diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/last-checkpoint/trainer_state.json"
@@ -0,0 +1,7052 @@
+{
+  "best_metric": 0.7326682209968567,
+  "best_model_checkpoint": "./w2v-bert-2.0-chichewa_34h/checkpoint-1000",
+  "epoch": 5.619718309859155,
+  "eval_steps": 1000,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.005633802816901409,
+      "grad_norm": 23.37450408935547,
+      "learning_rate": 3.0000000000000004e-09,
+      "loss": 10.0931,
+      "step": 1
+    },
+    {
+      "epoch": 0.011267605633802818,
+      "grad_norm": 23.055225372314453,
+      "learning_rate": 6.000000000000001e-09,
+      "loss": 9.7041,
+      "step": 2
+    },
+    {
+      "epoch": 0.016901408450704224,
+      "grad_norm": 21.591968536376953,
+      "learning_rate": 9e-09,
+      "loss": 9.2937,
+      "step": 3
+    },
+    {
+      "epoch": 0.022535211267605635,
+      "grad_norm": 19.100778579711914,
+      "learning_rate": 1.2000000000000002e-08,
+      "loss": 8.3783,
+      "step": 4
+    },
+    {
+      "epoch": 0.028169014084507043,
+      "grad_norm": 20.580049514770508,
+      "learning_rate": 1.5000000000000002e-08,
+      "loss": 8.7382,
+      "step": 5
+    },
+    {
+      "epoch": 0.03380281690140845,
+      "grad_norm": 19.053171157836914,
+      "learning_rate": 1.8e-08,
+      "loss": 8.1976,
+      "step": 6
+    },
+    {
+      "epoch": 0.03943661971830986,
+      "grad_norm": 20.025707244873047,
+      "learning_rate": 2.1e-08,
+      "loss": 8.5869,
+      "step": 7
+    },
+    {
+      "epoch": 0.04507042253521127,
+      "grad_norm": 19.60275650024414,
+      "learning_rate": 2.4000000000000003e-08,
+      "loss": 8.2925,
+      "step": 8
+    },
+    {
+      "epoch": 0.05070422535211268,
+      "grad_norm": 19.109025955200195,
+      "learning_rate": 2.7e-08,
+      "loss": 8.3882,
+      "step": 9
+    },
+    {
+      "epoch": 0.056338028169014086,
+      "grad_norm": 18.777395248413086,
+      "learning_rate": 3.0000000000000004e-08,
+      "loss": 8.0625,
+      "step": 10
+    },
+    {
+      "epoch": 0.061971830985915494,
+      "grad_norm": 19.636445999145508,
+      "learning_rate": 3.3000000000000004e-08,
+      "loss": 8.4087,
+      "step": 11
+    },
+    {
+      "epoch": 0.0676056338028169,
+      "grad_norm": 20.13718032836914,
+      "learning_rate": 3.6e-08,
+      "loss": 8.6273,
+      "step": 12
+    },
+    {
+      "epoch": 0.07323943661971831,
+      "grad_norm": 19.168004989624023,
+      "learning_rate": 3.9e-08,
+      "loss": 8.1722,
+      "step": 13
+    },
+    {
+      "epoch": 0.07887323943661972,
+      "grad_norm": 19.736047744750977,
+      "learning_rate": 4.2e-08,
+      "loss": 8.4533,
+      "step": 14
+    },
+    {
+      "epoch": 0.08450704225352113,
+      "grad_norm": 18.998886108398438,
+      "learning_rate": 4.5e-08,
+      "loss": 8.145,
+      "step": 15
+    },
+    {
+      "epoch": 0.09014084507042254,
+      "grad_norm": 19.929027557373047,
+      "learning_rate": 4.8000000000000006e-08,
+      "loss": 8.3833,
+      "step": 16
+    },
+    {
+      "epoch": 0.09577464788732394,
+      "grad_norm": 18.18242645263672,
+      "learning_rate": 5.1e-08,
+      "loss": 7.8769,
+      "step": 17
+    },
+    {
+      "epoch": 0.10140845070422536,
+      "grad_norm": 19.729808807373047,
+      "learning_rate": 5.4e-08,
+      "loss": 8.3034,
+      "step": 18
+    },
+    {
+      "epoch": 0.10704225352112676,
+      "grad_norm": 19.35500144958496,
+      "learning_rate": 5.7e-08,
+      "loss": 8.1648,
+      "step": 19
+    },
+    {
+      "epoch": 0.11267605633802817,
+      "grad_norm": 19.351375579833984,
+      "learning_rate": 6.000000000000001e-08,
+      "loss": 8.2531,
+      "step": 20
+    },
+    {
+      "epoch": 0.11830985915492957,
+      "grad_norm": 20.13310432434082,
+      "learning_rate": 6.3e-08,
+      "loss": 8.4955,
+      "step": 21
+    },
+    {
+      "epoch": 0.12394366197183099,
+      "grad_norm": 19.18810272216797,
+      "learning_rate": 6.600000000000001e-08,
+      "loss": 8.0951,
+      "step": 22
+    },
+    {
+      "epoch": 0.1295774647887324,
+      "grad_norm": 19.997060775756836,
+      "learning_rate": 6.9e-08,
+      "loss": 8.3085,
+      "step": 23
+    },
+    {
+      "epoch": 0.1352112676056338,
+      "grad_norm": 19.73467445373535,
+      "learning_rate": 7.2e-08,
+      "loss": 8.0012,
+      "step": 24
+    },
+    {
+      "epoch": 0.14084507042253522,
+      "grad_norm": 20.199962615966797,
+      "learning_rate": 7.500000000000001e-08,
+      "loss": 8.1924,
+      "step": 25
+    },
+    {
+      "epoch": 0.14647887323943662,
+      "grad_norm": 19.340503692626953,
+      "learning_rate": 7.8e-08,
+      "loss": 8.1655,
+      "step": 26
+    },
+    {
+      "epoch": 0.15211267605633802,
+      "grad_norm": 18.63260269165039,
+      "learning_rate": 8.100000000000001e-08,
+      "loss": 7.8248,
+      "step": 27
+    },
+    {
+      "epoch": 0.15774647887323945,
+      "grad_norm": 18.444799423217773,
+      "learning_rate": 8.4e-08,
+      "loss": 7.9568,
+      "step": 28
+    },
+    {
+      "epoch": 0.16338028169014085,
+      "grad_norm": 19.628917694091797,
+      "learning_rate": 8.7e-08,
+      "loss": 7.9813,
+      "step": 29
+    },
+    {
+      "epoch": 0.16901408450704225,
+      "grad_norm": 18.812955856323242,
+      "learning_rate": 9e-08,
+      "loss": 7.8513,
+      "step": 30
+    },
+    {
+      "epoch": 0.17464788732394365,
+      "grad_norm": 19.53274154663086,
+      "learning_rate": 9.3e-08,
+      "loss": 8.0772,
+      "step": 31
+    },
+    {
+      "epoch": 0.18028169014084508,
+      "grad_norm": 18.708633422851562,
+      "learning_rate": 9.600000000000001e-08,
+      "loss": 7.7278,
+      "step": 32
+    },
+    {
+      "epoch": 0.18591549295774648,
+      "grad_norm": 18.192834854125977,
+      "learning_rate": 9.9e-08,
+      "loss": 7.7268,
+      "step": 33
+    },
+    {
+      "epoch": 0.19154929577464788,
+      "grad_norm": 18.53217887878418,
+      "learning_rate": 1.02e-07,
+      "loss": 7.6227,
+      "step": 34
+    },
+    {
+      "epoch": 0.19718309859154928,
+      "grad_norm": 18.989917755126953,
+      "learning_rate": 1.05e-07,
+      "loss": 7.7894,
+      "step": 35
+    },
+    {
+      "epoch": 0.2028169014084507,
+      "grad_norm": 19.437061309814453,
+      "learning_rate": 1.08e-07,
+      "loss": 7.6945,
+      "step": 36
+    },
+    {
+      "epoch": 0.2084507042253521,
+      "grad_norm": 18.471364974975586,
+      "learning_rate": 1.11e-07,
+      "loss": 7.524,
+      "step": 37
+    },
+    {
+      "epoch": 0.2140845070422535,
+      "grad_norm": 19.86949920654297,
+      "learning_rate": 1.14e-07,
+      "loss": 7.8956,
+      "step": 38
+    },
+    {
+      "epoch": 0.21971830985915494,
+      "grad_norm": 19.169902801513672,
+      "learning_rate": 1.17e-07,
+      "loss": 7.6071,
+      "step": 39
+    },
+    {
+      "epoch": 0.22535211267605634,
+      "grad_norm": 18.673480987548828,
+      "learning_rate": 1.2000000000000002e-07,
+      "loss": 7.3223,
+      "step": 40
+    },
+    {
+      "epoch": 0.23098591549295774,
+      "grad_norm": 18.413576126098633,
+      "learning_rate": 1.23e-07,
+      "loss": 7.2614,
+      "step": 41
+    },
+    {
+      "epoch": 0.23661971830985915,
+      "grad_norm": 18.13873863220215,
+      "learning_rate": 1.26e-07,
+      "loss": 7.179,
+      "step": 42
+    },
+    {
+      "epoch": 0.24225352112676057,
+      "grad_norm": 18.680736541748047,
+      "learning_rate": 1.29e-07,
+      "loss": 7.36,
+      "step": 43
+    },
+    {
+      "epoch": 0.24788732394366197,
+      "grad_norm": 18.37202262878418,
+      "learning_rate": 1.3200000000000002e-07,
+      "loss": 7.1083,
+      "step": 44
+    },
+    {
+      "epoch": 0.2535211267605634,
+      "grad_norm": 28.554738998413086,
+      "learning_rate": 1.35e-07,
+      "loss": 10.9576,
+      "step": 45
+    },
+    {
+      "epoch": 0.2591549295774648,
+      "grad_norm": 23.22928810119629,
+      "learning_rate": 1.38e-07,
+      "loss": 9.4035,
+      "step": 46
+    },
+    {
+      "epoch": 0.2647887323943662,
+      "grad_norm": 22.669530868530273,
+      "learning_rate": 1.41e-07,
+      "loss": 9.0631,
+      "step": 47
+    },
+    {
+      "epoch": 0.2704225352112676,
+      "grad_norm": 21.29054069519043,
+      "learning_rate": 1.44e-07,
+      "loss": 8.5767,
+      "step": 48
+    },
+    {
+      "epoch": 0.27605633802816903,
+      "grad_norm": 21.267587661743164,
+      "learning_rate": 1.47e-07,
+      "loss": 8.5628,
+      "step": 49
+    },
+    {
+      "epoch": 0.28169014084507044,
+      "grad_norm": 20.70184898376465,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 8.4956,
+      "step": 50
+    },
+    {
+      "epoch": 0.28732394366197184,
+      "grad_norm": 20.579097747802734,
+      "learning_rate": 1.53e-07,
+      "loss": 8.4322,
+      "step": 51
+    },
+    {
+      "epoch": 0.29295774647887324,
+      "grad_norm": 20.76494026184082,
+      "learning_rate": 1.56e-07,
+      "loss": 8.3483,
+      "step": 52
+    },
+    {
+      "epoch": 0.29859154929577464,
+      "grad_norm": 22.010337829589844,
+      "learning_rate": 1.59e-07,
+      "loss": 8.6267,
+      "step": 53
+    },
+    {
+      "epoch": 0.30422535211267604,
+      "grad_norm": 20.0366268157959,
+      "learning_rate": 1.6200000000000002e-07,
+      "loss": 8.0702,
+      "step": 54
+    },
+    {
+      "epoch": 0.30985915492957744,
+      "grad_norm": 20.99293327331543,
+      "learning_rate": 1.6499999999999998e-07,
+      "loss": 8.232,
+      "step": 55
+    },
+    {
+      "epoch": 0.3154929577464789,
+      "grad_norm": 19.92287254333496,
+      "learning_rate": 1.68e-07,
+      "loss": 8.09,
+      "step": 56
+    },
+    {
+      "epoch": 0.3211267605633803,
+      "grad_norm": 20.472942352294922,
+      "learning_rate": 1.71e-07,
+      "loss": 8.0906,
+      "step": 57
+    },
+    {
+      "epoch": 0.3267605633802817,
+      "grad_norm": 21.892868041992188,
+      "learning_rate": 1.74e-07,
+      "loss": 8.4593,
+      "step": 58
+    },
+    {
+      "epoch": 0.3323943661971831,
+      "grad_norm": 21.296226501464844,
+      "learning_rate": 1.77e-07,
+      "loss": 8.2587,
+      "step": 59
+    },
+    {
+      "epoch": 0.3380281690140845,
+      "grad_norm": 21.45792579650879,
+      "learning_rate": 1.8e-07,
+      "loss": 8.2867,
+      "step": 60
+    },
+    {
+      "epoch": 0.3436619718309859,
+      "grad_norm": 20.300464630126953,
+      "learning_rate": 1.83e-07,
+      "loss": 7.7559,
+      "step": 61
+    },
+    {
+      "epoch": 0.3492957746478873,
+      "grad_norm": 21.698631286621094,
+      "learning_rate": 1.86e-07,
+      "loss": 8.199,
+      "step": 62
+    },
+    {
+      "epoch": 0.35492957746478876,
+      "grad_norm": 19.86581039428711,
+      "learning_rate": 1.89e-07,
+      "loss": 7.6167,
+      "step": 63
+    },
+    {
+      "epoch": 0.36056338028169016,
+      "grad_norm": 21.967016220092773,
+      "learning_rate": 1.9200000000000003e-07,
+      "loss": 8.1939,
+      "step": 64
+    },
+    {
+      "epoch": 0.36619718309859156,
+      "grad_norm": 21.37307357788086,
+      "learning_rate": 1.9499999999999999e-07,
+      "loss": 8.1185,
+      "step": 65
+    },
+    {
+      "epoch": 0.37183098591549296,
+      "grad_norm": 20.290283203125,
+      "learning_rate": 1.98e-07,
+      "loss": 7.8508,
+      "step": 66
+    },
+    {
+      "epoch": 0.37746478873239436,
+      "grad_norm": 21.446462631225586,
+      "learning_rate": 2.01e-07,
+      "loss": 7.8273,
+      "step": 67
+    },
+    {
+      "epoch": 0.38309859154929576,
+      "grad_norm": 21.343917846679688,
+      "learning_rate": 2.04e-07,
+      "loss": 7.8103,
+      "step": 68
+    },
+    {
+      "epoch": 0.38873239436619716,
+      "grad_norm": 20.38936996459961,
+      "learning_rate": 2.0700000000000001e-07,
+      "loss": 7.5509,
+      "step": 69
+    },
+    {
+      "epoch": 0.39436619718309857,
+      "grad_norm": 21.852344512939453,
+      "learning_rate": 2.1e-07,
+      "loss": 7.975,
+      "step": 70
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 21.029077529907227,
+      "learning_rate": 2.1300000000000001e-07,
+      "loss": 7.7329,
+      "step": 71
+    },
+    {
+      "epoch": 0.4056338028169014,
+      "grad_norm": 21.277189254760742,
+      "learning_rate": 2.16e-07,
+      "loss": 7.697,
+      "step": 72
+    },
+    {
+      "epoch": 0.4112676056338028,
+      "grad_norm": 20.803091049194336,
+      "learning_rate": 2.1900000000000002e-07,
+      "loss": 7.6135,
+      "step": 73
+    },
+    {
+      "epoch": 0.4169014084507042,
+      "grad_norm": 19.920984268188477,
+      "learning_rate": 2.22e-07,
+      "loss": 7.3491,
+      "step": 74
+    },
+    {
+      "epoch": 0.4225352112676056,
+      "grad_norm": 20.588159561157227,
+      "learning_rate": 2.25e-07,
+      "loss": 7.3246,
+      "step": 75
+    },
+    {
+      "epoch": 0.428169014084507,
+      "grad_norm": 20.864524841308594,
+      "learning_rate": 2.28e-07,
+      "loss": 7.4887,
+      "step": 76
+    },
+    {
+      "epoch": 0.43380281690140843,
+      "grad_norm": 20.144886016845703,
+      "learning_rate": 2.3100000000000002e-07,
+      "loss": 7.2445,
+      "step": 77
+    },
+    {
+      "epoch": 0.4394366197183099,
+      "grad_norm": 20.628171920776367,
+      "learning_rate": 2.34e-07,
+      "loss": 7.1465,
+      "step": 78
+    },
+    {
+      "epoch": 0.4450704225352113,
+      "grad_norm": 21.245912551879883,
+      "learning_rate": 2.3700000000000002e-07,
+      "loss": 7.4386,
+      "step": 79
+    },
+    {
+      "epoch": 0.4507042253521127,
+      "grad_norm": 22.11734390258789,
+      "learning_rate": 2.4000000000000003e-07,
+      "loss": 7.3555,
+      "step": 80
+    },
+    {
+      "epoch": 0.4563380281690141,
+      "grad_norm": 21.49259376525879,
+      "learning_rate": 2.43e-07,
+      "loss": 7.2534,
+      "step": 81
+    },
+    {
+      "epoch": 0.4619718309859155,
+      "grad_norm": 21.531265258789062,
+      "learning_rate": 2.46e-07,
+      "loss": 7.2902,
+      "step": 82
+    },
+    {
+      "epoch": 0.4676056338028169,
+      "grad_norm": 23.177501678466797,
+      "learning_rate": 2.49e-07,
+      "loss": 7.5368,
+      "step": 83
+    },
+    {
+      "epoch": 0.4732394366197183,
+      "grad_norm": 21.143033981323242,
+      "learning_rate": 2.52e-07,
+      "loss": 7.0782,
+      "step": 84
+    },
+    {
+      "epoch": 0.4788732394366197,
+      "grad_norm": 21.302194595336914,
+      "learning_rate": 2.5500000000000005e-07,
+      "loss": 6.9088,
+      "step": 85
+    },
+    {
+      "epoch": 0.48450704225352115,
+      "grad_norm": 20.92340660095215,
+      "learning_rate": 2.58e-07,
+      "loss": 6.8607,
+      "step": 86
+    },
+    {
+      "epoch": 0.49014084507042255,
+      "grad_norm": 21.352828979492188,
+      "learning_rate": 2.6099999999999997e-07,
+      "loss": 6.8558,
+      "step": 87
+    },
+    {
+      "epoch": 0.49577464788732395,
+      "grad_norm": 20.361351013183594,
+      "learning_rate": 2.6400000000000003e-07,
+      "loss": 6.574,
+      "step": 88
+    },
+    {
+      "epoch": 0.5014084507042254,
+      "grad_norm": 32.090736389160156,
+      "learning_rate": 2.67e-07,
+      "loss": 9.4598,
+      "step": 89
+    },
+    {
+      "epoch": 0.5070422535211268,
+      "grad_norm": 29.89952850341797,
+      "learning_rate": 2.7e-07,
+      "loss": 8.9964,
+      "step": 90
+    },
+    {
+      "epoch": 0.5126760563380282,
+      "grad_norm": 26.13751983642578,
+      "learning_rate": 2.73e-07,
+      "loss": 8.2028,
+      "step": 91
+    },
+    {
+      "epoch": 0.5183098591549296,
+      "grad_norm": 24.984628677368164,
+      "learning_rate": 2.76e-07,
+      "loss": 7.9064,
+      "step": 92
+    },
+    {
+      "epoch": 0.523943661971831,
+      "grad_norm": 26.014310836791992,
+      "learning_rate": 2.79e-07,
+      "loss": 7.9926,
+      "step": 93
+    },
+    {
+      "epoch": 0.5295774647887324,
+      "grad_norm": 24.743000030517578,
+      "learning_rate": 2.82e-07,
+      "loss": 7.7252,
+      "step": 94
+    },
+    {
+      "epoch": 0.5352112676056338,
+      "grad_norm": 26.368148803710938,
+      "learning_rate": 2.85e-07,
+      "loss": 7.9539,
+      "step": 95
+    },
+    {
+      "epoch": 0.5408450704225352,
+      "grad_norm": 27.547103881835938,
+      "learning_rate": 2.88e-07,
+      "loss": 8.0634,
+      "step": 96
+    },
+    {
+      "epoch": 0.5464788732394367,
+      "grad_norm": 25.676986694335938,
+      "learning_rate": 2.91e-07,
+      "loss": 7.727,
+      "step": 97
+    },
+    {
+      "epoch": 0.5521126760563381,
+      "grad_norm": 26.837203979492188,
+      "learning_rate": 2.94e-07,
+      "loss": 7.8317,
+      "step": 98
+    },
+    {
+      "epoch": 0.5577464788732395,
+      "grad_norm": 26.280786514282227,
+      "learning_rate": 2.97e-07,
+      "loss": 7.6846,
+      "step": 99
+    },
+    {
+      "epoch": 0.5633802816901409,
+      "grad_norm": 27.732624053955078,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 7.79,
+      "step": 100
+    },
+    {
+      "epoch": 0.5690140845070423,
+      "grad_norm": 27.480138778686523,
+      "learning_rate": 3.03e-07,
+      "loss": 7.6279,
+      "step": 101
+    },
+    {
+      "epoch": 0.5746478873239437,
+      "grad_norm": 25.806196212768555,
+      "learning_rate": 3.06e-07,
+      "loss": 7.3647,
+      "step": 102
+    },
+    {
+      "epoch": 0.5802816901408451,
+      "grad_norm": 27.241296768188477,
+      "learning_rate": 3.0900000000000003e-07,
+      "loss": 7.6385,
+      "step": 103
+    },
+    {
+      "epoch": 0.5859154929577465,
+      "grad_norm": 27.93297004699707,
+      "learning_rate": 3.12e-07,
+      "loss": 7.6403,
+      "step": 104
+    },
+    {
+      "epoch": 0.5915492957746479,
+      "grad_norm": 28.791919708251953,
+      "learning_rate": 3.15e-07,
+      "loss": 7.6192,
+      "step": 105
+    },
+    {
+      "epoch": 0.5971830985915493,
+      "grad_norm": 28.215551376342773,
+      "learning_rate": 3.18e-07,
+      "loss": 7.4607,
+      "step": 106
+    },
+    {
+      "epoch": 0.6028169014084507,
+      "grad_norm": 29.93889617919922,
+      "learning_rate": 3.21e-07,
+      "loss": 7.6467,
+      "step": 107
+    },
+    {
+      "epoch": 0.6084507042253521,
+      "grad_norm": 27.232234954833984,
+      "learning_rate": 3.2400000000000004e-07,
+      "loss": 7.1521,
+      "step": 108
+    },
+    {
+      "epoch": 0.6140845070422535,
+      "grad_norm": 28.14197540283203,
+      "learning_rate": 3.27e-07,
+      "loss": 7.2471,
+      "step": 109
+    },
+    {
+      "epoch": 0.6197183098591549,
+      "grad_norm": 27.78325653076172,
+      "learning_rate": 3.2999999999999996e-07,
+      "loss": 7.153,
+      "step": 110
+    },
+    {
+      "epoch": 0.6253521126760564,
+      "grad_norm": 27.265422821044922,
+      "learning_rate": 3.3300000000000003e-07,
+      "loss": 7.0405,
+      "step": 111
+    },
+    {
+      "epoch": 0.6309859154929578,
+      "grad_norm": 29.104612350463867,
+      "learning_rate": 3.36e-07,
+      "loss": 7.294,
+      "step": 112
+    },
+    {
+      "epoch": 0.6366197183098592,
+      "grad_norm": 28.48741912841797,
+      "learning_rate": 3.39e-07,
+      "loss": 7.0529,
+      "step": 113
+    },
+    {
+      "epoch": 0.6422535211267606,
+      "grad_norm": 29.633562088012695,
+      "learning_rate": 3.42e-07,
+      "loss": 7.0792,
+      "step": 114
+    },
+    {
+      "epoch": 0.647887323943662,
+      "grad_norm": 29.809619903564453,
+      "learning_rate": 3.45e-07,
+      "loss": 7.1284,
+      "step": 115
+    },
+    {
+      "epoch": 0.6535211267605634,
+      "grad_norm": 29.871963500976562,
+      "learning_rate": 3.48e-07,
+      "loss": 7.0132,
+      "step": 116
+    },
+    {
+      "epoch": 0.6591549295774648,
+      "grad_norm": 29.897907257080078,
+      "learning_rate": 3.51e-07,
+      "loss": 6.9608,
+      "step": 117
+    },
+    {
+      "epoch": 0.6647887323943662,
+      "grad_norm": 30.40986442565918,
+      "learning_rate": 3.54e-07,
+      "loss": 7.0252,
+      "step": 118
+    },
+    {
+      "epoch": 0.6704225352112676,
+      "grad_norm": 27.947280883789062,
+      "learning_rate": 3.5700000000000003e-07,
+      "loss": 6.595,
+      "step": 119
+    },
+    {
+      "epoch": 0.676056338028169,
+      "grad_norm": 27.550397872924805,
+      "learning_rate": 3.6e-07,
+      "loss": 6.506,
+      "step": 120
+    },
+    {
+      "epoch": 0.6816901408450704,
+      "grad_norm": 29.49580192565918,
+      "learning_rate": 3.63e-07,
+      "loss": 6.6519,
+      "step": 121
+    },
+    {
+      "epoch": 0.6873239436619718,
+      "grad_norm": 26.66489028930664,
+      "learning_rate": 3.66e-07,
+      "loss": 6.2367,
+      "step": 122
+    },
+    {
+      "epoch": 0.6929577464788732,
+      "grad_norm": 27.495731353759766,
+      "learning_rate": 3.6900000000000004e-07,
+      "loss": 6.3315,
+      "step": 123
+    },
+    {
+      "epoch": 0.6985915492957746,
+      "grad_norm": 28.700468063354492,
+      "learning_rate": 3.72e-07,
+      "loss": 6.445,
+      "step": 124
+    },
+    {
+      "epoch": 0.704225352112676,
+      "grad_norm": 29.51987075805664,
+      "learning_rate": 3.75e-07,
+      "loss": 6.4283,
+      "step": 125
+    },
+    {
+      "epoch": 0.7098591549295775,
+      "grad_norm": 28.541399002075195,
+      "learning_rate": 3.78e-07,
+      "loss": 6.2563,
+      "step": 126
+    },
+    {
+      "epoch": 0.7154929577464789,
+      "grad_norm": 30.63127899169922,
+      "learning_rate": 3.81e-07,
+      "loss": 6.4155,
+      "step": 127
+    },
+    {
+      "epoch": 0.7211267605633803,
+      "grad_norm": 28.95149803161621,
+      "learning_rate": 3.8400000000000005e-07,
+      "loss": 6.2745,
+      "step": 128
+    },
+    {
+      "epoch": 0.7267605633802817,
+      "grad_norm": 28.605030059814453,
+      "learning_rate": 3.87e-07,
+      "loss": 6.2393,
+      "step": 129
+    },
+    {
+      "epoch": 0.7323943661971831,
+      "grad_norm": 27.956668853759766,
+      "learning_rate": 3.8999999999999997e-07,
+      "loss": 6.0958,
+      "step": 130
+    },
+    {
+      "epoch": 0.7380281690140845,
+      "grad_norm": 28.11646270751953,
+      "learning_rate": 3.9300000000000004e-07,
+      "loss": 5.963,
+      "step": 131
+    },
+    {
+      "epoch": 0.7436619718309859,
+      "grad_norm": 24.97135353088379,
+      "learning_rate": 3.96e-07,
+      "loss": 5.6173,
+      "step": 132
+    },
+    {
+      "epoch": 0.7492957746478873,
+      "grad_norm": Infinity,
+      "learning_rate": 3.96e-07,
+      "loss": 8.1969,
+      "step": 133
+    },
+    {
+      "epoch": 0.7549295774647887,
+      "grad_norm": 42.452293395996094,
+      "learning_rate": 3.99e-07,
+      "loss": 7.3189,
+      "step": 134
+    },
+    {
+      "epoch": 0.7605633802816901,
+      "grad_norm": 39.87089157104492,
+      "learning_rate": 4.02e-07,
+      "loss": 6.9278,
+      "step": 135
+    },
+    {
+      "epoch": 0.7661971830985915,
+      "grad_norm": 36.937870025634766,
+      "learning_rate": 4.05e-07,
+      "loss": 6.6801,
+      "step": 136
+    },
+    {
+      "epoch": 0.7718309859154929,
+      "grad_norm": 36.147464752197266,
+      "learning_rate": 4.08e-07,
+      "loss": 6.6071,
+      "step": 137
+    },
+    {
+      "epoch": 0.7774647887323943,
+      "grad_norm": 39.3355598449707,
+      "learning_rate": 4.11e-07,
+      "loss": 6.7304,
+      "step": 138
+    },
+    {
+      "epoch": 0.7830985915492957,
+      "grad_norm": 34.039981842041016,
+      "learning_rate": 4.1400000000000003e-07,
+      "loss": 6.2375,
+      "step": 139
+    },
+    {
+      "epoch": 0.7887323943661971,
+      "grad_norm": 36.65914535522461,
+      "learning_rate": 4.17e-07,
+      "loss": 6.4087,
+      "step": 140
+    },
+    {
+      "epoch": 0.7943661971830986,
+      "grad_norm": 37.91633605957031,
+      "learning_rate": 4.2e-07,
+      "loss": 6.4484,
+      "step": 141
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 39.105125427246094,
+      "learning_rate": 4.23e-07,
+      "loss": 6.4929,
+      "step": 142
+    },
+    {
+      "epoch": 0.8056338028169014,
+      "grad_norm": 40.3162727355957,
+      "learning_rate": 4.2600000000000003e-07,
+      "loss": 6.5174,
+      "step": 143
+    },
+    {
+      "epoch": 0.8112676056338028,
+      "grad_norm": 35.70503234863281,
+      "learning_rate": 4.2900000000000004e-07,
+      "loss": 6.1367,
+      "step": 144
+    },
+    {
+      "epoch": 0.8169014084507042,
+      "grad_norm": 36.86891174316406,
+      "learning_rate": 4.32e-07,
+      "loss": 6.1179,
+      "step": 145
+    },
+    {
+      "epoch": 0.8225352112676056,
+      "grad_norm": 39.80533218383789,
+      "learning_rate": 4.35e-07,
+      "loss": 6.2952,
+      "step": 146
+    },
+    {
+      "epoch": 0.828169014084507,
+      "grad_norm": 36.694480895996094,
+      "learning_rate": 4.3800000000000003e-07,
+      "loss": 5.9916,
+      "step": 147
+    },
+    {
+      "epoch": 0.8338028169014085,
+      "grad_norm": 38.44818115234375,
+      "learning_rate": 4.41e-07,
+      "loss": 6.0445,
+      "step": 148
+    },
+    {
+      "epoch": 0.8394366197183099,
+      "grad_norm": 38.265995025634766,
+      "learning_rate": 4.44e-07,
+      "loss": 6.0292,
+      "step": 149
+    },
+    {
+      "epoch": 0.8450704225352113,
+      "grad_norm": 36.704952239990234,
+      "learning_rate": 4.47e-07,
+      "loss": 5.8879,
+      "step": 150
+    },
+    {
+      "epoch": 0.8507042253521127,
+      "grad_norm": 37.268917083740234,
+      "learning_rate": 4.5e-07,
+      "loss": 5.8587,
+      "step": 151
+    },
+    {
+      "epoch": 0.856338028169014,
+      "grad_norm": 35.31634521484375,
+      "learning_rate": 4.5300000000000005e-07,
+      "loss": 5.7078,
+      "step": 152
+    },
+    {
+      "epoch": 0.8619718309859155,
+      "grad_norm": 37.19781494140625,
+      "learning_rate": 4.56e-07,
+      "loss": 5.7555,
+      "step": 153
+    },
+    {
+      "epoch": 0.8676056338028169,
+      "grad_norm": 34.77775955200195,
+      "learning_rate": 4.5899999999999997e-07,
+      "loss": 5.6057,
+      "step": 154
+    },
+    {
+      "epoch": 0.8732394366197183,
+      "grad_norm": 33.57127380371094,
+      "learning_rate": 4.6200000000000003e-07,
+      "loss": 5.4405,
+      "step": 155
+    },
+    {
+      "epoch": 0.8788732394366198,
+      "grad_norm": 36.10496139526367,
+      "learning_rate": 4.65e-07,
+      "loss": 5.4799,
+      "step": 156
+    },
+    {
+      "epoch": 0.8845070422535212,
+      "grad_norm": 36.44635009765625,
+      "learning_rate": 4.68e-07,
+      "loss": 5.4978,
+      "step": 157
+    },
+    {
+      "epoch": 0.8901408450704226,
+      "grad_norm": 34.81025314331055,
+      "learning_rate": 4.7099999999999997e-07,
+      "loss": 5.351,
+      "step": 158
+    },
+    {
+      "epoch": 0.895774647887324,
+      "grad_norm": 34.08943557739258,
+      "learning_rate": 4.7400000000000004e-07,
+      "loss": 5.2605,
+      "step": 159
+    },
+    {
+      "epoch": 0.9014084507042254,
+      "grad_norm": 34.23727798461914,
+      "learning_rate": 4.77e-07,
+      "loss": 5.2195,
+      "step": 160
+    },
+    {
+      "epoch": 0.9070422535211268,
+      "grad_norm": 32.607879638671875,
+      "learning_rate": 4.800000000000001e-07,
+      "loss": 5.1298,
+      "step": 161
+    },
+    {
+      "epoch": 0.9126760563380282,
+      "grad_norm": 33.94306564331055,
+      "learning_rate": 4.83e-07,
+      "loss": 5.1769,
+      "step": 162
+    },
+    {
+      "epoch": 0.9183098591549296,
+      "grad_norm": 31.63994026184082,
+      "learning_rate": 4.86e-07,
+      "loss": 5.024,
+      "step": 163
+    },
+    {
+      "epoch": 0.923943661971831,
+      "grad_norm": 33.26905059814453,
+      "learning_rate": 4.89e-07,
+      "loss": 4.9806,
+      "step": 164
+    },
+    {
+      "epoch": 0.9295774647887324,
+      "grad_norm": 32.36957931518555,
+      "learning_rate": 4.92e-07,
+      "loss": 4.9891,
+      "step": 165
+    },
+    {
+      "epoch": 0.9352112676056338,
+      "grad_norm": 28.738908767700195,
+      "learning_rate": 4.95e-07,
+      "loss": 4.7682,
+      "step": 166
+    },
+    {
+      "epoch": 0.9408450704225352,
+      "grad_norm": 28.542316436767578,
+      "learning_rate": 4.98e-07,
+      "loss": 4.8045,
+      "step": 167
+    },
+    {
+      "epoch": 0.9464788732394366,
+      "grad_norm": 25.792383193969727,
+      "learning_rate": 5.01e-07,
+      "loss": 4.6496,
+      "step": 168
+    },
+    {
+      "epoch": 0.952112676056338,
+      "grad_norm": 27.16576385498047,
+      "learning_rate": 5.04e-07,
+      "loss": 4.6519,
+      "step": 169
+    },
+    {
+      "epoch": 0.9577464788732394,
+      "grad_norm": 25.441028594970703,
+      "learning_rate": 5.07e-07,
+      "loss": 4.5866,
+      "step": 170
+    },
+    {
+      "epoch": 0.9633802816901409,
+      "grad_norm": 25.167131423950195,
+      "learning_rate": 5.100000000000001e-07,
+      "loss": 4.6062,
+      "step": 171
+    },
+    {
+      "epoch": 0.9690140845070423,
+      "grad_norm": 23.295936584472656,
+      "learning_rate": 5.13e-07,
+      "loss": 4.5423,
+      "step": 172
+    },
+    {
+      "epoch": 0.9746478873239437,
+      "grad_norm": 21.83075523376465,
+      "learning_rate": 5.16e-07,
+      "loss": 4.4349,
+      "step": 173
+    },
+    {
+      "epoch": 0.9802816901408451,
+      "grad_norm": 18.890296936035156,
+      "learning_rate": 5.19e-07,
+      "loss": 4.3618,
+      "step": 174
+    },
+    {
+      "epoch": 0.9859154929577465,
+      "grad_norm": 17.86733627319336,
+      "learning_rate": 5.219999999999999e-07,
+      "loss": 4.3321,
+      "step": 175
+    },
+    {
+      "epoch": 0.9915492957746479,
+      "grad_norm": 15.089102745056152,
+      "learning_rate": 5.250000000000001e-07,
+      "loss": 4.2655,
+      "step": 176
+    },
+    {
+      "epoch": 0.9971830985915493,
+      "grad_norm": 20.50058937072754,
+      "learning_rate": 5.280000000000001e-07,
+      "loss": 4.4262,
+      "step": 177
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.529200553894043,
+      "learning_rate": 5.31e-07,
+      "loss": 2.1518,
+      "step": 178
+    },
+    {
+      "epoch": 1.0056338028169014,
+      "grad_norm": 28.378694534301758,
+      "learning_rate": 5.34e-07,
+      "loss": 4.7345,
+      "step": 179
+    },
+    {
+      "epoch": 1.0112676056338028,
+      "grad_norm": 24.153104782104492,
+      "learning_rate": 5.37e-07,
+      "loss": 4.617,
+      "step": 180
+    },
+    {
+      "epoch": 1.0169014084507042,
+      "grad_norm": 16.407052993774414,
+      "learning_rate": 5.4e-07,
+      "loss": 4.3972,
+      "step": 181
+    },
+    {
+      "epoch": 1.0225352112676056,
+      "grad_norm": 13.738978385925293,
+      "learning_rate": 5.43e-07,
+      "loss": 4.3588,
+      "step": 182
+    },
+    {
+      "epoch": 1.028169014084507,
+      "grad_norm": 11.753392219543457,
+      "learning_rate": 5.46e-07,
+      "loss": 4.3117,
+      "step": 183
+    },
+    {
+      "epoch": 1.0338028169014084,
+      "grad_norm": 10.56106185913086,
+      "learning_rate": 5.490000000000001e-07,
+      "loss": 4.2987,
+      "step": 184
+    },
+    {
+      "epoch": 1.0394366197183098,
+      "grad_norm": 8.984735488891602,
+      "learning_rate": 5.52e-07,
+      "loss": 4.2585,
+      "step": 185
+    },
+    {
+      "epoch": 1.0450704225352112,
+      "grad_norm": 7.201539516448975,
+      "learning_rate": 5.55e-07,
+      "loss": 4.2362,
+      "step": 186
+    },
+    {
+      "epoch": 1.0507042253521126,
+      "grad_norm": 6.696246147155762,
+      "learning_rate": 5.58e-07,
+      "loss": 4.2512,
+      "step": 187
+    },
+    {
+      "epoch": 1.056338028169014,
+      "grad_norm": 7.102692604064941,
+      "learning_rate": 5.61e-07,
+      "loss": 4.2358,
+      "step": 188
+    },
+    {
+      "epoch": 1.0619718309859154,
+      "grad_norm": 7.0099616050720215,
+      "learning_rate": 5.64e-07,
+      "loss": 4.252,
+      "step": 189
+    },
+    {
+      "epoch": 1.0676056338028168,
+      "grad_norm": 6.835728645324707,
+      "learning_rate": 5.67e-07,
+      "loss": 4.2108,
+      "step": 190
+    },
+    {
+      "epoch": 1.0732394366197182,
+      "grad_norm": 7.586347579956055,
+      "learning_rate": 5.7e-07,
+      "loss": 4.1935,
+      "step": 191
+    },
+    {
+      "epoch": 1.0788732394366196,
+      "grad_norm": 7.363047122955322,
+      "learning_rate": 5.73e-07,
+      "loss": 4.1578,
+      "step": 192
+    },
+    {
+      "epoch": 1.084507042253521,
+      "grad_norm": 7.040609836578369,
+      "learning_rate": 5.76e-07,
+      "loss": 4.136,
+      "step": 193
+    },
+    {
+      "epoch": 1.0901408450704226,
+      "grad_norm": 6.436159610748291,
+      "learning_rate": 5.790000000000001e-07,
+      "loss": 4.1612,
+      "step": 194
+    },
+    {
+      "epoch": 1.095774647887324,
+      "grad_norm": 7.356907367706299,
+      "learning_rate": 5.82e-07,
+      "loss": 4.096,
+      "step": 195
+    },
+    {
+      "epoch": 1.1014084507042254,
+      "grad_norm": 6.898291110992432,
+      "learning_rate": 5.85e-07,
+      "loss": 4.0987,
+      "step": 196
+    },
+    {
+      "epoch": 1.1070422535211268,
+      "grad_norm": 6.676201343536377,
+      "learning_rate": 5.88e-07,
+      "loss": 4.0404,
+      "step": 197
+    },
+    {
+      "epoch": 1.1126760563380282,
+      "grad_norm": 7.3079352378845215,
+      "learning_rate": 5.909999999999999e-07,
+      "loss": 4.0107,
+      "step": 198
+    },
+    {
+      "epoch": 1.1183098591549296,
+      "grad_norm": 6.069024562835693,
+      "learning_rate": 5.94e-07,
+      "loss": 4.0537,
+      "step": 199
+    },
+    {
+      "epoch": 1.123943661971831,
+      "grad_norm": 6.009862899780273,
+      "learning_rate": 5.970000000000001e-07,
+      "loss": 4.0273,
+      "step": 200
+    },
+    {
+      "epoch": 1.1295774647887324,
+      "grad_norm": 6.165462970733643,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 4.037,
+      "step": 201
+    },
+    {
+      "epoch": 1.1352112676056338,
+      "grad_norm": 6.048194408416748,
+      "learning_rate": 6.03e-07,
+      "loss": 3.9988,
+      "step": 202
+    },
+    {
+      "epoch": 1.1408450704225352,
+      "grad_norm": 6.066869258880615,
+      "learning_rate": 6.06e-07,
+      "loss": 3.9743,
+      "step": 203
+    },
+    {
+      "epoch": 1.1464788732394366,
+      "grad_norm": 5.647124290466309,
+      "learning_rate": 6.09e-07,
+      "loss": 3.9283,
+      "step": 204
+    },
+    {
+      "epoch": 1.152112676056338,
+      "grad_norm": 6.250226974487305,
+      "learning_rate": 6.12e-07,
+      "loss": 4.0008,
+      "step": 205
+    },
+    {
+      "epoch": 1.1577464788732394,
+      "grad_norm": 5.445071697235107,
+      "learning_rate": 6.15e-07,
+      "loss": 3.9175,
+      "step": 206
+    },
+    {
+      "epoch": 1.1633802816901408,
+      "grad_norm": 5.444154262542725,
+      "learning_rate": 6.180000000000001e-07,
+      "loss": 3.9345,
+      "step": 207
+    },
+    {
+      "epoch": 1.1690140845070423,
+      "grad_norm": 5.232634544372559,
+      "learning_rate": 6.21e-07,
+      "loss": 3.9629,
+      "step": 208
+    },
+    {
+      "epoch": 1.1746478873239437,
+      "grad_norm": 5.474876403808594,
+      "learning_rate": 6.24e-07,
+      "loss": 3.8974,
+      "step": 209
+    },
+    {
+      "epoch": 1.180281690140845,
+      "grad_norm": 5.5575642585754395,
+      "learning_rate": 6.27e-07,
+      "loss": 3.8296,
+      "step": 210
+    },
+    {
+      "epoch": 1.1859154929577465,
+      "grad_norm": 5.940123081207275,
+      "learning_rate": 6.3e-07,
+      "loss": 3.8652,
+      "step": 211
+    },
+    {
+      "epoch": 1.1915492957746479,
+      "grad_norm": 5.474195957183838,
+      "learning_rate": 6.33e-07,
+      "loss": 3.8572,
+      "step": 212
+    },
+    {
+      "epoch": 1.1971830985915493,
+      "grad_norm": 6.2269415855407715,
+      "learning_rate": 6.36e-07,
+      "loss": 3.7893,
+      "step": 213
+    },
+    {
+      "epoch": 1.2028169014084507,
+      "grad_norm": 5.20849609375,
+      "learning_rate": 6.39e-07,
+      "loss": 3.8468,
+      "step": 214
+    },
+    {
+      "epoch": 1.208450704225352,
+      "grad_norm": 6.209558963775635,
+      "learning_rate": 6.42e-07,
+      "loss": 3.7182,
+      "step": 215
+    },
+    {
+      "epoch": 1.2140845070422535,
+      "grad_norm": 5.492266654968262,
+      "learning_rate": 6.45e-07,
+      "loss": 3.7881,
+      "step": 216
+    },
+    {
+      "epoch": 1.2197183098591549,
+      "grad_norm": 4.731509208679199,
+      "learning_rate": 6.480000000000001e-07,
+      "loss": 3.7978,
+      "step": 217
+    },
+    {
+      "epoch": 1.2253521126760563,
+      "grad_norm": 5.000571250915527,
+      "learning_rate": 6.51e-07,
+      "loss": 3.7661,
+      "step": 218
+    },
+    {
+      "epoch": 1.2309859154929577,
+      "grad_norm": 5.254445552825928,
+      "learning_rate": 6.54e-07,
+      "loss": 3.7341,
+      "step": 219
+    },
+    {
+      "epoch": 1.236619718309859,
+      "grad_norm": 4.641867637634277,
+      "learning_rate": 6.57e-07,
+      "loss": 3.767,
+      "step": 220
+    },
+    {
+      "epoch": 1.2422535211267607,
+      "grad_norm": 5.31817626953125,
+      "learning_rate": 6.599999999999999e-07,
+      "loss": 3.7103,
+      "step": 221
+    },
+    {
+      "epoch": 1.247887323943662,
+      "grad_norm": 5.551826477050781,
+      "learning_rate": 6.63e-07,
+      "loss": 3.6299,
+      "step": 222
+    },
+    {
+      "epoch": 1.2535211267605635,
+      "grad_norm": 17.55986213684082,
+      "learning_rate": 6.660000000000001e-07,
+      "loss": 4.1343,
+      "step": 223
+    },
+    {
+      "epoch": 1.2591549295774649,
+      "grad_norm": 12.343875885009766,
+      "learning_rate": 6.690000000000001e-07,
+      "loss": 4.0199,
+      "step": 224
+    },
+    {
+      "epoch": 1.2647887323943663,
+      "grad_norm": 10.083864212036133,
+      "learning_rate": 6.72e-07,
+      "loss": 3.9426,
+      "step": 225
+    },
+    {
+      "epoch": 1.2704225352112677,
+      "grad_norm": 6.6908345222473145,
+      "learning_rate": 6.75e-07,
+      "loss": 3.8179,
+      "step": 226
+    },
+    {
+      "epoch": 1.276056338028169,
+      "grad_norm": 6.976649284362793,
+      "learning_rate": 6.78e-07,
+      "loss": 3.8272,
+      "step": 227
+    },
+    {
+      "epoch": 1.2816901408450705,
+      "grad_norm": 5.452723503112793,
+      "learning_rate": 6.81e-07,
+      "loss": 3.7949,
+      "step": 228
+    },
+    {
+      "epoch": 1.287323943661972,
+      "grad_norm": 4.782120704650879,
+      "learning_rate": 6.84e-07,
+      "loss": 3.739,
+      "step": 229
+    },
+    {
+      "epoch": 1.2929577464788733,
+      "grad_norm": 5.477733612060547,
+      "learning_rate": 6.87e-07,
+      "loss": 3.7878,
+      "step": 230
+    },
+    {
+      "epoch": 1.2985915492957747,
+      "grad_norm": 4.35331916809082,
+      "learning_rate": 6.9e-07,
+      "loss": 3.7312,
+      "step": 231
+    },
+    {
+      "epoch": 1.304225352112676,
+      "grad_norm": 4.688896656036377,
+      "learning_rate": 6.93e-07,
+      "loss": 3.6926,
+      "step": 232
+    },
+    {
+      "epoch": 1.3098591549295775,
+      "grad_norm": 4.093502998352051,
+      "learning_rate": 6.96e-07,
+      "loss": 3.7448,
+      "step": 233
+    },
+    {
+      "epoch": 1.315492957746479,
+      "grad_norm": 4.601402282714844,
+      "learning_rate": 6.990000000000001e-07,
+      "loss": 3.733,
+      "step": 234
+    },
+    {
+      "epoch": 1.3211267605633803,
+      "grad_norm": 5.0725555419921875,
+      "learning_rate": 7.02e-07,
+      "loss": 3.6794,
+      "step": 235
+    },
+    {
+      "epoch": 1.3267605633802817,
+      "grad_norm": 4.2628374099731445,
+      "learning_rate": 7.05e-07,
+      "loss": 3.6747,
+      "step": 236
+    },
+    {
+      "epoch": 1.332394366197183,
+      "grad_norm": 4.159196853637695,
+      "learning_rate": 7.08e-07,
+      "loss": 3.6834,
+      "step": 237
+    },
+    {
+      "epoch": 1.3380281690140845,
+      "grad_norm": 3.971897602081299,
+      "learning_rate": 7.11e-07,
+      "loss": 3.6498,
+      "step": 238
+    },
+    {
+      "epoch": 1.343661971830986,
+      "grad_norm": 4.230648517608643,
+      "learning_rate": 7.140000000000001e-07,
+      "loss": 3.6179,
+      "step": 239
+    },
+    {
+      "epoch": 1.3492957746478873,
+      "grad_norm": 4.850123405456543,
+      "learning_rate": 7.170000000000001e-07,
+      "loss": 3.5813,
+      "step": 240
+    },
+    {
+      "epoch": 1.3549295774647887,
+      "grad_norm": 3.619288206100464,
+      "learning_rate": 7.2e-07,
+      "loss": 3.6227,
+      "step": 241
+    },
+    {
+      "epoch": 1.36056338028169,
+      "grad_norm": 4.144800662994385,
+      "learning_rate": 7.23e-07,
+      "loss": 3.5781,
+      "step": 242
+    },
+    {
+      "epoch": 1.3661971830985915,
+      "grad_norm": 3.449338912963867,
+      "learning_rate": 7.26e-07,
+      "loss": 3.5961,
+      "step": 243
+    },
+    {
+      "epoch": 1.371830985915493,
+      "grad_norm": 3.306309938430786,
+      "learning_rate": 7.29e-07,
+      "loss": 3.542,
+      "step": 244
+    },
+    {
+      "epoch": 1.3774647887323943,
+      "grad_norm": 3.7249999046325684,
+      "learning_rate": 7.32e-07,
+      "loss": 3.5631,
+      "step": 245
+    },
+    {
+      "epoch": 1.3830985915492957,
+      "grad_norm": 3.403693914413452,
+      "learning_rate": 7.350000000000001e-07,
+      "loss": 3.4998,
+      "step": 246
+    },
+    {
+      "epoch": 1.388732394366197,
+      "grad_norm": 3.535109519958496,
+      "learning_rate": 7.380000000000001e-07,
+      "loss": 3.5187,
+      "step": 247
+    },
+    {
+      "epoch": 1.3943661971830985,
+      "grad_norm": 3.3754496574401855,
+      "learning_rate": 7.41e-07,
+      "loss": 3.5296,
+      "step": 248
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.4589614868164062,
+      "learning_rate": 7.44e-07,
+      "loss": 3.5365,
+      "step": 249
+    },
+    {
+      "epoch": 1.4056338028169013,
+      "grad_norm": 3.0371317863464355,
+      "learning_rate": 7.47e-07,
+      "loss": 3.4723,
+      "step": 250
+    },
+    {
+      "epoch": 1.4112676056338027,
+      "grad_norm": 3.40959095954895,
+      "learning_rate": 7.5e-07,
+      "loss": 3.47,
+      "step": 251
+    },
+    {
+      "epoch": 1.4169014084507041,
+      "grad_norm": 3.2615842819213867,
+      "learning_rate": 7.53e-07,
+      "loss": 3.5185,
+      "step": 252
+    },
+    {
+      "epoch": 1.4225352112676055,
+      "grad_norm": 2.8736321926116943,
+      "learning_rate": 7.56e-07,
+      "loss": 3.5047,
+      "step": 253
+    },
+    {
+      "epoch": 1.428169014084507,
+      "grad_norm": 3.355738639831543,
+      "learning_rate": 7.59e-07,
+      "loss": 3.4177,
+      "step": 254
+    },
+    {
+      "epoch": 1.4338028169014083,
+      "grad_norm": 3.714266300201416,
+      "learning_rate": 7.62e-07,
+      "loss": 3.4446,
+      "step": 255
+    },
+    {
+      "epoch": 1.43943661971831,
+      "grad_norm": 3.5116286277770996,
+      "learning_rate": 7.65e-07,
+      "loss": 3.4098,
+      "step": 256
+    },
+    {
+      "epoch": 1.4450704225352113,
+      "grad_norm": 3.016411542892456,
+      "learning_rate": 7.680000000000001e-07,
+      "loss": 3.3968,
+      "step": 257
+    },
+    {
+      "epoch": 1.4507042253521127,
+      "grad_norm": 3.350151538848877,
+      "learning_rate": 7.71e-07,
+      "loss": 3.3943,
+      "step": 258
+    },
+    {
+      "epoch": 1.4563380281690141,
+      "grad_norm": 3.121488571166992,
+      "learning_rate": 7.74e-07,
+      "loss": 3.3664,
+      "step": 259
+    },
+    {
+      "epoch": 1.4619718309859155,
+      "grad_norm": 2.776261329650879,
+      "learning_rate": 7.77e-07,
+      "loss": 3.3606,
+      "step": 260
+    },
+    {
+      "epoch": 1.467605633802817,
+      "grad_norm": 3.6956863403320312,
+      "learning_rate": 7.799999999999999e-07,
+      "loss": 3.3889,
+      "step": 261
+    },
+    {
+      "epoch": 1.4732394366197183,
+      "grad_norm": 5.630766868591309,
+      "learning_rate": 7.830000000000001e-07,
+      "loss": 3.3498,
+      "step": 262
+    },
+    {
+      "epoch": 1.4788732394366197,
+      "grad_norm": 4.4029693603515625,
+      "learning_rate": 7.860000000000001e-07,
+      "loss": 3.3344,
+      "step": 263
+    },
+    {
+      "epoch": 1.4845070422535211,
+      "grad_norm": 2.447305917739868,
+      "learning_rate": 7.89e-07,
+      "loss": 3.3012,
+      "step": 264
+    },
+    {
+      "epoch": 1.4901408450704225,
+      "grad_norm": 10.303879737854004,
+      "learning_rate": 7.92e-07,
+      "loss": 3.3133,
+      "step": 265
+    },
+    {
+      "epoch": 1.495774647887324,
+      "grad_norm": 3.3926241397857666,
+      "learning_rate": 7.95e-07,
+      "loss": 3.3414,
+      "step": 266
+    },
+    {
+      "epoch": 1.5014084507042254,
+      "grad_norm": 15.229345321655273,
+      "learning_rate": 7.98e-07,
+      "loss": 3.6492,
+      "step": 267
+    },
+    {
+      "epoch": 1.5070422535211268,
+      "grad_norm": 8.38891887664795,
+      "learning_rate": 8.01e-07,
+      "loss": 3.4816,
+      "step": 268
+    },
+    {
+      "epoch": 1.5126760563380282,
+      "grad_norm": 7.57686185836792,
+      "learning_rate": 8.04e-07,
+      "loss": 3.4408,
+      "step": 269
+    },
+    {
+      "epoch": 1.5183098591549296,
+      "grad_norm": 4.836228847503662,
+      "learning_rate": 8.070000000000001e-07,
+      "loss": 3.4019,
+      "step": 270
+    },
+    {
+      "epoch": 1.523943661971831,
+      "grad_norm": 4.142337322235107,
+      "learning_rate": 8.1e-07,
+      "loss": 3.3743,
+      "step": 271
+    },
+    {
+      "epoch": 1.5295774647887324,
+      "grad_norm": 3.1797330379486084,
+      "learning_rate": 8.13e-07,
+      "loss": 3.3709,
+      "step": 272
+    },
+    {
+      "epoch": 1.5352112676056338,
+      "grad_norm": 3.2781105041503906,
+      "learning_rate": 8.16e-07,
+      "loss": 3.3235,
+      "step": 273
+    },
+    {
+      "epoch": 1.5408450704225352,
+      "grad_norm": 2.989748239517212,
+      "learning_rate": 8.19e-07,
+      "loss": 3.3489,
+      "step": 274
+    },
+    {
+      "epoch": 1.5464788732394368,
+      "grad_norm": 3.2133727073669434,
+      "learning_rate": 8.22e-07,
+      "loss": 3.3769,
+      "step": 275
+    },
+    {
+      "epoch": 1.5521126760563382,
+      "grad_norm": 2.8232502937316895,
+      "learning_rate": 8.25e-07,
+      "loss": 3.3146,
+      "step": 276
+    },
+    {
+      "epoch": 1.5577464788732396,
+      "grad_norm": 3.972604751586914,
+      "learning_rate": 8.280000000000001e-07,
+      "loss": 3.3137,
+      "step": 277
+    },
+    {
+      "epoch": 1.563380281690141,
+      "grad_norm": 3.0173697471618652,
+      "learning_rate": 8.31e-07,
+      "loss": 3.2774,
+      "step": 278
+    },
+    {
+      "epoch": 1.5690140845070424,
+      "grad_norm": 2.3883769512176514,
+      "learning_rate": 8.34e-07,
+      "loss": 3.3073,
+      "step": 279
+    },
+    {
+      "epoch": 1.5746478873239438,
+      "grad_norm": 2.2080185413360596,
+      "learning_rate": 8.370000000000001e-07,
+      "loss": 3.2611,
+      "step": 280
+    },
+    {
+      "epoch": 1.5802816901408452,
+      "grad_norm": 3.058180570602417,
+      "learning_rate": 8.4e-07,
+      "loss": 3.2561,
+      "step": 281
+    },
+    {
+      "epoch": 1.5859154929577466,
+      "grad_norm": 3.721339225769043,
+      "learning_rate": 8.43e-07,
+      "loss": 3.2693,
+      "step": 282
+    },
+    {
+      "epoch": 1.591549295774648,
+      "grad_norm": 4.630777359008789,
+      "learning_rate": 8.46e-07,
+      "loss": 3.2459,
+      "step": 283
+    },
+    {
+      "epoch": 1.5971830985915494,
+      "grad_norm": 2.207216501235962,
+      "learning_rate": 8.489999999999999e-07,
+      "loss": 3.215,
+      "step": 284
+    },
+    {
+      "epoch": 1.6028169014084508,
+      "grad_norm": 2.0840320587158203,
+      "learning_rate": 8.520000000000001e-07,
+      "loss": 3.2255,
+      "step": 285
+    },
+    {
+      "epoch": 1.6084507042253522,
+      "grad_norm": 2.489400625228882,
+      "learning_rate": 8.550000000000001e-07,
+      "loss": 3.2098,
+      "step": 286
+    },
+    {
+      "epoch": 1.6140845070422536,
+      "grad_norm": 2.2128732204437256,
+      "learning_rate": 8.580000000000001e-07,
+      "loss": 3.236,
+      "step": 287
+    },
+    {
+      "epoch": 1.619718309859155,
+      "grad_norm": 1.8475383520126343,
+      "learning_rate": 8.61e-07,
+      "loss": 3.2143,
+      "step": 288
+    },
+    {
+      "epoch": 1.6253521126760564,
+      "grad_norm": 3.034592390060425,
+      "learning_rate": 8.64e-07,
+      "loss": 3.1599,
+      "step": 289
+    },
+    {
+      "epoch": 1.6309859154929578,
+      "grad_norm": 3.48130464553833,
+      "learning_rate": 8.67e-07,
+      "loss": 3.1656,
+      "step": 290
+    },
+    {
+      "epoch": 1.6366197183098592,
+      "grad_norm": 2.465686321258545,
+      "learning_rate": 8.7e-07,
+      "loss": 3.1281,
+      "step": 291
+    },
+    {
+      "epoch": 1.6422535211267606,
+      "grad_norm": 2.5849077701568604,
+      "learning_rate": 8.73e-07,
+      "loss": 3.1726,
+      "step": 292
+    },
+    {
+      "epoch": 1.647887323943662,
+      "grad_norm": 2.192965507507324,
+      "learning_rate": 8.760000000000001e-07,
+      "loss": 3.1238,
+      "step": 293
+    },
+    {
+      "epoch": 1.6535211267605634,
+      "grad_norm": 1.818284273147583,
+      "learning_rate": 8.79e-07,
+      "loss": 3.1451,
+      "step": 294
+    },
+    {
+      "epoch": 1.6591549295774648,
+      "grad_norm": 1.8378015756607056,
+      "learning_rate": 8.82e-07,
+      "loss": 3.1432,
+      "step": 295
+    },
+    {
+      "epoch": 1.6647887323943662,
+      "grad_norm": 4.533341407775879,
+      "learning_rate": 8.85e-07,
+      "loss": 3.1177,
+      "step": 296
+    },
+    {
+      "epoch": 1.6704225352112676,
+      "grad_norm": 3.2857859134674072,
+      "learning_rate": 8.88e-07,
+      "loss": 3.1139,
+      "step": 297
+    },
+    {
+      "epoch": 1.676056338028169,
+      "grad_norm": 2.6295394897460938,
+      "learning_rate": 8.91e-07,
+      "loss": 3.1375,
+      "step": 298
+    },
+    {
+      "epoch": 1.6816901408450704,
+      "grad_norm": 2.6304931640625,
+      "learning_rate": 8.94e-07,
+      "loss": 3.1008,
+      "step": 299
+    },
+    {
+      "epoch": 1.6873239436619718,
+      "grad_norm": 2.0864717960357666,
+      "learning_rate": 8.97e-07,
+      "loss": 3.0836,
+      "step": 300
+    },
+    {
+      "epoch": 1.6929577464788732,
+      "grad_norm": 1.9646295309066772,
+      "learning_rate": 9e-07,
+      "loss": 3.0995,
+      "step": 301
+    },
+    {
+      "epoch": 1.6985915492957746,
+      "grad_norm": 2.4765114784240723,
+      "learning_rate": 9.03e-07,
+      "loss": 3.0557,
+      "step": 302
+    },
+    {
+      "epoch": 1.704225352112676,
+      "grad_norm": 1.6692005395889282,
+      "learning_rate": 9.060000000000001e-07,
+      "loss": 3.0774,
+      "step": 303
+    },
+    {
+      "epoch": 1.7098591549295774,
+      "grad_norm": 3.1556854248046875,
+      "learning_rate": 9.09e-07,
+      "loss": 3.103,
+      "step": 304
+    },
+    {
+      "epoch": 1.7154929577464788,
+      "grad_norm": 4.288612365722656,
+      "learning_rate": 9.12e-07,
+      "loss": 3.0551,
+      "step": 305
+    },
+    {
+      "epoch": 1.7211267605633802,
+      "grad_norm": 1.729299783706665,
+      "learning_rate": 9.15e-07,
+      "loss": 3.0499,
+      "step": 306
+    },
+    {
+      "epoch": 1.7267605633802816,
+      "grad_norm": 3.1948986053466797,
+      "learning_rate": 9.179999999999999e-07,
+      "loss": 3.0367,
+      "step": 307
+    },
+    {
+      "epoch": 1.732394366197183,
+      "grad_norm": 2.3002986907958984,
+      "learning_rate": 9.210000000000001e-07,
+      "loss": 3.0475,
+      "step": 308
+    },
+    {
+      "epoch": 1.7380281690140844,
+      "grad_norm": 2.291055917739868,
+      "learning_rate": 9.240000000000001e-07,
+      "loss": 3.0386,
+      "step": 309
+    },
+    {
+      "epoch": 1.7436619718309858,
+      "grad_norm": 3.8127360343933105,
+      "learning_rate": 9.270000000000001e-07,
+      "loss": 3.0193,
+      "step": 310
+    },
+    {
+      "epoch": 1.7492957746478872,
+      "grad_norm": 8.73247241973877,
+      "learning_rate": 9.3e-07,
+      "loss": 3.3324,
+      "step": 311
+    },
+    {
+      "epoch": 1.7549295774647886,
+      "grad_norm": 5.212857723236084,
+      "learning_rate": 9.33e-07,
+      "loss": 3.2241,
+      "step": 312
+    },
+    {
+      "epoch": 1.76056338028169,
+      "grad_norm": 5.405500411987305,
+      "learning_rate": 9.36e-07,
+      "loss": 3.1716,
+      "step": 313
+    },
+    {
+      "epoch": 1.7661971830985914,
+      "grad_norm": 3.3166182041168213,
+      "learning_rate": 9.39e-07,
+      "loss": 3.1203,
+      "step": 314
+    },
+    {
+      "epoch": 1.7718309859154928,
+      "grad_norm": 4.144299507141113,
+      "learning_rate": 9.419999999999999e-07,
+      "loss": 3.1726,
+      "step": 315
+    },
+    {
+      "epoch": 1.7774647887323942,
+      "grad_norm": 3.576324224472046,
+      "learning_rate": 9.450000000000001e-07,
+      "loss": 3.1518,
+      "step": 316
+    },
+    {
+      "epoch": 1.7830985915492956,
+      "grad_norm": 3.0369715690612793,
+      "learning_rate": 9.480000000000001e-07,
+      "loss": 3.1077,
+      "step": 317
+    },
+    {
+      "epoch": 1.788732394366197,
+      "grad_norm": 3.1641204357147217,
+      "learning_rate": 9.51e-07,
+      "loss": 3.1223,
+      "step": 318
+    },
+    {
+      "epoch": 1.7943661971830986,
+      "grad_norm": 2.084167242050171,
+      "learning_rate": 9.54e-07,
+      "loss": 3.0577,
+      "step": 319
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.145345449447632,
+      "learning_rate": 9.57e-07,
+      "loss": 3.1057,
+      "step": 320
+    },
+    {
+      "epoch": 1.8056338028169014,
+      "grad_norm": 3.1458144187927246,
+      "learning_rate": 9.600000000000001e-07,
+      "loss": 3.1221,
+      "step": 321
+    },
+    {
+      "epoch": 1.8112676056338028,
+      "grad_norm": 2.8545215129852295,
+      "learning_rate": 9.63e-07,
+      "loss": 3.092,
+      "step": 322
+    },
+    {
+      "epoch": 1.8169014084507042,
+      "grad_norm": 15.60283374786377,
+      "learning_rate": 9.66e-07,
+      "loss": 3.0654,
+      "step": 323
+    },
+    {
+      "epoch": 1.8225352112676056,
+      "grad_norm": 2.4702093601226807,
+      "learning_rate": 9.690000000000002e-07,
+      "loss": 3.0606,
+      "step": 324
+    },
+    {
+      "epoch": 1.828169014084507,
+      "grad_norm": 3.160369396209717,
+      "learning_rate": 9.72e-07,
+      "loss": 3.0641,
+      "step": 325
+    },
+    {
+      "epoch": 1.8338028169014085,
+      "grad_norm": 2.9996860027313232,
+      "learning_rate": 9.75e-07,
+      "loss": 3.06,
+      "step": 326
+    },
+    {
+      "epoch": 1.8394366197183099,
+      "grad_norm": 4.691850662231445,
+      "learning_rate": 9.78e-07,
+      "loss": 3.0031,
+      "step": 327
+    },
+    {
+      "epoch": 1.8450704225352113,
+      "grad_norm": 2.0418202877044678,
+      "learning_rate": 9.81e-07,
+      "loss": 3.0525,
+      "step": 328
+    },
+    {
+      "epoch": 1.8507042253521127,
+      "grad_norm": 3.0207626819610596,
+      "learning_rate": 9.84e-07,
+      "loss": 3.0289,
+      "step": 329
+    },
+    {
+      "epoch": 1.856338028169014,
+      "grad_norm": 1.9885998964309692,
+      "learning_rate": 9.87e-07,
+      "loss": 3.0429,
+      "step": 330
+    },
+    {
+      "epoch": 1.8619718309859155,
+      "grad_norm": 2.9951462745666504,
+      "learning_rate": 9.9e-07,
+      "loss": 3.0391,
+      "step": 331
+    },
+    {
+      "epoch": 1.8676056338028169,
+      "grad_norm": 3.579716682434082,
+      "learning_rate": 9.929999999999999e-07,
+      "loss": 3.0725,
+      "step": 332
+    },
+    {
+      "epoch": 1.8732394366197183,
+      "grad_norm": 3.4275898933410645,
+      "learning_rate": 9.96e-07,
+      "loss": 3.0552,
+      "step": 333
+    },
+    {
+      "epoch": 1.8788732394366199,
+      "grad_norm": 3.915097951889038,
+      "learning_rate": 9.99e-07,
+      "loss": 3.0403,
+      "step": 334
+    },
+    {
+      "epoch": 1.8845070422535213,
+      "grad_norm": 3.5404648780822754,
+      "learning_rate": 1.002e-06,
+      "loss": 3.0184,
+      "step": 335
+    },
+    {
+      "epoch": 1.8901408450704227,
+      "grad_norm": 1.993233561515808,
+      "learning_rate": 1.0050000000000001e-06,
+      "loss": 3.0271,
+      "step": 336
+    },
+    {
+      "epoch": 1.895774647887324,
+      "grad_norm": 2.958256483078003,
+      "learning_rate": 1.008e-06,
+      "loss": 3.0009,
+      "step": 337
+    },
+    {
+      "epoch": 1.9014084507042255,
+      "grad_norm": 2.1944117546081543,
+      "learning_rate": 1.0110000000000001e-06,
+      "loss": 2.973,
+      "step": 338
+    },
+    {
+      "epoch": 1.9070422535211269,
+      "grad_norm": 4.562254905700684,
+      "learning_rate": 1.014e-06,
+      "loss": 3.0002,
+      "step": 339
+    },
+    {
+      "epoch": 1.9126760563380283,
+      "grad_norm": 2.5491044521331787,
+      "learning_rate": 1.017e-06,
+      "loss": 3.0079,
+      "step": 340
+    },
+    {
+      "epoch": 1.9183098591549297,
+      "grad_norm": 4.045207977294922,
+      "learning_rate": 1.0200000000000002e-06,
+      "loss": 2.9921,
+      "step": 341
+    },
+    {
+      "epoch": 1.923943661971831,
+      "grad_norm": 4.1953864097595215,
+      "learning_rate": 1.023e-06,
+      "loss": 2.9992,
+      "step": 342
+    },
+    {
+      "epoch": 1.9295774647887325,
+      "grad_norm": 2.195458173751831,
+      "learning_rate": 1.026e-06,
+      "loss": 3.0202,
+      "step": 343
+    },
+    {
+      "epoch": 1.935211267605634,
+      "grad_norm": 6.150205135345459,
+      "learning_rate": 1.029e-06,
+      "loss": 2.9895,
+      "step": 344
+    },
+    {
+      "epoch": 1.9408450704225353,
+      "grad_norm": 2.1709234714508057,
+      "learning_rate": 1.032e-06,
+      "loss": 3.026,
+      "step": 345
+    },
+    {
+      "epoch": 1.9464788732394367,
+      "grad_norm": 3.657919406890869,
+      "learning_rate": 1.035e-06,
+      "loss": 2.9522,
+      "step": 346
+    },
+    {
+      "epoch": 1.952112676056338,
+      "grad_norm": 5.459045886993408,
+      "learning_rate": 1.038e-06,
+      "loss": 2.9906,
+      "step": 347
+    },
+    {
+      "epoch": 1.9577464788732395,
+      "grad_norm": 2.215078592300415,
+      "learning_rate": 1.041e-06,
+      "loss": 2.9721,
+      "step": 348
+    },
+    {
+      "epoch": 1.963380281690141,
+      "grad_norm": 2.1503262519836426,
+      "learning_rate": 1.0439999999999999e-06,
+      "loss": 2.9818,
+      "step": 349
+    },
+    {
+      "epoch": 1.9690140845070423,
+      "grad_norm": 21.535140991210938,
+      "learning_rate": 1.047e-06,
+      "loss": 2.9945,
+      "step": 350
+    },
+    {
+      "epoch": 1.9746478873239437,
+      "grad_norm": 3.492640733718872,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 2.9585,
+      "step": 351
+    },
+    {
+      "epoch": 1.980281690140845,
+      "grad_norm": 10.216153144836426,
+      "learning_rate": 1.053e-06,
+      "loss": 3.0104,
+      "step": 352
+    },
+    {
+      "epoch": 1.9859154929577465,
+      "grad_norm": 3.894277334213257,
+      "learning_rate": 1.0560000000000001e-06,
+      "loss": 2.9685,
+      "step": 353
+    },
+    {
+      "epoch": 1.991549295774648,
+      "grad_norm": NaN,
+      "learning_rate": 1.0560000000000001e-06,
+      "loss": 3.016,
+      "step": 354
+    },
+    {
+      "epoch": 1.9971830985915493,
+      "grad_norm": 2.764295816421509,
+      "learning_rate": 1.059e-06,
+      "loss": 3.0144,
+      "step": 355
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.5533089637756348,
+      "learning_rate": 1.062e-06,
+      "loss": 1.4893,
+      "step": 356
+    },
+    {
+      "epoch": 2.0056338028169014,
+      "grad_norm": 6.689918518066406,
+      "learning_rate": 1.065e-06,
+      "loss": 3.156,
+      "step": 357
+    },
+    {
+      "epoch": 2.011267605633803,
+      "grad_norm": 4.828878879547119,
+      "learning_rate": 1.068e-06,
+      "loss": 3.1063,
+      "step": 358
+    },
+    {
+      "epoch": 2.016901408450704,
+      "grad_norm": 3.7969484329223633,
+      "learning_rate": 1.0710000000000002e-06,
+      "loss": 3.0965,
+      "step": 359
+    },
+    {
+      "epoch": 2.0225352112676056,
+      "grad_norm": 3.629793405532837,
+      "learning_rate": 1.074e-06,
+      "loss": 3.0502,
+      "step": 360
+    },
+    {
+      "epoch": 2.028169014084507,
+      "grad_norm": 4.136383533477783,
+      "learning_rate": 1.077e-06,
+      "loss": 3.0603,
+      "step": 361
+    },
+    {
+      "epoch": 2.0338028169014084,
+      "grad_norm": 4.763708591461182,
+      "learning_rate": 1.08e-06,
+      "loss": 3.035,
+      "step": 362
+    },
+    {
+      "epoch": 2.03943661971831,
+      "grad_norm": 2.662196397781372,
+      "learning_rate": 1.083e-06,
+      "loss": 2.979,
+      "step": 363
+    },
+    {
+      "epoch": 2.045070422535211,
+      "grad_norm": 2.799008369445801,
+      "learning_rate": 1.086e-06,
+      "loss": 3.0423,
+      "step": 364
+    },
+    {
+      "epoch": 2.0507042253521126,
+      "grad_norm": 3.0370540618896484,
+      "learning_rate": 1.089e-06,
+      "loss": 3.0191,
+      "step": 365
+    },
+    {
+      "epoch": 2.056338028169014,
+      "grad_norm": 3.5550408363342285,
+      "learning_rate": 1.092e-06,
+      "loss": 2.9998,
+      "step": 366
+    },
+    {
+      "epoch": 2.0619718309859154,
+      "grad_norm": 2.8191301822662354,
+      "learning_rate": 1.0949999999999999e-06,
+      "loss": 3.0373,
+      "step": 367
+    },
+    {
+      "epoch": 2.067605633802817,
+      "grad_norm": 4.018879413604736,
+      "learning_rate": 1.0980000000000001e-06,
+      "loss": 2.9933,
+      "step": 368
+    },
+    {
+      "epoch": 2.073239436619718,
+      "grad_norm": 3.813567876815796,
+      "learning_rate": 1.1010000000000001e-06,
+      "loss": 3.0079,
+      "step": 369
+    },
+    {
+      "epoch": 2.0788732394366196,
+      "grad_norm": 2.6536049842834473,
+      "learning_rate": 1.104e-06,
+      "loss": 2.9907,
+      "step": 370
+    },
+    {
+      "epoch": 2.084507042253521,
+      "grad_norm": 1.7705937623977661,
+      "learning_rate": 1.1070000000000002e-06,
+      "loss": 2.9996,
+      "step": 371
+    },
+    {
+      "epoch": 2.0901408450704224,
+      "grad_norm": 5.244455814361572,
+      "learning_rate": 1.11e-06,
+      "loss": 3.0193,
+      "step": 372
+    },
+    {
+      "epoch": 2.095774647887324,
+      "grad_norm": 4.140601634979248,
+      "learning_rate": 1.113e-06,
+      "loss": 2.9686,
+      "step": 373
+    },
+    {
+      "epoch": 2.101408450704225,
+      "grad_norm": 2.9329001903533936,
+      "learning_rate": 1.116e-06,
+      "loss": 2.9843,
+      "step": 374
+    },
+    {
+      "epoch": 2.1070422535211266,
+      "grad_norm": 1.8754574060440063,
+      "learning_rate": 1.119e-06,
+      "loss": 2.9626,
+      "step": 375
+    },
+    {
+      "epoch": 2.112676056338028,
+      "grad_norm": 4.379260540008545,
+      "learning_rate": 1.122e-06,
+      "loss": 2.9652,
+      "step": 376
+    },
+    {
+      "epoch": 2.1183098591549294,
+      "grad_norm": 3.4436750411987305,
+      "learning_rate": 1.125e-06,
+      "loss": 2.9727,
+      "step": 377
+    },
+    {
+      "epoch": 2.123943661971831,
+      "grad_norm": 3.883599281311035,
+      "learning_rate": 1.128e-06,
+      "loss": 2.9619,
+      "step": 378
+    },
+    {
+      "epoch": 2.129577464788732,
+      "grad_norm": 3.0950586795806885,
+      "learning_rate": 1.131e-06,
+      "loss": 2.9473,
+      "step": 379
+    },
+    {
+      "epoch": 2.1352112676056336,
+      "grad_norm": 1.735630989074707,
+      "learning_rate": 1.134e-06,
+      "loss": 2.9603,
+      "step": 380
+    },
+    {
+      "epoch": 2.140845070422535,
+      "grad_norm": 5.800061225891113,
+      "learning_rate": 1.137e-06,
+      "loss": 3.0003,
+      "step": 381
+    },
+    {
+      "epoch": 2.1464788732394364,
+      "grad_norm": 3.413560152053833,
+      "learning_rate": 1.14e-06,
+      "loss": 2.9284,
+      "step": 382
+    },
+    {
+      "epoch": 2.152112676056338,
+      "grad_norm": 2.977813482284546,
+      "learning_rate": 1.1430000000000001e-06,
+      "loss": 2.9381,
+      "step": 383
+    },
+    {
+      "epoch": 2.1577464788732392,
+      "grad_norm": 2.820646047592163,
+      "learning_rate": 1.146e-06,
+      "loss": 2.9239,
+      "step": 384
+    },
+    {
+      "epoch": 2.1633802816901406,
+      "grad_norm": 3.4326274394989014,
+      "learning_rate": 1.1490000000000001e-06,
+      "loss": 2.9261,
+      "step": 385
+    },
+    {
+      "epoch": 2.169014084507042,
+      "grad_norm": 6.641375541687012,
+      "learning_rate": 1.152e-06,
+      "loss": 2.9679,
+      "step": 386
+    },
+    {
+      "epoch": 2.1746478873239434,
+      "grad_norm": 5.253774166107178,
+      "learning_rate": 1.155e-06,
+      "loss": 2.9745,
+      "step": 387
+    },
+    {
+      "epoch": 2.1802816901408453,
+      "grad_norm": 3.8959267139434814,
+      "learning_rate": 1.1580000000000002e-06,
+      "loss": 2.9117,
+      "step": 388
+    },
+    {
+      "epoch": 2.1859154929577467,
+      "grad_norm": 1.9529342651367188,
+      "learning_rate": 1.161e-06,
+      "loss": 2.9216,
+      "step": 389
+    },
+    {
+      "epoch": 2.191549295774648,
+      "grad_norm": 6.595493793487549,
+      "learning_rate": 1.164e-06,
+      "loss": 2.9176,
+      "step": 390
+    },
+    {
+      "epoch": 2.1971830985915495,
+      "grad_norm": 3.104062557220459,
+      "learning_rate": 1.167e-06,
+      "loss": 2.9234,
+      "step": 391
+    },
+    {
+      "epoch": 2.202816901408451,
+      "grad_norm": 2.850505828857422,
+      "learning_rate": 1.17e-06,
+      "loss": 2.9371,
+      "step": 392
+    },
+    {
+      "epoch": 2.2084507042253523,
+      "grad_norm": 3.024420738220215,
+      "learning_rate": 1.173e-06,
+      "loss": 2.9335,
+      "step": 393
+    },
+    {
+      "epoch": 2.2140845070422537,
+      "grad_norm": 3.7631072998046875,
+      "learning_rate": 1.176e-06,
+      "loss": 2.9129,
+      "step": 394
+    },
+    {
+      "epoch": 2.219718309859155,
+      "grad_norm": 3.0405235290527344,
+      "learning_rate": 1.179e-06,
+      "loss": 2.9242,
+      "step": 395
+    },
+    {
+      "epoch": 2.2253521126760565,
+      "grad_norm": 4.891364097595215,
+      "learning_rate": 1.1819999999999999e-06,
+      "loss": 2.9131,
+      "step": 396
+    },
+    {
+      "epoch": 2.230985915492958,
+      "grad_norm": 2.57051420211792,
+      "learning_rate": 1.185e-06,
+      "loss": 2.8991,
+      "step": 397
+    },
+    {
+      "epoch": 2.2366197183098593,
+      "grad_norm": 3.683042049407959,
+      "learning_rate": 1.188e-06,
+      "loss": 2.9211,
+      "step": 398
+    },
+    {
+      "epoch": 2.2422535211267607,
+      "grad_norm": 2.8408870697021484,
+      "learning_rate": 1.191e-06,
+      "loss": 2.8893,
+      "step": 399
+    },
+    {
+      "epoch": 2.247887323943662,
+      "grad_norm": 6.567075252532959,
+      "learning_rate": 1.1940000000000001e-06,
+      "loss": 2.9126,
+      "step": 400
+    },
+    {
+      "epoch": 2.2535211267605635,
+      "grad_norm": 4.963144779205322,
+      "learning_rate": 1.197e-06,
+      "loss": 3.0861,
+      "step": 401
+    },
+    {
+      "epoch": 2.259154929577465,
+      "grad_norm": 5.517329692840576,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 3.0618,
+      "step": 402
+    },
+    {
+      "epoch": 2.2647887323943663,
+      "grad_norm": 3.461358070373535,
+      "learning_rate": 1.203e-06,
+      "loss": 2.9884,
+      "step": 403
+    },
+    {
+      "epoch": 2.2704225352112677,
+      "grad_norm": 5.536230087280273,
+      "learning_rate": 1.206e-06,
+      "loss": 2.9819,
+      "step": 404
+    },
+    {
+      "epoch": 2.276056338028169,
+      "grad_norm": 3.172144651412964,
+      "learning_rate": 1.2090000000000002e-06,
+      "loss": 2.965,
+      "step": 405
+    },
+    {
+      "epoch": 2.2816901408450705,
+      "grad_norm": 4.672169208526611,
+      "learning_rate": 1.212e-06,
+      "loss": 2.9963,
+      "step": 406
+    },
+    {
+      "epoch": 2.287323943661972,
+      "grad_norm": 2.5767641067504883,
+      "learning_rate": 1.215e-06,
+      "loss": 2.9119,
+      "step": 407
+    },
+    {
+      "epoch": 2.2929577464788733,
+      "grad_norm": 3.546799659729004,
+      "learning_rate": 1.218e-06,
+      "loss": 2.9451,
+      "step": 408
+    },
+    {
+      "epoch": 2.2985915492957747,
+      "grad_norm": 2.6019978523254395,
+      "learning_rate": 1.221e-06,
+      "loss": 2.9306,
+      "step": 409
+    },
+    {
+      "epoch": 2.304225352112676,
+      "grad_norm": 5.824201583862305,
+      "learning_rate": 1.224e-06,
+      "loss": 2.9445,
+      "step": 410
+    },
+    {
+      "epoch": 2.3098591549295775,
+      "grad_norm": 3.3293278217315674,
+      "learning_rate": 1.227e-06,
+      "loss": 2.9351,
+      "step": 411
+    },
+    {
+      "epoch": 2.315492957746479,
+      "grad_norm": 5.109325885772705,
+      "learning_rate": 1.23e-06,
+      "loss": 2.9134,
+      "step": 412
+    },
+    {
+      "epoch": 2.3211267605633803,
+      "grad_norm": 3.5838801860809326,
+      "learning_rate": 1.2329999999999999e-06,
+      "loss": 2.9145,
+      "step": 413
+    },
+    {
+      "epoch": 2.3267605633802817,
+      "grad_norm": 4.243381977081299,
+      "learning_rate": 1.2360000000000001e-06,
+      "loss": 2.905,
+      "step": 414
+    },
+    {
+      "epoch": 2.332394366197183,
+      "grad_norm": 6.325119972229004,
+      "learning_rate": 1.2390000000000001e-06,
+      "loss": 2.9102,
+      "step": 415
+    },
+    {
+      "epoch": 2.3380281690140845,
+      "grad_norm": 3.0729176998138428,
+      "learning_rate": 1.242e-06,
+      "loss": 2.8892,
+      "step": 416
+    },
+    {
+      "epoch": 2.343661971830986,
+      "grad_norm": 5.03809118270874,
+      "learning_rate": 1.2450000000000002e-06,
+      "loss": 2.8839,
+      "step": 417
+    },
+    {
+      "epoch": 2.3492957746478873,
+      "grad_norm": 2.4732346534729004,
+      "learning_rate": 1.248e-06,
+      "loss": 2.9064,
+      "step": 418
+    },
+    {
+      "epoch": 2.3549295774647887,
+      "grad_norm": 3.957566499710083,
+      "learning_rate": 1.251e-06,
+      "loss": 2.8872,
+      "step": 419
+    },
+    {
+      "epoch": 2.36056338028169,
+      "grad_norm": 3.1195902824401855,
+      "learning_rate": 1.254e-06,
+      "loss": 2.8886,
+      "step": 420
+    },
+    {
+      "epoch": 2.3661971830985915,
+      "grad_norm": 4.334029674530029,
+      "learning_rate": 1.257e-06,
+      "loss": 2.873,
+      "step": 421
+    },
+    {
+      "epoch": 2.371830985915493,
+      "grad_norm": 2.1096246242523193,
+      "learning_rate": 1.26e-06,
+      "loss": 2.8649,
+      "step": 422
+    },
+    {
+      "epoch": 2.3774647887323943,
+      "grad_norm": 4.480950355529785,
+      "learning_rate": 1.263e-06,
+      "loss": 2.8852,
+      "step": 423
+    },
+    {
+      "epoch": 2.3830985915492957,
+      "grad_norm": 4.877039909362793,
+      "learning_rate": 1.266e-06,
+      "loss": 2.8639,
+      "step": 424
+    },
+    {
+      "epoch": 2.388732394366197,
+      "grad_norm": 3.921311140060425,
+      "learning_rate": 1.269e-06,
+      "loss": 2.8589,
+      "step": 425
+    },
+    {
+      "epoch": 2.3943661971830985,
+      "grad_norm": 2.2594597339630127,
+      "learning_rate": 1.272e-06,
+      "loss": 2.8935,
+      "step": 426
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.150793075561523,
+      "learning_rate": 1.275e-06,
+      "loss": 2.8623,
+      "step": 427
+    },
+    {
+      "epoch": 2.4056338028169013,
+      "grad_norm": 2.5821540355682373,
+      "learning_rate": 1.278e-06,
+      "loss": 2.8597,
+      "step": 428
+    },
+    {
+      "epoch": 2.4112676056338027,
+      "grad_norm": 3.32155704498291,
+      "learning_rate": 1.281e-06,
+      "loss": 2.8644,
+      "step": 429
+    },
+    {
+      "epoch": 2.416901408450704,
+      "grad_norm": 2.0583064556121826,
+      "learning_rate": 1.284e-06,
+      "loss": 2.8809,
+      "step": 430
+    },
+    {
+      "epoch": 2.4225352112676055,
+      "grad_norm": 2.4684250354766846,
+      "learning_rate": 1.2870000000000001e-06,
+      "loss": 2.8583,
+      "step": 431
+    },
+    {
+      "epoch": 2.428169014084507,
+      "grad_norm": 2.8934361934661865,
+      "learning_rate": 1.29e-06,
+      "loss": 2.8812,
+      "step": 432
+    },
+    {
+      "epoch": 2.4338028169014083,
+      "grad_norm": 3.3147571086883545,
+      "learning_rate": 1.293e-06,
+      "loss": 2.8513,
+      "step": 433
+    },
+    {
+      "epoch": 2.4394366197183097,
+      "grad_norm": 3.118459701538086,
+      "learning_rate": 1.2960000000000002e-06,
+      "loss": 2.8737,
+      "step": 434
+    },
+    {
+      "epoch": 2.445070422535211,
+      "grad_norm": 5.725314140319824,
+      "learning_rate": 1.299e-06,
+      "loss": 2.8238,
+      "step": 435
+    },
+    {
+      "epoch": 2.4507042253521125,
+      "grad_norm": 2.1813156604766846,
+      "learning_rate": 1.302e-06,
+      "loss": 2.827,
+      "step": 436
+    },
+    {
+      "epoch": 2.456338028169014,
+      "grad_norm": 2.7385337352752686,
+      "learning_rate": 1.305e-06,
+      "loss": 2.8563,
+      "step": 437
+    },
+    {
+      "epoch": 2.4619718309859153,
+      "grad_norm": 4.738112926483154,
+      "learning_rate": 1.308e-06,
+      "loss": 2.822,
+      "step": 438
+    },
+    {
+      "epoch": 2.4676056338028167,
+      "grad_norm": 3.1284008026123047,
+      "learning_rate": 1.311e-06,
+      "loss": 2.8628,
+      "step": 439
+    },
+    {
+      "epoch": 2.473239436619718,
+      "grad_norm": 3.920558452606201,
+      "learning_rate": 1.314e-06,
+      "loss": 2.8456,
+      "step": 440
+    },
+    {
+      "epoch": 2.4788732394366195,
+      "grad_norm": 3.4516639709472656,
+      "learning_rate": 1.317e-06,
+      "loss": 2.8238,
+      "step": 441
+    },
+    {
+      "epoch": 2.4845070422535214,
+      "grad_norm": 2.7140746116638184,
+      "learning_rate": 1.3199999999999999e-06,
+      "loss": 2.8164,
+      "step": 442
+    },
+    {
+      "epoch": 2.4901408450704228,
+      "grad_norm": 3.0408647060394287,
+      "learning_rate": 1.323e-06,
+      "loss": 2.8268,
+      "step": 443
+    },
+    {
+      "epoch": 2.495774647887324,
+      "grad_norm": 5.611374378204346,
+      "learning_rate": 1.326e-06,
+      "loss": 2.8166,
+      "step": 444
+    },
+    {
+      "epoch": 2.5014084507042256,
+      "grad_norm": 4.1387619972229,
+      "learning_rate": 1.3290000000000001e-06,
+      "loss": 3.0038,
+      "step": 445
+    },
+    {
+      "epoch": 2.507042253521127,
+      "grad_norm": 3.045279026031494,
+      "learning_rate": 1.3320000000000001e-06,
+      "loss": 2.9354,
+      "step": 446
+    },
+    {
+      "epoch": 2.5126760563380284,
+      "grad_norm": 2.492701530456543,
+      "learning_rate": 1.335e-06,
+      "loss": 2.9068,
+      "step": 447
+    },
+    {
+      "epoch": 2.5183098591549298,
+      "grad_norm": 3.7552857398986816,
+      "learning_rate": 1.3380000000000001e-06,
+      "loss": 2.8962,
+      "step": 448
+    },
+    {
+      "epoch": 2.523943661971831,
+      "grad_norm": 3.1780288219451904,
+      "learning_rate": 1.341e-06,
+      "loss": 2.869,
+      "step": 449
+    },
+    {
+      "epoch": 2.5295774647887326,
+      "grad_norm": 3.949904680252075,
+      "learning_rate": 1.344e-06,
+      "loss": 2.8825,
+      "step": 450
+    },
+    {
+      "epoch": 2.535211267605634,
+      "grad_norm": 1.9001998901367188,
+      "learning_rate": 1.3470000000000002e-06,
+      "loss": 2.8544,
+      "step": 451
+    },
+    {
+      "epoch": 2.5408450704225354,
+      "grad_norm": 3.131744384765625,
+      "learning_rate": 1.35e-06,
+      "loss": 2.8851,
+      "step": 452
+    },
+    {
+      "epoch": 2.546478873239437,
+      "grad_norm": 3.6800670623779297,
+      "learning_rate": 1.353e-06,
+      "loss": 2.8998,
+      "step": 453
+    },
+    {
+      "epoch": 2.552112676056338,
+      "grad_norm": 3.9882307052612305,
+      "learning_rate": 1.356e-06,
+      "loss": 2.8831,
+      "step": 454
+    },
+    {
+      "epoch": 2.5577464788732396,
+      "grad_norm": 2.457385778427124,
+      "learning_rate": 1.359e-06,
+      "loss": 2.8616,
+      "step": 455
+    },
+    {
+      "epoch": 2.563380281690141,
+      "grad_norm": 2.2192134857177734,
+      "learning_rate": 1.362e-06,
+      "loss": 2.8799,
+      "step": 456
+    },
+    {
+      "epoch": 2.5690140845070424,
+      "grad_norm": 3.6589062213897705,
+      "learning_rate": 1.365e-06,
+      "loss": 2.8637,
+      "step": 457
+    },
+    {
+      "epoch": 2.574647887323944,
+      "grad_norm": 2.2132861614227295,
+      "learning_rate": 1.368e-06,
+      "loss": 2.8135,
+      "step": 458
+    },
+    {
+      "epoch": 2.580281690140845,
+      "grad_norm": 2.988835334777832,
+      "learning_rate": 1.3709999999999999e-06,
+      "loss": 2.8356,
+      "step": 459
+    },
+    {
+      "epoch": 2.5859154929577466,
+      "grad_norm": 2.2689504623413086,
+      "learning_rate": 1.374e-06,
+      "loss": 2.8513,
+      "step": 460
+    },
+    {
+      "epoch": 2.591549295774648,
+      "grad_norm": 2.1941077709198,
+      "learning_rate": 1.3770000000000001e-06,
+      "loss": 2.8369,
+      "step": 461
+    },
+    {
+      "epoch": 2.5971830985915494,
+      "grad_norm": 1.760879635810852,
+      "learning_rate": 1.38e-06,
+      "loss": 2.8273,
+      "step": 462
+    },
+    {
+      "epoch": 2.602816901408451,
+      "grad_norm": 4.897495269775391,
+      "learning_rate": 1.3830000000000001e-06,
+      "loss": 2.8198,
+      "step": 463
+    },
+    {
+      "epoch": 2.608450704225352,
+      "grad_norm": 4.290358543395996,
+      "learning_rate": 1.386e-06,
+      "loss": 2.8371,
+      "step": 464
+    },
+    {
+      "epoch": 2.6140845070422536,
+      "grad_norm": 3.481776714324951,
+      "learning_rate": 1.389e-06,
+      "loss": 2.8027,
+      "step": 465
+    },
+    {
+      "epoch": 2.619718309859155,
+      "grad_norm": 2.979186773300171,
+      "learning_rate": 1.392e-06,
+      "loss": 2.8286,
+      "step": 466
+    },
+    {
+      "epoch": 2.6253521126760564,
+      "grad_norm": 4.169500350952148,
+      "learning_rate": 1.395e-06,
+      "loss": 2.8278,
+      "step": 467
+    },
+    {
+      "epoch": 2.630985915492958,
+      "grad_norm": 5.333034515380859,
+      "learning_rate": 1.3980000000000002e-06,
+      "loss": 2.8301,
+      "step": 468
+    },
+    {
+      "epoch": 2.636619718309859,
+      "grad_norm": 6.221430778503418,
+      "learning_rate": 1.401e-06,
+      "loss": 2.8066,
+      "step": 469
+    },
+    {
+      "epoch": 2.6422535211267606,
+      "grad_norm": 4.619940757751465,
+      "learning_rate": 1.404e-06,
+      "loss": 2.7916,
+      "step": 470
+    },
+    {
+      "epoch": 2.647887323943662,
+      "grad_norm": 5.482257843017578,
+      "learning_rate": 1.407e-06,
+      "loss": 2.806,
+      "step": 471
+    },
+    {
+      "epoch": 2.6535211267605634,
+      "grad_norm": 4.32538366317749,
+      "learning_rate": 1.41e-06,
+      "loss": 2.799,
+      "step": 472
+    },
+    {
+      "epoch": 2.659154929577465,
+      "grad_norm": 5.186101913452148,
+      "learning_rate": 1.413e-06,
+      "loss": 2.7999,
+      "step": 473
+    },
+    {
+      "epoch": 2.664788732394366,
+      "grad_norm": 6.246681213378906,
+      "learning_rate": 1.416e-06,
+      "loss": 2.7924,
+      "step": 474
+    },
+    {
+      "epoch": 2.6704225352112676,
+      "grad_norm": 5.235856056213379,
+      "learning_rate": 1.419e-06,
+      "loss": 2.7918,
+      "step": 475
+    },
+    {
+      "epoch": 2.676056338028169,
+      "grad_norm": 2.9274845123291016,
+      "learning_rate": 1.422e-06,
+      "loss": 2.8112,
+      "step": 476
+    },
+    {
+      "epoch": 2.6816901408450704,
+      "grad_norm": 3.573071241378784,
+      "learning_rate": 1.4250000000000001e-06,
+      "loss": 2.7849,
+      "step": 477
+    },
+    {
+      "epoch": 2.687323943661972,
+      "grad_norm": 4.214164733886719,
+      "learning_rate": 1.4280000000000001e-06,
+      "loss": 2.7879,
+      "step": 478
+    },
+    {
+      "epoch": 2.692957746478873,
+      "grad_norm": 5.562944412231445,
+      "learning_rate": 1.431e-06,
+      "loss": 2.7879,
+      "step": 479
+    },
+    {
+      "epoch": 2.6985915492957746,
+      "grad_norm": 5.23793888092041,
+      "learning_rate": 1.4340000000000002e-06,
+      "loss": 2.7858,
+      "step": 480
+    },
+    {
+      "epoch": 2.704225352112676,
+      "grad_norm": 4.238168716430664,
+      "learning_rate": 1.437e-06,
+      "loss": 2.7769,
+      "step": 481
+    },
+    {
+      "epoch": 2.7098591549295774,
+      "grad_norm": 4.448953628540039,
+      "learning_rate": 1.44e-06,
+      "loss": 2.7617,
+      "step": 482
+    },
+    {
+      "epoch": 2.715492957746479,
+      "grad_norm": 7.470413684844971,
+      "learning_rate": 1.443e-06,
+      "loss": 2.7972,
+      "step": 483
+    },
+    {
+      "epoch": 2.72112676056338,
+      "grad_norm": 3.1278200149536133,
+      "learning_rate": 1.446e-06,
+      "loss": 2.7428,
+      "step": 484
+    },
+    {
+      "epoch": 2.7267605633802816,
+      "grad_norm": 3.5966947078704834,
+      "learning_rate": 1.449e-06,
+      "loss": 2.774,
+      "step": 485
+    },
+    {
+      "epoch": 2.732394366197183,
+      "grad_norm": 3.109821319580078,
+      "learning_rate": 1.452e-06,
+      "loss": 2.7867,
+      "step": 486
+    },
+    {
+      "epoch": 2.7380281690140844,
+      "grad_norm": 4.992013931274414,
+      "learning_rate": 1.455e-06,
+      "loss": 2.7909,
+      "step": 487
+    },
+    {
+      "epoch": 2.743661971830986,
+      "grad_norm": 5.205508232116699,
+      "learning_rate": 1.458e-06,
+      "loss": 2.742,
+      "step": 488
+    },
+    {
+      "epoch": 2.749295774647887,
+      "grad_norm": 3.8282198905944824,
+      "learning_rate": 1.461e-06,
+      "loss": 2.9147,
+      "step": 489
+    },
+    {
+      "epoch": 2.7549295774647886,
+      "grad_norm": 3.4786813259124756,
+      "learning_rate": 1.464e-06,
+      "loss": 2.9016,
+      "step": 490
+    },
+    {
+      "epoch": 2.76056338028169,
+      "grad_norm": 4.716727256774902,
+      "learning_rate": 1.467e-06,
+      "loss": 2.8472,
+      "step": 491
+    },
+    {
+      "epoch": 2.7661971830985914,
+      "grad_norm": 2.4909396171569824,
+      "learning_rate": 1.4700000000000001e-06,
+      "loss": 2.8292,
+      "step": 492
+    },
+    {
+      "epoch": 2.771830985915493,
+      "grad_norm": 2.098987102508545,
+      "learning_rate": 1.473e-06,
+      "loss": 2.8354,
+      "step": 493
+    },
+    {
+      "epoch": 2.777464788732394,
+      "grad_norm": 3.8950672149658203,
+      "learning_rate": 1.4760000000000001e-06,
+      "loss": 2.8584,
+      "step": 494
+    },
+    {
+      "epoch": 2.7830985915492956,
+      "grad_norm": 3.389186382293701,
+      "learning_rate": 1.479e-06,
+      "loss": 2.8431,
+      "step": 495
+    },
+    {
+      "epoch": 2.788732394366197,
+      "grad_norm": 2.1209559440612793,
+      "learning_rate": 1.482e-06,
+      "loss": 2.8189,
+      "step": 496
+    },
+    {
+      "epoch": 2.7943661971830984,
+      "grad_norm": 2.8957386016845703,
+      "learning_rate": 1.4850000000000002e-06,
+      "loss": 2.8003,
+      "step": 497
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.9849811792373657,
+      "learning_rate": 1.488e-06,
+      "loss": 2.7981,
+      "step": 498
+    },
+    {
+      "epoch": 2.8056338028169012,
+      "grad_norm": 2.376265287399292,
+      "learning_rate": 1.491e-06,
+      "loss": 2.8231,
+      "step": 499
+    },
+    {
+      "epoch": 2.8112676056338026,
+      "grad_norm": 3.4236159324645996,
+      "learning_rate": 1.494e-06,
+      "loss": 2.8096,
+      "step": 500
+    },
+    {
+      "epoch": 2.816901408450704,
+      "grad_norm": 2.7298974990844727,
+      "learning_rate": 1.497e-06,
+      "loss": 2.8237,
+      "step": 501
+    },
+    {
+      "epoch": 2.8225352112676054,
+      "grad_norm": 2.521082639694214,
+      "learning_rate": 1.5e-06,
+      "loss": 2.765,
+      "step": 502
+    },
+    {
+      "epoch": 2.828169014084507,
+      "grad_norm": 2.644693613052368,
+      "learning_rate": 1.503e-06,
+      "loss": 2.8049,
+      "step": 503
+    },
+    {
+      "epoch": 2.8338028169014082,
+      "grad_norm": 3.143251419067383,
+      "learning_rate": 1.506e-06,
+      "loss": 2.7773,
+      "step": 504
+    },
+    {
+      "epoch": 2.8394366197183096,
+      "grad_norm": 2.4421496391296387,
+      "learning_rate": 1.5089999999999999e-06,
+      "loss": 2.7613,
+      "step": 505
+    },
+    {
+      "epoch": 2.845070422535211,
+      "grad_norm": 3.1848864555358887,
+      "learning_rate": 1.512e-06,
+      "loss": 2.7795,
+      "step": 506
+    },
+    {
+      "epoch": 2.8507042253521124,
+      "grad_norm": 2.911860704421997,
+      "learning_rate": 1.5150000000000001e-06,
+      "loss": 2.7608,
+      "step": 507
+    },
+    {
+      "epoch": 2.856338028169014,
+      "grad_norm": 1.8944320678710938,
+      "learning_rate": 1.518e-06,
+      "loss": 2.7835,
+      "step": 508
+    },
+    {
+      "epoch": 2.8619718309859152,
+      "grad_norm": 2.056074380874634,
+      "learning_rate": 1.5210000000000001e-06,
+      "loss": 2.7737,
+      "step": 509
+    },
+    {
+      "epoch": 2.8676056338028166,
+      "grad_norm": 3.7578797340393066,
+      "learning_rate": 1.524e-06,
+      "loss": 2.7783,
+      "step": 510
+    },
+    {
+      "epoch": 2.873239436619718,
+      "grad_norm": 2.587336778640747,
+      "learning_rate": 1.5270000000000002e-06,
+      "loss": 2.7797,
+      "step": 511
+    },
+    {
+      "epoch": 2.87887323943662,
+      "grad_norm": 5.611008644104004,
+      "learning_rate": 1.53e-06,
+      "loss": 2.744,
+      "step": 512
+    },
+    {
+      "epoch": 2.8845070422535213,
+      "grad_norm": 10.416324615478516,
+      "learning_rate": 1.533e-06,
+      "loss": 2.7366,
+      "step": 513
+    },
+    {
+      "epoch": 2.8901408450704227,
+      "grad_norm": 4.064530849456787,
+      "learning_rate": 1.5360000000000002e-06,
+      "loss": 2.7416,
+      "step": 514
+    },
+    {
+      "epoch": 2.895774647887324,
+      "grad_norm": 6.801990032196045,
+      "learning_rate": 1.539e-06,
+      "loss": 2.7769,
+      "step": 515
+    },
+    {
+      "epoch": 2.9014084507042255,
+      "grad_norm": 2.629598379135132,
+      "learning_rate": 1.542e-06,
+      "loss": 2.744,
+      "step": 516
+    },
+    {
+      "epoch": 2.907042253521127,
+      "grad_norm": 4.535959720611572,
+      "learning_rate": 1.545e-06,
+      "loss": 2.7681,
+      "step": 517
+    },
+    {
+      "epoch": 2.9126760563380283,
+      "grad_norm": 5.370257377624512,
+      "learning_rate": 1.548e-06,
+      "loss": 2.7789,
+      "step": 518
+    },
+    {
+      "epoch": 2.9183098591549297,
+      "grad_norm": 5.6404852867126465,
+      "learning_rate": 1.551e-06,
+      "loss": 2.7694,
+      "step": 519
+    },
+    {
+      "epoch": 2.923943661971831,
+      "grad_norm": 3.5447475910186768,
+      "learning_rate": 1.554e-06,
+      "loss": 2.7573,
+      "step": 520
+    },
+    {
+      "epoch": 2.9295774647887325,
+      "grad_norm": 3.844684362411499,
+      "learning_rate": 1.557e-06,
+      "loss": 2.7499,
+      "step": 521
+    },
+    {
+      "epoch": 2.935211267605634,
+      "grad_norm": 39.249507904052734,
+      "learning_rate": 1.5599999999999999e-06,
+      "loss": 2.7408,
+      "step": 522
+    },
+    {
+      "epoch": 2.9408450704225353,
+      "grad_norm": 2.2351744174957275,
+      "learning_rate": 1.5630000000000001e-06,
+      "loss": 2.7732,
+      "step": 523
+    },
+    {
+      "epoch": 2.9464788732394367,
+      "grad_norm": 4.3039870262146,
+      "learning_rate": 1.5660000000000001e-06,
+      "loss": 2.7418,
+      "step": 524
+    },
+    {
+      "epoch": 2.952112676056338,
+      "grad_norm": 8.005955696105957,
+      "learning_rate": 1.569e-06,
+      "loss": 2.699,
+      "step": 525
+    },
+    {
+      "epoch": 2.9577464788732395,
+      "grad_norm": 4.523921489715576,
+      "learning_rate": 1.5720000000000002e-06,
+      "loss": 2.7148,
+      "step": 526
+    },
+    {
+      "epoch": 2.963380281690141,
+      "grad_norm": 8.109228134155273,
+      "learning_rate": 1.575e-06,
+      "loss": 2.7359,
+      "step": 527
+    },
+    {
+      "epoch": 2.9690140845070423,
+      "grad_norm": 54.502098083496094,
+      "learning_rate": 1.578e-06,
+      "loss": 2.753,
+      "step": 528
+    },
+    {
+      "epoch": 2.9746478873239437,
+      "grad_norm": 4.717286586761475,
+      "learning_rate": 1.581e-06,
+      "loss": 2.7562,
+      "step": 529
+    },
+    {
+      "epoch": 2.980281690140845,
+      "grad_norm": 17.204912185668945,
+      "learning_rate": 1.584e-06,
+      "loss": 2.8052,
+      "step": 530
+    },
+    {
+      "epoch": 2.9859154929577465,
+      "grad_norm": 6.698378562927246,
+      "learning_rate": 1.5870000000000002e-06,
+      "loss": 2.7627,
+      "step": 531
+    },
+    {
+      "epoch": 2.991549295774648,
+      "grad_norm": 9.332170486450195,
+      "learning_rate": 1.59e-06,
+      "loss": 2.7683,
+      "step": 532
+    },
+    {
+      "epoch": 2.9971830985915493,
+      "grad_norm": 2.9608309268951416,
+      "learning_rate": 1.593e-06,
+      "loss": 2.8084,
+      "step": 533
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 3.3101515769958496,
+      "learning_rate": 1.596e-06,
+      "loss": 1.3821,
+      "step": 534
+    },
+    {
+      "epoch": 3.0056338028169014,
+      "grad_norm": 4.513636112213135,
+      "learning_rate": 1.599e-06,
+      "loss": 2.8764,
+      "step": 535
+    },
+    {
+      "epoch": 3.011267605633803,
+      "grad_norm": 6.628481864929199,
+      "learning_rate": 1.602e-06,
+      "loss": 2.8876,
+      "step": 536
+    },
+    {
+      "epoch": 3.016901408450704,
+      "grad_norm": 2.4313488006591797,
+      "learning_rate": 1.605e-06,
+      "loss": 2.8239,
+      "step": 537
+    },
+    {
+      "epoch": 3.0225352112676056,
+      "grad_norm": 3.9404282569885254,
+      "learning_rate": 1.608e-06,
+      "loss": 2.8526,
+      "step": 538
+    },
+    {
+      "epoch": 3.028169014084507,
+      "grad_norm": 2.381598949432373,
+      "learning_rate": 1.611e-06,
+      "loss": 2.8307,
+      "step": 539
+    },
+    {
+      "epoch": 3.0338028169014084,
+      "grad_norm": 3.1737258434295654,
+      "learning_rate": 1.6140000000000001e-06,
+      "loss": 2.8175,
+      "step": 540
+    },
+    {
+      "epoch": 3.03943661971831,
+      "grad_norm": 2.4474072456359863,
+      "learning_rate": 1.6170000000000001e-06,
+      "loss": 2.792,
+      "step": 541
+    },
+    {
+      "epoch": 3.045070422535211,
+      "grad_norm": 4.117803573608398,
+      "learning_rate": 1.62e-06,
+      "loss": 2.7852,
+      "step": 542
+    },
+    {
+      "epoch": 3.0507042253521126,
+      "grad_norm": 2.7527599334716797,
+      "learning_rate": 1.6230000000000002e-06,
+      "loss": 2.7904,
+      "step": 543
+    },
+    {
+      "epoch": 3.056338028169014,
+      "grad_norm": 1.7446880340576172,
+      "learning_rate": 1.626e-06,
+      "loss": 2.7979,
+      "step": 544
+    },
+    {
+      "epoch": 3.0619718309859154,
+      "grad_norm": 2.72400164604187,
+      "learning_rate": 1.629e-06,
+      "loss": 2.7947,
+      "step": 545
+    },
+    {
+      "epoch": 3.067605633802817,
+      "grad_norm": 2.1043014526367188,
+      "learning_rate": 1.632e-06,
+      "loss": 2.7753,
+      "step": 546
+    },
+    {
+      "epoch": 3.073239436619718,
+      "grad_norm": 2.054772138595581,
+      "learning_rate": 1.635e-06,
+      "loss": 2.7834,
+      "step": 547
+    },
+    {
+      "epoch": 3.0788732394366196,
+      "grad_norm": 2.433123826980591,
+      "learning_rate": 1.638e-06,
+      "loss": 2.7527,
+      "step": 548
+    },
+    {
+      "epoch": 3.084507042253521,
+      "grad_norm": 4.187600135803223,
+      "learning_rate": 1.641e-06,
+      "loss": 2.7706,
+      "step": 549
+    },
+    {
+      "epoch": 3.0901408450704224,
+      "grad_norm": 1.6656733751296997,
+      "learning_rate": 1.644e-06,
+      "loss": 2.7446,
+      "step": 550
+    },
+    {
+      "epoch": 3.095774647887324,
+      "grad_norm": 1.757063388824463,
+      "learning_rate": 1.6469999999999999e-06,
+      "loss": 2.729,
+      "step": 551
+    },
+    {
+      "epoch": 3.101408450704225,
+      "grad_norm": 2.2468183040618896,
+      "learning_rate": 1.65e-06,
+      "loss": 2.7629,
+      "step": 552
+    },
+    {
+      "epoch": 3.1070422535211266,
+      "grad_norm": 2.3012444972991943,
+      "learning_rate": 1.653e-06,
+      "loss": 2.7679,
+      "step": 553
+    },
+    {
+      "epoch": 3.112676056338028,
+      "grad_norm": 3.9080018997192383,
+      "learning_rate": 1.6560000000000001e-06,
+      "loss": 2.7438,
+      "step": 554
+    },
+    {
+      "epoch": 3.1183098591549294,
+      "grad_norm": 2.3420796394348145,
+      "learning_rate": 1.6590000000000001e-06,
+      "loss": 2.7693,
+      "step": 555
+    },
+    {
+      "epoch": 3.123943661971831,
+      "grad_norm": 3.211016893386841,
+      "learning_rate": 1.662e-06,
+      "loss": 2.7565,
+      "step": 556
+    },
+    {
+      "epoch": 3.129577464788732,
+      "grad_norm": 4.163708209991455,
+      "learning_rate": 1.6650000000000002e-06,
+      "loss": 2.7658,
+      "step": 557
+    },
+    {
+      "epoch": 3.1352112676056336,
+      "grad_norm": 3.493562698364258,
+      "learning_rate": 1.668e-06,
+      "loss": 2.7368,
+      "step": 558
+    },
+    {
+      "epoch": 3.140845070422535,
+      "grad_norm": 2.458523988723755,
+      "learning_rate": 1.671e-06,
+      "loss": 2.7076,
+      "step": 559
+    },
+    {
+      "epoch": 3.1464788732394364,
+      "grad_norm": 3.26328182220459,
+      "learning_rate": 1.6740000000000002e-06,
+      "loss": 2.6912,
+      "step": 560
+    },
+    {
+      "epoch": 3.152112676056338,
+      "grad_norm": 2.4104485511779785,
+      "learning_rate": 1.677e-06,
+      "loss": 2.7585,
+      "step": 561
+    },
+    {
+      "epoch": 3.1577464788732392,
+      "grad_norm": 2.657015800476074,
+      "learning_rate": 1.68e-06,
+      "loss": 2.7099,
+      "step": 562
+    },
+    {
+      "epoch": 3.1633802816901406,
+      "grad_norm": 2.3434484004974365,
+      "learning_rate": 1.683e-06,
+      "loss": 2.744,
+      "step": 563
+    },
+    {
+      "epoch": 3.169014084507042,
+      "grad_norm": 3.666388988494873,
+      "learning_rate": 1.686e-06,
+      "loss": 2.7251,
+      "step": 564
+    },
+    {
+      "epoch": 3.1746478873239434,
+      "grad_norm": 3.7017822265625,
+      "learning_rate": 1.689e-06,
+      "loss": 2.7218,
+      "step": 565
+    },
+    {
+      "epoch": 3.1802816901408453,
+      "grad_norm": 6.779127597808838,
+      "learning_rate": 1.692e-06,
+      "loss": 2.7156,
+      "step": 566
+    },
+    {
+      "epoch": 3.1859154929577467,
+      "grad_norm": 4.765157222747803,
+      "learning_rate": 1.695e-06,
+      "loss": 2.7168,
+      "step": 567
+    },
+    {
+      "epoch": 3.191549295774648,
+      "grad_norm": 3.8683316707611084,
+      "learning_rate": 1.6979999999999999e-06,
+      "loss": 2.6945,
+      "step": 568
+    },
+    {
+      "epoch": 3.1971830985915495,
+      "grad_norm": 2.9999983310699463,
+      "learning_rate": 1.701e-06,
+      "loss": 2.7323,
+      "step": 569
+    },
+    {
+      "epoch": 3.202816901408451,
+      "grad_norm": 2.4626176357269287,
+      "learning_rate": 1.7040000000000001e-06,
+      "loss": 2.7142,
+      "step": 570
+    },
+    {
+      "epoch": 3.2084507042253523,
+      "grad_norm": 4.122666358947754,
+      "learning_rate": 1.707e-06,
+      "loss": 2.6754,
+      "step": 571
+    },
+    {
+      "epoch": 3.2140845070422537,
+      "grad_norm": 3.5713374614715576,
+      "learning_rate": 1.7100000000000001e-06,
+      "loss": 2.6513,
+      "step": 572
+    },
+    {
+      "epoch": 3.219718309859155,
+      "grad_norm": 4.797949314117432,
+      "learning_rate": 1.713e-06,
+      "loss": 2.688,
+      "step": 573
+    },
+    {
+      "epoch": 3.2253521126760565,
+      "grad_norm": 3.065605640411377,
+      "learning_rate": 1.7160000000000002e-06,
+      "loss": 2.6772,
+      "step": 574
+    },
+    {
+      "epoch": 3.230985915492958,
+      "grad_norm": 2.238821029663086,
+      "learning_rate": 1.719e-06,
+      "loss": 2.6712,
+      "step": 575
+    },
+    {
+      "epoch": 3.2366197183098593,
+      "grad_norm": 2.4874846935272217,
+      "learning_rate": 1.722e-06,
+      "loss": 2.7081,
+      "step": 576
+    },
+    {
+      "epoch": 3.2422535211267607,
+      "grad_norm": 2.7824885845184326,
+      "learning_rate": 1.7250000000000002e-06,
+      "loss": 2.7113,
+      "step": 577
+    },
+    {
+      "epoch": 3.247887323943662,
+      "grad_norm": 3.1023831367492676,
+      "learning_rate": 1.728e-06,
+      "loss": 2.6572,
+      "step": 578
+    },
+    {
+      "epoch": 3.2535211267605635,
+      "grad_norm": 5.776034832000732,
+      "learning_rate": 1.731e-06,
+      "loss": 2.8611,
+      "step": 579
+    },
+    {
+      "epoch": 3.259154929577465,
+      "grad_norm": 2.806821823120117,
+      "learning_rate": 1.734e-06,
+      "loss": 2.8481,
+      "step": 580
+    },
+    {
+      "epoch": 3.2647887323943663,
+      "grad_norm": 3.0129337310791016,
+      "learning_rate": 1.737e-06,
+      "loss": 2.8037,
+      "step": 581
+    },
+    {
+      "epoch": 3.2704225352112677,
+      "grad_norm": 5.805482387542725,
+      "learning_rate": 1.74e-06,
+      "loss": 2.8091,
+      "step": 582
+    },
+    {
+      "epoch": 3.276056338028169,
+      "grad_norm": 9.397015571594238,
+      "learning_rate": 1.743e-06,
+      "loss": 2.8086,
+      "step": 583
+    },
+    {
+      "epoch": 3.2816901408450705,
+      "grad_norm": 2.8330020904541016,
+      "learning_rate": 1.746e-06,
+      "loss": 2.7994,
+      "step": 584
+    },
+    {
+      "epoch": 3.287323943661972,
+      "grad_norm": 3.208578109741211,
+      "learning_rate": 1.749e-06,
+      "loss": 2.7591,
+      "step": 585
+    },
+    {
+      "epoch": 3.2929577464788733,
+      "grad_norm": 4.422690391540527,
+      "learning_rate": 1.7520000000000001e-06,
+      "loss": 2.7969,
+      "step": 586
+    },
+    {
+      "epoch": 3.2985915492957747,
+      "grad_norm": 2.6905901432037354,
+      "learning_rate": 1.7550000000000001e-06,
+      "loss": 2.7611,
+      "step": 587
+    },
+    {
+      "epoch": 3.304225352112676,
+      "grad_norm": 4.492880821228027,
+      "learning_rate": 1.758e-06,
+      "loss": 2.7692,
+      "step": 588
+    },
+    {
+      "epoch": 3.3098591549295775,
+      "grad_norm": 2.1210246086120605,
+      "learning_rate": 1.7610000000000002e-06,
+      "loss": 2.7765,
+      "step": 589
+    },
+    {
+      "epoch": 3.315492957746479,
+      "grad_norm": 2.7414491176605225,
+      "learning_rate": 1.764e-06,
+      "loss": 2.7318,
+      "step": 590
+    },
+    {
+      "epoch": 3.3211267605633803,
+      "grad_norm": 2.884840965270996,
+      "learning_rate": 1.767e-06,
+      "loss": 2.7482,
+      "step": 591
+    },
+    {
+      "epoch": 3.3267605633802817,
+      "grad_norm": 3.144826650619507,
+      "learning_rate": 1.77e-06,
+      "loss": 2.7188,
+      "step": 592
+    },
+    {
+      "epoch": 3.332394366197183,
+      "grad_norm": 2.618098020553589,
+      "learning_rate": 1.773e-06,
+      "loss": 2.7531,
+      "step": 593
+    },
+    {
+      "epoch": 3.3380281690140845,
+      "grad_norm": 1.7363290786743164,
+      "learning_rate": 1.776e-06,
+      "loss": 2.7328,
+      "step": 594
+    },
+    {
+      "epoch": 3.343661971830986,
+      "grad_norm": 3.7917160987854004,
+      "learning_rate": 1.779e-06,
+      "loss": 2.7158,
+      "step": 595
+    },
+    {
+      "epoch": 3.3492957746478873,
+      "grad_norm": 3.086524248123169,
+      "learning_rate": 1.782e-06,
+      "loss": 2.7073,
+      "step": 596
+    },
+    {
+      "epoch": 3.3549295774647887,
+      "grad_norm": 1.768385887145996,
+      "learning_rate": 1.785e-06,
+      "loss": 2.7167,
+      "step": 597
+    },
+    {
+      "epoch": 3.36056338028169,
+      "grad_norm": 3.622553586959839,
+      "learning_rate": 1.788e-06,
+      "loss": 2.7171,
+      "step": 598
+    },
+    {
+      "epoch": 3.3661971830985915,
+      "grad_norm": 2.9177920818328857,
+      "learning_rate": 1.791e-06,
+      "loss": 2.7344,
+      "step": 599
+    },
+    {
+      "epoch": 3.371830985915493,
+      "grad_norm": 2.422119140625,
+      "learning_rate": 1.794e-06,
+      "loss": 2.6836,
+      "step": 600
+    },
+    {
+      "epoch": 3.3774647887323943,
+      "grad_norm": 3.2521135807037354,
+      "learning_rate": 1.7970000000000001e-06,
+      "loss": 2.6564,
+      "step": 601
+    },
+    {
+      "epoch": 3.3830985915492957,
+      "grad_norm": 3.1682698726654053,
+      "learning_rate": 1.8e-06,
+      "loss": 2.6881,
+      "step": 602
+    },
+    {
+      "epoch": 3.388732394366197,
+      "grad_norm": 2.0565860271453857,
+      "learning_rate": 1.8030000000000001e-06,
+      "loss": 2.681,
+      "step": 603
+    },
+    {
+      "epoch": 3.3943661971830985,
+      "grad_norm": 5.913877964019775,
+      "learning_rate": 1.806e-06,
+      "loss": 2.66,
+      "step": 604
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 5.101724624633789,
+      "learning_rate": 1.809e-06,
+      "loss": 2.7031,
+      "step": 605
+    },
+    {
+      "epoch": 3.4056338028169013,
+      "grad_norm": 4.672321319580078,
+      "learning_rate": 1.8120000000000002e-06,
+      "loss": 2.6847,
+      "step": 606
+    },
+    {
+      "epoch": 3.4112676056338027,
+      "grad_norm": 3.59171462059021,
+      "learning_rate": 1.815e-06,
+      "loss": 2.6721,
+      "step": 607
+    },
+    {
+      "epoch": 3.416901408450704,
+      "grad_norm": 4.02622652053833,
+      "learning_rate": 1.818e-06,
+      "loss": 2.6493,
+      "step": 608
+    },
+    {
+      "epoch": 3.4225352112676055,
+      "grad_norm": 6.657413959503174,
+      "learning_rate": 1.821e-06,
+      "loss": 2.6754,
+      "step": 609
+    },
+    {
+      "epoch": 3.428169014084507,
+      "grad_norm": 6.219770908355713,
+      "learning_rate": 1.824e-06,
+      "loss": 2.6491,
+      "step": 610
+    },
+    {
+      "epoch": 3.4338028169014083,
+      "grad_norm": 4.524576187133789,
+      "learning_rate": 1.827e-06,
+      "loss": 2.676,
+      "step": 611
+    },
+    {
+      "epoch": 3.4394366197183097,
+      "grad_norm": 2.754516363143921,
+      "learning_rate": 1.83e-06,
+      "loss": 2.6256,
+      "step": 612
+    },
+    {
+      "epoch": 3.445070422535211,
+      "grad_norm": 7.4485979080200195,
+      "learning_rate": 1.833e-06,
+      "loss": 2.6068,
+      "step": 613
+    },
+    {
+      "epoch": 3.4507042253521125,
+      "grad_norm": 5.208416938781738,
+      "learning_rate": 1.8359999999999999e-06,
+      "loss": 2.6381,
+      "step": 614
+    },
+    {
+      "epoch": 3.456338028169014,
+      "grad_norm": 5.324192047119141,
+      "learning_rate": 1.839e-06,
+      "loss": 2.6656,
+      "step": 615
+    },
+    {
+      "epoch": 3.4619718309859153,
+      "grad_norm": 3.2816479206085205,
+      "learning_rate": 1.8420000000000001e-06,
+      "loss": 2.6474,
+      "step": 616
+    },
+    {
+      "epoch": 3.4676056338028167,
+      "grad_norm": 7.767096519470215,
+      "learning_rate": 1.8450000000000001e-06,
+      "loss": 2.6936,
+      "step": 617
+    },
+    {
+      "epoch": 3.473239436619718,
+      "grad_norm": 3.7725744247436523,
+      "learning_rate": 1.8480000000000001e-06,
+      "loss": 2.6174,
+      "step": 618
+    },
+    {
+      "epoch": 3.4788732394366195,
+      "grad_norm": 5.0832905769348145,
+      "learning_rate": 1.851e-06,
+      "loss": 2.6354,
+      "step": 619
+    },
+    {
+      "epoch": 3.4845070422535214,
+      "grad_norm": 5.80606746673584,
+      "learning_rate": 1.8540000000000002e-06,
+      "loss": 2.5779,
+      "step": 620
+    },
+    {
+      "epoch": 3.4901408450704228,
+      "grad_norm": 3.8659050464630127,
+      "learning_rate": 1.857e-06,
+      "loss": 2.5956,
+      "step": 621
+    },
+    {
+      "epoch": 3.495774647887324,
+      "grad_norm": 5.184934616088867,
+      "learning_rate": 1.86e-06,
+      "loss": 2.6341,
+      "step": 622
+    },
+    {
+      "epoch": 3.5014084507042256,
+      "grad_norm": 4.191617488861084,
+      "learning_rate": 1.8630000000000002e-06,
+      "loss": 2.8809,
+      "step": 623
+    },
+    {
+      "epoch": 3.507042253521127,
+      "grad_norm": 5.942683219909668,
+      "learning_rate": 1.866e-06,
+      "loss": 2.7936,
+      "step": 624
+    },
+    {
+      "epoch": 3.5126760563380284,
+      "grad_norm": 3.73496413230896,
+      "learning_rate": 1.869e-06,
+      "loss": 2.7653,
+      "step": 625
+    },
+    {
+      "epoch": 3.5183098591549298,
+      "grad_norm": 4.427114009857178,
+      "learning_rate": 1.872e-06,
+      "loss": 2.7579,
+      "step": 626
+    },
+    {
+      "epoch": 3.523943661971831,
+      "grad_norm": 7.355522632598877,
+      "learning_rate": 1.875e-06,
+      "loss": 2.7391,
+      "step": 627
+    },
+    {
+      "epoch": 3.5295774647887326,
+      "grad_norm": 5.933565616607666,
+      "learning_rate": 1.878e-06,
+      "loss": 2.7299,
+      "step": 628
+    },
+    {
+      "epoch": 3.535211267605634,
+      "grad_norm": 2.801149845123291,
+      "learning_rate": 1.8810000000000003e-06,
+      "loss": 2.7234,
+      "step": 629
+    },
+    {
+      "epoch": 3.5408450704225354,
+      "grad_norm": 3.7464067935943604,
+      "learning_rate": 1.8839999999999999e-06,
+      "loss": 2.7146,
+      "step": 630
+    },
+    {
+      "epoch": 3.546478873239437,
+      "grad_norm": 4.548725605010986,
+      "learning_rate": 1.8869999999999999e-06,
+      "loss": 2.7071,
+      "step": 631
+    },
+    {
+      "epoch": 3.552112676056338,
+      "grad_norm": 5.175673961639404,
+      "learning_rate": 1.8900000000000001e-06,
+      "loss": 2.7026,
+      "step": 632
+    },
+    {
+      "epoch": 3.5577464788732396,
+      "grad_norm": 2.5320608615875244,
+      "learning_rate": 1.8930000000000001e-06,
+      "loss": 2.7358,
+      "step": 633
+    },
+    {
+      "epoch": 3.563380281690141,
+      "grad_norm": 2.2658591270446777,
+      "learning_rate": 1.8960000000000001e-06,
+      "loss": 2.693,
+      "step": 634
+    },
+    {
+      "epoch": 3.5690140845070424,
+      "grad_norm": 2.9923315048217773,
+      "learning_rate": 1.899e-06,
+      "loss": 2.6586,
+      "step": 635
+    },
+    {
+      "epoch": 3.574647887323944,
+      "grad_norm": 1.8636302947998047,
+      "learning_rate": 1.902e-06,
+      "loss": 2.6489,
+      "step": 636
+    },
+    {
+      "epoch": 3.580281690140845,
+      "grad_norm": 2.93691349029541,
+      "learning_rate": 1.905e-06,
+      "loss": 2.6609,
+      "step": 637
+    },
+    {
+      "epoch": 3.5859154929577466,
+      "grad_norm": 3.5617032051086426,
+      "learning_rate": 1.908e-06,
+      "loss": 2.659,
+      "step": 638
+    },
+    {
+      "epoch": 3.591549295774648,
+      "grad_norm": 3.435441732406616,
+      "learning_rate": 1.9110000000000004e-06,
+      "loss": 2.6154,
+      "step": 639
+    },
+    {
+      "epoch": 3.5971830985915494,
+      "grad_norm": 4.196032524108887,
+      "learning_rate": 1.914e-06,
+      "loss": 2.616,
+      "step": 640
+    },
+    {
+      "epoch": 3.602816901408451,
+      "grad_norm": 3.23492169380188,
+      "learning_rate": 1.917e-06,
+      "loss": 2.6048,
+      "step": 641
+    },
+    {
+      "epoch": 3.608450704225352,
+      "grad_norm": 2.949294328689575,
+      "learning_rate": 1.9200000000000003e-06,
+      "loss": 2.598,
+      "step": 642
+    },
+    {
+      "epoch": 3.6140845070422536,
+      "grad_norm": 2.844964027404785,
+      "learning_rate": 1.923e-06,
+      "loss": 2.6041,
+      "step": 643
+    },
+    {
+      "epoch": 3.619718309859155,
+      "grad_norm": 3.5030791759490967,
+      "learning_rate": 1.926e-06,
+      "loss": 2.5685,
+      "step": 644
+    },
+    {
+      "epoch": 3.6253521126760564,
+      "grad_norm": 10.395556449890137,
+      "learning_rate": 1.929e-06,
+      "loss": 2.576,
+      "step": 645
+    },
+    {
+      "epoch": 3.630985915492958,
+      "grad_norm": 2.640410900115967,
+      "learning_rate": 1.932e-06,
+      "loss": 2.5623,
+      "step": 646
+    },
+    {
+      "epoch": 3.636619718309859,
+      "grad_norm": 3.267350673675537,
+      "learning_rate": 1.935e-06,
+      "loss": 2.5574,
+      "step": 647
+    },
+    {
+      "epoch": 3.6422535211267606,
+      "grad_norm": 9.421785354614258,
+      "learning_rate": 1.9380000000000003e-06,
+      "loss": 2.539,
+      "step": 648
+    },
+    {
+      "epoch": 3.647887323943662,
+      "grad_norm": 5.227574825286865,
+      "learning_rate": 1.9409999999999997e-06,
+      "loss": 2.5801,
+      "step": 649
+    },
+    {
+      "epoch": 3.6535211267605634,
+      "grad_norm": 2.5911121368408203,
+      "learning_rate": 1.944e-06,
+      "loss": 2.4754,
+      "step": 650
+    },
+    {
+      "epoch": 3.659154929577465,
+      "grad_norm": 3.5224828720092773,
+      "learning_rate": 1.947e-06,
+      "loss": 2.5001,
+      "step": 651
+    },
+    {
+      "epoch": 3.664788732394366,
+      "grad_norm": 4.783766269683838,
+      "learning_rate": 1.95e-06,
+      "loss": 2.5012,
+      "step": 652
+    },
+    {
+      "epoch": 3.6704225352112676,
+      "grad_norm": 3.3553149700164795,
+      "learning_rate": 1.953e-06,
+      "loss": 2.531,
+      "step": 653
+    },
+    {
+      "epoch": 3.676056338028169,
+      "grad_norm": 2.931241512298584,
+      "learning_rate": 1.956e-06,
+      "loss": 2.497,
+      "step": 654
+    },
+    {
+      "epoch": 3.6816901408450704,
+      "grad_norm": 4.4694695472717285,
+      "learning_rate": 1.959e-06,
+      "loss": 2.4942,
+      "step": 655
+    },
+    {
+      "epoch": 3.687323943661972,
+      "grad_norm": 3.122934103012085,
+      "learning_rate": 1.962e-06,
+      "loss": 2.4754,
+      "step": 656
+    },
+    {
+      "epoch": 3.692957746478873,
+      "grad_norm": 6.193077087402344,
+      "learning_rate": 1.9650000000000002e-06,
+      "loss": 2.4694,
+      "step": 657
+    },
+    {
+      "epoch": 3.6985915492957746,
+      "grad_norm": 2.414370059967041,
+      "learning_rate": 1.968e-06,
+      "loss": 2.4517,
+      "step": 658
+    },
+    {
+      "epoch": 3.704225352112676,
+      "grad_norm": 3.34991717338562,
+      "learning_rate": 1.971e-06,
+      "loss": 2.4296,
+      "step": 659
+    },
+    {
+      "epoch": 3.7098591549295774,
+      "grad_norm": 6.847357749938965,
+      "learning_rate": 1.974e-06,
+      "loss": 2.4321,
+      "step": 660
+    },
+    {
+      "epoch": 3.715492957746479,
+      "grad_norm": 3.4056034088134766,
+      "learning_rate": 1.977e-06,
+      "loss": 2.4373,
+      "step": 661
+    },
+    {
+      "epoch": 3.72112676056338,
+      "grad_norm": 2.5987486839294434,
+      "learning_rate": 1.98e-06,
+      "loss": 2.4231,
+      "step": 662
+    },
+    {
+      "epoch": 3.7267605633802816,
+      "grad_norm": 2.899972915649414,
+      "learning_rate": 1.9830000000000003e-06,
+      "loss": 2.4102,
+      "step": 663
+    },
+    {
+      "epoch": 3.732394366197183,
+      "grad_norm": 4.216456890106201,
+      "learning_rate": 1.9859999999999997e-06,
+      "loss": 2.4324,
+      "step": 664
+    },
+    {
+      "epoch": 3.7380281690140844,
+      "grad_norm": 3.7976484298706055,
+      "learning_rate": 1.989e-06,
+      "loss": 2.3899,
+      "step": 665
+    },
+    {
+      "epoch": 3.743661971830986,
+      "grad_norm": 13.57735538482666,
+      "learning_rate": 1.992e-06,
+      "loss": 2.3653,
+      "step": 666
+    },
+    {
+      "epoch": 3.749295774647887,
+      "grad_norm": 5.536102771759033,
+      "learning_rate": 1.995e-06,
+      "loss": 2.6115,
+      "step": 667
+    },
+    {
+      "epoch": 3.7549295774647886,
+      "grad_norm": 3.1471431255340576,
+      "learning_rate": 1.998e-06,
+      "loss": 2.6113,
+      "step": 668
+    },
+    {
+      "epoch": 3.76056338028169,
+      "grad_norm": 2.3586504459381104,
+      "learning_rate": 2.001e-06,
+      "loss": 2.5347,
+      "step": 669
+    },
+    {
+      "epoch": 3.7661971830985914,
+      "grad_norm": 3.342392683029175,
+      "learning_rate": 2.004e-06,
+      "loss": 2.5331,
+      "step": 670
+    },
+    {
+      "epoch": 3.771830985915493,
+      "grad_norm": 3.824932098388672,
+      "learning_rate": 2.007e-06,
+      "loss": 2.4711,
+      "step": 671
+    },
+    {
+      "epoch": 3.777464788732394,
+      "grad_norm": 2.8359029293060303,
+      "learning_rate": 2.0100000000000002e-06,
+      "loss": 2.454,
+      "step": 672
+    },
+    {
+      "epoch": 3.7830985915492956,
+      "grad_norm": 3.8589389324188232,
+      "learning_rate": 2.0130000000000005e-06,
+      "loss": 2.3785,
+      "step": 673
+    },
+    {
+      "epoch": 3.788732394366197,
+      "grad_norm": 4.179661750793457,
+      "learning_rate": 2.016e-06,
+      "loss": 2.4114,
+      "step": 674
+    },
+    {
+      "epoch": 3.7943661971830984,
+      "grad_norm": 4.384422302246094,
+      "learning_rate": 2.019e-06,
+      "loss": 2.399,
+      "step": 675
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 2.1641438007354736,
+      "learning_rate": 2.0220000000000003e-06,
+      "loss": 2.3519,
+      "step": 676
+    },
+    {
+      "epoch": 3.8056338028169012,
+      "grad_norm": 4.078607082366943,
+      "learning_rate": 2.025e-06,
+      "loss": 2.3549,
+      "step": 677
+    },
+    {
+      "epoch": 3.8112676056338026,
+      "grad_norm": 4.0506181716918945,
+      "learning_rate": 2.028e-06,
+      "loss": 2.3168,
+      "step": 678
+    },
+    {
+      "epoch": 3.816901408450704,
+      "grad_norm": 3.3806586265563965,
+      "learning_rate": 2.031e-06,
+      "loss": 2.3383,
+      "step": 679
+    },
+    {
+      "epoch": 3.8225352112676054,
+      "grad_norm": 2.962216854095459,
+      "learning_rate": 2.034e-06,
+      "loss": 2.2476,
+      "step": 680
+    },
+    {
+      "epoch": 3.828169014084507,
+      "grad_norm": 2.2728936672210693,
+      "learning_rate": 2.037e-06,
+      "loss": 2.2852,
+      "step": 681
+    },
+    {
+      "epoch": 3.8338028169014082,
+      "grad_norm": 2.4136433601379395,
+      "learning_rate": 2.0400000000000004e-06,
+      "loss": 2.2914,
+      "step": 682
+    },
+    {
+      "epoch": 3.8394366197183096,
+      "grad_norm": 2.7648229598999023,
+      "learning_rate": 2.0429999999999998e-06,
+      "loss": 2.1966,
+      "step": 683
+    },
+    {
+      "epoch": 3.845070422535211,
+      "grad_norm": 2.779562473297119,
+      "learning_rate": 2.046e-06,
+      "loss": 2.2622,
+      "step": 684
+    },
+    {
+      "epoch": 3.8507042253521124,
+      "grad_norm": 2.7937114238739014,
+      "learning_rate": 2.049e-06,
+      "loss": 2.1844,
+      "step": 685
+    },
+    {
+      "epoch": 3.856338028169014,
+      "grad_norm": 2.820657730102539,
+      "learning_rate": 2.052e-06,
+      "loss": 2.1237,
+      "step": 686
+    },
+    {
+      "epoch": 3.8619718309859152,
+      "grad_norm": 3.204495668411255,
+      "learning_rate": 2.0550000000000002e-06,
+      "loss": 2.1909,
+      "step": 687
+    },
+    {
+      "epoch": 3.8676056338028166,
+      "grad_norm": 3.271052598953247,
+      "learning_rate": 2.058e-06,
+      "loss": 2.1032,
+      "step": 688
+    },
+    {
+      "epoch": 3.873239436619718,
+      "grad_norm": 3.531216621398926,
+      "learning_rate": 2.061e-06,
+      "loss": 2.144,
+      "step": 689
+    },
+    {
+      "epoch": 3.87887323943662,
+      "grad_norm": 2.4491655826568604,
+      "learning_rate": 2.064e-06,
+      "loss": 2.0294,
+      "step": 690
+    },
+    {
+      "epoch": 3.8845070422535213,
+      "grad_norm": 4.347051620483398,
+      "learning_rate": 2.0670000000000003e-06,
+      "loss": 2.0617,
+      "step": 691
+    },
+    {
+      "epoch": 3.8901408450704227,
+      "grad_norm": 4.764668941497803,
+      "learning_rate": 2.07e-06,
+      "loss": 2.0337,
+      "step": 692
+    },
+    {
+      "epoch": 3.895774647887324,
+      "grad_norm": 7.034329891204834,
+      "learning_rate": 2.073e-06,
+      "loss": 1.9692,
+      "step": 693
+    },
+    {
+      "epoch": 3.9014084507042255,
+      "grad_norm": 2.9240710735321045,
+      "learning_rate": 2.076e-06,
+      "loss": 1.9619,
+      "step": 694
+    },
+    {
+      "epoch": 3.907042253521127,
+      "grad_norm": 4.533735752105713,
+      "learning_rate": 2.079e-06,
+      "loss": 1.9579,
+      "step": 695
+    },
+    {
+      "epoch": 3.9126760563380283,
+      "grad_norm": 2.520022392272949,
+      "learning_rate": 2.082e-06,
+      "loss": 2.0109,
+      "step": 696
+    },
+    {
+      "epoch": 3.9183098591549297,
+      "grad_norm": 2.678165912628174,
+      "learning_rate": 2.0850000000000004e-06,
+      "loss": 1.9549,
+      "step": 697
+    },
+    {
+      "epoch": 3.923943661971831,
+      "grad_norm": 3.2894513607025146,
+      "learning_rate": 2.0879999999999997e-06,
+      "loss": 1.8676,
+      "step": 698
+    },
+    {
+      "epoch": 3.9295774647887325,
+      "grad_norm": 2.6271660327911377,
+      "learning_rate": 2.091e-06,
+      "loss": 1.9041,
+      "step": 699
+    },
+    {
+      "epoch": 3.935211267605634,
+      "grad_norm": 4.272736549377441,
+      "learning_rate": 2.094e-06,
+      "loss": 1.8641,
+      "step": 700
+    },
+    {
+      "epoch": 3.9408450704225353,
+      "grad_norm": 3.6323680877685547,
+      "learning_rate": 2.097e-06,
+      "loss": 1.7602,
+      "step": 701
+    },
+    {
+      "epoch": 3.9464788732394367,
+      "grad_norm": 3.2892682552337646,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 1.8483,
+      "step": 702
+    },
+    {
+      "epoch": 3.952112676056338,
+      "grad_norm": 3.96871018409729,
+      "learning_rate": 2.103e-06,
+      "loss": 1.7396,
+      "step": 703
+    },
+    {
+      "epoch": 3.9577464788732395,
+      "grad_norm": 3.023526668548584,
+      "learning_rate": 2.106e-06,
+      "loss": 1.7653,
+      "step": 704
+    },
+    {
+      "epoch": 3.963380281690141,
+      "grad_norm": 3.367448091506958,
+      "learning_rate": 2.109e-06,
+      "loss": 1.7519,
+      "step": 705
+    },
+    {
+      "epoch": 3.9690140845070423,
+      "grad_norm": 3.7302796840667725,
+      "learning_rate": 2.1120000000000003e-06,
+      "loss": 1.6711,
+      "step": 706
+    },
+    {
+      "epoch": 3.9746478873239437,
+      "grad_norm": 3.7805285453796387,
+      "learning_rate": 2.1149999999999997e-06,
+      "loss": 1.7193,
+      "step": 707
+    },
+    {
+      "epoch": 3.980281690140845,
+      "grad_norm": 3.4666497707366943,
+      "learning_rate": 2.118e-06,
+      "loss": 1.669,
+      "step": 708
+    },
+    {
+      "epoch": 3.9859154929577465,
+      "grad_norm": 4.103410243988037,
+      "learning_rate": 2.121e-06,
+      "loss": 1.7194,
+      "step": 709
+    },
+    {
+      "epoch": 3.991549295774648,
+      "grad_norm": NaN,
+      "learning_rate": 2.121e-06,
+      "loss": 1.7002,
+      "step": 710
+    },
+    {
+      "epoch": 3.9971830985915493,
+      "grad_norm": 5.339576244354248,
+      "learning_rate": 2.124e-06,
+      "loss": 1.8009,
+      "step": 711
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 6.8820624351501465,
+      "learning_rate": 2.127e-06,
+      "loss": 0.8197,
+      "step": 712
+    },
+    {
+      "epoch": 4.005633802816901,
+      "grad_norm": 8.731258392333984,
+      "learning_rate": 2.13e-06,
+      "loss": 2.1075,
+      "step": 713
+    },
+    {
+      "epoch": 4.011267605633803,
+      "grad_norm": 4.658092498779297,
+      "learning_rate": 2.133e-06,
+      "loss": 2.0035,
+      "step": 714
+    },
+    {
+      "epoch": 4.016901408450704,
+      "grad_norm": 3.997706890106201,
+      "learning_rate": 2.136e-06,
+      "loss": 1.9042,
+      "step": 715
+    },
+    {
+      "epoch": 4.022535211267606,
+      "grad_norm": 5.904595375061035,
+      "learning_rate": 2.139e-06,
+      "loss": 1.8859,
+      "step": 716
+    },
+    {
+      "epoch": 4.028169014084507,
+      "grad_norm": 4.508280277252197,
+      "learning_rate": 2.1420000000000004e-06,
+      "loss": 1.7883,
+      "step": 717
+    },
+    {
+      "epoch": 4.033802816901408,
+      "grad_norm": 3.468778371810913,
+      "learning_rate": 2.145e-06,
+      "loss": 1.7626,
+      "step": 718
+    },
+    {
+      "epoch": 4.03943661971831,
+      "grad_norm": 3.4685027599334717,
+      "learning_rate": 2.148e-06,
+      "loss": 1.7001,
+      "step": 719
+    },
+    {
+      "epoch": 4.045070422535211,
+      "grad_norm": 3.296505928039551,
+      "learning_rate": 2.1510000000000002e-06,
+      "loss": 1.669,
+      "step": 720
+    },
+    {
+      "epoch": 4.050704225352113,
+      "grad_norm": 4.874464988708496,
+      "learning_rate": 2.154e-06,
+      "loss": 1.705,
+      "step": 721
+    },
+    {
+      "epoch": 4.056338028169014,
+      "grad_norm": 4.497097969055176,
+      "learning_rate": 2.1570000000000003e-06,
+      "loss": 1.6824,
+      "step": 722
+    },
+    {
+      "epoch": 4.061971830985915,
+      "grad_norm": 3.7446484565734863,
+      "learning_rate": 2.16e-06,
+      "loss": 1.7157,
+      "step": 723
+    },
+    {
+      "epoch": 4.067605633802817,
+      "grad_norm": 3.5548088550567627,
+      "learning_rate": 2.163e-06,
+      "loss": 1.6035,
+      "step": 724
+    },
+    {
+      "epoch": 4.073239436619718,
+      "grad_norm": 3.665282964706421,
+      "learning_rate": 2.166e-06,
+      "loss": 1.5725,
+      "step": 725
+    },
+    {
+      "epoch": 4.07887323943662,
+      "grad_norm": 3.388298749923706,
+      "learning_rate": 2.1690000000000003e-06,
+      "loss": 1.5237,
+      "step": 726
+    },
+    {
+      "epoch": 4.084507042253521,
+      "grad_norm": 2.3221206665039062,
+      "learning_rate": 2.172e-06,
+      "loss": 1.5083,
+      "step": 727
+    },
+    {
+      "epoch": 4.090140845070422,
+      "grad_norm": 14.295064926147461,
+      "learning_rate": 2.175e-06,
+      "loss": 1.4483,
+      "step": 728
+    },
+    {
+      "epoch": 4.095774647887324,
+      "grad_norm": 4.429758548736572,
+      "learning_rate": 2.178e-06,
+      "loss": 1.4727,
+      "step": 729
+    },
+    {
+      "epoch": 4.101408450704225,
+      "grad_norm": 4.3126959800720215,
+      "learning_rate": 2.181e-06,
+      "loss": 1.4766,
+      "step": 730
+    },
+    {
+      "epoch": 4.107042253521127,
+      "grad_norm": 2.687000036239624,
+      "learning_rate": 2.184e-06,
+      "loss": 1.4647,
+      "step": 731
+    },
+    {
+      "epoch": 4.112676056338028,
+      "grad_norm": 3.530883550643921,
+      "learning_rate": 2.1870000000000004e-06,
+      "loss": 1.4935,
+      "step": 732
+    },
+    {
+      "epoch": 4.118309859154929,
+      "grad_norm": 4.200030326843262,
+      "learning_rate": 2.1899999999999998e-06,
+      "loss": 1.4652,
+      "step": 733
+    },
+    {
+      "epoch": 4.123943661971831,
+      "grad_norm": 7.423676013946533,
+      "learning_rate": 2.193e-06,
+      "loss": 1.3852,
+      "step": 734
+    },
+    {
+      "epoch": 4.129577464788732,
+      "grad_norm": 2.8539798259735107,
+      "learning_rate": 2.1960000000000002e-06,
+      "loss": 1.3247,
+      "step": 735
+    },
+    {
+      "epoch": 4.135211267605634,
+      "grad_norm": 3.968510627746582,
+      "learning_rate": 2.199e-06,
+      "loss": 1.2985,
+      "step": 736
+    },
+    {
+      "epoch": 4.140845070422535,
+      "grad_norm": 2.972294807434082,
+      "learning_rate": 2.2020000000000003e-06,
+      "loss": 1.3183,
+      "step": 737
+    },
+    {
+      "epoch": 4.146478873239436,
+      "grad_norm": 3.018853187561035,
+      "learning_rate": 2.205e-06,
+      "loss": 1.2738,
+      "step": 738
+    },
+    {
+      "epoch": 4.152112676056338,
+      "grad_norm": 2.274994134902954,
+      "learning_rate": 2.208e-06,
+      "loss": 1.3257,
+      "step": 739
+    },
+    {
+      "epoch": 4.157746478873239,
+      "grad_norm": 2.3417112827301025,
+      "learning_rate": 2.211e-06,
+      "loss": 1.2046,
+      "step": 740
+    },
+    {
+      "epoch": 4.163380281690141,
+      "grad_norm": 3.256418228149414,
+      "learning_rate": 2.2140000000000003e-06,
+      "loss": 1.2625,
+      "step": 741
+    },
+    {
+      "epoch": 4.169014084507042,
+      "grad_norm": 2.557704210281372,
+      "learning_rate": 2.2169999999999997e-06,
+      "loss": 1.2124,
+      "step": 742
+    },
+    {
+      "epoch": 4.174647887323943,
+      "grad_norm": 3.7538537979125977,
+      "learning_rate": 2.22e-06,
+      "loss": 1.1492,
+      "step": 743
+    },
+    {
+      "epoch": 4.180281690140845,
+      "grad_norm": 2.9859628677368164,
+      "learning_rate": 2.223e-06,
+      "loss": 1.2012,
+      "step": 744
+    },
+    {
+      "epoch": 4.185915492957746,
+      "grad_norm": 3.2278072834014893,
+      "learning_rate": 2.226e-06,
+      "loss": 1.2356,
+      "step": 745
+    },
+    {
+      "epoch": 4.191549295774648,
+      "grad_norm": 3.32592511177063,
+      "learning_rate": 2.229e-06,
+      "loss": 1.1824,
+      "step": 746
+    },
+    {
+      "epoch": 4.197183098591549,
+      "grad_norm": 4.771190643310547,
+      "learning_rate": 2.232e-06,
+      "loss": 1.1087,
+      "step": 747
+    },
+    {
+      "epoch": 4.20281690140845,
+      "grad_norm": 2.638829231262207,
+      "learning_rate": 2.2349999999999998e-06,
+      "loss": 1.1162,
+      "step": 748
+    },
+    {
+      "epoch": 4.208450704225352,
+      "grad_norm": 6.826143264770508,
+      "learning_rate": 2.238e-06,
+      "loss": 1.1157,
+      "step": 749
+    },
+    {
+      "epoch": 4.214084507042253,
+      "grad_norm": 3.0462405681610107,
+      "learning_rate": 2.2410000000000002e-06,
+      "loss": 1.0576,
+      "step": 750
+    },
+    {
+      "epoch": 4.219718309859155,
+      "grad_norm": 4.8925909996032715,
+      "learning_rate": 2.244e-06,
+      "loss": 1.2062,
+      "step": 751
+    },
+    {
+      "epoch": 4.225352112676056,
+      "grad_norm": 5.137844562530518,
+      "learning_rate": 2.247e-06,
+      "loss": 1.1099,
+      "step": 752
+    },
+    {
+      "epoch": 4.230985915492957,
+      "grad_norm": 4.4226861000061035,
+      "learning_rate": 2.25e-06,
+      "loss": 1.0575,
+      "step": 753
+    },
+    {
+      "epoch": 4.236619718309859,
+      "grad_norm": 8.942727088928223,
+      "learning_rate": 2.253e-06,
+      "loss": 1.0234,
+      "step": 754
+    },
+    {
+      "epoch": 4.24225352112676,
+      "grad_norm": 5.536681652069092,
+      "learning_rate": 2.256e-06,
+      "loss": 1.0121,
+      "step": 755
+    },
+    {
+      "epoch": 4.247887323943662,
+      "grad_norm": 5.978938102722168,
+      "learning_rate": 2.2590000000000003e-06,
+      "loss": 0.9913,
+      "step": 756
+    },
+    {
+      "epoch": 4.253521126760563,
+      "grad_norm": 9.773181915283203,
+      "learning_rate": 2.262e-06,
+      "loss": 1.4302,
+      "step": 757
+    },
+    {
+      "epoch": 4.259154929577464,
+      "grad_norm": 8.560917854309082,
+      "learning_rate": 2.265e-06,
+      "loss": 1.4571,
+      "step": 758
+    },
+    {
+      "epoch": 4.264788732394366,
+      "grad_norm": 5.962185382843018,
+      "learning_rate": 2.268e-06,
+      "loss": 1.3484,
+      "step": 759
+    },
+    {
+      "epoch": 4.270422535211267,
+      "grad_norm": 3.354548454284668,
+      "learning_rate": 2.2710000000000004e-06,
+      "loss": 1.2123,
+      "step": 760
+    },
+    {
+      "epoch": 4.276056338028169,
+      "grad_norm": 3.654059410095215,
+      "learning_rate": 2.274e-06,
+      "loss": 1.3055,
+      "step": 761
+    },
+    {
+      "epoch": 4.28169014084507,
+      "grad_norm": 4.993838787078857,
+      "learning_rate": 2.277e-06,
+      "loss": 1.3121,
+      "step": 762
+    },
+    {
+      "epoch": 4.2873239436619714,
+      "grad_norm": 3.195439577102661,
+      "learning_rate": 2.28e-06,
+      "loss": 1.1986,
+      "step": 763
+    },
+    {
+      "epoch": 4.292957746478873,
+      "grad_norm": 2.968320608139038,
+      "learning_rate": 2.283e-06,
+      "loss": 1.2231,
+      "step": 764
+    },
+    {
+      "epoch": 4.298591549295774,
+      "grad_norm": 3.53824520111084,
+      "learning_rate": 2.2860000000000002e-06,
+      "loss": 1.2196,
+      "step": 765
+    },
+    {
+      "epoch": 4.304225352112676,
+      "grad_norm": 5.174617767333984,
+      "learning_rate": 2.2890000000000004e-06,
+      "loss": 1.1367,
+      "step": 766
+    },
+    {
+      "epoch": 4.309859154929577,
+      "grad_norm": 8.613746643066406,
+      "learning_rate": 2.292e-06,
+      "loss": 1.0579,
+      "step": 767
+    },
+    {
+      "epoch": 4.3154929577464785,
+      "grad_norm": 3.378936529159546,
+      "learning_rate": 2.295e-06,
+      "loss": 1.077,
+      "step": 768
+    },
+    {
+      "epoch": 4.321126760563381,
+      "grad_norm": 3.4128663539886475,
+      "learning_rate": 2.2980000000000003e-06,
+      "loss": 1.0403,
+      "step": 769
+    },
+    {
+      "epoch": 4.326760563380281,
+      "grad_norm": 4.564036846160889,
+      "learning_rate": 2.301e-06,
+      "loss": 1.0029,
+      "step": 770
+    },
+    {
+      "epoch": 4.3323943661971835,
+      "grad_norm": 4.310069561004639,
+      "learning_rate": 2.304e-06,
+      "loss": 1.0173,
+      "step": 771
+    },
+    {
+      "epoch": 4.338028169014084,
+      "grad_norm": 4.131349086761475,
+      "learning_rate": 2.307e-06,
+      "loss": 1.0233,
+      "step": 772
+    },
+    {
+      "epoch": 4.343661971830986,
+      "grad_norm": 2.9398694038391113,
+      "learning_rate": 2.31e-06,
+      "loss": 0.9647,
+      "step": 773
+    },
+    {
+      "epoch": 4.349295774647887,
+      "grad_norm": 2.1672401428222656,
+      "learning_rate": 2.313e-06,
+      "loss": 0.9424,
+      "step": 774
+    },
+    {
+      "epoch": 4.354929577464789,
+      "grad_norm": 3.017409324645996,
+      "learning_rate": 2.3160000000000004e-06,
+      "loss": 0.9195,
+      "step": 775
+    },
+    {
+      "epoch": 4.3605633802816905,
+      "grad_norm": 7.53600549697876,
+      "learning_rate": 2.3189999999999997e-06,
+      "loss": 0.9747,
+      "step": 776
+    },
+    {
+      "epoch": 4.366197183098592,
+      "grad_norm": 4.035458564758301,
+      "learning_rate": 2.322e-06,
+      "loss": 0.9055,
+      "step": 777
+    },
+    {
+      "epoch": 4.371830985915493,
+      "grad_norm": 2.679673671722412,
+      "learning_rate": 2.325e-06,
+      "loss": 0.911,
+      "step": 778
+    },
+    {
+      "epoch": 4.377464788732395,
+      "grad_norm": 2.053769588470459,
+      "learning_rate": 2.328e-06,
+      "loss": 0.8975,
+      "step": 779
+    },
+    {
+      "epoch": 4.383098591549296,
+      "grad_norm": 3.9188501834869385,
+      "learning_rate": 2.3310000000000002e-06,
+      "loss": 0.8792,
+      "step": 780
+    },
+    {
+      "epoch": 4.3887323943661976,
+      "grad_norm": 2.785120725631714,
+      "learning_rate": 2.334e-06,
+      "loss": 0.8571,
+      "step": 781
+    },
+    {
+      "epoch": 4.394366197183099,
+      "grad_norm": 2.4356346130371094,
+      "learning_rate": 2.337e-06,
+      "loss": 0.8674,
+      "step": 782
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 2.8814375400543213,
+      "learning_rate": 2.34e-06,
+      "loss": 0.8873,
+      "step": 783
+    },
+    {
+      "epoch": 4.405633802816902,
+      "grad_norm": 5.882245063781738,
+      "learning_rate": 2.3430000000000003e-06,
+      "loss": 0.846,
+      "step": 784
+    },
+    {
+      "epoch": 4.411267605633803,
+      "grad_norm": 4.481248378753662,
+      "learning_rate": 2.346e-06,
+      "loss": 0.7242,
+      "step": 785
+    },
+    {
+      "epoch": 4.416901408450705,
+      "grad_norm": 3.376279592514038,
+      "learning_rate": 2.349e-06,
+      "loss": 0.8909,
+      "step": 786
+    },
+    {
+      "epoch": 4.422535211267606,
+      "grad_norm": 5.917909622192383,
+      "learning_rate": 2.352e-06,
+      "loss": 0.8524,
+      "step": 787
+    },
+    {
+      "epoch": 4.428169014084507,
+      "grad_norm": 2.9090683460235596,
+      "learning_rate": 2.355e-06,
+      "loss": 0.7465,
+      "step": 788
+    },
+    {
+      "epoch": 4.433802816901409,
+      "grad_norm": 4.003406524658203,
+      "learning_rate": 2.358e-06,
+      "loss": 0.8354,
+      "step": 789
+    },
+    {
+      "epoch": 4.43943661971831,
+      "grad_norm": 14.42676067352295,
+      "learning_rate": 2.3610000000000003e-06,
+      "loss": 0.7535,
+      "step": 790
+    },
+    {
+      "epoch": 4.445070422535212,
+      "grad_norm": 3.3684017658233643,
+      "learning_rate": 2.3639999999999997e-06,
+      "loss": 0.8171,
+      "step": 791
+    },
+    {
+      "epoch": 4.450704225352113,
+      "grad_norm": 3.4548377990722656,
+      "learning_rate": 2.367e-06,
+      "loss": 0.6947,
+      "step": 792
+    },
+    {
+      "epoch": 4.456338028169014,
+      "grad_norm": 3.2732181549072266,
+      "learning_rate": 2.37e-06,
+      "loss": 0.7211,
+      "step": 793
+    },
+    {
+      "epoch": 4.461971830985916,
+      "grad_norm": 3.2089741230010986,
+      "learning_rate": 2.373e-06,
+      "loss": 0.656,
+      "step": 794
+    },
+    {
+      "epoch": 4.467605633802817,
+      "grad_norm": 3.5135998725891113,
+      "learning_rate": 2.376e-06,
+      "loss": 0.764,
+      "step": 795
+    },
+    {
+      "epoch": 4.473239436619719,
+      "grad_norm": 2.727151393890381,
+      "learning_rate": 2.379e-06,
+      "loss": 0.6935,
+      "step": 796
+    },
+    {
+      "epoch": 4.47887323943662,
+      "grad_norm": 6.873143196105957,
+      "learning_rate": 2.382e-06,
+      "loss": 0.6877,
+      "step": 797
+    },
+    {
+      "epoch": 4.484507042253521,
+      "grad_norm": 3.446615695953369,
+      "learning_rate": 2.385e-06,
+      "loss": 0.6804,
+      "step": 798
+    },
+    {
+      "epoch": 4.490140845070423,
+      "grad_norm": 5.030722618103027,
+      "learning_rate": 2.3880000000000003e-06,
+      "loss": 0.7444,
+      "step": 799
+    },
+    {
+      "epoch": 4.495774647887324,
+      "grad_norm": 6.555557727813721,
+      "learning_rate": 2.391e-06,
+      "loss": 0.7532,
+      "step": 800
+    },
+    {
+      "epoch": 4.501408450704226,
+      "grad_norm": 11.13542652130127,
+      "learning_rate": 2.394e-06,
+      "loss": 1.2519,
+      "step": 801
+    },
+    {
+      "epoch": 4.507042253521127,
+      "grad_norm": 6.424833297729492,
+      "learning_rate": 2.397e-06,
+      "loss": 1.1184,
+      "step": 802
+    },
+    {
+      "epoch": 4.512676056338028,
+      "grad_norm": 5.684906959533691,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.1603,
+      "step": 803
+    },
+    {
+      "epoch": 4.51830985915493,
+      "grad_norm": 6.032043933868408,
+      "learning_rate": 2.403e-06,
+      "loss": 0.9755,
+      "step": 804
+    },
+    {
+      "epoch": 4.523943661971831,
+      "grad_norm": 7.522682189941406,
+      "learning_rate": 2.406e-06,
+      "loss": 1.0374,
+      "step": 805
+    },
+    {
+      "epoch": 4.529577464788733,
+      "grad_norm": 4.559472560882568,
+      "learning_rate": 2.409e-06,
+      "loss": 0.963,
+      "step": 806
+    },
+    {
+      "epoch": 4.535211267605634,
+      "grad_norm": 8.044770240783691,
+      "learning_rate": 2.412e-06,
+      "loss": 0.8639,
+      "step": 807
+    },
+    {
+      "epoch": 4.540845070422535,
+      "grad_norm": 4.376523017883301,
+      "learning_rate": 2.415e-06,
+      "loss": 0.8814,
+      "step": 808
+    },
+    {
+      "epoch": 4.546478873239437,
+      "grad_norm": 5.2792534828186035,
+      "learning_rate": 2.4180000000000004e-06,
+      "loss": 0.9314,
+      "step": 809
+    },
+    {
+      "epoch": 4.552112676056338,
+      "grad_norm": 6.261806964874268,
+      "learning_rate": 2.4209999999999998e-06,
+      "loss": 0.8478,
+      "step": 810
+    },
+    {
+      "epoch": 4.55774647887324,
+      "grad_norm": 5.189519882202148,
+      "learning_rate": 2.424e-06,
+      "loss": 0.9977,
+      "step": 811
+    },
+    {
+      "epoch": 4.563380281690141,
+      "grad_norm": 5.815868854522705,
+      "learning_rate": 2.4270000000000002e-06,
+      "loss": 0.8007,
+      "step": 812
+    },
+    {
+      "epoch": 4.569014084507042,
+      "grad_norm": 5.150207042694092,
+      "learning_rate": 2.43e-06,
+      "loss": 0.9216,
+      "step": 813
+    },
+    {
+      "epoch": 4.574647887323944,
+      "grad_norm": 3.9940452575683594,
+      "learning_rate": 2.4330000000000003e-06,
+      "loss": 0.8655,
+      "step": 814
+    },
+    {
+      "epoch": 4.580281690140845,
+      "grad_norm": 2.9555556774139404,
+      "learning_rate": 2.436e-06,
+      "loss": 0.8565,
+      "step": 815
+    },
+    {
+      "epoch": 4.585915492957747,
+      "grad_norm": 3.326611280441284,
+      "learning_rate": 2.439e-06,
+      "loss": 0.7621,
+      "step": 816
+    },
+    {
+      "epoch": 4.591549295774648,
+      "grad_norm": 3.882078170776367,
+      "learning_rate": 2.442e-06,
+      "loss": 0.7741,
+      "step": 817
+    },
+    {
+      "epoch": 4.597183098591549,
+      "grad_norm": 11.13386344909668,
+      "learning_rate": 2.4450000000000003e-06,
+      "loss": 0.7245,
+      "step": 818
+    },
+    {
+      "epoch": 4.602816901408451,
+      "grad_norm": 4.292423248291016,
+      "learning_rate": 2.448e-06,
+      "loss": 0.7295,
+      "step": 819
+    },
+    {
+      "epoch": 4.608450704225352,
+      "grad_norm": 10.06823444366455,
+      "learning_rate": 2.451e-06,
+      "loss": 0.6212,
+      "step": 820
+    },
+    {
+      "epoch": 4.614084507042254,
+      "grad_norm": 4.313129901885986,
+      "learning_rate": 2.454e-06,
+      "loss": 0.6961,
+      "step": 821
+    },
+    {
+      "epoch": 4.619718309859155,
+      "grad_norm": 6.2645182609558105,
+      "learning_rate": 2.457e-06,
+      "loss": 0.7328,
+      "step": 822
+    },
+    {
+      "epoch": 4.625352112676056,
+      "grad_norm": 2.457139015197754,
+      "learning_rate": 2.46e-06,
+      "loss": 0.7715,
+      "step": 823
+    },
+    {
+      "epoch": 4.630985915492958,
+      "grad_norm": 3.2750065326690674,
+      "learning_rate": 2.4630000000000004e-06,
+      "loss": 0.7052,
+      "step": 824
+    },
+    {
+      "epoch": 4.636619718309859,
+      "grad_norm": 4.161436557769775,
+      "learning_rate": 2.4659999999999998e-06,
+      "loss": 0.6182,
+      "step": 825
+    },
+    {
+      "epoch": 4.642253521126761,
+      "grad_norm": 6.171960353851318,
+      "learning_rate": 2.469e-06,
+      "loss": 0.6678,
+      "step": 826
+    },
+    {
+      "epoch": 4.647887323943662,
+      "grad_norm": 4.594521999359131,
+      "learning_rate": 2.4720000000000002e-06,
+      "loss": 0.6498,
+      "step": 827
+    },
+    {
+      "epoch": 4.653521126760563,
+      "grad_norm": 5.677289009094238,
+      "learning_rate": 2.475e-06,
+      "loss": 0.6358,
+      "step": 828
+    },
+    {
+      "epoch": 4.659154929577465,
+      "grad_norm": 7.616566181182861,
+      "learning_rate": 2.4780000000000002e-06,
+      "loss": 0.6303,
+      "step": 829
+    },
+    {
+      "epoch": 4.664788732394366,
+      "grad_norm": 3.6935417652130127,
+      "learning_rate": 2.481e-06,
+      "loss": 0.695,
+      "step": 830
+    },
+    {
+      "epoch": 4.670422535211268,
+      "grad_norm": 6.680566310882568,
+      "learning_rate": 2.484e-06,
+      "loss": 0.6622,
+      "step": 831
+    },
+    {
+      "epoch": 4.676056338028169,
+      "grad_norm": 4.233198165893555,
+      "learning_rate": 2.487e-06,
+      "loss": 0.5588,
+      "step": 832
+    },
+    {
+      "epoch": 4.68169014084507,
+      "grad_norm": 2.316086769104004,
+      "learning_rate": 2.4900000000000003e-06,
+      "loss": 0.6056,
+      "step": 833
+    },
+    {
+      "epoch": 4.687323943661972,
+      "grad_norm": 4.981112480163574,
+      "learning_rate": 2.4929999999999997e-06,
+      "loss": 0.615,
+      "step": 834
+    },
+    {
+      "epoch": 4.692957746478873,
+      "grad_norm": 2.577280044555664,
+      "learning_rate": 2.496e-06,
+      "loss": 0.6051,
+      "step": 835
+    },
+    {
+      "epoch": 4.698591549295775,
+      "grad_norm": 3.450861930847168,
+      "learning_rate": 2.499e-06,
+      "loss": 0.6528,
+      "step": 836
+    },
+    {
+      "epoch": 4.704225352112676,
+      "grad_norm": 2.7186598777770996,
+      "learning_rate": 2.502e-06,
+      "loss": 0.6306,
+      "step": 837
+    },
+    {
+      "epoch": 4.709859154929577,
+      "grad_norm": 2.859384298324585,
+      "learning_rate": 2.505e-06,
+      "loss": 0.6805,
+      "step": 838
+    },
+    {
+      "epoch": 4.715492957746479,
+      "grad_norm": 3.7891767024993896,
+      "learning_rate": 2.508e-06,
+      "loss": 0.6214,
+      "step": 839
+    },
+    {
+      "epoch": 4.72112676056338,
+      "grad_norm": 3.8955564498901367,
+      "learning_rate": 2.5109999999999998e-06,
+      "loss": 0.5786,
+      "step": 840
+    },
+    {
+      "epoch": 4.726760563380282,
+      "grad_norm": 6.762790203094482,
+      "learning_rate": 2.514e-06,
+      "loss": 0.5551,
+      "step": 841
+    },
+    {
+      "epoch": 4.732394366197183,
+      "grad_norm": 3.186650276184082,
+      "learning_rate": 2.517e-06,
+      "loss": 0.5718,
+      "step": 842
+    },
+    {
+      "epoch": 4.738028169014084,
+      "grad_norm": 4.856934547424316,
+      "learning_rate": 2.52e-06,
+      "loss": 0.5954,
+      "step": 843
+    },
+    {
+      "epoch": 4.743661971830986,
+      "grad_norm": 5.6527791023254395,
+      "learning_rate": 2.523e-06,
+      "loss": 0.6891,
+      "step": 844
+    },
+    {
+      "epoch": 4.749295774647887,
+      "grad_norm": 6.431458473205566,
+      "learning_rate": 2.526e-06,
+      "loss": 1.0238,
+      "step": 845
+    },
+    {
+      "epoch": 4.754929577464789,
+      "grad_norm": 3.19467830657959,
+      "learning_rate": 2.5290000000000003e-06,
+      "loss": 0.9253,
+      "step": 846
+    },
+    {
+      "epoch": 4.76056338028169,
+      "grad_norm": 3.688904285430908,
+      "learning_rate": 2.532e-06,
+      "loss": 0.8768,
+      "step": 847
+    },
+    {
+      "epoch": 4.766197183098591,
+      "grad_norm": 3.788217067718506,
+      "learning_rate": 2.5350000000000003e-06,
+      "loss": 0.9304,
+      "step": 848
+    },
+    {
+      "epoch": 4.771830985915493,
+      "grad_norm": 3.510573625564575,
+      "learning_rate": 2.538e-06,
+      "loss": 0.8295,
+      "step": 849
+    },
+    {
+      "epoch": 4.777464788732394,
+      "grad_norm": 3.672971248626709,
+      "learning_rate": 2.541e-06,
+      "loss": 0.8227,
+      "step": 850
+    },
+    {
+      "epoch": 4.783098591549296,
+      "grad_norm": 9.765240669250488,
+      "learning_rate": 2.544e-06,
+      "loss": 0.8331,
+      "step": 851
+    },
+    {
+      "epoch": 4.788732394366197,
+      "grad_norm": 4.7542924880981445,
+      "learning_rate": 2.5470000000000003e-06,
+      "loss": 0.8178,
+      "step": 852
+    },
+    {
+      "epoch": 4.794366197183098,
+      "grad_norm": 2.981187105178833,
+      "learning_rate": 2.55e-06,
+      "loss": 0.8533,
+      "step": 853
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 5.025519847869873,
+      "learning_rate": 2.553e-06,
+      "loss": 0.8352,
+      "step": 854
+    },
+    {
+      "epoch": 4.805633802816901,
+      "grad_norm": 3.1089770793914795,
+      "learning_rate": 2.556e-06,
+      "loss": 0.7018,
+      "step": 855
+    },
+    {
+      "epoch": 4.811267605633803,
+      "grad_norm": 2.9379541873931885,
+      "learning_rate": 2.559e-06,
+      "loss": 0.7346,
+      "step": 856
+    },
+    {
+      "epoch": 4.816901408450704,
+      "grad_norm": 3.395998001098633,
+      "learning_rate": 2.562e-06,
+      "loss": 0.7934,
+      "step": 857
+    },
+    {
+      "epoch": 4.822535211267605,
+      "grad_norm": 3.98203182220459,
+      "learning_rate": 2.5650000000000004e-06,
+      "loss": 0.7462,
+      "step": 858
+    },
+    {
+      "epoch": 4.828169014084507,
+      "grad_norm": 2.7102365493774414,
+      "learning_rate": 2.568e-06,
+      "loss": 0.6715,
+      "step": 859
+    },
+    {
+      "epoch": 4.833802816901408,
+      "grad_norm": 6.256322860717773,
+      "learning_rate": 2.571e-06,
+      "loss": 0.8342,
+      "step": 860
+    },
+    {
+      "epoch": 4.83943661971831,
+      "grad_norm": 2.3925929069519043,
+      "learning_rate": 2.5740000000000003e-06,
+      "loss": 0.6078,
+      "step": 861
+    },
+    {
+      "epoch": 4.845070422535211,
+      "grad_norm": 2.342928647994995,
+      "learning_rate": 2.577e-06,
+      "loss": 0.6536,
+      "step": 862
+    },
+    {
+      "epoch": 4.850704225352112,
+      "grad_norm": 2.7884795665740967,
+      "learning_rate": 2.58e-06,
+      "loss": 0.6163,
+      "step": 863
+    },
+    {
+      "epoch": 4.856338028169014,
+      "grad_norm": 2.60550856590271,
+      "learning_rate": 2.583e-06,
+      "loss": 0.6322,
+      "step": 864
+    },
+    {
+      "epoch": 4.861971830985915,
+      "grad_norm": 3.2051239013671875,
+      "learning_rate": 2.586e-06,
+      "loss": 0.5733,
+      "step": 865
+    },
+    {
+      "epoch": 4.867605633802817,
+      "grad_norm": 2.349964141845703,
+      "learning_rate": 2.589e-06,
+      "loss": 0.5361,
+      "step": 866
+    },
+    {
+      "epoch": 4.873239436619718,
+      "grad_norm": 7.408186435699463,
+      "learning_rate": 2.5920000000000003e-06,
+      "loss": 0.6847,
+      "step": 867
+    },
+    {
+      "epoch": 4.878873239436619,
+      "grad_norm": 108.75530242919922,
+      "learning_rate": 2.5949999999999997e-06,
+      "loss": 0.5442,
+      "step": 868
+    },
+    {
+      "epoch": 4.884507042253521,
+      "grad_norm": 3.8446199893951416,
+      "learning_rate": 2.598e-06,
+      "loss": 0.6033,
+      "step": 869
+    },
+    {
+      "epoch": 4.890140845070422,
+      "grad_norm": 2.7630155086517334,
+      "learning_rate": 2.601e-06,
+      "loss": 0.5898,
+      "step": 870
+    },
+    {
+      "epoch": 4.895774647887324,
+      "grad_norm": 2.9433813095092773,
+      "learning_rate": 2.604e-06,
+      "loss": 0.6187,
+      "step": 871
+    },
+    {
+      "epoch": 4.901408450704225,
+      "grad_norm": 3.2574610710144043,
+      "learning_rate": 2.607e-06,
+      "loss": 0.4792,
+      "step": 872
+    },
+    {
+      "epoch": 4.907042253521126,
+      "grad_norm": 2.240130662918091,
+      "learning_rate": 2.61e-06,
+      "loss": 0.4657,
+      "step": 873
+    },
+    {
+      "epoch": 4.912676056338028,
+      "grad_norm": 3.9430322647094727,
+      "learning_rate": 2.613e-06,
+      "loss": 0.6533,
+      "step": 874
+    },
+    {
+      "epoch": 4.918309859154929,
+      "grad_norm": 2.186457633972168,
+      "learning_rate": 2.616e-06,
+      "loss": 0.4592,
+      "step": 875
+    },
+    {
+      "epoch": 4.923943661971831,
+      "grad_norm": 6.3133955001831055,
+      "learning_rate": 2.6190000000000003e-06,
+      "loss": 0.6135,
+      "step": 876
+    },
+    {
+      "epoch": 4.929577464788732,
+      "grad_norm": 2.4026143550872803,
+      "learning_rate": 2.622e-06,
+      "loss": 0.6044,
+      "step": 877
+    },
+    {
+      "epoch": 4.9352112676056334,
+      "grad_norm": 2.5374703407287598,
+      "learning_rate": 2.625e-06,
+      "loss": 0.5245,
+      "step": 878
+    },
+    {
+      "epoch": 4.940845070422535,
+      "grad_norm": 2.3137059211730957,
+      "learning_rate": 2.628e-06,
+      "loss": 0.6356,
+      "step": 879
+    },
+    {
+      "epoch": 4.946478873239436,
+      "grad_norm": 3.0200490951538086,
+      "learning_rate": 2.631e-06,
+      "loss": 0.5608,
+      "step": 880
+    },
+    {
+      "epoch": 4.952112676056338,
+      "grad_norm": 2.407966375350952,
+      "learning_rate": 2.634e-06,
+      "loss": 0.5017,
+      "step": 881
+    },
+    {
+      "epoch": 4.957746478873239,
+      "grad_norm": 4.552791118621826,
+      "learning_rate": 2.6370000000000003e-06,
+      "loss": 0.5556,
+      "step": 882
+    },
+    {
+      "epoch": 4.9633802816901404,
+      "grad_norm": 2.422546863555908,
+      "learning_rate": 2.6399999999999997e-06,
+      "loss": 0.4539,
+      "step": 883
+    },
+    {
+      "epoch": 4.969014084507043,
+      "grad_norm": 2.9516642093658447,
+      "learning_rate": 2.643e-06,
+      "loss": 0.5157,
+      "step": 884
+    },
+    {
+      "epoch": 4.974647887323943,
+      "grad_norm": 3.9323694705963135,
+      "learning_rate": 2.646e-06,
+      "loss": 0.5258,
+      "step": 885
+    },
+    {
+      "epoch": 4.9802816901408455,
+      "grad_norm": 4.032447338104248,
+      "learning_rate": 2.649e-06,
+      "loss": 0.6014,
+      "step": 886
+    },
+    {
+      "epoch": 4.985915492957746,
+      "grad_norm": 4.838427543640137,
+      "learning_rate": 2.652e-06,
+      "loss": 0.4939,
+      "step": 887
+    },
+    {
+      "epoch": 4.991549295774648,
+      "grad_norm": 7.549025535583496,
+      "learning_rate": 2.655e-06,
+      "loss": 0.5463,
+      "step": 888
+    },
+    {
+      "epoch": 4.997183098591549,
+      "grad_norm": 2.98043155670166,
+      "learning_rate": 2.6580000000000002e-06,
+      "loss": 0.6438,
+      "step": 889
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 3.0263454914093018,
+      "learning_rate": 2.661e-06,
+      "loss": 0.2365,
+      "step": 890
+    },
+    {
+      "epoch": 5.005633802816901,
+      "grad_norm": 3.7286951541900635,
+      "learning_rate": 2.6640000000000002e-06,
+      "loss": 0.8897,
+      "step": 891
+    },
+    {
+      "epoch": 5.011267605633803,
+      "grad_norm": 3.884514570236206,
+      "learning_rate": 2.6670000000000005e-06,
+      "loss": 0.7991,
+      "step": 892
+    },
+    {
+      "epoch": 5.016901408450704,
+      "grad_norm": 3.591808319091797,
+      "learning_rate": 2.67e-06,
+      "loss": 0.7984,
+      "step": 893
+    },
+    {
+      "epoch": 5.022535211267606,
+      "grad_norm": 2.895336866378784,
+      "learning_rate": 2.673e-06,
+      "loss": 0.8905,
+      "step": 894
+    },
+    {
+      "epoch": 5.028169014084507,
+      "grad_norm": 4.619030952453613,
+      "learning_rate": 2.6760000000000003e-06,
+      "loss": 0.7814,
+      "step": 895
+    },
+    {
+      "epoch": 5.033802816901408,
+      "grad_norm": 3.87629771232605,
+      "learning_rate": 2.679e-06,
+      "loss": 0.7244,
+      "step": 896
+    },
+    {
+      "epoch": 5.03943661971831,
+      "grad_norm": 3.3396666049957275,
+      "learning_rate": 2.682e-06,
+      "loss": 0.681,
+      "step": 897
+    },
+    {
+      "epoch": 5.045070422535211,
+      "grad_norm": 4.312685489654541,
+      "learning_rate": 2.685e-06,
+      "loss": 0.6109,
+      "step": 898
+    },
+    {
+      "epoch": 5.050704225352113,
+      "grad_norm": 3.3607585430145264,
+      "learning_rate": 2.688e-06,
+      "loss": 0.6679,
+      "step": 899
+    },
+    {
+      "epoch": 5.056338028169014,
+      "grad_norm": 4.933312892913818,
+      "learning_rate": 2.691e-06,
+      "loss": 0.6408,
+      "step": 900
+    },
+    {
+      "epoch": 5.061971830985915,
+      "grad_norm": 7.953579425811768,
+      "learning_rate": 2.6940000000000004e-06,
+      "loss": 0.704,
+      "step": 901
+    },
+    {
+      "epoch": 5.067605633802817,
+      "grad_norm": 2.636009931564331,
+      "learning_rate": 2.6969999999999998e-06,
+      "loss": 0.6687,
+      "step": 902
+    },
+    {
+      "epoch": 5.073239436619718,
+      "grad_norm": 2.501701831817627,
+      "learning_rate": 2.7e-06,
+      "loss": 0.6241,
+      "step": 903
+    },
+    {
+      "epoch": 5.07887323943662,
+      "grad_norm": 3.6819188594818115,
+      "learning_rate": 2.703e-06,
+      "loss": 0.7167,
+      "step": 904
+    },
+    {
+      "epoch": 5.084507042253521,
+      "grad_norm": 2.071209192276001,
+      "learning_rate": 2.706e-06,
+      "loss": 0.5678,
+      "step": 905
+    },
+    {
+      "epoch": 5.090140845070422,
+      "grad_norm": 2.5370724201202393,
+      "learning_rate": 2.7090000000000002e-06,
+      "loss": 0.6386,
+      "step": 906
+    },
+    {
+      "epoch": 5.095774647887324,
+      "grad_norm": 5.760916233062744,
+      "learning_rate": 2.712e-06,
+      "loss": 0.6145,
+      "step": 907
+    },
+    {
+      "epoch": 5.101408450704225,
+      "grad_norm": 2.104323148727417,
+      "learning_rate": 2.715e-06,
+      "loss": 0.5532,
+      "step": 908
+    },
+    {
+      "epoch": 5.107042253521127,
+      "grad_norm": 2.2896480560302734,
+      "learning_rate": 2.718e-06,
+      "loss": 0.5573,
+      "step": 909
+    },
+    {
+      "epoch": 5.112676056338028,
+      "grad_norm": 4.796744346618652,
+      "learning_rate": 2.7210000000000003e-06,
+      "loss": 0.4646,
+      "step": 910
+    },
+    {
+      "epoch": 5.118309859154929,
+      "grad_norm": 2.22436785697937,
+      "learning_rate": 2.724e-06,
+      "loss": 0.5938,
+      "step": 911
+    },
+    {
+      "epoch": 5.123943661971831,
+      "grad_norm": 5.623243808746338,
+      "learning_rate": 2.727e-06,
+      "loss": 0.5198,
+      "step": 912
+    },
+    {
+      "epoch": 5.129577464788732,
+      "grad_norm": 2.2928428649902344,
+      "learning_rate": 2.73e-06,
+      "loss": 0.5933,
+      "step": 913
+    },
+    {
+      "epoch": 5.135211267605634,
+      "grad_norm": 3.24739408493042,
+      "learning_rate": 2.733e-06,
+      "loss": 0.4768,
+      "step": 914
+    },
+    {
+      "epoch": 5.140845070422535,
+      "grad_norm": 5.295236110687256,
+      "learning_rate": 2.736e-06,
+      "loss": 0.5845,
+      "step": 915
+    },
+    {
+      "epoch": 5.146478873239436,
+      "grad_norm": 5.906332015991211,
+      "learning_rate": 2.7390000000000004e-06,
+      "loss": 0.536,
+      "step": 916
+    },
+    {
+      "epoch": 5.152112676056338,
+      "grad_norm": 2.0323636531829834,
+      "learning_rate": 2.7419999999999998e-06,
+      "loss": 0.598,
+      "step": 917
+    },
+    {
+      "epoch": 5.157746478873239,
+      "grad_norm": 2.347740650177002,
+      "learning_rate": 2.745e-06,
+      "loss": 0.4687,
+      "step": 918
+    },
+    {
+      "epoch": 5.163380281690141,
+      "grad_norm": 5.7692999839782715,
+      "learning_rate": 2.748e-06,
+      "loss": 0.4441,
+      "step": 919
+    },
+    {
+      "epoch": 5.169014084507042,
+      "grad_norm": 2.8650996685028076,
+      "learning_rate": 2.751e-06,
+      "loss": 0.6155,
+      "step": 920
+    },
+    {
+      "epoch": 5.174647887323943,
+      "grad_norm": 2.7199292182922363,
+      "learning_rate": 2.7540000000000002e-06,
+      "loss": 0.5006,
+      "step": 921
+    },
+    {
+      "epoch": 5.180281690140845,
+      "grad_norm": 2.181318521499634,
+      "learning_rate": 2.757e-06,
+      "loss": 0.4766,
+      "step": 922
+    },
+    {
+      "epoch": 5.185915492957746,
+      "grad_norm": 2.543658971786499,
+      "learning_rate": 2.76e-06,
+      "loss": 0.5543,
+      "step": 923
+    },
+    {
+      "epoch": 5.191549295774648,
+      "grad_norm": 7.329205513000488,
+      "learning_rate": 2.763e-06,
+      "loss": 0.4265,
+      "step": 924
+    },
+    {
+      "epoch": 5.197183098591549,
+      "grad_norm": 2.5346627235412598,
+      "learning_rate": 2.7660000000000003e-06,
+      "loss": 0.4808,
+      "step": 925
+    },
+    {
+      "epoch": 5.20281690140845,
+      "grad_norm": 2.384472370147705,
+      "learning_rate": 2.7689999999999997e-06,
+      "loss": 0.4395,
+      "step": 926
+    },
+    {
+      "epoch": 5.208450704225352,
+      "grad_norm": 3.4939169883728027,
+      "learning_rate": 2.772e-06,
+      "loss": 0.4119,
+      "step": 927
+    },
+    {
+      "epoch": 5.214084507042253,
+      "grad_norm": 5.23539924621582,
+      "learning_rate": 2.775e-06,
+      "loss": 0.4431,
+      "step": 928
+    },
+    {
+      "epoch": 5.219718309859155,
+      "grad_norm": 2.3811333179473877,
+      "learning_rate": 2.778e-06,
+      "loss": 0.4725,
+      "step": 929
+    },
+    {
+      "epoch": 5.225352112676056,
+      "grad_norm": 6.978809356689453,
+      "learning_rate": 2.781e-06,
+      "loss": 0.4696,
+      "step": 930
+    },
+    {
+      "epoch": 5.230985915492957,
+      "grad_norm": 5.870415687561035,
+      "learning_rate": 2.784e-06,
+      "loss": 0.381,
+      "step": 931
+    },
+    {
+      "epoch": 5.236619718309859,
+      "grad_norm": 4.045934677124023,
+      "learning_rate": 2.787e-06,
+      "loss": 0.6097,
+      "step": 932
+    },
+    {
+      "epoch": 5.24225352112676,
+      "grad_norm": 21.88624382019043,
+      "learning_rate": 2.79e-06,
+      "loss": 0.3777,
+      "step": 933
+    },
+    {
+      "epoch": 5.247887323943662,
+      "grad_norm": 3.4917380809783936,
+      "learning_rate": 2.793e-06,
+      "loss": 0.523,
+      "step": 934
+    },
+    {
+      "epoch": 5.253521126760563,
+      "grad_norm": 2.8486669063568115,
+      "learning_rate": 2.7960000000000004e-06,
+      "loss": 0.7939,
+      "step": 935
+    },
+    {
+      "epoch": 5.259154929577464,
+      "grad_norm": 3.822283983230591,
+      "learning_rate": 2.799e-06,
+      "loss": 0.8049,
+      "step": 936
+    },
+    {
+      "epoch": 5.264788732394366,
+      "grad_norm": 3.4523532390594482,
+      "learning_rate": 2.802e-06,
+      "loss": 0.7975,
+      "step": 937
+    },
+    {
+      "epoch": 5.270422535211267,
+      "grad_norm": 2.5025782585144043,
+      "learning_rate": 2.8050000000000002e-06,
+      "loss": 0.6506,
+      "step": 938
+    },
+    {
+      "epoch": 5.276056338028169,
+      "grad_norm": 2.210494041442871,
+      "learning_rate": 2.808e-06,
+      "loss": 0.6887,
+      "step": 939
+    },
+    {
+      "epoch": 5.28169014084507,
+      "grad_norm": 2.1525866985321045,
+      "learning_rate": 2.8110000000000003e-06,
+      "loss": 0.7456,
+      "step": 940
+    },
+    {
+      "epoch": 5.2873239436619714,
+      "grad_norm": 6.6180338859558105,
+      "learning_rate": 2.814e-06,
+      "loss": 0.6736,
+      "step": 941
+    },
+    {
+      "epoch": 5.292957746478873,
+      "grad_norm": 2.8995516300201416,
+      "learning_rate": 2.817e-06,
+      "loss": 0.6474,
+      "step": 942
+    },
+    {
+      "epoch": 5.298591549295774,
+      "grad_norm": 2.2757349014282227,
+      "learning_rate": 2.82e-06,
+      "loss": 0.6859,
+      "step": 943
+    },
+    {
+      "epoch": 5.304225352112676,
+      "grad_norm": 2.54069185256958,
+      "learning_rate": 2.8230000000000003e-06,
+      "loss": 0.5716,
+      "step": 944
+    },
+    {
+      "epoch": 5.309859154929577,
+      "grad_norm": 3.2585108280181885,
+      "learning_rate": 2.826e-06,
+      "loss": 0.5769,
+      "step": 945
+    },
+    {
+      "epoch": 5.3154929577464785,
+      "grad_norm": 2.3954012393951416,
+      "learning_rate": 2.829e-06,
+      "loss": 0.5716,
+      "step": 946
+    },
+    {
+      "epoch": 5.321126760563381,
+      "grad_norm": 2.4710357189178467,
+      "learning_rate": 2.832e-06,
+      "loss": 0.5858,
+      "step": 947
+    },
+    {
+      "epoch": 5.326760563380281,
+      "grad_norm": 2.4542315006256104,
+      "learning_rate": 2.835e-06,
+      "loss": 0.5595,
+      "step": 948
+    },
+    {
+      "epoch": 5.3323943661971835,
+      "grad_norm": 4.040088176727295,
+      "learning_rate": 2.838e-06,
+      "loss": 0.4993,
+      "step": 949
+    },
+    {
+      "epoch": 5.338028169014084,
+      "grad_norm": 2.6751508712768555,
+      "learning_rate": 2.8410000000000004e-06,
+      "loss": 0.5521,
+      "step": 950
+    },
+    {
+      "epoch": 5.343661971830986,
+      "grad_norm": 1.8362388610839844,
+      "learning_rate": 2.844e-06,
+      "loss": 0.4828,
+      "step": 951
+    },
+    {
+      "epoch": 5.349295774647887,
+      "grad_norm": 3.870666980743408,
+      "learning_rate": 2.847e-06,
+      "loss": 0.5741,
+      "step": 952
+    },
+    {
+      "epoch": 5.354929577464789,
+      "grad_norm": 2.98319411277771,
+      "learning_rate": 2.8500000000000002e-06,
+      "loss": 0.4582,
+      "step": 953
+    },
+    {
+      "epoch": 5.3605633802816905,
+      "grad_norm": 2.085663080215454,
+      "learning_rate": 2.853e-06,
+      "loss": 0.4475,
+      "step": 954
+    },
+    {
+      "epoch": 5.366197183098592,
+      "grad_norm": 2.988067626953125,
+      "learning_rate": 2.8560000000000003e-06,
+      "loss": 0.5461,
+      "step": 955
+    },
+    {
+      "epoch": 5.371830985915493,
+      "grad_norm": 2.501213312149048,
+      "learning_rate": 2.859e-06,
+      "loss": 0.5009,
+      "step": 956
+    },
+    {
+      "epoch": 5.377464788732395,
+      "grad_norm": 5.369483947753906,
+      "learning_rate": 2.862e-06,
+      "loss": 0.4742,
+      "step": 957
+    },
+    {
+      "epoch": 5.383098591549296,
+      "grad_norm": 3.1307785511016846,
+      "learning_rate": 2.865e-06,
+      "loss": 0.4614,
+      "step": 958
+    },
+    {
+      "epoch": 5.3887323943661976,
+      "grad_norm": 2.3280138969421387,
+      "learning_rate": 2.8680000000000003e-06,
+      "loss": 0.4165,
+      "step": 959
+    },
+    {
+      "epoch": 5.394366197183099,
+      "grad_norm": 3.9419243335723877,
+      "learning_rate": 2.8709999999999997e-06,
+      "loss": 0.4646,
+      "step": 960
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 1.6847516298294067,
+      "learning_rate": 2.874e-06,
+      "loss": 0.4,
+      "step": 961
+    },
+    {
+      "epoch": 5.405633802816902,
+      "grad_norm": 2.56376576423645,
+      "learning_rate": 2.877e-06,
+      "loss": 0.4022,
+      "step": 962
+    },
+    {
+      "epoch": 5.411267605633803,
+      "grad_norm": 2.88046932220459,
+      "learning_rate": 2.88e-06,
+      "loss": 0.4373,
+      "step": 963
+    },
+    {
+      "epoch": 5.416901408450705,
+      "grad_norm": 2.5140233039855957,
+      "learning_rate": 2.883e-06,
+      "loss": 0.476,
+      "step": 964
+    },
+    {
+      "epoch": 5.422535211267606,
+      "grad_norm": 3.400726079940796,
+      "learning_rate": 2.886e-06,
+      "loss": 0.3556,
+      "step": 965
+    },
+    {
+      "epoch": 5.428169014084507,
+      "grad_norm": 2.35408091545105,
+      "learning_rate": 2.8889999999999998e-06,
+      "loss": 0.3892,
+      "step": 966
+    },
+    {
+      "epoch": 5.433802816901409,
+      "grad_norm": 2.032426357269287,
+      "learning_rate": 2.892e-06,
+      "loss": 0.4039,
+      "step": 967
+    },
+    {
+      "epoch": 5.43943661971831,
+      "grad_norm": 2.5404016971588135,
+      "learning_rate": 2.8950000000000002e-06,
+      "loss": 0.3942,
+      "step": 968
+    },
+    {
+      "epoch": 5.445070422535212,
+      "grad_norm": 2.5330841541290283,
+      "learning_rate": 2.898e-06,
+      "loss": 0.3797,
+      "step": 969
+    },
+    {
+      "epoch": 5.450704225352113,
+      "grad_norm": 6.841852188110352,
+      "learning_rate": 2.901e-06,
+      "loss": 0.4562,
+      "step": 970
+    },
+    {
+      "epoch": 5.456338028169014,
+      "grad_norm": 2.308804988861084,
+      "learning_rate": 2.904e-06,
+      "loss": 0.5075,
+      "step": 971
+    },
+    {
+      "epoch": 5.461971830985916,
+      "grad_norm": 3.835463523864746,
+      "learning_rate": 2.907e-06,
+      "loss": 0.3515,
+      "step": 972
+    },
+    {
+      "epoch": 5.467605633802817,
+      "grad_norm": 8.132750511169434,
+      "learning_rate": 2.91e-06,
+      "loss": 0.4574,
+      "step": 973
+    },
+    {
+      "epoch": 5.473239436619719,
+      "grad_norm": 2.29156756401062,
+      "learning_rate": 2.9130000000000003e-06,
+      "loss": 0.3985,
+      "step": 974
+    },
+    {
+      "epoch": 5.47887323943662,
+      "grad_norm": 3.6257643699645996,
+      "learning_rate": 2.916e-06,
+      "loss": 0.4139,
+      "step": 975
+    },
+    {
+      "epoch": 5.484507042253521,
+      "grad_norm": 2.5936882495880127,
+      "learning_rate": 2.919e-06,
+      "loss": 0.4021,
+      "step": 976
+    },
+    {
+      "epoch": 5.490140845070423,
+      "grad_norm": 3.8019628524780273,
+      "learning_rate": 2.922e-06,
+      "loss": 0.476,
+      "step": 977
+    },
+    {
+      "epoch": 5.495774647887324,
+      "grad_norm": 4.725766658782959,
+      "learning_rate": 2.9250000000000004e-06,
+      "loss": 0.4575,
+      "step": 978
+    },
+    {
+      "epoch": 5.501408450704226,
+      "grad_norm": 5.978580951690674,
+      "learning_rate": 2.928e-06,
+      "loss": 0.8134,
+      "step": 979
+    },
+    {
+      "epoch": 5.507042253521127,
+      "grad_norm": 2.873779773712158,
+      "learning_rate": 2.931e-06,
+      "loss": 0.74,
+      "step": 980
+    },
+    {
+      "epoch": 5.512676056338028,
+      "grad_norm": 2.560568332672119,
+      "learning_rate": 2.934e-06,
+      "loss": 0.7099,
+      "step": 981
+    },
+    {
+      "epoch": 5.51830985915493,
+      "grad_norm": 2.4004578590393066,
+      "learning_rate": 2.937e-06,
+      "loss": 0.6652,
+      "step": 982
+    },
+    {
+      "epoch": 5.523943661971831,
+      "grad_norm": 3.904620409011841,
+      "learning_rate": 2.9400000000000002e-06,
+      "loss": 0.7448,
+      "step": 983
+    },
+    {
+      "epoch": 5.529577464788733,
+      "grad_norm": 1.8310728073120117,
+      "learning_rate": 2.9430000000000005e-06,
+      "loss": 0.6161,
+      "step": 984
+    },
+    {
+      "epoch": 5.535211267605634,
+      "grad_norm": 8.018976211547852,
+      "learning_rate": 2.946e-06,
+      "loss": 0.5445,
+      "step": 985
+    },
+    {
+      "epoch": 5.540845070422535,
+      "grad_norm": 2.1463332176208496,
+      "learning_rate": 2.949e-06,
+      "loss": 0.6716,
+      "step": 986
+    },
+    {
+      "epoch": 5.546478873239437,
+      "grad_norm": 2.329373359680176,
+      "learning_rate": 2.9520000000000003e-06,
+      "loss": 0.5893,
+      "step": 987
+    },
+    {
+      "epoch": 5.552112676056338,
+      "grad_norm": 2.2067439556121826,
+      "learning_rate": 2.955e-06,
+      "loss": 0.5582,
+      "step": 988
+    },
+    {
+      "epoch": 5.55774647887324,
+      "grad_norm": 2.535163164138794,
+      "learning_rate": 2.958e-06,
+      "loss": 0.6627,
+      "step": 989
+    },
+    {
+      "epoch": 5.563380281690141,
+      "grad_norm": 2.639993190765381,
+      "learning_rate": 2.961e-06,
+      "loss": 0.4677,
+      "step": 990
+    },
+    {
+      "epoch": 5.569014084507042,
+      "grad_norm": 2.374748945236206,
+      "learning_rate": 2.964e-06,
+      "loss": 0.5139,
+      "step": 991
+    },
+    {
+      "epoch": 5.574647887323944,
+      "grad_norm": 7.6541314125061035,
+      "learning_rate": 2.967e-06,
+      "loss": 0.4792,
+      "step": 992
+    },
+    {
+      "epoch": 5.580281690140845,
+      "grad_norm": 1.8786181211471558,
+      "learning_rate": 2.9700000000000004e-06,
+      "loss": 0.4939,
+      "step": 993
+    },
+    {
+      "epoch": 5.585915492957747,
+      "grad_norm": 2.8175878524780273,
+      "learning_rate": 2.9729999999999997e-06,
+      "loss": 0.5425,
+      "step": 994
+    },
+    {
+      "epoch": 5.591549295774648,
+      "grad_norm": 4.114474773406982,
+      "learning_rate": 2.976e-06,
+      "loss": 0.4222,
+      "step": 995
+    },
+    {
+      "epoch": 5.597183098591549,
+      "grad_norm": 7.6624860763549805,
+      "learning_rate": 2.979e-06,
+      "loss": 0.4513,
+      "step": 996
+    },
+    {
+      "epoch": 5.602816901408451,
+      "grad_norm": 1.9664232730865479,
+      "learning_rate": 2.982e-06,
+      "loss": 0.4046,
+      "step": 997
+    },
+    {
+      "epoch": 5.608450704225352,
+      "grad_norm": 2.2487237453460693,
+      "learning_rate": 2.9850000000000002e-06,
+      "loss": 0.4825,
+      "step": 998
+    },
+    {
+      "epoch": 5.614084507042254,
+      "grad_norm": 2.0302340984344482,
+      "learning_rate": 2.988e-06,
+      "loss": 0.5438,
+      "step": 999
+    },
+    {
+      "epoch": 5.619718309859155,
+      "grad_norm": 2.2003347873687744,
+      "learning_rate": 2.991e-06,
+      "loss": 0.4609,
+      "step": 1000
+    },
+    {
+      "epoch": 5.619718309859155,
+      "eval_cer": 0.1952717940741893,
+      "eval_loss": 0.7326682209968567,
+      "eval_runtime": 17.4904,
+      "eval_samples_per_second": 17.381,
+      "eval_steps_per_second": 0.572,
+      "eval_wer": 0.6745970836531082,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 100000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 565,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 5,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.055486320296336e+19,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}