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5886995436608541,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.927727567489065e-06,
+      "loss": 0.8397,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5890522695707766,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.916320673312872e-06,
+      "loss": 0.9032,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5894049954806992,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.904916612716143e-06,
+      "loss": 0.9542,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5897577213906219,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.893515401207172e-06,
+      "loss": 0.9849,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5901104473005445,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.882117054290375e-06,
+      "loss": 0.8784,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5904631732104672,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.870721587466285e-06,
+      "loss": 0.982,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5908158991203898,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.859329016231498e-06,
+      "loss": 1.0103,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5911686250303124,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.847939356078693e-06,
+      "loss": 0.9492,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5915213509402351,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.83655262249658e-06,
+      "loss": 1.0139,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5918740768501576,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.825168830969886e-06,
+      "loss": 1.01,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5922268027600802,
+      "grad_norm": 0.75,
+      "learning_rate": 7.813787996979343e-06,
+      "loss": 0.948,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5925795286700029,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.802410136001659e-06,
+      "loss": 0.9442,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5929322545799255,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.791035263509503e-06,
+      "loss": 0.8754,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5932849804898481,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.779663394971475e-06,
+      "loss": 0.9341,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5936377063997708,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.768294545852087e-06,
+      "loss": 0.9547,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5939904323096934,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.756928731611756e-06,
+      "loss": 0.9288,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5943431582196159,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.745565967706757e-06,
+      "loss": 0.9516,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5946958841295386,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.734206269589234e-06,
+      "loss": 0.8655,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5950486100394612,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.72284965270715e-06,
+      "loss": 0.9356,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5954013359493838,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.711496132504276e-06,
+      "loss": 0.9111,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5957540618593065,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.700145724420184e-06,
+      "loss": 0.9869,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5961067877692291,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.688798443890199e-06,
+      "loss": 0.9201,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5964595136791517,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.677454306345408e-06,
+      "loss": 0.9528,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5968122395890744,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.666113327212614e-06,
+      "loss": 0.8902,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5971649654989969,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.65477552191432e-06,
+      "loss": 1.0119,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5975176914089195,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.643440905868731e-06,
+      "loss": 0.9157,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5978704173188422,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.632109494489695e-06,
+      "loss": 0.8841,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5982231432287648,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.620781303186719e-06,
+      "loss": 0.9271,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5985758691386874,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.609456347364919e-06,
+      "loss": 0.9996,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5989285950486101,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.598134642425009e-06,
+      "loss": 0.7926,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5992813209585327,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.586816203763301e-06,
+      "loss": 0.9208,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5996340468684552,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.575501046771645e-06,
+      "loss": 0.8484,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5999867727783779,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.5641891868374326e-06,
+      "loss": 0.8836,
+      "step": 1701
+    },
+    {
+      "epoch": 0.6003394986883005,
+      "grad_norm": 0.77734375,
+      "learning_rate": 7.5528806393435836e-06,
+      "loss": 0.9489,
+      "step": 1702
+    },
+    {
+      "epoch": 0.6006922245982231,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.541575419668497e-06,
+      "loss": 0.9899,
+      "step": 1703
+    },
+    {
+      "epoch": 0.6010449505081458,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.530273543186059e-06,
+      "loss": 0.8277,
+      "step": 1704
+    },
+    {
+      "epoch": 0.6013976764180684,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.518975025265605e-06,
+      "loss": 0.922,
+      "step": 1705
+    },
+    {
+      "epoch": 0.601750402327991,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.5076798812718945e-06,
+      "loss": 1.004,
+      "step": 1706
+    },
+    {
+      "epoch": 0.6021031282379137,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.496388126565117e-06,
+      "loss": 0.8686,
+      "step": 1707
+    },
+    {
+      "epoch": 0.6024558541478362,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.4850997765008325e-06,
+      "loss": 1.0188,
+      "step": 1708
+    },
+    {
+      "epoch": 0.6028085800577588,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.473814846429993e-06,
+      "loss": 0.8266,
+      "step": 1709
+    },
+    {
+      "epoch": 0.6031613059676815,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.462533351698878e-06,
+      "loss": 0.9709,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6035140318776041,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.45125530764911e-06,
+      "loss": 0.9371,
+      "step": 1711
+    },
+    {
+      "epoch": 0.6038667577875267,
+      "grad_norm": 0.75,
+      "learning_rate": 7.439980729617613e-06,
+      "loss": 1.0354,
+      "step": 1712
+    },
+    {
+      "epoch": 0.6042194836974494,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.428709632936599e-06,
+      "loss": 0.9962,
+      "step": 1713
+    },
+    {
+      "epoch": 0.604572209607372,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.417442032933548e-06,
+      "loss": 0.8555,
+      "step": 1714
+    },
+    {
+      "epoch": 0.6049249355172945,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.406177944931179e-06,
+      "loss": 0.9341,
+      "step": 1715
+    },
+    {
+      "epoch": 0.6052776614272172,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.3949173842474375e-06,
+      "loss": 0.8445,
+      "step": 1716
+    },
+    {
+      "epoch": 0.6056303873371398,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.383660366195476e-06,
+      "loss": 1.0009,
+      "step": 1717
+    },
+    {
+      "epoch": 0.6059831132470624,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.372406906083627e-06,
+      "loss": 0.984,
+      "step": 1718
+    },
+    {
+      "epoch": 0.6063358391569851,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.361157019215382e-06,
+      "loss": 1.0302,
+      "step": 1719
+    },
+    {
+      "epoch": 0.6066885650669077,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.349910720889378e-06,
+      "loss": 0.8822,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6070412909768303,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.338668026399365e-06,
+      "loss": 0.8877,
+      "step": 1721
+    },
+    {
+      "epoch": 0.607394016886753,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.327428951034203e-06,
+      "loss": 1.0129,
+      "step": 1722
+    },
+    {
+      "epoch": 0.6077467427966755,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.31619351007782e-06,
+      "loss": 0.9462,
+      "step": 1723
+    },
+    {
+      "epoch": 0.6080994687065981,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.304961718809205e-06,
+      "loss": 0.9591,
+      "step": 1724
+    },
+    {
+      "epoch": 0.6084521946165208,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.293733592502388e-06,
+      "loss": 0.9666,
+      "step": 1725
+    },
+    {
+      "epoch": 0.6088049205264434,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.282509146426404e-06,
+      "loss": 1.0716,
+      "step": 1726
+    },
+    {
+      "epoch": 0.609157646436366,
+      "grad_norm": 0.75,
+      "learning_rate": 7.271288395845302e-06,
+      "loss": 0.8845,
+      "step": 1727
+    },
+    {
+      "epoch": 0.6095103723462887,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.260071356018088e-06,
+      "loss": 1.0288,
+      "step": 1728
+    },
+    {
+      "epoch": 0.6098630982562113,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.2488580421987235e-06,
+      "loss": 0.8951,
+      "step": 1729
+    },
+    {
+      "epoch": 0.610215824166134,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.237648469636115e-06,
+      "loss": 0.8788,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6105685500760565,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.2264426535740685e-06,
+      "loss": 0.8609,
+      "step": 1731
+    },
+    {
+      "epoch": 0.6109212759859791,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.215240609251293e-06,
+      "loss": 0.915,
+      "step": 1732
+    },
+    {
+      "epoch": 0.6112740018959018,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.204042351901359e-06,
+      "loss": 0.951,
+      "step": 1733
+    },
+    {
+      "epoch": 0.6116267278058244,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.192847896752686e-06,
+      "loss": 1.0195,
+      "step": 1734
+    },
+    {
+      "epoch": 0.611979453715747,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.181657259028536e-06,
+      "loss": 0.7927,
+      "step": 1735
+    },
+    {
+      "epoch": 0.6123321796256697,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.170470453946963e-06,
+      "loss": 0.9241,
+      "step": 1736
+    },
+    {
+      "epoch": 0.6126849055355923,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.159287496720824e-06,
+      "loss": 0.9476,
+      "step": 1737
+    },
+    {
+      "epoch": 0.6130376314455148,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.148108402557732e-06,
+      "loss": 0.9846,
+      "step": 1738
+    },
+    {
+      "epoch": 0.6133903573554375,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.136933186660049e-06,
+      "loss": 1.0396,
+      "step": 1739
+    },
+    {
+      "epoch": 0.6137430832653601,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.125761864224871e-06,
+      "loss": 1.0677,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6140958091752827,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.114594450443988e-06,
+      "loss": 0.9884,
+      "step": 1741
+    },
+    {
+      "epoch": 0.6144485350852054,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.1034309605038865e-06,
+      "loss": 0.8882,
+      "step": 1742
+    },
+    {
+      "epoch": 0.614801260995128,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.0922714095857044e-06,
+      "loss": 0.8674,
+      "step": 1743
+    },
+    {
+      "epoch": 0.6151539869050506,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.081115812865229e-06,
+      "loss": 0.9664,
+      "step": 1744
+    },
+    {
+      "epoch": 0.6155067128149733,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.069964185512874e-06,
+      "loss": 0.9441,
+      "step": 1745
+    },
+    {
+      "epoch": 0.6158594387248958,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.058816542693647e-06,
+      "loss": 1.0343,
+      "step": 1746
+    },
+    {
+      "epoch": 0.6162121646348184,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.047672899567148e-06,
+      "loss": 0.9889,
+      "step": 1747
+    },
+    {
+      "epoch": 0.6165648905447411,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.036533271287529e-06,
+      "loss": 0.9463,
+      "step": 1748
+    },
+    {
+      "epoch": 0.6169176164546637,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.025397673003478e-06,
+      "loss": 0.9197,
+      "step": 1749
+    },
+    {
+      "epoch": 0.6172703423645863,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.014266119858219e-06,
+      "loss": 0.8792,
+      "step": 1750
+    },
+    {
+      "epoch": 0.617623068274509,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.003138626989457e-06,
+      "loss": 0.8428,
+      "step": 1751
+    },
+    {
+      "epoch": 0.6179757941844316,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.992015209529393e-06,
+      "loss": 0.9971,
+      "step": 1752
+    },
+    {
+      "epoch": 0.6183285200943541,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.98089588260467e-06,
+      "loss": 1.0313,
+      "step": 1753
+    },
+    {
+      "epoch": 0.6186812460042768,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.9697806613363735e-06,
+      "loss": 0.845,
+      "step": 1754
+    },
+    {
+      "epoch": 0.6190339719141994,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.958669560840015e-06,
+      "loss": 0.8572,
+      "step": 1755
+    },
+    {
+      "epoch": 0.619386697824122,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.947562596225487e-06,
+      "loss": 0.9324,
+      "step": 1756
+    },
+    {
+      "epoch": 0.6197394237340447,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.936459782597069e-06,
+      "loss": 0.8433,
+      "step": 1757
+    },
+    {
+      "epoch": 0.6200921496439673,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.925361135053392e-06,
+      "loss": 0.9234,
+      "step": 1758
+    },
+    {
+      "epoch": 0.6204448755538899,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.914266668687416e-06,
+      "loss": 0.8681,
+      "step": 1759
+    },
+    {
+      "epoch": 0.6207976014638126,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.903176398586428e-06,
+      "loss": 1.0204,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6211503273737351,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.892090339831996e-06,
+      "loss": 0.9662,
+      "step": 1761
+    },
+    {
+      "epoch": 0.6215030532836577,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.881008507499967e-06,
+      "loss": 0.8788,
+      "step": 1762
+    },
+    {
+      "epoch": 0.6218557791935804,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.86993091666044e-06,
+      "loss": 1.027,
+      "step": 1763
+    },
+    {
+      "epoch": 0.622208505103503,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.858857582377745e-06,
+      "loss": 0.8526,
+      "step": 1764
+    },
+    {
+      "epoch": 0.6225612310134256,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.847788519710422e-06,
+      "loss": 0.979,
+      "step": 1765
+    },
+    {
+      "epoch": 0.6229139569233483,
+      "grad_norm": 0.75,
+      "learning_rate": 6.83672374371121e-06,
+      "loss": 1.0253,
+      "step": 1766
+    },
+    {
+      "epoch": 0.6232666828332709,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.825663269427006e-06,
+      "loss": 1.0206,
+      "step": 1767
+    },
+    {
+      "epoch": 0.6236194087431934,
+      "grad_norm": 0.859375,
+      "learning_rate": 6.8146071118988654e-06,
+      "loss": 0.9781,
+      "step": 1768
+    },
+    {
+      "epoch": 0.6239721346531161,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.803555286161973e-06,
+      "loss": 0.9615,
+      "step": 1769
+    },
+    {
+      "epoch": 0.6243248605630387,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.792507807245623e-06,
+      "loss": 0.9189,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6246775864729613,
+      "grad_norm": 0.75,
+      "learning_rate": 6.781464690173196e-06,
+      "loss": 0.8781,
+      "step": 1771
+    },
+    {
+      "epoch": 0.625030312382884,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.770425949962136e-06,
+      "loss": 0.9966,
+      "step": 1772
+    },
+    {
+      "epoch": 0.6253830382928066,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.75939160162395e-06,
+      "loss": 1.0336,
+      "step": 1773
+    },
+    {
+      "epoch": 0.6257357642027292,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.748361660164157e-06,
+      "loss": 0.862,
+      "step": 1774
+    },
+    {
+      "epoch": 0.6260884901126519,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.737336140582291e-06,
+      "loss": 0.9895,
+      "step": 1775
+    },
+    {
+      "epoch": 0.6264412160225744,
+      "grad_norm": 0.75,
+      "learning_rate": 6.726315057871875e-06,
+      "loss": 1.0075,
+      "step": 1776
+    },
+    {
+      "epoch": 0.626793941932497,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.715298427020385e-06,
+      "loss": 0.9776,
+      "step": 1777
+    },
+    {
+      "epoch": 0.6271466678424197,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.704286263009263e-06,
+      "loss": 0.9652,
+      "step": 1778
+    },
+    {
+      "epoch": 0.6274993937523423,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.693278580813861e-06,
+      "loss": 0.8963,
+      "step": 1779
+    },
+    {
+      "epoch": 0.627852119662265,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.682275395403438e-06,
+      "loss": 0.8845,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6282048455721876,
+      "grad_norm": 0.77734375,
+      "learning_rate": 6.671276721741149e-06,
+      "loss": 1.0097,
+      "step": 1781
+    },
+    {
+      "epoch": 0.6285575714821102,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.660282574783996e-06,
+      "loss": 0.9903,
+      "step": 1782
+    },
+    {
+      "epoch": 0.6289102973920329,
+      "grad_norm": 0.78125,
+      "learning_rate": 6.649292969482845e-06,
+      "loss": 0.976,
+      "step": 1783
+    },
+    {
+      "epoch": 0.6292630233019554,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.63830792078237e-06,
+      "loss": 0.9392,
+      "step": 1784
+    },
+    {
+      "epoch": 0.629615749211878,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.62732744362105e-06,
+      "loss": 0.8908,
+      "step": 1785
+    },
+    {
+      "epoch": 0.6299684751218007,
+      "grad_norm": 0.75,
+      "learning_rate": 6.6163515529311606e-06,
+      "loss": 1.0142,
+      "step": 1786
+    },
+    {
+      "epoch": 0.6303212010317233,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.605380263638722e-06,
+      "loss": 0.952,
+      "step": 1787
+    },
+    {
+      "epoch": 0.6306739269416459,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.594413590663515e-06,
+      "loss": 0.8827,
+      "step": 1788
+    },
+    {
+      "epoch": 0.6310266528515686,
+      "grad_norm": 0.828125,
+      "learning_rate": 6.583451548919029e-06,
+      "loss": 0.9823,
+      "step": 1789
+    },
+    {
+      "epoch": 0.6313793787614912,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.572494153312454e-06,
+      "loss": 0.9612,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6317321046714137,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.561541418744682e-06,
+      "loss": 0.9612,
+      "step": 1791
+    },
+    {
+      "epoch": 0.6320848305813364,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.5505933601102355e-06,
+      "loss": 0.9129,
+      "step": 1792
+    },
+    {
+      "epoch": 0.632437556491259,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.539649992297311e-06,
+      "loss": 0.9441,
+      "step": 1793
+    },
+    {
+      "epoch": 0.6327902824011816,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.528711330187703e-06,
+      "loss": 0.9767,
+      "step": 1794
+    },
+    {
+      "epoch": 0.6331430083111043,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.517777388656808e-06,
+      "loss": 0.8682,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6334957342210269,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.50684818257362e-06,
+      "loss": 0.8698,
+      "step": 1796
+    },
+    {
+      "epoch": 0.6338484601309495,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.495923726800672e-06,
+      "loss": 0.9279,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6342011860408722,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.485004036194058e-06,
+      "loss": 0.8965,
+      "step": 1798
+    },
+    {
+      "epoch": 0.6345539119507947,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.4740891256033736e-06,
+      "loss": 0.9413,
+      "step": 1799
+    },
+    {
+      "epoch": 0.6349066378607173,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.463179009871719e-06,
+      "loss": 0.9469,
+      "step": 1800
+    },
+    {
+      "epoch": 0.63525936377064,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.452273703835685e-06,
+      "loss": 0.999,
+      "step": 1801
+    },
+    {
+      "epoch": 0.6356120896805626,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.441373222325306e-06,
+      "loss": 0.9171,
+      "step": 1802
+    },
+    {
+      "epoch": 0.6359648155904852,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.430477580164069e-06,
+      "loss": 1.0689,
+      "step": 1803
+    },
+    {
+      "epoch": 0.6363175415004079,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.41958679216887e-06,
+      "loss": 1.0189,
+      "step": 1804
+    },
+    {
+      "epoch": 0.6366702674103305,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.408700873150005e-06,
+      "loss": 0.8533,
+      "step": 1805
+    },
+    {
+      "epoch": 0.637022993320253,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.397819837911159e-06,
+      "loss": 0.8829,
+      "step": 1806
+    },
+    {
+      "epoch": 0.6373757192301757,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.386943701249362e-06,
+      "loss": 0.9481,
+      "step": 1807
+    },
+    {
+      "epoch": 0.6377284451400983,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.376072477954989e-06,
+      "loss": 0.9594,
+      "step": 1808
+    },
+    {
+      "epoch": 0.6380811710500209,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.365206182811735e-06,
+      "loss": 0.9677,
+      "step": 1809
+    },
+    {
+      "epoch": 0.6384338969599436,
+      "grad_norm": 0.86328125,
+      "learning_rate": 6.35434483059659e-06,
+      "loss": 0.9418,
+      "step": 1810
+    },
+    {
+      "epoch": 0.6387866228698662,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.3434884360798255e-06,
+      "loss": 0.9859,
+      "step": 1811
+    },
+    {
+      "epoch": 0.6391393487797888,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.332637014024964e-06,
+      "loss": 0.9344,
+      "step": 1812
+    },
+    {
+      "epoch": 0.6394920746897115,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.321790579188773e-06,
+      "loss": 0.9901,
+      "step": 1813
+    },
+    {
+      "epoch": 0.639844800599634,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.310949146321237e-06,
+      "loss": 1.0277,
+      "step": 1814
+    },
+    {
+      "epoch": 0.6401975265095566,
+      "grad_norm": 0.78125,
+      "learning_rate": 6.300112730165535e-06,
+      "loss": 1.0069,
+      "step": 1815
+    },
+    {
+      "epoch": 0.6405502524194793,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.289281345458027e-06,
+      "loss": 0.9152,
+      "step": 1816
+    },
+    {
+      "epoch": 0.6409029783294019,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.278455006928233e-06,
+      "loss": 1.0057,
+      "step": 1817
+    },
+    {
+      "epoch": 0.6412557042393245,
+      "grad_norm": 0.796875,
+      "learning_rate": 6.267633729298799e-06,
+      "loss": 0.8714,
+      "step": 1818
+    },
+    {
+      "epoch": 0.6416084301492472,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.256817527285504e-06,
+      "loss": 0.8974,
+      "step": 1819
+    },
+    {
+      "epoch": 0.6419611560591698,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.246006415597214e-06,
+      "loss": 0.9737,
+      "step": 1820
+    },
+    {
+      "epoch": 0.6423138819690923,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.23520040893588e-06,
+      "loss": 0.9981,
+      "step": 1821
+    },
+    {
+      "epoch": 0.642666607879015,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.224399521996507e-06,
+      "loss": 1.0068,
+      "step": 1822
+    },
+    {
+      "epoch": 0.6430193337889376,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.213603769467132e-06,
+      "loss": 0.998,
+      "step": 1823
+    },
+    {
+      "epoch": 0.6433720596988602,
+      "grad_norm": 1.328125,
+      "learning_rate": 6.202813166028823e-06,
+      "loss": 1.0679,
+      "step": 1824
+    },
+    {
+      "epoch": 0.6437247856087829,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.192027726355633e-06,
+      "loss": 0.9798,
+      "step": 1825
+    },
+    {
+      "epoch": 0.6440775115187055,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.1812474651146035e-06,
+      "loss": 0.9486,
+      "step": 1826
+    },
+    {
+      "epoch": 0.6444302374286281,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.1704723969657275e-06,
+      "loss": 1.0464,
+      "step": 1827
+    },
+    {
+      "epoch": 0.6447829633385508,
+      "grad_norm": 0.80859375,
+      "learning_rate": 6.159702536561931e-06,
+      "loss": 1.0928,
+      "step": 1828
+    },
+    {
+      "epoch": 0.6451356892484733,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.148937898549072e-06,
+      "loss": 0.8814,
+      "step": 1829
+    },
+    {
+      "epoch": 0.645488415158396,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.138178497565892e-06,
+      "loss": 0.9329,
+      "step": 1830
+    },
+    {
+      "epoch": 0.6458411410683186,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.127424348244027e-06,
+      "loss": 0.9132,
+      "step": 1831
+    },
+    {
+      "epoch": 0.6461938669782412,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.116675465207954e-06,
+      "loss": 0.988,
+      "step": 1832
+    },
+    {
+      "epoch": 0.6465465928881639,
+      "grad_norm": 0.75,
+      "learning_rate": 6.105931863074995e-06,
+      "loss": 1.0272,
+      "step": 1833
+    },
+    {
+      "epoch": 0.6468993187980865,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.0951935564552984e-06,
+      "loss": 0.9736,
+      "step": 1834
+    },
+    {
+      "epoch": 0.6472520447080091,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.084460559951802e-06,
+      "loss": 0.9231,
+      "step": 1835
+    },
+    {
+      "epoch": 0.6476047706179318,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.0737328881602195e-06,
+      "loss": 0.8497,
+      "step": 1836
+    },
+    {
+      "epoch": 0.6479574965278543,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.06301055566904e-06,
+      "loss": 0.8481,
+      "step": 1837
+    },
+    {
+      "epoch": 0.6483102224377769,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.052293577059468e-06,
+      "loss": 0.8343,
+      "step": 1838
+    },
+    {
+      "epoch": 0.6486629483476996,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.041581966905452e-06,
+      "loss": 0.9737,
+      "step": 1839
+    },
+    {
+      "epoch": 0.6490156742576222,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.030875739773623e-06,
+      "loss": 0.9162,
+      "step": 1840
+    },
+    {
+      "epoch": 0.6493684001675448,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.020174910223293e-06,
+      "loss": 0.8284,
+      "step": 1841
+    },
+    {
+      "epoch": 0.6497211260774675,
+      "grad_norm": 5.1875,
+      "learning_rate": 6.009479492806446e-06,
+      "loss": 1.0187,
+      "step": 1842
+    },
+    {
+      "epoch": 0.6500738519873901,
+      "grad_norm": 0.8359375,
+      "learning_rate": 5.9987895020676876e-06,
+      "loss": 0.9811,
+      "step": 1843
+    },
+    {
+      "epoch": 0.6504265778973126,
+      "grad_norm": 0.7578125,
+      "learning_rate": 5.988104952544266e-06,
+      "loss": 0.8204,
+      "step": 1844
+    },
+    {
+      "epoch": 0.6507793038072353,
+      "grad_norm": 0.84765625,
+      "learning_rate": 5.977425858766009e-06,
+      "loss": 1.1132,
+      "step": 1845
+    },
+    {
+      "epoch": 0.6511320297171579,
+      "grad_norm": 0.75,
+      "learning_rate": 5.966752235255333e-06,
+      "loss": 0.8922,
+      "step": 1846
+    },
+    {
+      "epoch": 0.6514847556270805,
+      "grad_norm": 0.75,
+      "learning_rate": 5.956084096527224e-06,
+      "loss": 0.9526,
+      "step": 1847
+    },
+    {
+      "epoch": 0.6518374815370032,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.9454214570891935e-06,
+      "loss": 0.8888,
+      "step": 1848
+    },
+    {
+      "epoch": 0.6521902074469258,
+      "grad_norm": 0.78125,
+      "learning_rate": 5.9347643314412896e-06,
+      "loss": 0.9028,
+      "step": 1849
+    },
+    {
+      "epoch": 0.6525429333568484,
+      "grad_norm": 0.828125,
+      "learning_rate": 5.924112734076053e-06,
+      "loss": 0.9504,
+      "step": 1850
+    },
+    {
+      "epoch": 0.6528956592667711,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.9134666794785035e-06,
+      "loss": 0.9359,
+      "step": 1851
+    },
+    {
+      "epoch": 0.6532483851766936,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.9028261821261376e-06,
+      "loss": 0.9732,
+      "step": 1852
+    },
+    {
+      "epoch": 0.6536011110866162,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.8921912564888775e-06,
+      "loss": 0.9044,
+      "step": 1853
+    },
+    {
+      "epoch": 0.6539538369965389,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.881561917029085e-06,
+      "loss": 0.9551,
+      "step": 1854
+    },
+    {
+      "epoch": 0.6543065629064615,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.870938178201512e-06,
+      "loss": 0.9294,
+      "step": 1855
+    },
+    {
+      "epoch": 0.6546592888163841,
+      "grad_norm": 0.76171875,
+      "learning_rate": 5.860320054453295e-06,
+      "loss": 0.8946,
+      "step": 1856
+    },
+    {
+      "epoch": 0.6550120147263068,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.849707560223949e-06,
+      "loss": 0.7946,
+      "step": 1857
+    },
+    {
+      "epoch": 0.6553647406362294,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.839100709945312e-06,
+      "loss": 0.8893,
+      "step": 1858
+    },
+    {
+      "epoch": 0.6557174665461519,
+      "grad_norm": 0.78125,
+      "learning_rate": 5.8284995180415685e-06,
+      "loss": 0.8684,
+      "step": 1859
+    },
+    {
+      "epoch": 0.6560701924560746,
+      "grad_norm": 0.7890625,
+      "learning_rate": 5.817903998929191e-06,
+      "loss": 0.9198,
+      "step": 1860
+    },
+    {
+      "epoch": 0.6564229183659972,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.807314167016945e-06,
+      "loss": 0.963,
+      "step": 1861
+    },
+    {
+      "epoch": 0.6567756442759198,
+      "grad_norm": 0.87890625,
+      "learning_rate": 5.796730036705862e-06,
+      "loss": 1.0666,
+      "step": 1862
+    },
+    {
+      "epoch": 0.6571283701858425,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.786151622389223e-06,
+      "loss": 0.8261,
+      "step": 1863
+    },
+    {
+      "epoch": 0.6574810960957651,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.775578938452522e-06,
+      "loss": 0.9886,
+      "step": 1864
+    },
+    {
+      "epoch": 0.6578338220056877,
+      "grad_norm": 0.7734375,
+      "learning_rate": 5.765011999273484e-06,
+      "loss": 0.9978,
+      "step": 1865
+    },
+    {
+      "epoch": 0.6581865479156104,
+      "grad_norm": 0.734375,
+      "learning_rate": 5.754450819221995e-06,
+      "loss": 0.9465,
+      "step": 1866
+    },
+    {
+      "epoch": 0.6585392738255329,
+      "grad_norm": 0.81640625,
+      "learning_rate": 5.743895412660135e-06,
+      "loss": 0.9959,
+      "step": 1867
+    },
+    {
+      "epoch": 0.6588919997354555,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.733345793942111e-06,
+      "loss": 1.0649,
+      "step": 1868
+    },
+    {
+      "epoch": 0.6592447256453782,
+      "grad_norm": 0.8125,
+      "learning_rate": 5.72280197741427e-06,
+      "loss": 0.8907,
+      "step": 1869
+    },
+    {
+      "epoch": 0.6595974515553008,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.712263977415069e-06,
+      "loss": 0.9537,
+      "step": 1870
+    },
+    {
+      "epoch": 0.6599501774652234,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.70173180827505e-06,
+      "loss": 0.8997,
+      "step": 1871
+    },
+    {
+      "epoch": 0.6603029033751461,
+      "grad_norm": 0.75,
+      "learning_rate": 5.691205484316832e-06,
+      "loss": 0.9367,
+      "step": 1872
+    },
+    {
+      "epoch": 0.6606556292850687,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.680685019855084e-06,
+      "loss": 0.9875,
+      "step": 1873
+    },
+    {
+      "epoch": 0.6610083551949912,
+      "grad_norm": 0.79296875,
+      "learning_rate": 5.6701704291964936e-06,
+      "loss": 0.9307,
+      "step": 1874
+    },
+    {
+      "epoch": 0.6613610811049139,
+      "grad_norm": 0.796875,
+      "learning_rate": 5.659661726639785e-06,
+      "loss": 0.9333,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6617138070148365,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.649158926475648e-06,
+      "loss": 0.9925,
+      "step": 1876
+    },
+    {
+      "epoch": 0.6620665329247591,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.638662042986777e-06,
+      "loss": 1.0038,
+      "step": 1877
+    },
+    {
+      "epoch": 0.6624192588346818,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.62817109044779e-06,
+      "loss": 0.911,
+      "step": 1878
+    },
+    {
+      "epoch": 0.6627719847446044,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.617686083125253e-06,
+      "loss": 1.0576,
+      "step": 1879
+    },
+    {
+      "epoch": 0.663124710654527,
+      "grad_norm": 0.75,
+      "learning_rate": 5.607207035277653e-06,
+      "loss": 0.9868,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6634774365644497,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.59673396115536e-06,
+      "loss": 0.8623,
+      "step": 1881
+    },
+    {
+      "epoch": 0.6638301624743722,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.586266875000632e-06,
+      "loss": 0.9335,
+      "step": 1882
+    },
+    {
+      "epoch": 0.6641828883842948,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.575805791047577e-06,
+      "loss": 0.9848,
+      "step": 1883
+    },
+    {
+      "epoch": 0.6645356142942175,
+      "grad_norm": 0.75,
+      "learning_rate": 5.56535072352214e-06,
+      "loss": 0.9484,
+      "step": 1884
+    },
+    {
+      "epoch": 0.6648883402041401,
+      "grad_norm": 0.828125,
+      "learning_rate": 5.554901686642095e-06,
+      "loss": 0.9177,
+      "step": 1885
+    },
+    {
+      "epoch": 0.6652410661140628,
+      "grad_norm": 0.84765625,
+      "learning_rate": 5.544458694616992e-06,
+      "loss": 0.9031,
+      "step": 1886
+    },
+    {
+      "epoch": 0.6655937920239854,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.5340217616481915e-06,
+      "loss": 0.9534,
+      "step": 1887
+    },
+    {
+      "epoch": 0.665946517933908,
+      "grad_norm": 0.7734375,
+      "learning_rate": 5.5235909019287884e-06,
+      "loss": 0.8923,
+      "step": 1888
+    },
+    {
+      "epoch": 0.6662992438438307,
+      "grad_norm": 0.76171875,
+      "learning_rate": 5.51316612964363e-06,
+      "loss": 0.9891,
+      "step": 1889
+    },
+    {
+      "epoch": 0.6666519697537532,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.502747458969284e-06,
+      "loss": 0.9111,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6670046956636758,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.492334904074022e-06,
+      "loss": 0.924,
+      "step": 1891
+    },
+    {
+      "epoch": 0.6673574215735985,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.4819284791177965e-06,
+      "loss": 1.0345,
+      "step": 1892
+    },
+    {
+      "epoch": 0.6677101474835211,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.471528198252224e-06,
+      "loss": 1.0086,
+      "step": 1893
+    },
+    {
+      "epoch": 0.6680628733934437,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.461134075620567e-06,
+      "loss": 0.8608,
+      "step": 1894
+    },
+    {
+      "epoch": 0.6684155993033664,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.450746125357712e-06,
+      "loss": 1.0111,
+      "step": 1895
+    },
+    {
+      "epoch": 0.668768325213289,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.440364361590158e-06,
+      "loss": 0.9262,
+      "step": 1896
+    },
+    {
+      "epoch": 0.6691210511232115,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.429988798435972e-06,
+      "loss": 0.9187,
+      "step": 1897
+    },
+    {
+      "epoch": 0.6694737770331342,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.4196194500048186e-06,
+      "loss": 0.8924,
+      "step": 1898
+    },
+    {
+      "epoch": 0.6698265029430568,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.409256330397881e-06,
+      "loss": 0.9031,
+      "step": 1899
+    },
+    {
+      "epoch": 0.6701792288529794,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.398899453707897e-06,
+      "loss": 0.9272,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6705319547629021,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.388548834019097e-06,
+      "loss": 0.9838,
+      "step": 1901
+    },
+    {
+      "epoch": 0.6708846806728247,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.37820448540721e-06,
+      "loss": 0.8946,
+      "step": 1902
+    },
+    {
+      "epoch": 0.6712374065827473,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.367866421939436e-06,
+      "loss": 0.8903,
+      "step": 1903
+    },
+    {
+      "epoch": 0.67159013249267,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.357534657674428e-06,
+      "loss": 0.9004,
+      "step": 1904
+    },
+    {
+      "epoch": 0.6719428584025925,
+      "grad_norm": 0.83203125,
+      "learning_rate": 5.347209206662274e-06,
+      "loss": 0.9201,
+      "step": 1905
+    },
+    {
+      "epoch": 0.6722955843125151,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.336890082944477e-06,
+      "loss": 1.0138,
+      "step": 1906
+    },
+    {
+      "epoch": 0.6726483102224378,
+      "grad_norm": 0.7578125,
+      "learning_rate": 5.326577300553923e-06,
+      "loss": 0.8314,
+      "step": 1907
+    },
+    {
+      "epoch": 0.6730010361323604,
+      "grad_norm": 0.75390625,
+      "learning_rate": 5.316270873514902e-06,
+      "loss": 0.9047,
+      "step": 1908
+    },
+    {
+      "epoch": 0.673353762042283,
+      "grad_norm": 0.80859375,
+      "learning_rate": 5.305970815843027e-06,
+      "loss": 0.8955,
+      "step": 1909
+    },
+    {
+      "epoch": 0.6737064879522057,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.295677141545284e-06,
+      "loss": 0.9684,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6740592138621283,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.285389864619951e-06,
+      "loss": 0.9303,
+      "step": 1911
+    },
+    {
+      "epoch": 0.6744119397720508,
+      "grad_norm": 0.75390625,
+      "learning_rate": 5.275108999056618e-06,
+      "loss": 0.9838,
+      "step": 1912
+    },
+    {
+      "epoch": 0.6747646656819735,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.264834558836156e-06,
+      "loss": 0.9836,
+      "step": 1913
+    },
+    {
+      "epoch": 0.6751173915918961,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.254566557930698e-06,
+      "loss": 1.0211,
+      "step": 1914
+    },
+    {
+      "epoch": 0.6754701175018187,
+      "grad_norm": 0.82421875,
+      "learning_rate": 5.244305010303618e-06,
+      "loss": 0.8236,
+      "step": 1915
+    },
+    {
+      "epoch": 0.6758228434117414,
+      "grad_norm": 0.80078125,
+      "learning_rate": 5.234049929909519e-06,
+      "loss": 1.0188,
+      "step": 1916
+    },
+    {
+      "epoch": 0.676175569321664,
+      "grad_norm": 0.76171875,
+      "learning_rate": 5.223801330694196e-06,
+      "loss": 1.0312,
+      "step": 1917
+    },
+    {
+      "epoch": 0.6765282952315866,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.213559226594656e-06,
+      "loss": 0.9332,
+      "step": 1918
+    },
+    {
+      "epoch": 0.6768810211415093,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.203323631539042e-06,
+      "loss": 0.982,
+      "step": 1919
+    },
+    {
+      "epoch": 0.6772337470514318,
+      "grad_norm": 0.80078125,
+      "learning_rate": 5.193094559446669e-06,
+      "loss": 1.0488,
+      "step": 1920
+    },
+    {
+      "epoch": 0.6775864729613544,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.1828720242279706e-06,
+      "loss": 1.0159,
+      "step": 1921
+    },
+    {
+      "epoch": 0.6779391988712771,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.172656039784495e-06,
+      "loss": 0.9306,
+      "step": 1922
+    },
+    {
+      "epoch": 0.6782919247811997,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.16244662000888e-06,
+      "loss": 0.8763,
+      "step": 1923
+    },
+    {
+      "epoch": 0.6786446506911223,
+      "grad_norm": 0.78125,
+      "learning_rate": 5.152243778784842e-06,
+      "loss": 0.9485,
+      "step": 1924
+    },
+    {
+      "epoch": 0.678997376601045,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.142047529987133e-06,
+      "loss": 0.9241,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6793501025109676,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.13185788748157e-06,
+      "loss": 0.9768,
+      "step": 1926
+    },
+    {
+      "epoch": 0.6797028284208901,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.121674865124953e-06,
+      "loss": 0.9762,
+      "step": 1927
+    },
+    {
+      "epoch": 0.6800555543308128,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.111498476765111e-06,
+      "loss": 0.924,
+      "step": 1928
+    },
+    {
+      "epoch": 0.6804082802407354,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.101328736240825e-06,
+      "loss": 0.9955,
+      "step": 1929
+    },
+    {
+      "epoch": 0.680761006150658,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.091165657381852e-06,
+      "loss": 0.8696,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6811137320605807,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.081009254008882e-06,
+      "loss": 0.917,
+      "step": 1931
+    },
+    {
+      "epoch": 0.6814664579705033,
+      "grad_norm": 0.75390625,
+      "learning_rate": 5.07085953993353e-06,
+      "loss": 0.7783,
+      "step": 1932
+    },
+    {
+      "epoch": 0.681819183880426,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.060716528958316e-06,
+      "loss": 0.917,
+      "step": 1933
+    },
+    {
+      "epoch": 0.6821719097903486,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.050580234876641e-06,
+      "loss": 0.9131,
+      "step": 1934
+    },
+    {
+      "epoch": 0.6825246357002711,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.040450671472774e-06,
+      "loss": 0.9363,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6828773616101937,
+      "grad_norm": 0.75390625,
+      "learning_rate": 5.0303278525218305e-06,
+      "loss": 0.9595,
+      "step": 1936
+    },
+    {
+      "epoch": 0.6832300875201164,
+      "grad_norm": 0.8515625,
+      "learning_rate": 5.020211791789753e-06,
+      "loss": 1.0016,
+      "step": 1937
+    },
+    {
+      "epoch": 0.683582813430039,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.0101025030332985e-06,
+      "loss": 0.8816,
+      "step": 1938
+    },
+    {
+      "epoch": 0.6839355393399617,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.9466,
+      "step": 1939
+    },
+    {
+      "epoch": 0.6842882652498843,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.989904296428186e-06,
+      "loss": 0.9813,
+      "step": 1940
+    },
+    {
+      "epoch": 0.6846409911598069,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.979815406046919e-06,
+      "loss": 0.9489,
+      "step": 1941
+    },
+    {
+      "epoch": 0.6849937170697296,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.969733342576005e-06,
+      "loss": 0.9784,
+      "step": 1942
+    },
+    {
+      "epoch": 0.6853464429796521,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.959658119725965e-06,
+      "loss": 0.9137,
+      "step": 1943
+    },
+    {
+      "epoch": 0.6856991688895747,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.9495897511980175e-06,
+      "loss": 0.9341,
+      "step": 1944
+    },
+    {
+      "epoch": 0.6860518947994974,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.939528250684057e-06,
+      "loss": 0.8575,
+      "step": 1945
+    },
+    {
+      "epoch": 0.68640462070942,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.9294736318666415e-06,
+      "loss": 1.0977,
+      "step": 1946
+    },
+    {
+      "epoch": 0.6867573466193426,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.919425908418974e-06,
+      "loss": 0.9055,
+      "step": 1947
+    },
+    {
+      "epoch": 0.6871100725292653,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.909385094004865e-06,
+      "loss": 1.0076,
+      "step": 1948
+    },
+    {
+      "epoch": 0.6874627984391879,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.899351202278756e-06,
+      "loss": 1.0106,
+      "step": 1949
+    },
+    {
+      "epoch": 0.6878155243491104,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.889324246885642e-06,
+      "loss": 0.9541,
+      "step": 1950
+    },
+    {
+      "epoch": 0.688168250259033,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.87930424146112e-06,
+      "loss": 1.046,
+      "step": 1951
+    },
+    {
+      "epoch": 0.6885209761689557,
+      "grad_norm": 0.75,
+      "learning_rate": 4.869291199631306e-06,
+      "loss": 0.9091,
+      "step": 1952
+    },
+    {
+      "epoch": 0.6888737020788783,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.8592851350128615e-06,
+      "loss": 0.9743,
+      "step": 1953
+    },
+    {
+      "epoch": 0.689226427988801,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.849286061212956e-06,
+      "loss": 0.9996,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6895791538987236,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.839293991829256e-06,
+      "loss": 0.9903,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6899318798086462,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.829308940449897e-06,
+      "loss": 0.9782,
+      "step": 1956
+    },
+    {
+      "epoch": 0.6902846057185689,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.8193309206534776e-06,
+      "loss": 1.0214,
+      "step": 1957
+    },
+    {
+      "epoch": 0.6906373316284914,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.80935994600902e-06,
+      "loss": 0.9456,
+      "step": 1958
+    },
+    {
+      "epoch": 0.690990057538414,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.799396030075989e-06,
+      "loss": 0.8896,
+      "step": 1959
+    },
+    {
+      "epoch": 0.6913427834483367,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.7894391864042235e-06,
+      "loss": 0.875,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6916955093582593,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.779489428533973e-06,
+      "loss": 1.031,
+      "step": 1961
+    },
+    {
+      "epoch": 0.6920482352681819,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.769546769995826e-06,
+      "loss": 0.9648,
+      "step": 1962
+    },
+    {
+      "epoch": 0.6924009611781046,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.759611224310729e-06,
+      "loss": 0.8847,
+      "step": 1963
+    },
+    {
+      "epoch": 0.6927536870880272,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.7496828049899565e-06,
+      "loss": 0.9676,
+      "step": 1964
+    },
+    {
+      "epoch": 0.6931064129979497,
+      "grad_norm": 0.75,
+      "learning_rate": 4.7397615255350885e-06,
+      "loss": 0.984,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6934591389078724,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.729847399437995e-06,
+      "loss": 0.8774,
+      "step": 1966
+    },
+    {
+      "epoch": 0.693811864817795,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.719940440180827e-06,
+      "loss": 0.9656,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6941645907277176,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.710040661235967e-06,
+      "loss": 0.948,
+      "step": 1968
+    },
+    {
+      "epoch": 0.6945173166376403,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.700148076066065e-06,
+      "loss": 0.8614,
+      "step": 1969
+    },
+    {
+      "epoch": 0.6948700425475629,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.690262698123956e-06,
+      "loss": 0.8608,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6952227684574855,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.680384540852705e-06,
+      "loss": 0.8716,
+      "step": 1971
+    },
+    {
+      "epoch": 0.6955754943674082,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.670513617685529e-06,
+      "loss": 0.9845,
+      "step": 1972
+    },
+    {
+      "epoch": 0.6959282202773307,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.660649942045826e-06,
+      "loss": 0.9356,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6962809461872533,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.650793527347133e-06,
+      "loss": 0.9975,
+      "step": 1974
+    },
+    {
+      "epoch": 0.696633672097176,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.640944386993114e-06,
+      "loss": 0.9242,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6969863980070986,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.631102534377527e-06,
+      "loss": 0.9503,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6973391239170212,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.62126798288425e-06,
+      "loss": 0.9461,
+      "step": 1977
+    },
+    {
+      "epoch": 0.6976918498269439,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.611440745887197e-06,
+      "loss": 0.9074,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6980445757368665,
+      "grad_norm": 0.75,
+      "learning_rate": 4.601620836750367e-06,
+      "loss": 0.8866,
+      "step": 1979
+    },
+    {
+      "epoch": 0.698397301646789,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.591808268827768e-06,
+      "loss": 0.9518,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6987500275567117,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.582003055463437e-06,
+      "loss": 1.0997,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6991027534666343,
+      "grad_norm": 0.75,
+      "learning_rate": 4.572205209991412e-06,
+      "loss": 0.944,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6994554793765569,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.562414745735705e-06,
+      "loss": 0.9336,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6998082052864796,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.552631676010293e-06,
+      "loss": 0.8488,
+      "step": 1984
+    },
+    {
+      "epoch": 0.7001609311964022,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.542856014119098e-06,
+      "loss": 0.9581,
+      "step": 1985
+    },
+    {
+      "epoch": 0.7005136571063248,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.533087773355967e-06,
+      "loss": 1.0036,
+      "step": 1986
+    },
+    {
+      "epoch": 0.7008663830162475,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.523326967004653e-06,
+      "loss": 1.0278,
+      "step": 1987
+    },
+    {
+      "epoch": 0.70121910892617,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.513573608338806e-06,
+      "loss": 0.9386,
+      "step": 1988
+    },
+    {
+      "epoch": 0.7015718348360926,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.503827710621944e-06,
+      "loss": 0.8122,
+      "step": 1989
+    },
+    {
+      "epoch": 0.7019245607460153,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.4940892871074305e-06,
+      "loss": 0.9522,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7022772866559379,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.484358351038478e-06,
+      "loss": 0.9873,
+      "step": 1991
+    },
+    {
+      "epoch": 0.7026300125658606,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.474634915648111e-06,
+      "loss": 1.0533,
+      "step": 1992
+    },
+    {
+      "epoch": 0.7029827384757832,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.464918994159154e-06,
+      "loss": 0.9092,
+      "step": 1993
+    },
+    {
+      "epoch": 0.7033354643857058,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.455210599784213e-06,
+      "loss": 0.8943,
+      "step": 1994
+    },
+    {
+      "epoch": 0.7036881902956285,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.44550974572566e-06,
+      "loss": 0.9165,
+      "step": 1995
+    },
+    {
+      "epoch": 0.704040916205551,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.435816445175611e-06,
+      "loss": 0.9078,
+      "step": 1996
+    },
+    {
+      "epoch": 0.7043936421154736,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.426130711315913e-06,
+      "loss": 0.9557,
+      "step": 1997
+    },
+    {
+      "epoch": 0.7047463680253963,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.416452557318116e-06,
+      "loss": 1.0424,
+      "step": 1998
+    },
+    {
+      "epoch": 0.7050990939353189,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.406781996343472e-06,
+      "loss": 1.037,
+      "step": 1999
+    },
+    {
+      "epoch": 0.7054518198452415,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.397119041542903e-06,
+      "loss": 0.9234,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7058045457551642,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.387463706056977e-06,
+      "loss": 0.9431,
+      "step": 2001
+    },
+    {
+      "epoch": 0.7061572716650868,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.377816003015925e-06,
+      "loss": 0.9396,
+      "step": 2002
+    },
+    {
+      "epoch": 0.7065099975750093,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.368175945539572e-06,
+      "loss": 0.9508,
+      "step": 2003
+    },
+    {
+      "epoch": 0.706862723484932,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.3585435467373626e-06,
+      "loss": 1.0059,
+      "step": 2004
+    },
+    {
+      "epoch": 0.7072154493948546,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.348918819708322e-06,
+      "loss": 0.9335,
+      "step": 2005
+    },
+    {
+      "epoch": 0.7075681753047772,
+      "grad_norm": 0.75,
+      "learning_rate": 4.339301777541041e-06,
+      "loss": 0.9852,
+      "step": 2006
+    },
+    {
+      "epoch": 0.7079209012146999,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.3296924333136624e-06,
+      "loss": 1.0415,
+      "step": 2007
+    },
+    {
+      "epoch": 0.7082736271246225,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.320090800093862e-06,
+      "loss": 0.8873,
+      "step": 2008
+    },
+    {
+      "epoch": 0.7086263530345451,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.3104968909388174e-06,
+      "loss": 0.8064,
+      "step": 2009
+    },
+    {
+      "epoch": 0.7089790789444678,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.300910718895226e-06,
+      "loss": 0.9755,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7093318048543903,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.291332296999233e-06,
+      "loss": 0.8816,
+      "step": 2011
+    },
+    {
+      "epoch": 0.7096845307643129,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.281761638276476e-06,
+      "loss": 0.987,
+      "step": 2012
+    },
+    {
+      "epoch": 0.7100372566742356,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.272198755742011e-06,
+      "loss": 1.0442,
+      "step": 2013
+    },
+    {
+      "epoch": 0.7103899825841582,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.262643662400329e-06,
+      "loss": 0.9037,
+      "step": 2014
+    },
+    {
+      "epoch": 0.7107427084940808,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.253096371245329e-06,
+      "loss": 0.9093,
+      "step": 2015
+    },
+    {
+      "epoch": 0.7110954344040035,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.243556895260299e-06,
+      "loss": 0.983,
+      "step": 2016
+    },
+    {
+      "epoch": 0.7114481603139261,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.2340252474179e-06,
+      "loss": 0.9163,
+      "step": 2017
+    },
+    {
+      "epoch": 0.7118008862238486,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.224501440680148e-06,
+      "loss": 0.9921,
+      "step": 2018
+    },
+    {
+      "epoch": 0.7121536121337713,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.214985487998384e-06,
+      "loss": 1.0831,
+      "step": 2019
+    },
+    {
+      "epoch": 0.7125063380436939,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.2054774023132935e-06,
+      "loss": 0.9177,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7128590639536165,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.195977196554835e-06,
+      "loss": 0.9535,
+      "step": 2021
+    },
+    {
+      "epoch": 0.7132117898635392,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.186484883642279e-06,
+      "loss": 0.9384,
+      "step": 2022
+    },
+    {
+      "epoch": 0.7135645157734618,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.177000476484139e-06,
+      "loss": 1.0122,
+      "step": 2023
+    },
+    {
+      "epoch": 0.7139172416833844,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.167523987978188e-06,
+      "loss": 0.9283,
+      "step": 2024
+    },
+    {
+      "epoch": 0.7142699675933071,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.158055431011433e-06,
+      "loss": 0.8982,
+      "step": 2025
+    },
+    {
+      "epoch": 0.7146226935032296,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.14859481846009e-06,
+      "loss": 0.869,
+      "step": 2026
+    },
+    {
+      "epoch": 0.7149754194131522,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.139142163189573e-06,
+      "loss": 0.9747,
+      "step": 2027
+    },
+    {
+      "epoch": 0.7153281453230749,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.1296974780544795e-06,
+      "loss": 0.9785,
+      "step": 2028
+    },
+    {
+      "epoch": 0.7156808712329975,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.120260775898554e-06,
+      "loss": 0.8936,
+      "step": 2029
+    },
+    {
+      "epoch": 0.7160335971429201,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.1108320695547065e-06,
+      "loss": 0.8349,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7163863230528428,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.1014113718449536e-06,
+      "loss": 0.9677,
+      "step": 2031
+    },
+    {
+      "epoch": 0.7167390489627654,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.091998695580433e-06,
+      "loss": 0.9037,
+      "step": 2032
+    },
+    {
+      "epoch": 0.7170917748726879,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.082594053561369e-06,
+      "loss": 0.8993,
+      "step": 2033
+    },
+    {
+      "epoch": 0.7174445007826106,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.073197458577064e-06,
+      "loss": 0.96,
+      "step": 2034
+    },
+    {
+      "epoch": 0.7177972266925332,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.063808923405874e-06,
+      "loss": 1.0648,
+      "step": 2035
+    },
+    {
+      "epoch": 0.7181499526024558,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.054428460815196e-06,
+      "loss": 0.7732,
+      "step": 2036
+    },
+    {
+      "epoch": 0.7185026785123785,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.0450560835614495e-06,
+      "loss": 0.878,
+      "step": 2037
+    },
+    {
+      "epoch": 0.7188554044223011,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.035691804390056e-06,
+      "loss": 0.891,
+      "step": 2038
+    },
+    {
+      "epoch": 0.7192081303322237,
+      "grad_norm": 0.75,
+      "learning_rate": 4.026335636035429e-06,
+      "loss": 0.9218,
+      "step": 2039
+    },
+    {
+      "epoch": 0.7195608562421464,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.016987591220948e-06,
+      "loss": 0.9351,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7199135821520689,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.0076476826589505e-06,
+      "loss": 0.958,
+      "step": 2041
+    },
+    {
+      "epoch": 0.7202663080619915,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.9983159230507014e-06,
+      "loss": 0.9566,
+      "step": 2042
+    },
+    {
+      "epoch": 0.7206190339719142,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.988992325086391e-06,
+      "loss": 0.8945,
+      "step": 2043
+    },
+    {
+      "epoch": 0.7209717598818368,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.979676901445107e-06,
+      "loss": 0.9552,
+      "step": 2044
+    },
+    {
+      "epoch": 0.7213244857917595,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.970369664794823e-06,
+      "loss": 1.0024,
+      "step": 2045
+    },
+    {
+      "epoch": 0.7216772117016821,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.961070627792379e-06,
+      "loss": 0.9954,
+      "step": 2046
+    },
+    {
+      "epoch": 0.7220299376116047,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.9517798030834655e-06,
+      "loss": 0.9828,
+      "step": 2047
+    },
+    {
+      "epoch": 0.7223826635215274,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.9424972033026e-06,
+      "loss": 1.0237,
+      "step": 2048
+    },
+    {
+      "epoch": 0.7227353894314499,
+      "grad_norm": 0.75,
+      "learning_rate": 3.93322284107312e-06,
+      "loss": 0.9111,
+      "step": 2049
+    },
+    {
+      "epoch": 0.7230881153413725,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.923956729007159e-06,
+      "loss": 0.8459,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7234408412512952,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.914698879705635e-06,
+      "loss": 0.8622,
+      "step": 2051
+    },
+    {
+      "epoch": 0.7237935671612178,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.905449305758215e-06,
+      "loss": 0.9452,
+      "step": 2052
+    },
+    {
+      "epoch": 0.7241462930711404,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.89620801974334e-06,
+      "loss": 0.9755,
+      "step": 2053
+    },
+    {
+      "epoch": 0.7244990189810631,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.88697503422815e-06,
+      "loss": 0.9793,
+      "step": 2054
+    },
+    {
+      "epoch": 0.7248517448909857,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.877750361768514e-06,
+      "loss": 1.0269,
+      "step": 2055
+    },
+    {
+      "epoch": 0.7252044708009082,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.868534014908997e-06,
+      "loss": 0.9183,
+      "step": 2056
+    },
+    {
+      "epoch": 0.7255571967108309,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.859326006182833e-06,
+      "loss": 0.9183,
+      "step": 2057
+    },
+    {
+      "epoch": 0.7259099226207535,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.850126348111925e-06,
+      "loss": 0.9373,
+      "step": 2058
+    },
+    {
+      "epoch": 0.7262626485306761,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.8409350532068225e-06,
+      "loss": 0.8735,
+      "step": 2059
+    },
+    {
+      "epoch": 0.7266153744405988,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.831752133966682e-06,
+      "loss": 0.9677,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7269681003505214,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.8225776028793e-06,
+      "loss": 0.8911,
+      "step": 2061
+    },
+    {
+      "epoch": 0.727320826260444,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.813411472421039e-06,
+      "loss": 0.8841,
+      "step": 2062
+    },
+    {
+      "epoch": 0.7276735521703667,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.8042537550568603e-06,
+      "loss": 0.8349,
+      "step": 2063
+    },
+    {
+      "epoch": 0.7280262780802892,
+      "grad_norm": 0.75,
+      "learning_rate": 3.7951044632402644e-06,
+      "loss": 0.9919,
+      "step": 2064
+    },
+    {
+      "epoch": 0.7283790039902118,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.7859636094133045e-06,
+      "loss": 0.9523,
+      "step": 2065
+    },
+    {
+      "epoch": 0.7287317299001345,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.7768312060065593e-06,
+      "loss": 0.8798,
+      "step": 2066
+    },
+    {
+      "epoch": 0.7290844558100571,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.7677072654391124e-06,
+      "loss": 0.9269,
+      "step": 2067
+    },
+    {
+      "epoch": 0.7294371817199797,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.7585918001185416e-06,
+      "loss": 0.8599,
+      "step": 2068
+    },
+    {
+      "epoch": 0.7297899076299024,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.7494848224408998e-06,
+      "loss": 1.0203,
+      "step": 2069
+    },
+    {
+      "epoch": 0.730142633539825,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.7403863447906873e-06,
+      "loss": 0.9969,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7304953594497475,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.731296379540865e-06,
+      "loss": 1.0394,
+      "step": 2071
+    },
+    {
+      "epoch": 0.7308480853596702,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.722214939052796e-06,
+      "loss": 0.9245,
+      "step": 2072
+    },
+    {
+      "epoch": 0.7312008112695928,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.7131420356762726e-06,
+      "loss": 0.9485,
+      "step": 2073
+    },
+    {
+      "epoch": 0.7315535371795154,
+      "grad_norm": 0.75,
+      "learning_rate": 3.704077681749456e-06,
+      "loss": 0.9502,
+      "step": 2074
+    },
+    {
+      "epoch": 0.7319062630894381,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.6950218895988966e-06,
+      "loss": 0.9209,
+      "step": 2075
+    },
+    {
+      "epoch": 0.7322589889993607,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.6859746715394953e-06,
+      "loss": 0.9896,
+      "step": 2076
+    },
+    {
+      "epoch": 0.7326117149092833,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.6769360398744924e-06,
+      "loss": 0.9089,
+      "step": 2077
+    },
+    {
+      "epoch": 0.732964440819206,
+      "grad_norm": 0.75,
+      "learning_rate": 3.6679060068954566e-06,
+      "loss": 0.9698,
+      "step": 2078
+    },
+    {
+      "epoch": 0.7333171667291285,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.6588845848822617e-06,
+      "loss": 0.9589,
+      "step": 2079
+    },
+    {
+      "epoch": 0.7336698926390511,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.649871786103061e-06,
+      "loss": 0.9419,
+      "step": 2080
+    },
+    {
+      "epoch": 0.7340226185489738,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.6408676228143013e-06,
+      "loss": 1.0287,
+      "step": 2081
+    },
+    {
+      "epoch": 0.7343753444588964,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.6318721072606677e-06,
+      "loss": 1.0023,
+      "step": 2082
+    },
+    {
+      "epoch": 0.734728070368819,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.6228852516750946e-06,
+      "loss": 0.8804,
+      "step": 2083
+    },
+    {
+      "epoch": 0.7350807962787417,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.613907068278737e-06,
+      "loss": 0.9584,
+      "step": 2084
+    },
+    {
+      "epoch": 0.7354335221886643,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.6049375692809572e-06,
+      "loss": 0.8883,
+      "step": 2085
+    },
+    {
+      "epoch": 0.7357862480985868,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.59597676687931e-06,
+      "loss": 0.9442,
+      "step": 2086
+    },
+    {
+      "epoch": 0.7361389740085095,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.587024673259519e-06,
+      "loss": 0.8876,
+      "step": 2087
+    },
+    {
+      "epoch": 0.7364916999184321,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.5780813005954697e-06,
+      "loss": 0.842,
+      "step": 2088
+    },
+    {
+      "epoch": 0.7368444258283547,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.569146661049183e-06,
+      "loss": 0.912,
+      "step": 2089
+    },
+    {
+      "epoch": 0.7371971517382774,
+      "grad_norm": 4.1875,
+      "learning_rate": 3.560220766770809e-06,
+      "loss": 0.9417,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7375498776482,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.5513036298986004e-06,
+      "loss": 0.9619,
+      "step": 2091
+    },
+    {
+      "epoch": 0.7379026035581226,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.5423952625589085e-06,
+      "loss": 0.8802,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7382553294680453,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.533495676866141e-06,
+      "loss": 0.9788,
+      "step": 2093
+    },
+    {
+      "epoch": 0.7386080553779678,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.5246048849227908e-06,
+      "loss": 0.9981,
+      "step": 2094
+    },
+    {
+      "epoch": 0.7389607812878904,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.515722898819367e-06,
+      "loss": 0.9843,
+      "step": 2095
+    },
+    {
+      "epoch": 0.7393135071978131,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.5068497306344173e-06,
+      "loss": 0.9037,
+      "step": 2096
+    },
+    {
+      "epoch": 0.7396662331077357,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.497985392434494e-06,
+      "loss": 0.9239,
+      "step": 2097
+    },
+    {
+      "epoch": 0.7400189590176584,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.489129896274145e-06,
+      "loss": 0.8407,
+      "step": 2098
+    },
+    {
+      "epoch": 0.740371684927581,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.4802832541958886e-06,
+      "loss": 1.0143,
+      "step": 2099
+    },
+    {
+      "epoch": 0.7407244108375036,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.4714454782302053e-06,
+      "loss": 0.9938,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7410771367474263,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.4626165803955204e-06,
+      "loss": 0.8901,
+      "step": 2101
+    },
+    {
+      "epoch": 0.7414298626573488,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.453796572698186e-06,
+      "loss": 0.9707,
+      "step": 2102
+    },
+    {
+      "epoch": 0.7417825885672714,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.444985467132452e-06,
+      "loss": 0.9564,
+      "step": 2103
+    },
+    {
+      "epoch": 0.742135314477194,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.436183275680488e-06,
+      "loss": 0.9624,
+      "step": 2104
+    },
+    {
+      "epoch": 0.7424880403871167,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.4273900103123125e-06,
+      "loss": 0.9463,
+      "step": 2105
+    },
+    {
+      "epoch": 0.7428407662970393,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.4186056829858314e-06,
+      "loss": 0.9711,
+      "step": 2106
+    },
+    {
+      "epoch": 0.743193492206962,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.4098303056467762e-06,
+      "loss": 0.9342,
+      "step": 2107
+    },
+    {
+      "epoch": 0.7435462181168846,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.4010638902287163e-06,
+      "loss": 0.9353,
+      "step": 2108
+    },
+    {
+      "epoch": 0.7438989440268071,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.392306448653032e-06,
+      "loss": 0.8748,
+      "step": 2109
+    },
+    {
+      "epoch": 0.7442516699367298,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.383557992828906e-06,
+      "loss": 0.8976,
+      "step": 2110
+    },
+    {
+      "epoch": 0.7446043958466524,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.3748185346532847e-06,
+      "loss": 0.9053,
+      "step": 2111
+    },
+    {
+      "epoch": 0.744957121756575,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.3660880860109035e-06,
+      "loss": 1.0585,
+      "step": 2112
+    },
+    {
+      "epoch": 0.7453098476664977,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.3573666587742192e-06,
+      "loss": 1.0747,
+      "step": 2113
+    },
+    {
+      "epoch": 0.7456625735764203,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.348654264803446e-06,
+      "loss": 0.9508,
+      "step": 2114
+    },
+    {
+      "epoch": 0.7460152994863429,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.3399509159464925e-06,
+      "loss": 0.9391,
+      "step": 2115
+    },
+    {
+      "epoch": 0.7463680253962656,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.3312566240389787e-06,
+      "loss": 0.9832,
+      "step": 2116
+    },
+    {
+      "epoch": 0.7467207513061881,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.3225714009042045e-06,
+      "loss": 0.9523,
+      "step": 2117
+    },
+    {
+      "epoch": 0.7470734772161107,
+      "grad_norm": 0.75,
+      "learning_rate": 3.313895258353138e-06,
+      "loss": 0.9614,
+      "step": 2118
+    },
+    {
+      "epoch": 0.7474262031260334,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.3052282081844e-06,
+      "loss": 0.9533,
+      "step": 2119
+    },
+    {
+      "epoch": 0.747778929035956,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.2965702621842464e-06,
+      "loss": 0.9447,
+      "step": 2120
+    },
+    {
+      "epoch": 0.7481316549458786,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.287921432126544e-06,
+      "loss": 0.9182,
+      "step": 2121
+    },
+    {
+      "epoch": 0.7484843808558013,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.279281729772781e-06,
+      "loss": 0.954,
+      "step": 2122
+    },
+    {
+      "epoch": 0.7488371067657239,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.27065116687201e-06,
+      "loss": 0.9285,
+      "step": 2123
+    },
+    {
+      "epoch": 0.7491898326756464,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.2620297551608803e-06,
+      "loss": 0.9419,
+      "step": 2124
+    },
+    {
+      "epoch": 0.7495425585855691,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.253417506363572e-06,
+      "loss": 0.9805,
+      "step": 2125
+    },
+    {
+      "epoch": 0.7498952844954917,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.2448144321918217e-06,
+      "loss": 0.8983,
+      "step": 2126
+    },
+    {
+      "epoch": 0.7502480104054143,
+      "grad_norm": 0.75,
+      "learning_rate": 3.2362205443448813e-06,
+      "loss": 0.9586,
+      "step": 2127
+    },
+    {
+      "epoch": 0.750600736315337,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.2276358545095145e-06,
+      "loss": 0.9549,
+      "step": 2128
+    },
+    {
+      "epoch": 0.7509534622252596,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.2190603743599723e-06,
+      "loss": 1.0034,
+      "step": 2129
+    },
+    {
+      "epoch": 0.7513061881351822,
+      "grad_norm": 1.828125,
+      "learning_rate": 3.2104941155579905e-06,
+      "loss": 0.9746,
+      "step": 2130
+    },
+    {
+      "epoch": 0.7516589140451049,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.2019370897527467e-06,
+      "loss": 0.9625,
+      "step": 2131
+    },
+    {
+      "epoch": 0.7520116399550274,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.1933893085808875e-06,
+      "loss": 0.9381,
+      "step": 2132
+    },
+    {
+      "epoch": 0.75236436586495,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.1848507836664634e-06,
+      "loss": 0.9215,
+      "step": 2133
+    },
+    {
+      "epoch": 0.7527170917748727,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.176321526620959e-06,
+      "loss": 0.9265,
+      "step": 2134
+    },
+    {
+      "epoch": 0.7530698176847953,
+      "grad_norm": 1.265625,
+      "learning_rate": 3.1678015490432368e-06,
+      "loss": 1.0109,
+      "step": 2135
+    },
+    {
+      "epoch": 0.7534225435947179,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.159290862519552e-06,
+      "loss": 1.0761,
+      "step": 2136
+    },
+    {
+      "epoch": 0.7537752695046406,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.1507894786235214e-06,
+      "loss": 0.8636,
+      "step": 2137
+    },
+    {
+      "epoch": 0.7541279954145632,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.142297408916113e-06,
+      "loss": 0.9164,
+      "step": 2138
+    },
+    {
+      "epoch": 0.7544807213244857,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.1338146649456236e-06,
+      "loss": 0.8944,
+      "step": 2139
+    },
+    {
+      "epoch": 0.7548334472344084,
+      "grad_norm": 0.75,
+      "learning_rate": 3.125341258247675e-06,
+      "loss": 1.0459,
+      "step": 2140
+    },
+    {
+      "epoch": 0.755186173144331,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.1168772003451855e-06,
+      "loss": 0.9658,
+      "step": 2141
+    },
+    {
+      "epoch": 0.7555388990542536,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.1084225027483628e-06,
+      "loss": 0.8652,
+      "step": 2142
+    },
+    {
+      "epoch": 0.7558916249641763,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.099977176954687e-06,
+      "loss": 0.8926,
+      "step": 2143
+    },
+    {
+      "epoch": 0.7562443508740989,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.091541234448885e-06,
+      "loss": 0.8945,
+      "step": 2144
+    },
+    {
+      "epoch": 0.7565970767840215,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.0831146867029407e-06,
+      "loss": 0.8927,
+      "step": 2145
+    },
+    {
+      "epoch": 0.7569498026939442,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.074697545176041e-06,
+      "loss": 0.8888,
+      "step": 2146
+    },
+    {
+      "epoch": 0.7573025286038667,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.0662898213145965e-06,
+      "loss": 0.9453,
+      "step": 2147
+    },
+    {
+      "epoch": 0.7576552545137893,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.057891526552207e-06,
+      "loss": 0.9295,
+      "step": 2148
+    },
+    {
+      "epoch": 0.758007980423712,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.0495026723096466e-06,
+      "loss": 0.9004,
+      "step": 2149
+    },
+    {
+      "epoch": 0.7583607063336346,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.0411232699948546e-06,
+      "loss": 0.815,
+      "step": 2150
+    },
+    {
+      "epoch": 0.7587134322435573,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.0327533310029157e-06,
+      "loss": 0.8649,
+      "step": 2151
+    },
+    {
+      "epoch": 0.7590661581534799,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.0243928667160448e-06,
+      "loss": 0.9221,
+      "step": 2152
+    },
+    {
+      "epoch": 0.7594188840634025,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.016041888503578e-06,
+      "loss": 1.012,
+      "step": 2153
+    },
+    {
+      "epoch": 0.7597716099733252,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.007700407721934e-06,
+      "loss": 0.9353,
+      "step": 2154
+    },
+    {
+      "epoch": 0.7601243358832477,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.9993684357146434e-06,
+      "loss": 0.9293,
+      "step": 2155
+    },
+    {
+      "epoch": 0.7604770617931703,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.991045983812276e-06,
+      "loss": 1.0261,
+      "step": 2156
+    },
+    {
+      "epoch": 0.760829787703093,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.9827330633324846e-06,
+      "loss": 0.9439,
+      "step": 2157
+    },
+    {
+      "epoch": 0.7611825136130156,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.974429685579935e-06,
+      "loss": 0.9072,
+      "step": 2158
+    },
+    {
+      "epoch": 0.7615352395229382,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.9661358618463297e-06,
+      "loss": 0.9,
+      "step": 2159
+    },
+    {
+      "epoch": 0.7618879654328609,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.9578516034103764e-06,
+      "loss": 0.9063,
+      "step": 2160
+    },
+    {
+      "epoch": 0.7622406913427835,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.9495769215377733e-06,
+      "loss": 0.8937,
+      "step": 2161
+    },
+    {
+      "epoch": 0.762593417252706,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.941311827481198e-06,
+      "loss": 0.9233,
+      "step": 2162
+    },
+    {
+      "epoch": 0.7629461431626287,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.933056332480291e-06,
+      "loss": 0.9884,
+      "step": 2163
+    },
+    {
+      "epoch": 0.7632988690725513,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.9248104477616256e-06,
+      "loss": 0.9418,
+      "step": 2164
+    },
+    {
+      "epoch": 0.7636515949824739,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.916574184538731e-06,
+      "loss": 0.8817,
+      "step": 2165
+    },
+    {
+      "epoch": 0.7640043208923966,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.9083475540120275e-06,
+      "loss": 0.9814,
+      "step": 2166
+    },
+    {
+      "epoch": 0.7643570468023192,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.9001305673688507e-06,
+      "loss": 0.9496,
+      "step": 2167
+    },
+    {
+      "epoch": 0.7647097727122418,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.891923235783416e-06,
+      "loss": 0.9311,
+      "step": 2168
+    },
+    {
+      "epoch": 0.7650624986221645,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.883725570416812e-06,
+      "loss": 0.8799,
+      "step": 2169
+    },
+    {
+      "epoch": 0.765415224532087,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.8755375824169808e-06,
+      "loss": 0.8621,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7657679504420096,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.8673592829187057e-06,
+      "loss": 0.9006,
+      "step": 2171
+    },
+    {
+      "epoch": 0.7661206763519323,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.8591906830435856e-06,
+      "loss": 0.9455,
+      "step": 2172
+    },
+    {
+      "epoch": 0.7664734022618549,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.8510317939000474e-06,
+      "loss": 0.8358,
+      "step": 2173
+    },
+    {
+      "epoch": 0.7668261281717775,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.84288262658329e-06,
+      "loss": 1.089,
+      "step": 2174
+    },
+    {
+      "epoch": 0.7671788540817002,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.8347431921753156e-06,
+      "loss": 0.9974,
+      "step": 2175
+    },
+    {
+      "epoch": 0.7675315799916228,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.8266135017448693e-06,
+      "loss": 0.8132,
+      "step": 2176
+    },
+    {
+      "epoch": 0.7678843059015453,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.8184935663474567e-06,
+      "loss": 1.0032,
+      "step": 2177
+    },
+    {
+      "epoch": 0.768237031811468,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.810383397025317e-06,
+      "loss": 0.9547,
+      "step": 2178
+    },
+    {
+      "epoch": 0.7685897577213906,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.8022830048074045e-06,
+      "loss": 0.869,
+      "step": 2179
+    },
+    {
+      "epoch": 0.7689424836313132,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.7941924007093823e-06,
+      "loss": 0.9755,
+      "step": 2180
+    },
+    {
+      "epoch": 0.7692952095412359,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.7861115957336005e-06,
+      "loss": 0.9077,
+      "step": 2181
+    },
+    {
+      "epoch": 0.7696479354511585,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.778040600869082e-06,
+      "loss": 0.8934,
+      "step": 2182
+    },
+    {
+      "epoch": 0.7700006613610811,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.769979427091515e-06,
+      "loss": 0.9522,
+      "step": 2183
+    },
+    {
+      "epoch": 0.7703533872710038,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.7619280853632178e-06,
+      "loss": 0.8061,
+      "step": 2184
+    },
+    {
+      "epoch": 0.7707061131809263,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.753886586633161e-06,
+      "loss": 0.8908,
+      "step": 2185
+    },
+    {
+      "epoch": 0.7710588390908489,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.7458549418369064e-06,
+      "loss": 0.9428,
+      "step": 2186
+    },
+    {
+      "epoch": 0.7714115650007716,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.737833161896629e-06,
+      "loss": 0.9457,
+      "step": 2187
+    },
+    {
+      "epoch": 0.7717642909106942,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.729821257721087e-06,
+      "loss": 0.9315,
+      "step": 2188
+    },
+    {
+      "epoch": 0.7721170168206168,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.7218192402056052e-06,
+      "loss": 0.9929,
+      "step": 2189
+    },
+    {
+      "epoch": 0.7724697427305395,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.7138271202320674e-06,
+      "loss": 0.957,
+      "step": 2190
+    },
+    {
+      "epoch": 0.7728224686404621,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.7058449086688954e-06,
+      "loss": 0.8968,
+      "step": 2191
+    },
+    {
+      "epoch": 0.7731751945503846,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.6978726163710354e-06,
+      "loss": 1.0319,
+      "step": 2192
+    },
+    {
+      "epoch": 0.7735279204603073,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.689910254179949e-06,
+      "loss": 0.9418,
+      "step": 2193
+    },
+    {
+      "epoch": 0.7738806463702299,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.681957832923594e-06,
+      "loss": 0.9447,
+      "step": 2194
+    },
+    {
+      "epoch": 0.7742333722801525,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.6740153634163955e-06,
+      "loss": 0.8921,
+      "step": 2195
+    },
+    {
+      "epoch": 0.7745860981900752,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.66608285645927e-06,
+      "loss": 0.9304,
+      "step": 2196
+    },
+    {
+      "epoch": 0.7749388240999978,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.6581603228395614e-06,
+      "loss": 0.875,
+      "step": 2197
+    },
+    {
+      "epoch": 0.7752915500099204,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.6502477733310726e-06,
+      "loss": 0.8009,
+      "step": 2198
+    },
+    {
+      "epoch": 0.7756442759198431,
+      "grad_norm": 0.75,
+      "learning_rate": 2.64234521869401e-06,
+      "loss": 1.0709,
+      "step": 2199
+    },
+    {
+      "epoch": 0.7759970018297656,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.6344526696750006e-06,
+      "loss": 0.9688,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7763497277396882,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.626570137007061e-06,
+      "loss": 0.9174,
+      "step": 2201
+    },
+    {
+      "epoch": 0.7767024536496109,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.6186976314095857e-06,
+      "loss": 0.9784,
+      "step": 2202
+    },
+    {
+      "epoch": 0.7770551795595335,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.6108351635883353e-06,
+      "loss": 0.9985,
+      "step": 2203
+    },
+    {
+      "epoch": 0.7774079054694562,
+      "grad_norm": 1.75,
+      "learning_rate": 2.6029827442354218e-06,
+      "loss": 1.0155,
+      "step": 2204
+    },
+    {
+      "epoch": 0.7777606313793788,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.59514038402928e-06,
+      "loss": 0.8858,
+      "step": 2205
+    },
+    {
+      "epoch": 0.7781133572893014,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.5873080936346896e-06,
+      "loss": 0.8848,
+      "step": 2206
+    },
+    {
+      "epoch": 0.7784660831992241,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.579485883702707e-06,
+      "loss": 0.878,
+      "step": 2207
+    },
+    {
+      "epoch": 0.7788188091091466,
+      "grad_norm": 0.75,
+      "learning_rate": 2.571673764870709e-06,
+      "loss": 0.9377,
+      "step": 2208
+    },
+    {
+      "epoch": 0.7791715350190692,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.5638717477623252e-06,
+      "loss": 0.9,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7795242609289919,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.5560798429874646e-06,
+      "loss": 0.9362,
+      "step": 2210
+    },
+    {
+      "epoch": 0.7798769868389145,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.548298061142276e-06,
+      "loss": 0.9108,
+      "step": 2211
+    },
+    {
+      "epoch": 0.7802297127488371,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.5405264128091445e-06,
+      "loss": 0.9336,
+      "step": 2212
+    },
+    {
+      "epoch": 0.7805824386587598,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.532764908556675e-06,
+      "loss": 0.9312,
+      "step": 2213
+    },
+    {
+      "epoch": 0.7809351645686824,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.525013558939683e-06,
+      "loss": 0.9502,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7812878904786049,
+      "grad_norm": 1.265625,
+      "learning_rate": 2.5172723744991556e-06,
+      "loss": 0.9491,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7816406163885276,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.509541365762286e-06,
+      "loss": 0.9917,
+      "step": 2216
+    },
+    {
+      "epoch": 0.7819933422984502,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.501820543242398e-06,
+      "loss": 1.0721,
+      "step": 2217
+    },
+    {
+      "epoch": 0.7823460682083728,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.4941099174389937e-06,
+      "loss": 1.0189,
+      "step": 2218
+    },
+    {
+      "epoch": 0.7826987941182955,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.486409498837683e-06,
+      "loss": 0.9713,
+      "step": 2219
+    },
+    {
+      "epoch": 0.7830515200282181,
+      "grad_norm": 0.75,
+      "learning_rate": 2.478719297910208e-06,
+      "loss": 0.9673,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7834042459381407,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.471039325114416e-06,
+      "loss": 1.0263,
+      "step": 2221
+    },
+    {
+      "epoch": 0.7837569718480634,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.4633695908942445e-06,
+      "loss": 0.985,
+      "step": 2222
+    },
+    {
+      "epoch": 0.7841096977579859,
+      "grad_norm": 0.75,
+      "learning_rate": 2.455710105679697e-06,
+      "loss": 0.8101,
+      "step": 2223
+    },
+    {
+      "epoch": 0.7844624236679085,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.448060879886861e-06,
+      "loss": 0.9891,
+      "step": 2224
+    },
+    {
+      "epoch": 0.7848151495778312,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.440421923917846e-06,
+      "loss": 1.0159,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7851678754877538,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.432793248160824e-06,
+      "loss": 0.9677,
+      "step": 2226
+    },
+    {
+      "epoch": 0.7855206013976764,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.425174862989962e-06,
+      "loss": 0.9951,
+      "step": 2227
+    },
+    {
+      "epoch": 0.7858733273075991,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.4175667787654454e-06,
+      "loss": 0.9348,
+      "step": 2228
+    },
+    {
+      "epoch": 0.7862260532175217,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.409969005833451e-06,
+      "loss": 0.9422,
+      "step": 2229
+    },
+    {
+      "epoch": 0.7865787791274442,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.402381554526132e-06,
+      "loss": 1.0151,
+      "step": 2230
+    },
+    {
+      "epoch": 0.7869315050373669,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.394804435161603e-06,
+      "loss": 0.9087,
+      "step": 2231
+    },
+    {
+      "epoch": 0.7872842309472895,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.3872376580439303e-06,
+      "loss": 0.9146,
+      "step": 2232
+    },
+    {
+      "epoch": 0.7876369568572121,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.379681233463118e-06,
+      "loss": 0.8696,
+      "step": 2233
+    },
+    {
+      "epoch": 0.7879896827671348,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.3721351716950904e-06,
+      "loss": 0.8659,
+      "step": 2234
+    },
+    {
+      "epoch": 0.7883424086770574,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.3645994830016717e-06,
+      "loss": 0.9192,
+      "step": 2235
+    },
+    {
+      "epoch": 0.78869513458698,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.357074177630597e-06,
+      "loss": 0.9592,
+      "step": 2236
+    },
+    {
+      "epoch": 0.7890478604969027,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.349559265815463e-06,
+      "loss": 0.7897,
+      "step": 2237
+    },
+    {
+      "epoch": 0.7894005864068252,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.3420547577757446e-06,
+      "loss": 0.94,
+      "step": 2238
+    },
+    {
+      "epoch": 0.7897533123167478,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.334560663716763e-06,
+      "loss": 0.9215,
+      "step": 2239
+    },
+    {
+      "epoch": 0.7901060382266705,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.327076993829679e-06,
+      "loss": 0.9822,
+      "step": 2240
+    },
+    {
+      "epoch": 0.7904587641365931,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.3196037582914777e-06,
+      "loss": 1.0127,
+      "step": 2241
+    },
+    {
+      "epoch": 0.7908114900465157,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.3121409672649544e-06,
+      "loss": 0.901,
+      "step": 2242
+    },
+    {
+      "epoch": 0.7911642159564384,
+      "grad_norm": 0.75,
+      "learning_rate": 2.3046886308987017e-06,
+      "loss": 0.9682,
+      "step": 2243
+    },
+    {
+      "epoch": 0.791516941866361,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.297246759327094e-06,
+      "loss": 0.932,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7918696677762835,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.289815362670276e-06,
+      "loss": 0.9344,
+      "step": 2245
+    },
+    {
+      "epoch": 0.7922223936862062,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.2823944510341454e-06,
+      "loss": 0.9067,
+      "step": 2246
+    },
+    {
+      "epoch": 0.7925751195961288,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.274984034510347e-06,
+      "loss": 0.9411,
+      "step": 2247
+    },
+    {
+      "epoch": 0.7929278455060514,
+      "grad_norm": 0.75,
+      "learning_rate": 2.26758412317624e-06,
+      "loss": 0.992,
+      "step": 2248
+    },
+    {
+      "epoch": 0.7932805714159741,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.2601947270949164e-06,
+      "loss": 0.9205,
+      "step": 2249
+    },
+    {
+      "epoch": 0.7936332973258967,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.2528158563151536e-06,
+      "loss": 0.8626,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7939860232358193,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.245447520871421e-06,
+      "loss": 0.8831,
+      "step": 2251
+    },
+    {
+      "epoch": 0.794338749145742,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.2380897307838646e-06,
+      "loss": 0.9606,
+      "step": 2252
+    },
+    {
+      "epoch": 0.7946914750556645,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.2307424960582836e-06,
+      "loss": 1.0202,
+      "step": 2253
+    },
+    {
+      "epoch": 0.7950442009655871,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.223405826686126e-06,
+      "loss": 0.8335,
+      "step": 2254
+    },
+    {
+      "epoch": 0.7953969268755098,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.216079732644476e-06,
+      "loss": 0.8814,
+      "step": 2255
+    },
+    {
+      "epoch": 0.7957496527854324,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.2087642238960248e-06,
+      "loss": 0.9883,
+      "step": 2256
+    },
+    {
+      "epoch": 0.796102378695355,
+      "grad_norm": 0.75,
+      "learning_rate": 2.201459310389086e-06,
+      "loss": 0.9427,
+      "step": 2257
+    },
+    {
+      "epoch": 0.7964551046052777,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.194165002057543e-06,
+      "loss": 0.9771,
+      "step": 2258
+    },
+    {
+      "epoch": 0.7968078305152003,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.186881308820884e-06,
+      "loss": 0.9758,
+      "step": 2259
+    },
+    {
+      "epoch": 0.797160556425123,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.1796082405841356e-06,
+      "loss": 0.9206,
+      "step": 2260
+    },
+    {
+      "epoch": 0.7975132823350455,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.1723458072378924e-06,
+      "loss": 0.9446,
+      "step": 2261
+    },
+    {
+      "epoch": 0.7978660082449681,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.1650940186582805e-06,
+      "loss": 1.0506,
+      "step": 2262
+    },
+    {
+      "epoch": 0.7982187341548908,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.1578528847069514e-06,
+      "loss": 1.0231,
+      "step": 2263
+    },
+    {
+      "epoch": 0.7985714600648134,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.150622415231067e-06,
+      "loss": 0.9253,
+      "step": 2264
+    },
+    {
+      "epoch": 0.798924185974736,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.1434026200632906e-06,
+      "loss": 0.9094,
+      "step": 2265
+    },
+    {
+      "epoch": 0.7992769118846587,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.1361935090217578e-06,
+      "loss": 0.9882,
+      "step": 2266
+    },
+    {
+      "epoch": 0.7996296377945813,
+      "grad_norm": 11.375,
+      "learning_rate": 2.1289950919100932e-06,
+      "loss": 0.9279,
+      "step": 2267
+    },
+    {
+      "epoch": 0.7999823637045038,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.121807378517359e-06,
+      "loss": 0.8675,
+      "step": 2268
+    },
+    {
+      "epoch": 0.8003350896144265,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.114630378618081e-06,
+      "loss": 0.9353,
+      "step": 2269
+    },
+    {
+      "epoch": 0.8006878155243491,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.1074641019721997e-06,
+      "loss": 0.9911,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8010405414342717,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.1003085583250793e-06,
+      "loss": 0.8733,
+      "step": 2271
+    },
+    {
+      "epoch": 0.8013932673441944,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.0931637574074913e-06,
+      "loss": 0.9275,
+      "step": 2272
+    },
+    {
+      "epoch": 0.801745993254117,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.0860297089355943e-06,
+      "loss": 0.8946,
+      "step": 2273
+    },
+    {
+      "epoch": 0.8020987191640396,
+      "grad_norm": 0.75,
+      "learning_rate": 2.0789064226109245e-06,
+      "loss": 0.9835,
+      "step": 2274
+    },
+    {
+      "epoch": 0.8024514450739623,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.0717939081203898e-06,
+      "loss": 1.001,
+      "step": 2275
+    },
+    {
+      "epoch": 0.8028041709838848,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.064692175136234e-06,
+      "loss": 0.8393,
+      "step": 2276
+    },
+    {
+      "epoch": 0.8031568968938074,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.057601233316059e-06,
+      "loss": 0.9698,
+      "step": 2277
+    },
+    {
+      "epoch": 0.8035096228037301,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.0505210923027753e-06,
+      "loss": 0.9147,
+      "step": 2278
+    },
+    {
+      "epoch": 0.8038623487136527,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.043451761724614e-06,
+      "loss": 0.9397,
+      "step": 2279
+    },
+    {
+      "epoch": 0.8042150746235753,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.0363932511951046e-06,
+      "loss": 0.948,
+      "step": 2280
+    },
+    {
+      "epoch": 0.804567800533498,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.0293455703130593e-06,
+      "loss": 0.9263,
+      "step": 2281
+    },
+    {
+      "epoch": 0.8049205264434206,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.0223087286625665e-06,
+      "loss": 0.8624,
+      "step": 2282
+    },
+    {
+      "epoch": 0.8052732523533431,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.0152827358129714e-06,
+      "loss": 1.0427,
+      "step": 2283
+    },
+    {
+      "epoch": 0.8056259782632658,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.0082676013188694e-06,
+      "loss": 0.9259,
+      "step": 2284
+    },
+    {
+      "epoch": 0.8059787041731884,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.0012633347200873e-06,
+      "loss": 1.0189,
+      "step": 2285
+    },
+    {
+      "epoch": 0.806331430083111,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9942699455416723e-06,
+      "loss": 0.9508,
+      "step": 2286
+    },
+    {
+      "epoch": 0.8066841559930337,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9872874432938847e-06,
+      "loss": 0.9293,
+      "step": 2287
+    },
+    {
+      "epoch": 0.8070368819029563,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.980315837472169e-06,
+      "loss": 0.7787,
+      "step": 2288
+    },
+    {
+      "epoch": 0.8073896078128789,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.9733551375571603e-06,
+      "loss": 0.9434,
+      "step": 2289
+    },
+    {
+      "epoch": 0.8077423337228016,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.966405353014662e-06,
+      "loss": 0.9641,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8080950596327241,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.959466493295633e-06,
+      "loss": 0.9635,
+      "step": 2291
+    },
+    {
+      "epoch": 0.8084477855426467,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.952538567836174e-06,
+      "loss": 0.979,
+      "step": 2292
+    },
+    {
+      "epoch": 0.8088005114525694,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.945621586057519e-06,
+      "loss": 0.9816,
+      "step": 2293
+    },
+    {
+      "epoch": 0.809153237362492,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9387155573660154e-06,
+      "loss": 0.893,
+      "step": 2294
+    },
+    {
+      "epoch": 0.8095059632724146,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9318204911531213e-06,
+      "loss": 0.8325,
+      "step": 2295
+    },
+    {
+      "epoch": 0.8098586891823373,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.924936396795383e-06,
+      "loss": 0.9765,
+      "step": 2296
+    },
+    {
+      "epoch": 0.8102114150922599,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9180632836544268e-06,
+      "loss": 0.9647,
+      "step": 2297
+    },
+    {
+      "epoch": 0.8105641410021824,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.91120116107695e-06,
+      "loss": 0.9744,
+      "step": 2298
+    },
+    {
+      "epoch": 0.8109168669121051,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.904350038394691e-06,
+      "loss": 0.9029,
+      "step": 2299
+    },
+    {
+      "epoch": 0.8112695928220277,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.89750992492445e-06,
+      "loss": 1.0246,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8116223187319503,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.890680829968038e-06,
+      "loss": 0.9285,
+      "step": 2301
+    },
+    {
+      "epoch": 0.811975044641873,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.8838627628122897e-06,
+      "loss": 1.0038,
+      "step": 2302
+    },
+    {
+      "epoch": 0.8123277705517956,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.8770557327290418e-06,
+      "loss": 0.9209,
+      "step": 2303
+    },
+    {
+      "epoch": 0.8126804964617182,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.8702597489751228e-06,
+      "loss": 0.818,
+      "step": 2304
+    },
+    {
+      "epoch": 0.8130332223716409,
+      "grad_norm": 0.75,
+      "learning_rate": 1.8634748207923391e-06,
+      "loss": 0.9468,
+      "step": 2305
+    },
+    {
+      "epoch": 0.8133859482815634,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.8567009574074645e-06,
+      "loss": 0.8961,
+      "step": 2306
+    },
+    {
+      "epoch": 0.813738674191486,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.8499381680322159e-06,
+      "loss": 0.9225,
+      "step": 2307
+    },
+    {
+      "epoch": 0.8140914001014087,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8431864618632677e-06,
+      "loss": 0.8622,
+      "step": 2308
+    },
+    {
+      "epoch": 0.8144441260113313,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8364458480822045e-06,
+      "loss": 0.943,
+      "step": 2309
+    },
+    {
+      "epoch": 0.814796851921254,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.8297163358555436e-06,
+      "loss": 1.0217,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8151495778311766,
+      "grad_norm": 0.75,
+      "learning_rate": 1.8229979343346903e-06,
+      "loss": 0.9088,
+      "step": 2311
+    },
+    {
+      "epoch": 0.8155023037410992,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8162906526559498e-06,
+      "loss": 0.9591,
+      "step": 2312
+    },
+    {
+      "epoch": 0.8158550296510219,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8095944999405025e-06,
+      "loss": 0.893,
+      "step": 2313
+    },
+    {
+      "epoch": 0.8162077555609444,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.8029094852943951e-06,
+      "loss": 0.9126,
+      "step": 2314
+    },
+    {
+      "epoch": 0.816560481470867,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.796235617808526e-06,
+      "loss": 0.8841,
+      "step": 2315
+    },
+    {
+      "epoch": 0.8169132073807897,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.7895729065586408e-06,
+      "loss": 0.8965,
+      "step": 2316
+    },
+    {
+      "epoch": 0.8172659332907123,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7829213606053009e-06,
+      "loss": 0.9537,
+      "step": 2317
+    },
+    {
+      "epoch": 0.8176186592006349,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.7762809889939014e-06,
+      "loss": 0.8294,
+      "step": 2318
+    },
+    {
+      "epoch": 0.8179713851105576,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.7696518007546238e-06,
+      "loss": 0.93,
+      "step": 2319
+    },
+    {
+      "epoch": 0.8183241110204802,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.7630338049024597e-06,
+      "loss": 1.0184,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8186768369304027,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.7564270104371638e-06,
+      "loss": 0.9221,
+      "step": 2321
+    },
+    {
+      "epoch": 0.8190295628403254,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.7498314263432658e-06,
+      "loss": 0.8992,
+      "step": 2322
+    },
+    {
+      "epoch": 0.819382288750248,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.743247061590051e-06,
+      "loss": 1.0099,
+      "step": 2323
+    },
+    {
+      "epoch": 0.8197350146601706,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.7366739251315468e-06,
+      "loss": 0.8791,
+      "step": 2324
+    },
+    {
+      "epoch": 0.8200877405700933,
+      "grad_norm": 0.75,
+      "learning_rate": 1.7301120259065095e-06,
+      "loss": 0.9278,
+      "step": 2325
+    },
+    {
+      "epoch": 0.8204404664800159,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.7235613728384182e-06,
+      "loss": 0.9991,
+      "step": 2326
+    },
+    {
+      "epoch": 0.8207931923899385,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.7170219748354488e-06,
+      "loss": 0.989,
+      "step": 2327
+    },
+    {
+      "epoch": 0.8211459182998612,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.710493840790487e-06,
+      "loss": 0.8759,
+      "step": 2328
+    },
+    {
+      "epoch": 0.8214986442097837,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.7039769795810857e-06,
+      "loss": 0.9323,
+      "step": 2329
+    },
+    {
+      "epoch": 0.8218513701197063,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.6974714000694758e-06,
+      "loss": 1.024,
+      "step": 2330
+    },
+    {
+      "epoch": 0.822204096029629,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.6909771111025453e-06,
+      "loss": 1.0295,
+      "step": 2331
+    },
+    {
+      "epoch": 0.8225568219395516,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.6844941215118272e-06,
+      "loss": 0.9515,
+      "step": 2332
+    },
+    {
+      "epoch": 0.8229095478494742,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.6780224401134903e-06,
+      "loss": 0.8781,
+      "step": 2333
+    },
+    {
+      "epoch": 0.8232622737593969,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.6715620757083228e-06,
+      "loss": 0.9011,
+      "step": 2334
+    },
+    {
+      "epoch": 0.8236149996693195,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.6651130370817259e-06,
+      "loss": 1.0459,
+      "step": 2335
+    },
+    {
+      "epoch": 0.823967725579242,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.6586753330036963e-06,
+      "loss": 1.1045,
+      "step": 2336
+    },
+    {
+      "epoch": 0.8243204514891647,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.6522489722288181e-06,
+      "loss": 0.942,
+      "step": 2337
+    },
+    {
+      "epoch": 0.8246731773990873,
+      "grad_norm": 0.75,
+      "learning_rate": 1.6458339634962517e-06,
+      "loss": 1.005,
+      "step": 2338
+    },
+    {
+      "epoch": 0.8250259033090099,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.6394303155297186e-06,
+      "loss": 0.9344,
+      "step": 2339
+    },
+    {
+      "epoch": 0.8253786292189326,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.6330380370374853e-06,
+      "loss": 0.8877,
+      "step": 2340
+    },
+    {
+      "epoch": 0.8257313551288552,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.6266571367123663e-06,
+      "loss": 0.9074,
+      "step": 2341
+    },
+    {
+      "epoch": 0.8260840810387778,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6202876232316977e-06,
+      "loss": 0.9464,
+      "step": 2342
+    },
+    {
+      "epoch": 0.8264368069487005,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.6139295052573313e-06,
+      "loss": 1.0272,
+      "step": 2343
+    },
+    {
+      "epoch": 0.826789532858623,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.6075827914356236e-06,
+      "loss": 1.044,
+      "step": 2344
+    },
+    {
+      "epoch": 0.8271422587685456,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.6012474903974218e-06,
+      "loss": 0.9305,
+      "step": 2345
+    },
+    {
+      "epoch": 0.8274949846784683,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5949236107580524e-06,
+      "loss": 1.0065,
+      "step": 2346
+    },
+    {
+      "epoch": 0.8278477105883909,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5886111611173105e-06,
+      "loss": 0.9854,
+      "step": 2347
+    },
+    {
+      "epoch": 0.8282004364983135,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5823101500594495e-06,
+      "loss": 0.8508,
+      "step": 2348
+    },
+    {
+      "epoch": 0.8285531624082362,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5760205861531653e-06,
+      "loss": 1.0246,
+      "step": 2349
+    },
+    {
+      "epoch": 0.8289058883181588,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.5697424779515824e-06,
+      "loss": 0.8801,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8292586142280813,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5634758339922619e-06,
+      "loss": 0.9534,
+      "step": 2351
+    },
+    {
+      "epoch": 0.829611340138004,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.557220662797152e-06,
+      "loss": 0.9038,
+      "step": 2352
+    },
+    {
+      "epoch": 0.8299640660479266,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5509769728726243e-06,
+      "loss": 0.987,
+      "step": 2353
+    },
+    {
+      "epoch": 0.8303167919578492,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5447447727094168e-06,
+      "loss": 0.8404,
+      "step": 2354
+    },
+    {
+      "epoch": 0.8306695178677719,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.538524070782651e-06,
+      "loss": 0.8417,
+      "step": 2355
+    },
+    {
+      "epoch": 0.8310222437776945,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5323148755518136e-06,
+      "loss": 0.9575,
+      "step": 2356
+    },
+    {
+      "epoch": 0.8313749696876171,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.5261171954607435e-06,
+      "loss": 1.1097,
+      "step": 2357
+    },
+    {
+      "epoch": 0.8317276955975398,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.519931038937611e-06,
+      "loss": 0.9137,
+      "step": 2358
+    },
+    {
+      "epoch": 0.8320804215074623,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5137564143949313e-06,
+      "loss": 0.8995,
+      "step": 2359
+    },
+    {
+      "epoch": 0.8324331474173849,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.5075933302295198e-06,
+      "loss": 0.9177,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8327858733273076,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.5014417948225168e-06,
+      "loss": 0.9197,
+      "step": 2361
+    },
+    {
+      "epoch": 0.8331385992372302,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4953018165393407e-06,
+      "loss": 0.9106,
+      "step": 2362
+    },
+    {
+      "epoch": 0.8334913251471529,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.4891734037297035e-06,
+      "loss": 0.9001,
+      "step": 2363
+    },
+    {
+      "epoch": 0.8338440510570755,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.4830565647275852e-06,
+      "loss": 0.9199,
+      "step": 2364
+    },
+    {
+      "epoch": 0.8341967769669981,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4769513078512288e-06,
+      "loss": 0.9149,
+      "step": 2365
+    },
+    {
+      "epoch": 0.8345495028769208,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.4708576414031273e-06,
+      "loss": 0.9678,
+      "step": 2366
+    },
+    {
+      "epoch": 0.8349022287868433,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.4647755736700108e-06,
+      "loss": 0.9558,
+      "step": 2367
+    },
+    {
+      "epoch": 0.8352549546967659,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.4587051129228301e-06,
+      "loss": 0.9527,
+      "step": 2368
+    },
+    {
+      "epoch": 0.8356076806066886,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4526462674167685e-06,
+      "loss": 0.8477,
+      "step": 2369
+    },
+    {
+      "epoch": 0.8359604065166112,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.4465990453911938e-06,
+      "loss": 0.9128,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8363131324265338,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4405634550696845e-06,
+      "loss": 1.0072,
+      "step": 2371
+    },
+    {
+      "epoch": 0.8366658583364565,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4345395046599885e-06,
+      "loss": 1.0266,
+      "step": 2372
+    },
+    {
+      "epoch": 0.8370185842463791,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4285272023540297e-06,
+      "loss": 0.944,
+      "step": 2373
+    },
+    {
+      "epoch": 0.8373713101563016,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4225265563278923e-06,
+      "loss": 0.8746,
+      "step": 2374
+    },
+    {
+      "epoch": 0.8377240360662243,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4165375747418097e-06,
+      "loss": 1.0138,
+      "step": 2375
+    },
+    {
+      "epoch": 0.8380767619761469,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.4105602657401508e-06,
+      "loss": 1.0138,
+      "step": 2376
+    },
+    {
+      "epoch": 0.8384294878860695,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4045946374514152e-06,
+      "loss": 0.9105,
+      "step": 2377
+    },
+    {
+      "epoch": 0.8387822137959922,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.3986406979882073e-06,
+      "loss": 0.9527,
+      "step": 2378
+    },
+    {
+      "epoch": 0.8391349397059148,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.3926984554472523e-06,
+      "loss": 0.9336,
+      "step": 2379
+    },
+    {
+      "epoch": 0.8394876656158374,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.3867679179093506e-06,
+      "loss": 0.9862,
+      "step": 2380
+    },
+    {
+      "epoch": 0.8398403915257601,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3808490934394058e-06,
+      "loss": 0.9471,
+      "step": 2381
+    },
+    {
+      "epoch": 0.8401931174356826,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3749419900863737e-06,
+      "loss": 0.8788,
+      "step": 2382
+    },
+    {
+      "epoch": 0.8405458433456052,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3690466158832793e-06,
+      "loss": 0.9198,
+      "step": 2383
+    },
+    {
+      "epoch": 0.8408985692555279,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.3631629788471988e-06,
+      "loss": 0.8915,
+      "step": 2384
+    },
+    {
+      "epoch": 0.8412512951654505,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.3572910869792444e-06,
+      "loss": 0.9627,
+      "step": 2385
+    },
+    {
+      "epoch": 0.8416040210753731,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.351430948264556e-06,
+      "loss": 0.8133,
+      "step": 2386
+    },
+    {
+      "epoch": 0.8419567469852958,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.3455825706722913e-06,
+      "loss": 0.8533,
+      "step": 2387
+    },
+    {
+      "epoch": 0.8423094728952184,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.339745962155613e-06,
+      "loss": 0.9347,
+      "step": 2388
+    },
+    {
+      "epoch": 0.8426621988051409,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.333921130651682e-06,
+      "loss": 0.9712,
+      "step": 2389
+    },
+    {
+      "epoch": 0.8430149247150636,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.328108084081643e-06,
+      "loss": 0.8761,
+      "step": 2390
+    },
+    {
+      "epoch": 0.8433676506249862,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.3223068303506092e-06,
+      "loss": 0.8209,
+      "step": 2391
+    },
+    {
+      "epoch": 0.8437203765349088,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.3165173773476624e-06,
+      "loss": 0.9414,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8440731024448315,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.3107397329458348e-06,
+      "loss": 0.9755,
+      "step": 2393
+    },
+    {
+      "epoch": 0.8444258283547541,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.3049739050021026e-06,
+      "loss": 0.9034,
+      "step": 2394
+    },
+    {
+      "epoch": 0.8447785542646767,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2992199013573681e-06,
+      "loss": 0.9271,
+      "step": 2395
+    },
+    {
+      "epoch": 0.8451312801745994,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2934777298364575e-06,
+      "loss": 0.8209,
+      "step": 2396
+    },
+    {
+      "epoch": 0.8454840060845219,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.287747398248107e-06,
+      "loss": 0.9449,
+      "step": 2397
+    },
+    {
+      "epoch": 0.8458367319944445,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.2820289143849496e-06,
+      "loss": 0.8411,
+      "step": 2398
+    },
+    {
+      "epoch": 0.8461894579043672,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.276322286023506e-06,
+      "loss": 0.923,
+      "step": 2399
+    },
+    {
+      "epoch": 0.8465421838142898,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.2706275209241803e-06,
+      "loss": 0.8966,
+      "step": 2400
+    },
+    {
+      "epoch": 0.8468949097242124,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2649446268312316e-06,
+      "loss": 0.9347,
+      "step": 2401
+    },
+    {
+      "epoch": 0.8472476356341351,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2592736114727933e-06,
+      "loss": 0.9966,
+      "step": 2402
+    },
+    {
+      "epoch": 0.8476003615440577,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.2536144825608277e-06,
+      "loss": 1.0322,
+      "step": 2403
+    },
+    {
+      "epoch": 0.8479530874539802,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2479672477911464e-06,
+      "loss": 0.8977,
+      "step": 2404
+    },
+    {
+      "epoch": 0.8483058133639029,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2423319148433777e-06,
+      "loss": 0.9161,
+      "step": 2405
+    },
+    {
+      "epoch": 0.8486585392738255,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2367084913809668e-06,
+      "loss": 1.0717,
+      "step": 2406
+    },
+    {
+      "epoch": 0.8490112651837481,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.2310969850511656e-06,
+      "loss": 1.0017,
+      "step": 2407
+    },
+    {
+      "epoch": 0.8493639910936708,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.2254974034850176e-06,
+      "loss": 0.9227,
+      "step": 2408
+    },
+    {
+      "epoch": 0.8497167170035934,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2199097542973505e-06,
+      "loss": 0.9252,
+      "step": 2409
+    },
+    {
+      "epoch": 0.850069442913516,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.214334045086767e-06,
+      "loss": 0.9807,
+      "step": 2410
+    },
+    {
+      "epoch": 0.8504221688234387,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2087702834356264e-06,
+      "loss": 0.9032,
+      "step": 2411
+    },
+    {
+      "epoch": 0.8507748947333612,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2032184769100519e-06,
+      "loss": 0.94,
+      "step": 2412
+    },
+    {
+      "epoch": 0.8511276206432838,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.1976786330598978e-06,
+      "loss": 0.909,
+      "step": 2413
+    },
+    {
+      "epoch": 0.8514803465532065,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.1921507594187553e-06,
+      "loss": 0.9042,
+      "step": 2414
+    },
+    {
+      "epoch": 0.8518330724631291,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.1866348635039392e-06,
+      "loss": 0.9106,
+      "step": 2415
+    },
+    {
+      "epoch": 0.8521857983730518,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.1811309528164727e-06,
+      "loss": 0.9861,
+      "step": 2416
+    },
+    {
+      "epoch": 0.8525385242829744,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.1756390348410819e-06,
+      "loss": 0.93,
+      "step": 2417
+    },
+    {
+      "epoch": 0.852891250192897,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.1701591170461856e-06,
+      "loss": 1.0432,
+      "step": 2418
+    },
+    {
+      "epoch": 0.8532439761028197,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.1646912068838756e-06,
+      "loss": 0.989,
+      "step": 2419
+    },
+    {
+      "epoch": 0.8535967020127422,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.159235311789929e-06,
+      "loss": 0.9196,
+      "step": 2420
+    },
+    {
+      "epoch": 0.8539494279226648,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.1537914391837668e-06,
+      "loss": 1.0161,
+      "step": 2421
+    },
+    {
+      "epoch": 0.8543021538325875,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.148359596468479e-06,
+      "loss": 0.9012,
+      "step": 2422
+    },
+    {
+      "epoch": 0.8546548797425101,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.1429397910307794e-06,
+      "loss": 1.0419,
+      "step": 2423
+    },
+    {
+      "epoch": 0.8550076056524327,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.137532030241022e-06,
+      "loss": 0.8564,
+      "step": 2424
+    },
+    {
+      "epoch": 0.8553603315623554,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.13213632145318e-06,
+      "loss": 0.9359,
+      "step": 2425
+    },
+    {
+      "epoch": 0.855713057472278,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.1267526720048338e-06,
+      "loss": 0.9173,
+      "step": 2426
+    },
+    {
+      "epoch": 0.8560657833822005,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.1213810892171694e-06,
+      "loss": 0.9548,
+      "step": 2427
+    },
+    {
+      "epoch": 0.8564185092921232,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.1160215803949614e-06,
+      "loss": 0.9011,
+      "step": 2428
+    },
+    {
+      "epoch": 0.8567712352020458,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.1106741528265585e-06,
+      "loss": 1.0026,
+      "step": 2429
+    },
+    {
+      "epoch": 0.8571239611119684,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.105338813783895e-06,
+      "loss": 0.8731,
+      "step": 2430
+    },
+    {
+      "epoch": 0.8574766870218911,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.100015570522449e-06,
+      "loss": 0.9651,
+      "step": 2431
+    },
+    {
+      "epoch": 0.8578294129318137,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.094704430281266e-06,
+      "loss": 0.9793,
+      "step": 2432
+    },
+    {
+      "epoch": 0.8581821388417363,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.0894054002829192e-06,
+      "loss": 0.9695,
+      "step": 2433
+    },
+    {
+      "epoch": 0.858534864751659,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.0841184877335186e-06,
+      "loss": 0.9317,
+      "step": 2434
+    },
+    {
+      "epoch": 0.8588875906615815,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.0788436998226991e-06,
+      "loss": 0.9787,
+      "step": 2435
+    },
+    {
+      "epoch": 0.8592403165715041,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.0735810437236015e-06,
+      "loss": 0.9645,
+      "step": 2436
+    },
+    {
+      "epoch": 0.8595930424814268,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.0683305265928722e-06,
+      "loss": 0.8969,
+      "step": 2437
+    },
+    {
+      "epoch": 0.8599457683913494,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.0630921555706498e-06,
+      "loss": 0.9414,
+      "step": 2438
+    },
+    {
+      "epoch": 0.860298494301272,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.0578659377805534e-06,
+      "loss": 0.9327,
+      "step": 2439
+    },
+    {
+      "epoch": 0.8606512202111947,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.0526518803296771e-06,
+      "loss": 1.0531,
+      "step": 2440
+    },
+    {
+      "epoch": 0.8610039461211173,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.047449990308579e-06,
+      "loss": 0.9573,
+      "step": 2441
+    },
+    {
+      "epoch": 0.8613566720310398,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.0422602747912635e-06,
+      "loss": 0.8956,
+      "step": 2442
+    },
+    {
+      "epoch": 0.8617093979409625,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.0370827408351925e-06,
+      "loss": 0.9488,
+      "step": 2443
+    },
+    {
+      "epoch": 0.8620621238508851,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.0319173954812489e-06,
+      "loss": 0.9312,
+      "step": 2444
+    },
+    {
+      "epoch": 0.8624148497608077,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.0267642457537486e-06,
+      "loss": 1.0703,
+      "step": 2445
+    },
+    {
+      "epoch": 0.8627675756707304,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.021623298660419e-06,
+      "loss": 0.9181,
+      "step": 2446
+    },
+    {
+      "epoch": 0.863120301580653,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.0164945611923948e-06,
+      "loss": 1.0363,
+      "step": 2447
+    },
+    {
+      "epoch": 0.8634730274905756,
+      "grad_norm": 0.75,
+      "learning_rate": 1.0113780403242069e-06,
+      "loss": 1.0151,
+      "step": 2448
+    },
+    {
+      "epoch": 0.8638257534004983,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.0062737430137714e-06,
+      "loss": 0.9903,
+      "step": 2449
+    },
+    {
+      "epoch": 0.8641784793104208,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.0011816762023852e-06,
+      "loss": 0.9014,
+      "step": 2450
+    },
+    {
+      "epoch": 0.8645312052203434,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.96101846814711e-07,
+      "loss": 0.8469,
+      "step": 2451
+    },
+    {
+      "epoch": 0.8648839311302661,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.910342617587643e-07,
+      "loss": 0.9972,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8652366570401887,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.859789279259225e-07,
+      "loss": 1.0181,
+      "step": 2453
+    },
+    {
+      "epoch": 0.8655893829501113,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.809358521908884e-07,
+      "loss": 1.022,
+      "step": 2454
+    },
+    {
+      "epoch": 0.865942108860034,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.759050414117078e-07,
+      "loss": 0.9991,
+      "step": 2455
+    },
+    {
+      "epoch": 0.8662948347699566,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.708865024297365e-07,
+      "loss": 0.8729,
+      "step": 2456
+    },
+    {
+      "epoch": 0.8666475606798791,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.6588024206965e-07,
+      "loss": 0.808,
+      "step": 2457
+    },
+    {
+      "epoch": 0.8670002865898018,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.608862671394214e-07,
+      "loss": 1.0023,
+      "step": 2458
+    },
+    {
+      "epoch": 0.8673530124997244,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.559045844303193e-07,
+      "loss": 0.9267,
+      "step": 2459
+    },
+    {
+      "epoch": 0.867705738409647,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.50935200716897e-07,
+      "loss": 0.9294,
+      "step": 2460
+    },
+    {
+      "epoch": 0.8680584643195697,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.45978122756982e-07,
+      "loss": 0.9218,
+      "step": 2461
+    },
+    {
+      "epoch": 0.8684111902294923,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.410333572916597e-07,
+      "loss": 0.9414,
+      "step": 2462
+    },
+    {
+      "epoch": 0.868763916139415,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.361009110452879e-07,
+      "loss": 0.9726,
+      "step": 2463
+    },
+    {
+      "epoch": 0.8691166420493376,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.311807907254533e-07,
+      "loss": 0.9358,
+      "step": 2464
+    },
+    {
+      "epoch": 0.8694693679592601,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.262730030229983e-07,
+      "loss": 0.9751,
+      "step": 2465
+    },
+    {
+      "epoch": 0.8698220938691827,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.213775546119797e-07,
+      "loss": 0.8992,
+      "step": 2466
+    },
+    {
+      "epoch": 0.8701748197791054,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.16494452149681e-07,
+      "loss": 0.9129,
+      "step": 2467
+    },
+    {
+      "epoch": 0.870527545689028,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.116237022765972e-07,
+      "loss": 0.9838,
+      "step": 2468
+    },
+    {
+      "epoch": 0.8708802715989507,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.067653116164232e-07,
+      "loss": 0.9217,
+      "step": 2469
+    },
+    {
+      "epoch": 0.8712329975088733,
+      "grad_norm": 0.75,
+      "learning_rate": 9.01919286776044e-07,
+      "loss": 0.9451,
+      "step": 2470
+    },
+    {
+      "epoch": 0.8715857234187959,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.97085634345537e-07,
+      "loss": 0.9299,
+      "step": 2471
+    },
+    {
+      "epoch": 0.8719384493287186,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.922643608981419e-07,
+      "loss": 0.9293,
+      "step": 2472
+    },
+    {
+      "epoch": 0.8722911752386411,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.874554729902796e-07,
+      "loss": 0.868,
+      "step": 2473
+    },
+    {
+      "epoch": 0.8726439011485637,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.826589771615124e-07,
+      "loss": 0.9606,
+      "step": 2474
+    },
+    {
+      "epoch": 0.8729966270584864,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.778748799345615e-07,
+      "loss": 0.8969,
+      "step": 2475
+    },
+    {
+      "epoch": 0.873349352968409,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.73103187815284e-07,
+      "loss": 0.9859,
+      "step": 2476
+    },
+    {
+      "epoch": 0.8737020788783316,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.683439072926658e-07,
+      "loss": 0.9553,
+      "step": 2477
+    },
+    {
+      "epoch": 0.8740548047882543,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.635970448388175e-07,
+      "loss": 0.9004,
+      "step": 2478
+    },
+    {
+      "epoch": 0.8744075306981769,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.588626069089601e-07,
+      "loss": 1.0231,
+      "step": 2479
+    },
+    {
+      "epoch": 0.8747602566080994,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.541405999414166e-07,
+      "loss": 0.9099,
+      "step": 2480
+    },
+    {
+      "epoch": 0.875112982518022,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.494310303576137e-07,
+      "loss": 0.8567,
+      "step": 2481
+    },
+    {
+      "epoch": 0.8754657084279447,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.447339045620518e-07,
+      "loss": 0.9371,
+      "step": 2482
+    },
+    {
+      "epoch": 0.8758184343378673,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.400492289423235e-07,
+      "loss": 0.9943,
+      "step": 2483
+    },
+    {
+      "epoch": 0.87617116024779,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.353770098690772e-07,
+      "loss": 0.9459,
+      "step": 2484
+    },
+    {
+      "epoch": 0.8765238861577126,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.307172536960306e-07,
+      "loss": 1.0216,
+      "step": 2485
+    },
+    {
+      "epoch": 0.8768766120676352,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.260699667599481e-07,
+      "loss": 0.922,
+      "step": 2486
+    },
+    {
+      "epoch": 0.8772293379775579,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.214351553806421e-07,
+      "loss": 0.8772,
+      "step": 2487
+    },
+    {
+      "epoch": 0.8775820638874804,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.168128258609542e-07,
+      "loss": 0.8779,
+      "step": 2488
+    },
+    {
+      "epoch": 0.877934789797403,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.122029844867562e-07,
+      "loss": 0.9222,
+      "step": 2489
+    },
+    {
+      "epoch": 0.8782875157073257,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.076056375269359e-07,
+      "loss": 1.0084,
+      "step": 2490
+    },
+    {
+      "epoch": 0.8786402416172483,
+      "grad_norm": 0.73828125,
+      "learning_rate": 8.030207912333899e-07,
+      "loss": 1.0411,
+      "step": 2491
+    },
+    {
+      "epoch": 0.8789929675271709,
+      "grad_norm": 0.75,
+      "learning_rate": 7.984484518410141e-07,
+      "loss": 1.0542,
+      "step": 2492
+    },
+    {
+      "epoch": 0.8793456934370936,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.938886255676992e-07,
+      "loss": 0.9926,
+      "step": 2493
+    },
+    {
+      "epoch": 0.8796984193470162,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.893413186143195e-07,
+      "loss": 0.9152,
+      "step": 2494
+    },
+    {
+      "epoch": 0.8800511452569387,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.848065371647162e-07,
+      "loss": 0.9209,
+      "step": 2495
+    },
+    {
+      "epoch": 0.8804038711668614,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.80284287385713e-07,
+      "loss": 0.9683,
+      "step": 2496
+    },
+    {
+      "epoch": 0.880756597076784,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.757745754270741e-07,
+      "loss": 0.9442,
+      "step": 2497
+    },
+    {
+      "epoch": 0.8811093229867066,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.712774074215257e-07,
+      "loss": 0.9114,
+      "step": 2498
+    },
+    {
+      "epoch": 0.8814620488966293,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.667927894847327e-07,
+      "loss": 1.0404,
+      "step": 2499
+    },
+    {
+      "epoch": 0.8818147748065519,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.623207277152911e-07,
+      "loss": 0.9646,
+      "step": 2500
+    },
+    {
+      "epoch": 0.8821675007164745,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.578612281947229e-07,
+      "loss": 0.9176,
+      "step": 2501
+    },
+    {
+      "epoch": 0.8825202266263972,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.534142969874691e-07,
+      "loss": 0.9479,
+      "step": 2502
+    },
+    {
+      "epoch": 0.8828729525363197,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.489799401408737e-07,
+      "loss": 0.9277,
+      "step": 2503
+    },
+    {
+      "epoch": 0.8832256784462423,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.445581636851895e-07,
+      "loss": 1.0763,
+      "step": 2504
+    },
+    {
+      "epoch": 0.883578404356165,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.40148973633551e-07,
+      "loss": 1.0039,
+      "step": 2505
+    },
+    {
+      "epoch": 0.8839311302660876,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.357523759819862e-07,
+      "loss": 0.9549,
+      "step": 2506
+    },
+    {
+      "epoch": 0.8842838561760102,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.313683767093905e-07,
+      "loss": 0.926,
+      "step": 2507
+    },
+    {
+      "epoch": 0.8846365820859329,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.269969817775335e-07,
+      "loss": 0.9588,
+      "step": 2508
+    },
+    {
+      "epoch": 0.8849893079958555,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.226381971310404e-07,
+      "loss": 0.9853,
+      "step": 2509
+    },
+    {
+      "epoch": 0.885342033905778,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.182920286973882e-07,
+      "loss": 0.9865,
+      "step": 2510
+    },
+    {
+      "epoch": 0.8856947598157007,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.139584823868995e-07,
+      "loss": 0.855,
+      "step": 2511
+    },
+    {
+      "epoch": 0.8860474857256233,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.096375640927311e-07,
+      "loss": 0.9544,
+      "step": 2512
+    },
+    {
+      "epoch": 0.8864002116355459,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.053292796908629e-07,
+      "loss": 0.9277,
+      "step": 2513
+    },
+    {
+      "epoch": 0.8867529375454686,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.010336350401015e-07,
+      "loss": 0.8785,
+      "step": 2514
+    },
+    {
+      "epoch": 0.8871056634553912,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.967506359820564e-07,
+      "loss": 0.8914,
+      "step": 2515
+    },
+    {
+      "epoch": 0.8874583893653138,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.924802883411518e-07,
+      "loss": 0.934,
+      "step": 2516
+    },
+    {
+      "epoch": 0.8878111152752365,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.882225979245949e-07,
+      "loss": 0.96,
+      "step": 2517
+    },
+    {
+      "epoch": 0.888163841185159,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.839775705223883e-07,
+      "loss": 0.9952,
+      "step": 2518
+    },
+    {
+      "epoch": 0.8885165670950816,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.797452119073111e-07,
+      "loss": 0.9574,
+      "step": 2519
+    },
+    {
+      "epoch": 0.8888692930050043,
+      "grad_norm": 0.75,
+      "learning_rate": 6.75525527834916e-07,
+      "loss": 0.9689,
+      "step": 2520
+    },
+    {
+      "epoch": 0.8892220189149269,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.713185240435182e-07,
+      "loss": 0.8857,
+      "step": 2521
+    },
+    {
+      "epoch": 0.8895747448248496,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.671242062541928e-07,
+      "loss": 0.9698,
+      "step": 2522
+    },
+    {
+      "epoch": 0.8899274707347722,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.629425801707534e-07,
+      "loss": 0.948,
+      "step": 2523
+    },
+    {
+      "epoch": 0.8902801966446948,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.587736514797694e-07,
+      "loss": 0.9508,
+      "step": 2524
+    },
+    {
+      "epoch": 0.8906329225546175,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.546174258505289e-07,
+      "loss": 0.9501,
+      "step": 2525
+    },
+    {
+      "epoch": 0.89098564846454,
+      "grad_norm": 0.75,
+      "learning_rate": 6.50473908935052e-07,
+      "loss": 1.0083,
+      "step": 2526
+    },
+    {
+      "epoch": 0.8913383743744626,
+      "grad_norm": 0.75,
+      "learning_rate": 6.463431063680769e-07,
+      "loss": 1.0107,
+      "step": 2527
+    },
+    {
+      "epoch": 0.8916911002843853,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.422250237670502e-07,
+      "loss": 0.9926,
+      "step": 2528
+    },
+    {
+      "epoch": 0.8920438261943079,
+      "grad_norm": 0.796875,
+      "learning_rate": 6.381196667321199e-07,
+      "loss": 0.9092,
+      "step": 2529
+    },
+    {
+      "epoch": 0.8923965521042305,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.340270408461291e-07,
+      "loss": 0.8718,
+      "step": 2530
+    },
+    {
+      "epoch": 0.8927492780141532,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.299471516746092e-07,
+      "loss": 0.9328,
+      "step": 2531
+    },
+    {
+      "epoch": 0.8931020039240758,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.258800047657709e-07,
+      "loss": 0.9829,
+      "step": 2532
+    },
+    {
+      "epoch": 0.8934547298339983,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.218256056504923e-07,
+      "loss": 1.0285,
+      "step": 2533
+    },
+    {
+      "epoch": 0.893807455743921,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.177839598423252e-07,
+      "loss": 0.9905,
+      "step": 2534
+    },
+    {
+      "epoch": 0.8941601816538436,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.137550728374664e-07,
+      "loss": 1.0036,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8945129075637662,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.097389501147721e-07,
+      "loss": 0.9975,
+      "step": 2536
+    },
+    {
+      "epoch": 0.8948656334736889,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.057355971357359e-07,
+      "loss": 0.917,
+      "step": 2537
+    },
+    {
+      "epoch": 0.8952183593836115,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.017450193444852e-07,
+      "loss": 0.9841,
+      "step": 2538
+    },
+    {
+      "epoch": 0.8955710852935341,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.977672221677788e-07,
+      "loss": 0.922,
+      "step": 2539
+    },
+    {
+      "epoch": 0.8959238112034568,
+      "grad_norm": 0.734375,
+      "learning_rate": 5.9380221101499e-07,
+      "loss": 0.9831,
+      "step": 2540
+    },
+    {
+      "epoch": 0.8962765371133793,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.898499912781086e-07,
+      "loss": 1.0037,
+      "step": 2541
+    },
+    {
+      "epoch": 0.8966292630233019,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.859105683317268e-07,
+      "loss": 1.0269,
+      "step": 2542
+    },
+    {
+      "epoch": 0.8969819889332246,
+      "grad_norm": 0.7578125,
+      "learning_rate": 5.819839475330358e-07,
+      "loss": 0.8547,
+      "step": 2543
+    },
+    {
+      "epoch": 0.8973347148431472,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.780701342218164e-07,
+      "loss": 0.917,
+      "step": 2544
+    },
+    {
+      "epoch": 0.8976874407530698,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.741691337204358e-07,
+      "loss": 1.042,
+      "step": 2545
+    },
+    {
+      "epoch": 0.8980401666629925,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.702809513338292e-07,
+      "loss": 0.9671,
+      "step": 2546
+    },
+    {
+      "epoch": 0.8983928925729151,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.664055923495116e-07,
+      "loss": 1.001,
+      "step": 2547
+    },
+    {
+      "epoch": 0.8987456184828376,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.625430620375505e-07,
+      "loss": 0.9723,
+      "step": 2548
+    },
+    {
+      "epoch": 0.8990983443927603,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.586933656505711e-07,
+      "loss": 0.9921,
+      "step": 2549
+    },
+    {
+      "epoch": 0.8994510703026829,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.548565084237478e-07,
+      "loss": 0.9863,
+      "step": 2550
+    },
+    {
+      "epoch": 0.8998037962126055,
+      "grad_norm": 0.828125,
+      "learning_rate": 5.510324955747914e-07,
+      "loss": 0.9593,
+      "step": 2551
+    },
+    {
+      "epoch": 0.9001565221225282,
+      "grad_norm": 0.90625,
+      "learning_rate": 5.472213323039488e-07,
+      "loss": 0.9484,
+      "step": 2552
+    },
+    {
+      "epoch": 0.9005092480324508,
+      "grad_norm": 0.78125,
+      "learning_rate": 5.434230237939919e-07,
+      "loss": 0.9136,
+      "step": 2553
+    },
+    {
+      "epoch": 0.9008619739423734,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.396375752102068e-07,
+      "loss": 0.9867,
+      "step": 2554
+    },
+    {
+      "epoch": 0.9012146998522961,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.358649917004033e-07,
+      "loss": 0.8836,
+      "step": 2555
+    },
+    {
+      "epoch": 0.9015674257622186,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.32105278394881e-07,
+      "loss": 0.955,
+      "step": 2556
+    },
+    {
+      "epoch": 0.9019201516721412,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.283584404064512e-07,
+      "loss": 0.9861,
+      "step": 2557
+    },
+    {
+      "epoch": 0.9022728775820639,
+      "grad_norm": 0.80859375,
+      "learning_rate": 5.24624482830407e-07,
+      "loss": 0.8969,
+      "step": 2558
+    },
+    {
+      "epoch": 0.9026256034919865,
+      "grad_norm": 0.87109375,
+      "learning_rate": 5.209034107445288e-07,
+      "loss": 0.874,
+      "step": 2559
+    },
+    {
+      "epoch": 0.9029783294019091,
+      "grad_norm": 0.8203125,
+      "learning_rate": 5.171952292090743e-07,
+      "loss": 0.9559,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9033310553118318,
+      "grad_norm": 0.75390625,
+      "learning_rate": 5.134999432667708e-07,
+      "loss": 1.0259,
+      "step": 2561
+    },
+    {
+      "epoch": 0.9036837812217544,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.098175579428089e-07,
+      "loss": 0.9956,
+      "step": 2562
+    },
+    {
+      "epoch": 0.9040365071316769,
+      "grad_norm": 0.75,
+      "learning_rate": 5.06148078244838e-07,
+      "loss": 0.8884,
+      "step": 2563
+    },
+    {
+      "epoch": 0.9043892330415996,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.024915091629512e-07,
+      "loss": 0.961,
+      "step": 2564
+    },
+    {
+      "epoch": 0.9047419589515222,
+      "grad_norm": 0.75,
+      "learning_rate": 4.98847855669693e-07,
+      "loss": 1.0043,
+      "step": 2565
+    },
+    {
+      "epoch": 0.9050946848614448,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.952171227200353e-07,
+      "loss": 0.9396,
+      "step": 2566
+    },
+    {
+      "epoch": 0.9054474107713675,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.915993152513887e-07,
+      "loss": 0.8951,
+      "step": 2567
+    },
+    {
+      "epoch": 0.9058001366812901,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.879944381835788e-07,
+      "loss": 0.9284,
+      "step": 2568
+    },
+    {
+      "epoch": 0.9061528625912127,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.844024964188499e-07,
+      "loss": 0.9619,
+      "step": 2569
+    },
+    {
+      "epoch": 0.9065055885011354,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.808234948418578e-07,
+      "loss": 0.9021,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9068583144110579,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.77257438319657e-07,
+      "loss": 0.924,
+      "step": 2571
+    },
+    {
+      "epoch": 0.9072110403209805,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.7370433170170205e-07,
+      "loss": 0.9442,
+      "step": 2572
+    },
+    {
+      "epoch": 0.9075637662309032,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.701641798198353e-07,
+      "loss": 0.8605,
+      "step": 2573
+    },
+    {
+      "epoch": 0.9079164921408258,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.666369874882792e-07,
+      "loss": 0.9375,
+      "step": 2574
+    },
+    {
+      "epoch": 0.9082692180507485,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.6312275950364027e-07,
+      "loss": 0.8444,
+      "step": 2575
+    },
+    {
+      "epoch": 0.9086219439606711,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.5962150064488565e-07,
+      "loss": 0.9896,
+      "step": 2576
+    },
+    {
+      "epoch": 0.9089746698705937,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.5613321567335113e-07,
+      "loss": 0.8801,
+      "step": 2577
+    },
+    {
+      "epoch": 0.9093273957805164,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.52657909332731e-07,
+      "loss": 0.9393,
+      "step": 2578
+    },
+    {
+      "epoch": 0.9096801216904389,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.491955863490649e-07,
+      "loss": 1.0094,
+      "step": 2579
+    },
+    {
+      "epoch": 0.9100328476003615,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.457462514307409e-07,
+      "loss": 0.9011,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9103855735102842,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.423099092684824e-07,
+      "loss": 0.9745,
+      "step": 2581
+    },
+    {
+      "epoch": 0.9107382994202068,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.3888656453534463e-07,
+      "loss": 0.9195,
+      "step": 2582
+    },
+    {
+      "epoch": 0.9110910253301294,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.35476221886707e-07,
+      "loss": 0.9368,
+      "step": 2583
+    },
+    {
+      "epoch": 0.9114437512400521,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.320788859602687e-07,
+      "loss": 0.9436,
+      "step": 2584
+    },
+    {
+      "epoch": 0.9117964771499747,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.286945613760429e-07,
+      "loss": 0.9931,
+      "step": 2585
+    },
+    {
+      "epoch": 0.9121492030598972,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.253232527363438e-07,
+      "loss": 0.978,
+      "step": 2586
+    },
+    {
+      "epoch": 0.9125019289698199,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.2196496462578726e-07,
+      "loss": 1.0781,
+      "step": 2587
+    },
+    {
+      "epoch": 0.9128546548797425,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.1861970161128584e-07,
+      "loss": 0.9327,
+      "step": 2588
+    },
+    {
+      "epoch": 0.9132073807896651,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.1528746824203605e-07,
+      "loss": 1.0404,
+      "step": 2589
+    },
+    {
+      "epoch": 0.9135601066995878,
+      "grad_norm": 0.75,
+      "learning_rate": 4.1196826904951524e-07,
+      "loss": 0.881,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9139128326095104,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.0866210854747956e-07,
+      "loss": 0.973,
+      "step": 2591
+    },
+    {
+      "epoch": 0.914265558519433,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.0536899123194804e-07,
+      "loss": 1.0013,
+      "step": 2592
+    },
+    {
+      "epoch": 0.9146182844293557,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.020889215812085e-07,
+      "loss": 0.9795,
+      "step": 2593
+    },
+    {
+      "epoch": 0.9149710103392782,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.9882190405579945e-07,
+      "loss": 0.8498,
+      "step": 2594
+    },
+    {
+      "epoch": 0.9153237362492008,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.955679430985149e-07,
+      "loss": 0.8582,
+      "step": 2595
+    },
+    {
+      "epoch": 0.9156764621591235,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.9232704313439283e-07,
+      "loss": 0.9022,
+      "step": 2596
+    },
+    {
+      "epoch": 0.9160291880690461,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.8909920857070237e-07,
+      "loss": 0.9334,
+      "step": 2597
+    },
+    {
+      "epoch": 0.9163819139789687,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.8588444379695777e-07,
+      "loss": 0.9551,
+      "step": 2598
+    },
+    {
+      "epoch": 0.9167346398888914,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.8268275318488645e-07,
+      "loss": 0.9924,
+      "step": 2599
+    },
+    {
+      "epoch": 0.917087365798814,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.7949414108844893e-07,
+      "loss": 1.0303,
+      "step": 2600
+    },
+    {
+      "epoch": 0.9174400917087365,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.7631861184381e-07,
+      "loss": 0.9339,
+      "step": 2601
+    },
+    {
+      "epoch": 0.9177928176186592,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.7315616976934866e-07,
+      "loss": 1.1079,
+      "step": 2602
+    },
+    {
+      "epoch": 0.9181455435285818,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.7000681916564583e-07,
+      "loss": 0.8583,
+      "step": 2603
+    },
+    {
+      "epoch": 0.9184982694385044,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.6687056431548016e-07,
+      "loss": 0.9049,
+      "step": 2604
+    },
+    {
+      "epoch": 0.9188509953484271,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.637474094838167e-07,
+      "loss": 0.9746,
+      "step": 2605
+    },
+    {
+      "epoch": 0.9192037212583497,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.606373589178136e-07,
+      "loss": 1.0706,
+      "step": 2606
+    },
+    {
+      "epoch": 0.9195564471682723,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.5754041684680217e-07,
+      "loss": 0.962,
+      "step": 2607
+    },
+    {
+      "epoch": 0.919909173078195,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.544565874822925e-07,
+      "loss": 0.9269,
+      "step": 2608
+    },
+    {
+      "epoch": 0.9202618989881175,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.5138587501795884e-07,
+      "loss": 0.8512,
+      "step": 2609
+    },
+    {
+      "epoch": 0.9206146248980401,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.483282836296387e-07,
+      "loss": 1.0198,
+      "step": 2610
+    },
+    {
+      "epoch": 0.9209673508079628,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.452838174753281e-07,
+      "loss": 0.9932,
+      "step": 2611
+    },
+    {
+      "epoch": 0.9213200767178854,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.4225248069517326e-07,
+      "loss": 1.0073,
+      "step": 2612
+    },
+    {
+      "epoch": 0.921672802627808,
+      "grad_norm": 0.75,
+      "learning_rate": 3.392342774114643e-07,
+      "loss": 0.9317,
+      "step": 2613
+    },
+    {
+      "epoch": 0.9220255285377307,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.362292117286359e-07,
+      "loss": 0.9408,
+      "step": 2614
+    },
+    {
+      "epoch": 0.9223782544476533,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.3323728773324904e-07,
+      "loss": 0.9205,
+      "step": 2615
+    },
+    {
+      "epoch": 0.9227309803575758,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.302585094940025e-07,
+      "loss": 1.0273,
+      "step": 2616
+    },
+    {
+      "epoch": 0.9230837062674985,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.272928810617104e-07,
+      "loss": 0.934,
+      "step": 2617
+    },
+    {
+      "epoch": 0.9234364321774211,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.243404064693112e-07,
+      "loss": 0.8662,
+      "step": 2618
+    },
+    {
+      "epoch": 0.9237891580873437,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.214010897318498e-07,
+      "loss": 0.9278,
+      "step": 2619
+    },
+    {
+      "epoch": 0.9241418839972664,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.184749348464811e-07,
+      "loss": 0.9143,
+      "step": 2620
+    },
+    {
+      "epoch": 0.924494609907189,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.155619457924608e-07,
+      "loss": 0.8815,
+      "step": 2621
+    },
+    {
+      "epoch": 0.9248473358171116,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.1266212653114025e-07,
+      "loss": 0.9577,
+      "step": 2622
+    },
+    {
+      "epoch": 0.9252000617270343,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.0977548100596166e-07,
+      "loss": 0.9126,
+      "step": 2623
+    },
+    {
+      "epoch": 0.9255527876369568,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.069020131424527e-07,
+      "loss": 0.9186,
+      "step": 2624
+    },
+    {
+      "epoch": 0.9259055135468794,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.0404172684821655e-07,
+      "loss": 0.9724,
+      "step": 2625
+    },
+    {
+      "epoch": 0.9262582394568021,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.011946260129395e-07,
+      "loss": 0.9027,
+      "step": 2626
+    },
+    {
+      "epoch": 0.9266109653667247,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.9836071450836776e-07,
+      "loss": 0.8821,
+      "step": 2627
+    },
+    {
+      "epoch": 0.9269636912766474,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.9553999618831877e-07,
+      "loss": 0.9539,
+      "step": 2628
+    },
+    {
+      "epoch": 0.92731641718657,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.9273247488866507e-07,
+      "loss": 0.8536,
+      "step": 2629
+    },
+    {
+      "epoch": 0.9276691430964926,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.8993815442733166e-07,
+      "loss": 1.0014,
+      "step": 2630
+    },
+    {
+      "epoch": 0.9280218690064153,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.8715703860429436e-07,
+      "loss": 0.9174,
+      "step": 2631
+    },
+    {
+      "epoch": 0.9283745949163378,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.843891312015712e-07,
+      "loss": 0.8607,
+      "step": 2632
+    },
+    {
+      "epoch": 0.9287273208262604,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.81634435983219e-07,
+      "loss": 0.8692,
+      "step": 2633
+    },
+    {
+      "epoch": 0.929080046736183,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.788929566953258e-07,
+      "loss": 0.9514,
+      "step": 2634
+    },
+    {
+      "epoch": 0.9294327726461057,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.76164697066007e-07,
+      "loss": 0.9931,
+      "step": 2635
+    },
+    {
+      "epoch": 0.9297854985560283,
+      "grad_norm": 0.75,
+      "learning_rate": 2.734496608054049e-07,
+      "loss": 0.8869,
+      "step": 2636
+    },
+    {
+      "epoch": 0.930138224465951,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.7074785160567276e-07,
+      "loss": 0.8901,
+      "step": 2637
+    },
+    {
+      "epoch": 0.9304909503758736,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.680592731409826e-07,
+      "loss": 0.9104,
+      "step": 2638
+    },
+    {
+      "epoch": 0.9308436762857961,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.653839290675109e-07,
+      "loss": 0.9323,
+      "step": 2639
+    },
+    {
+      "epoch": 0.9311964021957188,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.627218230234363e-07,
+      "loss": 0.9363,
+      "step": 2640
+    },
+    {
+      "epoch": 0.9315491281056414,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.6007295862893743e-07,
+      "loss": 0.9753,
+      "step": 2641
+    },
+    {
+      "epoch": 0.931901854015564,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.574373394861851e-07,
+      "loss": 0.9311,
+      "step": 2642
+    },
+    {
+      "epoch": 0.9322545799254867,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.548149691793367e-07,
+      "loss": 0.9551,
+      "step": 2643
+    },
+    {
+      "epoch": 0.9326073058354093,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.522058512745329e-07,
+      "loss": 0.8658,
+      "step": 2644
+    },
+    {
+      "epoch": 0.9329600317453319,
+      "grad_norm": 1.2421875,
+      "learning_rate": 2.496099893198944e-07,
+      "loss": 0.9528,
+      "step": 2645
+    },
+    {
+      "epoch": 0.9333127576552546,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.470273868455131e-07,
+      "loss": 0.8896,
+      "step": 2646
+    },
+    {
+      "epoch": 0.9336654835651771,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.444580473634506e-07,
+      "loss": 0.8355,
+      "step": 2647
+    },
+    {
+      "epoch": 0.9340182094750997,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.419019743677298e-07,
+      "loss": 0.9187,
+      "step": 2648
+    },
+    {
+      "epoch": 0.9343709353850224,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.3935917133433905e-07,
+      "loss": 0.9612,
+      "step": 2649
+    },
+    {
+      "epoch": 0.934723661294945,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.3682964172121126e-07,
+      "loss": 0.9149,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9350763872048676,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.3431338896823917e-07,
+      "loss": 0.8325,
+      "step": 2651
+    },
+    {
+      "epoch": 0.9354291131147903,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.3181041649725234e-07,
+      "loss": 0.8344,
+      "step": 2652
+    },
+    {
+      "epoch": 0.9357818390247129,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.2932072771202464e-07,
+      "loss": 0.8965,
+      "step": 2653
+    },
+    {
+      "epoch": 0.9361345649346354,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.2684432599826444e-07,
+      "loss": 0.9105,
+      "step": 2654
+    },
+    {
+      "epoch": 0.9364872908445581,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.2438121472361128e-07,
+      "loss": 0.9306,
+      "step": 2655
+    },
+    {
+      "epoch": 0.9368400167544807,
+      "grad_norm": 0.75,
+      "learning_rate": 2.2193139723763246e-07,
+      "loss": 0.9715,
+      "step": 2656
+    },
+    {
+      "epoch": 0.9371927426644033,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.1949487687181525e-07,
+      "loss": 0.9541,
+      "step": 2657
+    },
+    {
+      "epoch": 0.937545468574326,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.1707165693956367e-07,
+      "loss": 0.8494,
+      "step": 2658
+    },
+    {
+      "epoch": 0.9378981944842486,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.146617407361984e-07,
+      "loss": 0.8822,
+      "step": 2659
+    },
+    {
+      "epoch": 0.9382509203941712,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.1226513153894345e-07,
+      "loss": 0.9268,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9386036463040939,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.0988183260692963e-07,
+      "loss": 0.9012,
+      "step": 2661
+    },
+    {
+      "epoch": 0.9389563722140164,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.0751184718118767e-07,
+      "loss": 0.8935,
+      "step": 2662
+    },
+    {
+      "epoch": 0.939309098123939,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.0515517848464284e-07,
+      "loss": 0.9399,
+      "step": 2663
+    },
+    {
+      "epoch": 0.9396618240338617,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.0281182972210934e-07,
+      "loss": 0.9741,
+      "step": 2664
+    },
+    {
+      "epoch": 0.9400145499437843,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.0048180408029026e-07,
+      "loss": 0.9454,
+      "step": 2665
+    },
+    {
+      "epoch": 0.9403672758537069,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9816510472776773e-07,
+      "loss": 0.9036,
+      "step": 2666
+    },
+    {
+      "epoch": 0.9407200017636296,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9586173481500602e-07,
+      "loss": 0.9744,
+      "step": 2667
+    },
+    {
+      "epoch": 0.9410727276735522,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9357169747433736e-07,
+      "loss": 0.9604,
+      "step": 2668
+    },
+    {
+      "epoch": 0.9414254535834747,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9129499581996945e-07,
+      "loss": 0.8836,
+      "step": 2669
+    },
+    {
+      "epoch": 0.9417781794933974,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.89031632947968e-07,
+      "loss": 0.8183,
+      "step": 2670
+    },
+    {
+      "epoch": 0.94213090540332,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.867816119362631e-07,
+      "loss": 0.9765,
+      "step": 2671
+    },
+    {
+      "epoch": 0.9424836313132426,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.8454493584464272e-07,
+      "loss": 0.9761,
+      "step": 2672
+    },
+    {
+      "epoch": 0.9428363572231653,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.8232160771474494e-07,
+      "loss": 0.9262,
+      "step": 2673
+    },
+    {
+      "epoch": 0.9431890831330879,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8011163057005566e-07,
+      "loss": 0.8921,
+      "step": 2674
+    },
+    {
+      "epoch": 0.9435418090430105,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.779150074159064e-07,
+      "loss": 0.9949,
+      "step": 2675
+    },
+    {
+      "epoch": 0.9438945349529332,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.7573174123946658e-07,
+      "loss": 0.9989,
+      "step": 2676
+    },
+    {
+      "epoch": 0.9442472608628557,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.7356183500974677e-07,
+      "loss": 0.9341,
+      "step": 2677
+    },
+    {
+      "epoch": 0.9445999867727783,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.7140529167758103e-07,
+      "loss": 0.9937,
+      "step": 2678
+    },
+    {
+      "epoch": 0.944952712682701,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.692621141756401e-07,
+      "loss": 0.9903,
+      "step": 2679
+    },
+    {
+      "epoch": 0.9453054385926236,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.6713230541841264e-07,
+      "loss": 1.0448,
+      "step": 2680
+    },
+    {
+      "epoch": 0.9456581645025463,
+      "grad_norm": 0.75,
+      "learning_rate": 1.6501586830220852e-07,
+      "loss": 0.8702,
+      "step": 2681
+    },
+    {
+      "epoch": 0.9460108904124689,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.6291280570515544e-07,
+      "loss": 0.9692,
+      "step": 2682
+    },
+    {
+      "epoch": 0.9463636163223915,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.608231204871924e-07,
+      "loss": 1.039,
+      "step": 2683
+    },
+    {
+      "epoch": 0.9467163422323142,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.58746815490064e-07,
+      "loss": 0.851,
+      "step": 2684
+    },
+    {
+      "epoch": 0.9470690681422367,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.5668389353732383e-07,
+      "loss": 1.0238,
+      "step": 2685
+    },
+    {
+      "epoch": 0.9474217940521593,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.5463435743432232e-07,
+      "loss": 1.1137,
+      "step": 2686
+    },
+    {
+      "epoch": 0.947774519962082,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5259820996820884e-07,
+      "loss": 0.9209,
+      "step": 2687
+    },
+    {
+      "epoch": 0.9481272458720046,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.5057545390792404e-07,
+      "loss": 0.9831,
+      "step": 2688
+    },
+    {
+      "epoch": 0.9484799717819272,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.4856609200419758e-07,
+      "loss": 0.8621,
+      "step": 2689
+    },
+    {
+      "epoch": 0.9488326976918499,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4657012698954587e-07,
+      "loss": 0.9328,
+      "step": 2690
+    },
+    {
+      "epoch": 0.9491854236017725,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4458756157826658e-07,
+      "loss": 1.069,
+      "step": 2691
+    },
+    {
+      "epoch": 0.949538149511695,
+      "grad_norm": 0.75,
+      "learning_rate": 1.426183984664342e-07,
+      "loss": 0.9421,
+      "step": 2692
+    },
+    {
+      "epoch": 0.9498908754216177,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.4066264033190002e-07,
+      "loss": 0.9482,
+      "step": 2693
+    },
+    {
+      "epoch": 0.9502436013315403,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.387202898342832e-07,
+      "loss": 0.8461,
+      "step": 2694
+    },
+    {
+      "epoch": 0.9505963272414629,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3679134961497199e-07,
+      "loss": 0.8931,
+      "step": 2695
+    },
+    {
+      "epoch": 0.9509490531513856,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3487582229711694e-07,
+      "loss": 1.0551,
+      "step": 2696
+    },
+    {
+      "epoch": 0.9513017790613082,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.3297371048562878e-07,
+      "loss": 1.0212,
+      "step": 2697
+    },
+    {
+      "epoch": 0.9516545049712308,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.3108501676717512e-07,
+      "loss": 0.8828,
+      "step": 2698
+    },
+    {
+      "epoch": 0.9520072308811535,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.292097437101747e-07,
+      "loss": 1.0192,
+      "step": 2699
+    },
+    {
+      "epoch": 0.952359956791076,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2734789386479985e-07,
+      "loss": 0.8549,
+      "step": 2700
+    },
+    {
+      "epoch": 0.9527126827009986,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.254994697629619e-07,
+      "loss": 0.9564,
+      "step": 2701
+    },
+    {
+      "epoch": 0.9530654086109213,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.2366447391832237e-07,
+      "loss": 0.8325,
+      "step": 2702
+    },
+    {
+      "epoch": 0.9534181345208439,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.2184290882627513e-07,
+      "loss": 0.9243,
+      "step": 2703
+    },
+    {
+      "epoch": 0.9537708604307665,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2003477696395315e-07,
+      "loss": 0.9374,
+      "step": 2704
+    },
+    {
+      "epoch": 0.9541235863406892,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.1824008079022286e-07,
+      "loss": 0.971,
+      "step": 2705
+    },
+    {
+      "epoch": 0.9544763122506118,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.1645882274567533e-07,
+      "loss": 0.8609,
+      "step": 2706
+    },
+    {
+      "epoch": 0.9548290381605343,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.1469100525263066e-07,
+      "loss": 0.9468,
+      "step": 2707
+    },
+    {
+      "epoch": 0.955181764070457,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.1293663071513138e-07,
+      "loss": 0.9587,
+      "step": 2708
+    },
+    {
+      "epoch": 0.9555344899803796,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.1119570151893467e-07,
+      "loss": 0.9078,
+      "step": 2709
+    },
+    {
+      "epoch": 0.9558872158903022,
+      "grad_norm": 0.75,
+      "learning_rate": 1.0946822003152113e-07,
+      "loss": 0.8958,
+      "step": 2710
+    },
+    {
+      "epoch": 0.9562399418002249,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.0775418860207498e-07,
+      "loss": 1.0371,
+      "step": 2711
+    },
+    {
+      "epoch": 0.9565926677101475,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.0605360956149835e-07,
+      "loss": 1.0139,
+      "step": 2712
+    },
+    {
+      "epoch": 0.9569453936200701,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.0436648522239245e-07,
+      "loss": 0.9656,
+      "step": 2713
+    },
+    {
+      "epoch": 0.9572981195299928,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.0269281787906759e-07,
+      "loss": 0.9615,
+      "step": 2714
+    },
+    {
+      "epoch": 0.9576508454399153,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.0103260980752761e-07,
+      "loss": 0.9732,
+      "step": 2715
+    },
+    {
+      "epoch": 0.9580035713498379,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.938586326548094e-08,
+      "loss": 0.9105,
+      "step": 2716
+    },
+    {
+      "epoch": 0.9583562972597606,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.775258049232072e-08,
+      "loss": 0.9941,
+      "step": 2717
+    },
+    {
+      "epoch": 0.9587090231696832,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.613276370914026e-08,
+      "loss": 0.8908,
+      "step": 2718
+    },
+    {
+      "epoch": 0.9590617490796058,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.452641511871196e-08,
+      "loss": 1.0087,
+      "step": 2719
+    },
+    {
+      "epoch": 0.9594144749895285,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.293353690550067e-08,
+      "loss": 0.883,
+      "step": 2720
+    },
+    {
+      "epoch": 0.9597672008994511,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.135413123564585e-08,
+      "loss": 0.9131,
+      "step": 2721
+    },
+    {
+      "epoch": 0.9601199268093736,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.978820025697055e-08,
+      "loss": 0.9455,
+      "step": 2722
+    },
+    {
+      "epoch": 0.9604726527192963,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.823574609897134e-08,
+      "loss": 0.8695,
+      "step": 2723
+    },
+    {
+      "epoch": 0.9608253786292189,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.669677087282059e-08,
+      "loss": 1.0351,
+      "step": 2724
+    },
+    {
+      "epoch": 0.9611781045391415,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.517127667135749e-08,
+      "loss": 0.8644,
+      "step": 2725
+    },
+    {
+      "epoch": 0.9615308304490642,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.365926556909043e-08,
+      "loss": 0.8321,
+      "step": 2726
+    },
+    {
+      "epoch": 0.9618835563589868,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.216073962219129e-08,
+      "loss": 0.9547,
+      "step": 2727
+    },
+    {
+      "epoch": 0.9622362822689094,
+      "grad_norm": 1.25,
+      "learning_rate": 8.067570086849442e-08,
+      "loss": 0.9104,
+      "step": 2728
+    },
+    {
+      "epoch": 0.9625890081788321,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.920415132748993e-08,
+      "loss": 0.946,
+      "step": 2729
+    },
+    {
+      "epoch": 0.9629417340887546,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.774609300033154e-08,
+      "loss": 0.9173,
+      "step": 2730
+    },
+    {
+      "epoch": 0.9632944599986772,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.63015278698176e-08,
+      "loss": 0.8927,
+      "step": 2731
+    },
+    {
+      "epoch": 0.9636471859085999,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.487045790040337e-08,
+      "loss": 0.9777,
+      "step": 2732
+    },
+    {
+      "epoch": 0.9639999118185225,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.345288503818771e-08,
+      "loss": 0.9372,
+      "step": 2733
+    },
+    {
+      "epoch": 0.9643526377284452,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.204881121091855e-08,
+      "loss": 0.8442,
+      "step": 2734
+    },
+    {
+      "epoch": 0.9647053636383678,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.065823832798524e-08,
+      "loss": 0.9702,
+      "step": 2735
+    },
+    {
+      "epoch": 0.9650580895482904,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.928116828041731e-08,
+      "loss": 0.8273,
+      "step": 2736
+    },
+    {
+      "epoch": 0.9654108154582131,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.791760294088012e-08,
+      "loss": 0.947,
+      "step": 2737
+    },
+    {
+      "epoch": 0.9657635413681356,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.656754416367705e-08,
+      "loss": 1.0153,
+      "step": 2738
+    },
+    {
+      "epoch": 0.9661162672780582,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.523099378474173e-08,
+      "loss": 0.9344,
+      "step": 2739
+    },
+    {
+      "epoch": 0.9664689931879809,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.390795362163915e-08,
+      "loss": 0.8576,
+      "step": 2740
+    },
+    {
+      "epoch": 0.9668217190979035,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.259842547356231e-08,
+      "loss": 0.7481,
+      "step": 2741
+    },
+    {
+      "epoch": 0.9671744450078261,
+      "grad_norm": 0.796875,
+      "learning_rate": 6.130241112132673e-08,
+      "loss": 0.9527,
+      "step": 2742
+    },
+    {
+      "epoch": 0.9675271709177488,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.00199123273737e-08,
+      "loss": 0.9166,
+      "step": 2743
+    },
+    {
+      "epoch": 0.9678798968276714,
+      "grad_norm": 2.25,
+      "learning_rate": 5.875093083576366e-08,
+      "loss": 1.0527,
+      "step": 2744
+    },
+    {
+      "epoch": 0.9682326227375939,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.74954683721729e-08,
+      "loss": 0.999,
+      "step": 2745
+    },
+    {
+      "epoch": 0.9685853486475166,
+      "grad_norm": 0.80078125,
+      "learning_rate": 5.625352664389683e-08,
+      "loss": 0.9765,
+      "step": 2746
+    },
+    {
+      "epoch": 0.9689380745574392,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.5025107339842234e-08,
+      "loss": 0.9173,
+      "step": 2747
+    },
+    {
+      "epoch": 0.9692908004673618,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.3810212130526174e-08,
+      "loss": 0.9623,
+      "step": 2748
+    },
+    {
+      "epoch": 0.9696435263772845,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.2608842668075976e-08,
+      "loss": 0.8944,
+      "step": 2749
+    },
+    {
+      "epoch": 0.9699962522872071,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.142100058622479e-08,
+      "loss": 0.9033,
+      "step": 2750
+    },
+    {
+      "epoch": 0.9703489781971297,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.0246687500310475e-08,
+      "loss": 0.9427,
+      "step": 2751
+    },
+    {
+      "epoch": 0.9707017041070524,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.9085905007270064e-08,
+      "loss": 0.9266,
+      "step": 2752
+    },
+    {
+      "epoch": 0.9710544300169749,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.79386546856464e-08,
+      "loss": 0.9336,
+      "step": 2753
+    },
+    {
+      "epoch": 0.9714071559268975,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.680493809557263e-08,
+      "loss": 0.9764,
+      "step": 2754
+    },
+    {
+      "epoch": 0.9717598818368202,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.568475677878326e-08,
+      "loss": 0.9176,
+      "step": 2755
+    },
+    {
+      "epoch": 0.9721126077467428,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.457811225860309e-08,
+      "loss": 0.956,
+      "step": 2756
+    },
+    {
+      "epoch": 0.9724653336566654,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.348500603995165e-08,
+      "loss": 0.8578,
+      "step": 2757
+    },
+    {
+      "epoch": 0.9728180595665881,
+      "grad_norm": 0.75,
+      "learning_rate": 4.240543960933319e-08,
+      "loss": 0.9266,
+      "step": 2758
+    },
+    {
+      "epoch": 0.9731707854765107,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.133941443484335e-08,
+      "loss": 0.945,
+      "step": 2759
+    },
+    {
+      "epoch": 0.9735235113864332,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.028693196616029e-08,
+      "loss": 0.9876,
+      "step": 2760
+    },
+    {
+      "epoch": 0.9738762372963559,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.924799363454801e-08,
+      "loss": 1.0118,
+      "step": 2761
+    },
+    {
+      "epoch": 0.9742289632062785,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.822260085284968e-08,
+      "loss": 0.9088,
+      "step": 2762
+    },
+    {
+      "epoch": 0.9745816891162011,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.721075501549098e-08,
+      "loss": 0.9893,
+      "step": 2763
+    },
+    {
+      "epoch": 0.9749344150261238,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.621245749847013e-08,
+      "loss": 0.9728,
+      "step": 2764
+    },
+    {
+      "epoch": 0.9752871409360464,
+      "grad_norm": 0.75,
+      "learning_rate": 3.5227709659367834e-08,
+      "loss": 0.9807,
+      "step": 2765
+    },
+    {
+      "epoch": 0.975639866845969,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.4256512837331777e-08,
+      "loss": 0.8587,
+      "step": 2766
+    },
+    {
+      "epoch": 0.9759925927558917,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.32988683530866e-08,
+      "loss": 0.981,
+      "step": 2767
+    },
+    {
+      "epoch": 0.9763453186658142,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.235477750892613e-08,
+      "loss": 0.9243,
+      "step": 2768
+    },
+    {
+      "epoch": 0.9766980445757368,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.142424158871227e-08,
+      "loss": 1.0283,
+      "step": 2769
+    },
+    {
+      "epoch": 0.9770507704856595,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.050726185787167e-08,
+      "loss": 0.9332,
+      "step": 2770
+    },
+    {
+      "epoch": 0.9774034963955821,
+      "grad_norm": 0.75,
+      "learning_rate": 2.9603839563400183e-08,
+      "loss": 0.9373,
+      "step": 2771
+    },
+    {
+      "epoch": 0.9777562223055047,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.8713975933853944e-08,
+      "loss": 0.9143,
+      "step": 2772
+    },
+    {
+      "epoch": 0.9781089482154274,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.7837672179351625e-08,
+      "loss": 0.8364,
+      "step": 2773
+    },
+    {
+      "epoch": 0.97846167412535,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.697492949157221e-08,
+      "loss": 0.8992,
+      "step": 2774
+    },
+    {
+      "epoch": 0.9788144000352725,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.6125749043752757e-08,
+      "loss": 0.922,
+      "step": 2775
+    },
+    {
+      "epoch": 0.9791671259451952,
+      "grad_norm": 0.75,
+      "learning_rate": 2.5290131990687305e-08,
+      "loss": 0.9773,
+      "step": 2776
+    },
+    {
+      "epoch": 0.9795198518551178,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.4468079468724647e-08,
+      "loss": 0.9814,
+      "step": 2777
+    },
+    {
+      "epoch": 0.9798725777650404,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.3659592595768333e-08,
+      "loss": 0.939,
+      "step": 2778
+    },
+    {
+      "epoch": 0.9802253036749631,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.286467247127222e-08,
+      "loss": 0.9617,
+      "step": 2779
+    },
+    {
+      "epoch": 0.9805780295848857,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.2083320176242707e-08,
+      "loss": 0.9929,
+      "step": 2780
+    },
+    {
+      "epoch": 0.9809307554948083,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.13155367732365e-08,
+      "loss": 0.9684,
+      "step": 2781
+    },
+    {
+      "epoch": 0.981283481404731,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.0561323306352853e-08,
+      "loss": 0.9055,
+      "step": 2782
+    },
+    {
+      "epoch": 0.9816362073146535,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9820680801243554e-08,
+      "loss": 1.0144,
+      "step": 2783
+    },
+    {
+      "epoch": 0.9819889332245761,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9093610265101814e-08,
+      "loss": 1.0117,
+      "step": 2784
+    },
+    {
+      "epoch": 0.9823416591344988,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8380112686665618e-08,
+      "loss": 1.0389,
+      "step": 2785
+    },
+    {
+      "epoch": 0.9826943850444214,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.768018903621438e-08,
+      "loss": 0.9488,
+      "step": 2786
+    },
+    {
+      "epoch": 0.983047110954344,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.6993840265568938e-08,
+      "loss": 0.8575,
+      "step": 2787
+    },
+    {
+      "epoch": 0.9833998368642667,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.6321067308089354e-08,
+      "loss": 0.9957,
+      "step": 2788
+    },
+    {
+      "epoch": 0.9837525627741893,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.5661871078674895e-08,
+      "loss": 0.8667,
+      "step": 2789
+    },
+    {
+      "epoch": 0.984105288684112,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.5016252473760706e-08,
+      "loss": 0.9261,
+      "step": 2790
+    },
+    {
+      "epoch": 0.9844580145940345,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4384212371320038e-08,
+      "loss": 0.9912,
+      "step": 2791
+    },
+    {
+      "epoch": 0.9848107405039571,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.3765751630858693e-08,
+      "loss": 0.9092,
+      "step": 2792
+    },
+    {
+      "epoch": 0.9851634664138798,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3160871093416127e-08,
+      "loss": 0.9351,
+      "step": 2793
+    },
+    {
+      "epoch": 0.9855161923238024,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2569571581564355e-08,
+      "loss": 0.9179,
+      "step": 2794
+    },
+    {
+      "epoch": 0.985868918233725,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1991853899409044e-08,
+      "loss": 1.0914,
+      "step": 2795
+    },
+    {
+      "epoch": 0.9862216441436477,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.1427718832581757e-08,
+      "loss": 0.8972,
+      "step": 2796
+    },
+    {
+      "epoch": 0.9865743700535703,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.0877167148246604e-08,
+      "loss": 1.0503,
+      "step": 2797
+    },
+    {
+      "epoch": 0.9869270959634928,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.0340199595093581e-08,
+      "loss": 0.9311,
+      "step": 2798
+    },
+    {
+      "epoch": 0.9872798218734155,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.816816903343018e-09,
+      "loss": 1.0582,
+      "step": 2799
+    },
+    {
+      "epoch": 0.9876325477833381,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.307019784736693e-09,
+      "loss": 0.8929,
+      "step": 2800
+    },
+    {
+      "epoch": 0.9879852736932607,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.81080893254449e-09,
+      "loss": 0.9421,
+      "step": 2801
+    },
+    {
+      "epoch": 0.9883379996031834,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.328185021559964e-09,
+      "loss": 1.0815,
+      "step": 2802
+    },
+    {
+      "epoch": 0.988690725513106,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.859148708099229e-09,
+      "loss": 0.9408,
+      "step": 2803
+    },
+    {
+      "epoch": 0.9890434514230286,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.403700630000954e-09,
+      "loss": 0.9924,
+      "step": 2804
+    },
+    {
+      "epoch": 0.9893961773329513,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.961841406626368e-09,
+      "loss": 0.8633,
+      "step": 2805
+    },
+    {
+      "epoch": 0.9897489032428738,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.533571638855929e-09,
+      "loss": 0.9447,
+      "step": 2806
+    },
+    {
+      "epoch": 0.9901016291527964,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.11889190909043e-09,
+      "loss": 1.031,
+      "step": 2807
+    },
+    {
+      "epoch": 0.9904543550627191,
+      "grad_norm": 0.81640625,
+      "learning_rate": 5.717802781249893e-09,
+      "loss": 0.9998,
+      "step": 2808
+    },
+    {
+      "epoch": 0.9908070809726417,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.330304800772457e-09,
+      "loss": 1.0197,
+      "step": 2809
+    },
+    {
+      "epoch": 0.9911598068825643,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.956398494613268e-09,
+      "loss": 0.937,
+      "step": 2810
+    },
+    {
+      "epoch": 0.991512532792487,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.59608437124337e-09,
+      "loss": 0.8803,
+      "step": 2811
+    },
+    {
+      "epoch": 0.9918652587024096,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.249362920654143e-09,
+      "loss": 1.0503,
+      "step": 2812
+    },
+    {
+      "epoch": 0.9922179846123321,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.916234614346204e-09,
+      "loss": 1.1253,
+      "step": 2813
+    },
+    {
+      "epoch": 0.9925707105222548,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.596699905340506e-09,
+      "loss": 0.9828,
+      "step": 2814
+    },
+    {
+      "epoch": 0.9929234364321774,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.29075922816946e-09,
+      "loss": 0.9711,
+      "step": 2815
+    },
+    {
+      "epoch": 0.9932761623421,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.998412998878042e-09,
+      "loss": 0.9604,
+      "step": 2816
+    },
+    {
+      "epoch": 0.9936288882520227,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.719661615027125e-09,
+      "loss": 0.9532,
+      "step": 2817
+    },
+    {
+      "epoch": 0.9939816141619453,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.4545054556868177e-09,
+      "loss": 0.8978,
+      "step": 2818
+    },
+    {
+      "epoch": 0.9943343400718679,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.2029448814431254e-09,
+      "loss": 0.8918,
+      "step": 2819
+    },
+    {
+      "epoch": 0.9946870659817906,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9649802343912892e-09,
+      "loss": 1.0232,
+      "step": 2820
+    },
+    {
+      "epoch": 0.9950397918917131,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7406118381346758e-09,
+      "loss": 0.9382,
+      "step": 2821
+    },
+    {
+      "epoch": 0.9953925178016357,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.5298399977936584e-09,
+      "loss": 0.9539,
+      "step": 2822
+    },
+    {
+      "epoch": 0.9957452437115584,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.3326649999934049e-09,
+      "loss": 1.0185,
+      "step": 2823
+    },
+    {
+      "epoch": 0.996097969621481,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.1490871128705394e-09,
+      "loss": 0.9304,
+      "step": 2824
+    },
+    {
+      "epoch": 0.9964506955314036,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.791065860720316e-10,
+      "loss": 0.9598,
+      "step": 2825
+    },
+    {
+      "epoch": 0.9968034214413263,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.227236507529768e-10,
+      "loss": 0.9346,
+      "step": 2826
+    },
+    {
+      "epoch": 0.9971561473512489,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.79938519576595e-10,
+      "loss": 1.0315,
+      "step": 2827
+    },
+    {
+      "epoch": 0.9975088732611714,
+      "grad_norm": 0.8046875,
+      "learning_rate": 5.507513867153425e-10,
+      "loss": 0.8747,
+      "step": 2828
+    },
+    {
+      "epoch": 0.9978615991710941,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.351624278509103e-10,
+      "loss": 0.962,
+      "step": 2829
+    },
+    {
+      "epoch": 0.9982143250810167,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.33171800170895e-10,
+      "loss": 0.8635,
+      "step": 2830
+    },
+    {
+      "epoch": 0.9985670509909393,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.4477964237212823e-10,
+      "loss": 0.9282,
+      "step": 2831
+    },
+    {
+      "epoch": 0.998919776900862,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6998607465734673e-10,
+      "loss": 0.9397,
+      "step": 2832
+    },
+    {
+      "epoch": 0.9992725028107846,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.0879119873852262e-10,
+      "loss": 0.9042,
+      "step": 2833
+    },
+    {
+      "epoch": 0.9996252287207072,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.119509783464316e-11,
+      "loss": 0.9092,
+      "step": 2834
+    },
+    {
+      "epoch": 0.9999779546306299,
+      "grad_norm": 0.75,
+      "learning_rate": 2.7197836670600497e-11,
+      "loss": 1.0714,
+      "step": 2835
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.90625,
+      "learning_rate": 6.799461479412018e-12,
+      "loss": 0.8105,
+      "step": 2836
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2836,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 10,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9.529617451893916e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}