diff --git "a/checkpoint-579/trainer_state.json" "b/checkpoint-579/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-579/trainer_state.json"
@@ -0,0 +1,4135 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.5,
+  "eval_steps": 97,
+  "global_step": 579,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0025906735751295338,
+      "grad_norm": 758.2562349755826,
+      "learning_rate": 0.0,
+      "loss": 1.3719,
+      "step": 1
+    },
+    {
+      "epoch": 0.0025906735751295338,
+      "eval_loss": 1.3159157037734985,
+      "eval_runtime": 36.907,
+      "eval_samples_per_second": 20.159,
+      "eval_steps_per_second": 1.273,
+      "step": 1
+    },
+    {
+      "epoch": 0.0051813471502590676,
+      "grad_norm": 666.308184823038,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.36,
+      "step": 2
+    },
+    {
+      "epoch": 0.007772020725388601,
+      "grad_norm": 211.0771195353068,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.3746,
+      "step": 3
+    },
+    {
+      "epoch": 0.010362694300518135,
+      "grad_norm": 431.5114709683218,
+      "learning_rate": 3e-06,
+      "loss": 1.3412,
+      "step": 4
+    },
+    {
+      "epoch": 0.012953367875647668,
+      "grad_norm": 230.87468433791625,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.3837,
+      "step": 5
+    },
+    {
+      "epoch": 0.015544041450777202,
+      "grad_norm": 635.1636587738542,
+      "learning_rate": 5e-06,
+      "loss": 1.3761,
+      "step": 6
+    },
+    {
+      "epoch": 0.018134715025906734,
+      "grad_norm": 791.5536958334704,
+      "learning_rate": 6e-06,
+      "loss": 1.2855,
+      "step": 7
+    },
+    {
+      "epoch": 0.02072538860103627,
+      "grad_norm": 667.7197994216477,
+      "learning_rate": 7e-06,
+      "loss": 1.3267,
+      "step": 8
+    },
+    {
+      "epoch": 0.023316062176165803,
+      "grad_norm": 254.3855973692125,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.2977,
+      "step": 9
+    },
+    {
+      "epoch": 0.025906735751295335,
+      "grad_norm": 162.29347257682093,
+      "learning_rate": 9e-06,
+      "loss": 1.3522,
+      "step": 10
+    },
+    {
+      "epoch": 0.02849740932642487,
+      "grad_norm": 352.6352930651456,
+      "learning_rate": 1e-05,
+      "loss": 1.2688,
+      "step": 11
+    },
+    {
+      "epoch": 0.031088082901554404,
+      "grad_norm": 148.2629265526552,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 1.3342,
+      "step": 12
+    },
+    {
+      "epoch": 0.03367875647668394,
+      "grad_norm": 249.88753789723657,
+      "learning_rate": 1.2e-05,
+      "loss": 1.2983,
+      "step": 13
+    },
+    {
+      "epoch": 0.03626943005181347,
+      "grad_norm": 184.03358422636597,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 1.3291,
+      "step": 14
+    },
+    {
+      "epoch": 0.038860103626943004,
+      "grad_norm": 198.4491469860763,
+      "learning_rate": 1.4e-05,
+      "loss": 1.4014,
+      "step": 15
+    },
+    {
+      "epoch": 0.04145077720207254,
+      "grad_norm": 680.9537058769038,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.3775,
+      "step": 16
+    },
+    {
+      "epoch": 0.04404145077720207,
+      "grad_norm": 563.0247638614801,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.3228,
+      "step": 17
+    },
+    {
+      "epoch": 0.046632124352331605,
+      "grad_norm": 271.985463813746,
+      "learning_rate": 1.7e-05,
+      "loss": 1.3695,
+      "step": 18
+    },
+    {
+      "epoch": 0.04922279792746114,
+      "grad_norm": 399.51218452223316,
+      "learning_rate": 1.8e-05,
+      "loss": 1.2556,
+      "step": 19
+    },
+    {
+      "epoch": 0.05181347150259067,
+      "grad_norm": 160.70697055826656,
+      "learning_rate": 1.9e-05,
+      "loss": 1.2982,
+      "step": 20
+    },
+    {
+      "epoch": 0.054404145077720206,
+      "grad_norm": 227.8927504687491,
+      "learning_rate": 2e-05,
+      "loss": 1.3532,
+      "step": 21
+    },
+    {
+      "epoch": 0.05699481865284974,
+      "grad_norm": 550.1538868076032,
+      "learning_rate": 2.1000000000000002e-05,
+      "loss": 1.2603,
+      "step": 22
+    },
+    {
+      "epoch": 0.05958549222797927,
+      "grad_norm": 291.8994359919024,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 1.3663,
+      "step": 23
+    },
+    {
+      "epoch": 0.06217616580310881,
+      "grad_norm": 120.60677833129643,
+      "learning_rate": 2.3e-05,
+      "loss": 1.3129,
+      "step": 24
+    },
+    {
+      "epoch": 0.06476683937823834,
+      "grad_norm": 414.4006662101242,
+      "learning_rate": 2.4e-05,
+      "loss": 1.3037,
+      "step": 25
+    },
+    {
+      "epoch": 0.06735751295336788,
+      "grad_norm": 141.48324465317884,
+      "learning_rate": 2.5e-05,
+      "loss": 1.3095,
+      "step": 26
+    },
+    {
+      "epoch": 0.06994818652849741,
+      "grad_norm": 147.86066819937994,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 1.2372,
+      "step": 27
+    },
+    {
+      "epoch": 0.07253886010362694,
+      "grad_norm": 214.47337614964576,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 1.3384,
+      "step": 28
+    },
+    {
+      "epoch": 0.07512953367875648,
+      "grad_norm": 898.4324889241673,
+      "learning_rate": 2.8e-05,
+      "loss": 1.2003,
+      "step": 29
+    },
+    {
+      "epoch": 0.07772020725388601,
+      "grad_norm": 128.83026557596128,
+      "learning_rate": 2.9e-05,
+      "loss": 1.2172,
+      "step": 30
+    },
+    {
+      "epoch": 0.08031088082901554,
+      "grad_norm": 183.0777862405529,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 1.2674,
+      "step": 31
+    },
+    {
+      "epoch": 0.08290155440414508,
+      "grad_norm": 119.01841833358732,
+      "learning_rate": 3.1e-05,
+      "loss": 1.2554,
+      "step": 32
+    },
+    {
+      "epoch": 0.08549222797927461,
+      "grad_norm": 117.65980267542858,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.2716,
+      "step": 33
+    },
+    {
+      "epoch": 0.08808290155440414,
+      "grad_norm": 82.40151099433953,
+      "learning_rate": 3.3e-05,
+      "loss": 1.2019,
+      "step": 34
+    },
+    {
+      "epoch": 0.09067357512953368,
+      "grad_norm": 82.61816783653785,
+      "learning_rate": 3.4e-05,
+      "loss": 1.2424,
+      "step": 35
+    },
+    {
+      "epoch": 0.09326424870466321,
+      "grad_norm": 136.42743433868276,
+      "learning_rate": 3.5000000000000004e-05,
+      "loss": 1.2066,
+      "step": 36
+    },
+    {
+      "epoch": 0.09585492227979274,
+      "grad_norm": 36.775911657584444,
+      "learning_rate": 3.6e-05,
+      "loss": 1.2485,
+      "step": 37
+    },
+    {
+      "epoch": 0.09844559585492228,
+      "grad_norm": 56.55022603284064,
+      "learning_rate": 3.7000000000000005e-05,
+      "loss": 1.2112,
+      "step": 38
+    },
+    {
+      "epoch": 0.10103626943005181,
+      "grad_norm": 50.09896932886107,
+      "learning_rate": 3.8e-05,
+      "loss": 1.2027,
+      "step": 39
+    },
+    {
+      "epoch": 0.10362694300518134,
+      "grad_norm": 54.2661481198025,
+      "learning_rate": 3.9e-05,
+      "loss": 1.2673,
+      "step": 40
+    },
+    {
+      "epoch": 0.10621761658031088,
+      "grad_norm": 60.04145981731815,
+      "learning_rate": 4e-05,
+      "loss": 1.1648,
+      "step": 41
+    },
+    {
+      "epoch": 0.10880829015544041,
+      "grad_norm": 169.47741055545822,
+      "learning_rate": 3.999981580539036e-05,
+      "loss": 1.2393,
+      "step": 42
+    },
+    {
+      "epoch": 0.11139896373056994,
+      "grad_norm": 43.64716987307323,
+      "learning_rate": 3.9999263224954204e-05,
+      "loss": 1.2906,
+      "step": 43
+    },
+    {
+      "epoch": 0.11398963730569948,
+      "grad_norm": 51.3206609767585,
+      "learning_rate": 3.999834226886976e-05,
+      "loss": 1.1807,
+      "step": 44
+    },
+    {
+      "epoch": 0.11658031088082901,
+      "grad_norm": 38.95055887413869,
+      "learning_rate": 3.999705295410054e-05,
+      "loss": 1.1825,
+      "step": 45
+    },
+    {
+      "epoch": 0.11917098445595854,
+      "grad_norm": 40.59968974426338,
+      "learning_rate": 3.999539530439504e-05,
+      "loss": 1.193,
+      "step": 46
+    },
+    {
+      "epoch": 0.12176165803108809,
+      "grad_norm": 34.5796571445333,
+      "learning_rate": 3.9993369350286265e-05,
+      "loss": 1.2127,
+      "step": 47
+    },
+    {
+      "epoch": 0.12435233160621761,
+      "grad_norm": 37.97693356149241,
+      "learning_rate": 3.99909751290912e-05,
+      "loss": 1.1543,
+      "step": 48
+    },
+    {
+      "epoch": 0.12694300518134716,
+      "grad_norm": 82.9217015858092,
+      "learning_rate": 3.9988212684910107e-05,
+      "loss": 1.2329,
+      "step": 49
+    },
+    {
+      "epoch": 0.12953367875647667,
+      "grad_norm": 49.256542144400214,
+      "learning_rate": 3.9985082068625724e-05,
+      "loss": 1.212,
+      "step": 50
+    },
+    {
+      "epoch": 0.13212435233160622,
+      "grad_norm": 45.025980435259484,
+      "learning_rate": 3.998158333790231e-05,
+      "loss": 1.2129,
+      "step": 51
+    },
+    {
+      "epoch": 0.13471502590673576,
+      "grad_norm": 45.98465689592428,
+      "learning_rate": 3.99777165571846e-05,
+      "loss": 1.1709,
+      "step": 52
+    },
+    {
+      "epoch": 0.13730569948186527,
+      "grad_norm": 43.481241408477906,
+      "learning_rate": 3.997348179769661e-05,
+      "loss": 1.1614,
+      "step": 53
+    },
+    {
+      "epoch": 0.13989637305699482,
+      "grad_norm": 82.17633750834132,
+      "learning_rate": 3.996887913744033e-05,
+      "loss": 1.2205,
+      "step": 54
+    },
+    {
+      "epoch": 0.14248704663212436,
+      "grad_norm": 53.0176514970764,
+      "learning_rate": 3.9963908661194285e-05,
+      "loss": 1.1204,
+      "step": 55
+    },
+    {
+      "epoch": 0.14507772020725387,
+      "grad_norm": 67.86382426995611,
+      "learning_rate": 3.995857046051196e-05,
+      "loss": 1.1839,
+      "step": 56
+    },
+    {
+      "epoch": 0.14766839378238342,
+      "grad_norm": 31.282407703790597,
+      "learning_rate": 3.995286463372013e-05,
+      "loss": 1.2126,
+      "step": 57
+    },
+    {
+      "epoch": 0.15025906735751296,
+      "grad_norm": 52.200764429265604,
+      "learning_rate": 3.994679128591706e-05,
+      "loss": 1.2036,
+      "step": 58
+    },
+    {
+      "epoch": 0.15284974093264247,
+      "grad_norm": 60.706608653531895,
+      "learning_rate": 3.9940350528970535e-05,
+      "loss": 1.1848,
+      "step": 59
+    },
+    {
+      "epoch": 0.15544041450777202,
+      "grad_norm": 47.31754062899529,
+      "learning_rate": 3.993354248151583e-05,
+      "loss": 1.0869,
+      "step": 60
+    },
+    {
+      "epoch": 0.15803108808290156,
+      "grad_norm": 49.42450836392811,
+      "learning_rate": 3.9926367268953514e-05,
+      "loss": 1.2651,
+      "step": 61
+    },
+    {
+      "epoch": 0.16062176165803108,
+      "grad_norm": 38.791167030088886,
+      "learning_rate": 3.991882502344712e-05,
+      "loss": 1.1881,
+      "step": 62
+    },
+    {
+      "epoch": 0.16321243523316062,
+      "grad_norm": 56.16339499737216,
+      "learning_rate": 3.991091588392077e-05,
+      "loss": 1.1518,
+      "step": 63
+    },
+    {
+      "epoch": 0.16580310880829016,
+      "grad_norm": 861.8559063020828,
+      "learning_rate": 3.990263999605652e-05,
+      "loss": 1.1614,
+      "step": 64
+    },
+    {
+      "epoch": 0.16839378238341968,
+      "grad_norm": 50.92822786500888,
+      "learning_rate": 3.989399751229179e-05,
+      "loss": 1.1998,
+      "step": 65
+    },
+    {
+      "epoch": 0.17098445595854922,
+      "grad_norm": 31.04121324055666,
+      "learning_rate": 3.988498859181645e-05,
+      "loss": 1.1795,
+      "step": 66
+    },
+    {
+      "epoch": 0.17357512953367876,
+      "grad_norm": 50.33061983380845,
+      "learning_rate": 3.9875613400569975e-05,
+      "loss": 1.1742,
+      "step": 67
+    },
+    {
+      "epoch": 0.17616580310880828,
+      "grad_norm": 75.20462514003519,
+      "learning_rate": 3.986587211123833e-05,
+      "loss": 1.1856,
+      "step": 68
+    },
+    {
+      "epoch": 0.17875647668393782,
+      "grad_norm": 38.82139317052205,
+      "learning_rate": 3.98557649032508e-05,
+      "loss": 1.1529,
+      "step": 69
+    },
+    {
+      "epoch": 0.18134715025906736,
+      "grad_norm": 36.55988806615175,
+      "learning_rate": 3.984529196277674e-05,
+      "loss": 1.1884,
+      "step": 70
+    },
+    {
+      "epoch": 0.18393782383419688,
+      "grad_norm": 104.8931793971097,
+      "learning_rate": 3.983445348272203e-05,
+      "loss": 1.2182,
+      "step": 71
+    },
+    {
+      "epoch": 0.18652849740932642,
+      "grad_norm": 36.50395409234617,
+      "learning_rate": 3.982324966272566e-05,
+      "loss": 1.1609,
+      "step": 72
+    },
+    {
+      "epoch": 0.18911917098445596,
+      "grad_norm": 35.019191693448626,
+      "learning_rate": 3.981168070915594e-05,
+      "loss": 1.173,
+      "step": 73
+    },
+    {
+      "epoch": 0.19170984455958548,
+      "grad_norm": 33.378390048053596,
+      "learning_rate": 3.979974683510677e-05,
+      "loss": 1.173,
+      "step": 74
+    },
+    {
+      "epoch": 0.19430051813471502,
+      "grad_norm": 43.356840136984154,
+      "learning_rate": 3.978744826039366e-05,
+      "loss": 1.2032,
+      "step": 75
+    },
+    {
+      "epoch": 0.19689119170984457,
+      "grad_norm": 31.285725922510768,
+      "learning_rate": 3.977478521154974e-05,
+      "loss": 1.1569,
+      "step": 76
+    },
+    {
+      "epoch": 0.19948186528497408,
+      "grad_norm": 35.19264482867074,
+      "learning_rate": 3.9761757921821544e-05,
+      "loss": 1.1365,
+      "step": 77
+    },
+    {
+      "epoch": 0.20207253886010362,
+      "grad_norm": 44.66037256551279,
+      "learning_rate": 3.974836663116472e-05,
+      "loss": 1.164,
+      "step": 78
+    },
+    {
+      "epoch": 0.20466321243523317,
+      "grad_norm": 68.91101457952654,
+      "learning_rate": 3.973461158623963e-05,
+      "loss": 1.2256,
+      "step": 79
+    },
+    {
+      "epoch": 0.20725388601036268,
+      "grad_norm": 45.866521854583,
+      "learning_rate": 3.9720493040406786e-05,
+      "loss": 1.1697,
+      "step": 80
+    },
+    {
+      "epoch": 0.20984455958549222,
+      "grad_norm": 59.63095169617338,
+      "learning_rate": 3.970601125372218e-05,
+      "loss": 1.2094,
+      "step": 81
+    },
+    {
+      "epoch": 0.21243523316062177,
+      "grad_norm": 39.085597271064216,
+      "learning_rate": 3.9691166492932535e-05,
+      "loss": 1.1048,
+      "step": 82
+    },
+    {
+      "epoch": 0.21502590673575128,
+      "grad_norm": 36.40256073477861,
+      "learning_rate": 3.9675959031470336e-05,
+      "loss": 1.248,
+      "step": 83
+    },
+    {
+      "epoch": 0.21761658031088082,
+      "grad_norm": 29.846921716586085,
+      "learning_rate": 3.966038914944881e-05,
+      "loss": 1.1718,
+      "step": 84
+    },
+    {
+      "epoch": 0.22020725388601037,
+      "grad_norm": 50.87052190327881,
+      "learning_rate": 3.964445713365682e-05,
+      "loss": 1.1529,
+      "step": 85
+    },
+    {
+      "epoch": 0.22279792746113988,
+      "grad_norm": 35.32915760431302,
+      "learning_rate": 3.9628163277553486e-05,
+      "loss": 1.1767,
+      "step": 86
+    },
+    {
+      "epoch": 0.22538860103626943,
+      "grad_norm": 157.5587514654703,
+      "learning_rate": 3.961150788126286e-05,
+      "loss": 1.2194,
+      "step": 87
+    },
+    {
+      "epoch": 0.22797927461139897,
+      "grad_norm": 25.03485489120971,
+      "learning_rate": 3.9594491251568376e-05,
+      "loss": 1.1392,
+      "step": 88
+    },
+    {
+      "epoch": 0.23056994818652848,
+      "grad_norm": 80.55933867045263,
+      "learning_rate": 3.957711370190716e-05,
+      "loss": 1.1819,
+      "step": 89
+    },
+    {
+      "epoch": 0.23316062176165803,
+      "grad_norm": 272.22874004071406,
+      "learning_rate": 3.9559375552364325e-05,
+      "loss": 1.0998,
+      "step": 90
+    },
+    {
+      "epoch": 0.23575129533678757,
+      "grad_norm": 91.94671663482514,
+      "learning_rate": 3.954127712966702e-05,
+      "loss": 1.2494,
+      "step": 91
+    },
+    {
+      "epoch": 0.23834196891191708,
+      "grad_norm": 54.31533598131098,
+      "learning_rate": 3.952281876717843e-05,
+      "loss": 1.1385,
+      "step": 92
+    },
+    {
+      "epoch": 0.24093264248704663,
+      "grad_norm": 103.20789745908105,
+      "learning_rate": 3.950400080489165e-05,
+      "loss": 1.1398,
+      "step": 93
+    },
+    {
+      "epoch": 0.24352331606217617,
+      "grad_norm": 45.14746362545893,
+      "learning_rate": 3.94848235894234e-05,
+      "loss": 1.2697,
+      "step": 94
+    },
+    {
+      "epoch": 0.24611398963730569,
+      "grad_norm": 21.271923336142002,
+      "learning_rate": 3.9465287474007654e-05,
+      "loss": 1.1397,
+      "step": 95
+    },
+    {
+      "epoch": 0.24870466321243523,
+      "grad_norm": 93.89786795431422,
+      "learning_rate": 3.944539281848912e-05,
+      "loss": 1.1542,
+      "step": 96
+    },
+    {
+      "epoch": 0.25129533678756477,
+      "grad_norm": 32.38768349342839,
+      "learning_rate": 3.942513998931663e-05,
+      "loss": 1.1693,
+      "step": 97
+    },
+    {
+      "epoch": 0.25129533678756477,
+      "eval_loss": 1.1344976425170898,
+      "eval_runtime": 37.8807,
+      "eval_samples_per_second": 19.641,
+      "eval_steps_per_second": 1.241,
+      "step": 97
+    },
+    {
+      "epoch": 0.2538860103626943,
+      "grad_norm": 91.41293468177638,
+      "learning_rate": 3.940452935953639e-05,
+      "loss": 1.1724,
+      "step": 98
+    },
+    {
+      "epoch": 0.25647668393782386,
+      "grad_norm": 39.20645478419229,
+      "learning_rate": 3.9383561308785075e-05,
+      "loss": 1.1583,
+      "step": 99
+    },
+    {
+      "epoch": 0.25906735751295334,
+      "grad_norm": 35.32804513153546,
+      "learning_rate": 3.9362236223282885e-05,
+      "loss": 1.158,
+      "step": 100
+    },
+    {
+      "epoch": 0.2616580310880829,
+      "grad_norm": 35.24783762804842,
+      "learning_rate": 3.934055449582641e-05,
+      "loss": 1.1552,
+      "step": 101
+    },
+    {
+      "epoch": 0.26424870466321243,
+      "grad_norm": 33.743808031979775,
+      "learning_rate": 3.931851652578137e-05,
+      "loss": 1.264,
+      "step": 102
+    },
+    {
+      "epoch": 0.266839378238342,
+      "grad_norm": 113.49798793226394,
+      "learning_rate": 3.92961227190753e-05,
+      "loss": 1.2361,
+      "step": 103
+    },
+    {
+      "epoch": 0.2694300518134715,
+      "grad_norm": 31.813807349410364,
+      "learning_rate": 3.9273373488190036e-05,
+      "loss": 1.1246,
+      "step": 104
+    },
+    {
+      "epoch": 0.27202072538860106,
+      "grad_norm": 29.391695486306187,
+      "learning_rate": 3.925026925215417e-05,
+      "loss": 1.1142,
+      "step": 105
+    },
+    {
+      "epoch": 0.27461139896373055,
+      "grad_norm": 33.79933331839905,
+      "learning_rate": 3.922681043653526e-05,
+      "loss": 1.1401,
+      "step": 106
+    },
+    {
+      "epoch": 0.2772020725388601,
+      "grad_norm": 39.09509012730907,
+      "learning_rate": 3.920299747343204e-05,
+      "loss": 1.1822,
+      "step": 107
+    },
+    {
+      "epoch": 0.27979274611398963,
+      "grad_norm": 37.81471938433609,
+      "learning_rate": 3.9178830801466465e-05,
+      "loss": 1.1592,
+      "step": 108
+    },
+    {
+      "epoch": 0.2823834196891192,
+      "grad_norm": 69.07753778460207,
+      "learning_rate": 3.915431086577561e-05,
+      "loss": 1.1683,
+      "step": 109
+    },
+    {
+      "epoch": 0.2849740932642487,
+      "grad_norm": 28.864787246081605,
+      "learning_rate": 3.912943811800347e-05,
+      "loss": 1.1179,
+      "step": 110
+    },
+    {
+      "epoch": 0.28756476683937826,
+      "grad_norm": 28.842042951717836,
+      "learning_rate": 3.910421301629264e-05,
+      "loss": 1.1317,
+      "step": 111
+    },
+    {
+      "epoch": 0.29015544041450775,
+      "grad_norm": 51.475482074695506,
+      "learning_rate": 3.9078636025275904e-05,
+      "loss": 1.1451,
+      "step": 112
+    },
+    {
+      "epoch": 0.2927461139896373,
+      "grad_norm": 33.48279556713943,
+      "learning_rate": 3.9052707616067654e-05,
+      "loss": 1.1554,
+      "step": 113
+    },
+    {
+      "epoch": 0.29533678756476683,
+      "grad_norm": 21.279603575929844,
+      "learning_rate": 3.9026428266255205e-05,
+      "loss": 1.1636,
+      "step": 114
+    },
+    {
+      "epoch": 0.2979274611398964,
+      "grad_norm": 36.226178034876675,
+      "learning_rate": 3.899979845989003e-05,
+      "loss": 1.1966,
+      "step": 115
+    },
+    {
+      "epoch": 0.3005181347150259,
+      "grad_norm": 29.90506353145981,
+      "learning_rate": 3.897281868747878e-05,
+      "loss": 1.1888,
+      "step": 116
+    },
+    {
+      "epoch": 0.30310880829015546,
+      "grad_norm": 36.04602777809767,
+      "learning_rate": 3.894548944597434e-05,
+      "loss": 1.2066,
+      "step": 117
+    },
+    {
+      "epoch": 0.30569948186528495,
+      "grad_norm": 36.42793844948301,
+      "learning_rate": 3.8917811238766606e-05,
+      "loss": 1.1712,
+      "step": 118
+    },
+    {
+      "epoch": 0.3082901554404145,
+      "grad_norm": 58.788967662325696,
+      "learning_rate": 3.888978457567323e-05,
+      "loss": 1.1225,
+      "step": 119
+    },
+    {
+      "epoch": 0.31088082901554404,
+      "grad_norm": 29.357299816022326,
+      "learning_rate": 3.886140997293024e-05,
+      "loss": 1.1315,
+      "step": 120
+    },
+    {
+      "epoch": 0.3134715025906736,
+      "grad_norm": 95.08345317107502,
+      "learning_rate": 3.883268795318252e-05,
+      "loss": 1.1852,
+      "step": 121
+    },
+    {
+      "epoch": 0.3160621761658031,
+      "grad_norm": 33.6623824593179,
+      "learning_rate": 3.88036190454742e-05,
+      "loss": 1.16,
+      "step": 122
+    },
+    {
+      "epoch": 0.31865284974093266,
+      "grad_norm": 42.587546987131105,
+      "learning_rate": 3.8774203785238886e-05,
+      "loss": 1.1374,
+      "step": 123
+    },
+    {
+      "epoch": 0.32124352331606215,
+      "grad_norm": 33.360649853064245,
+      "learning_rate": 3.8744442714289816e-05,
+      "loss": 1.1757,
+      "step": 124
+    },
+    {
+      "epoch": 0.3238341968911917,
+      "grad_norm": 49.09256643961471,
+      "learning_rate": 3.8714336380809874e-05,
+      "loss": 1.1782,
+      "step": 125
+    },
+    {
+      "epoch": 0.32642487046632124,
+      "grad_norm": 31.505007051172793,
+      "learning_rate": 3.86838853393415e-05,
+      "loss": 1.195,
+      "step": 126
+    },
+    {
+      "epoch": 0.3290155440414508,
+      "grad_norm": 34.36735417254799,
+      "learning_rate": 3.865309015077645e-05,
+      "loss": 1.1078,
+      "step": 127
+    },
+    {
+      "epoch": 0.3316062176165803,
+      "grad_norm": 36.63220606142181,
+      "learning_rate": 3.862195138234551e-05,
+      "loss": 1.1319,
+      "step": 128
+    },
+    {
+      "epoch": 0.33419689119170987,
+      "grad_norm": 53.324986862513676,
+      "learning_rate": 3.859046960760801e-05,
+      "loss": 1.2301,
+      "step": 129
+    },
+    {
+      "epoch": 0.33678756476683935,
+      "grad_norm": 47.41445409144979,
+      "learning_rate": 3.855864540644126e-05,
+      "loss": 1.2366,
+      "step": 130
+    },
+    {
+      "epoch": 0.3393782383419689,
+      "grad_norm": 32.57355122427366,
+      "learning_rate": 3.8526479365029906e-05,
+      "loss": 1.142,
+      "step": 131
+    },
+    {
+      "epoch": 0.34196891191709844,
+      "grad_norm": 28.445824333644715,
+      "learning_rate": 3.849397207585508e-05,
+      "loss": 1.0847,
+      "step": 132
+    },
+    {
+      "epoch": 0.344559585492228,
+      "grad_norm": 49.23062726715889,
+      "learning_rate": 3.846112413768353e-05,
+      "loss": 1.2241,
+      "step": 133
+    },
+    {
+      "epoch": 0.3471502590673575,
+      "grad_norm": 53.424206543788074,
+      "learning_rate": 3.842793615555657e-05,
+      "loss": 1.2392,
+      "step": 134
+    },
+    {
+      "epoch": 0.34974093264248707,
+      "grad_norm": 38.19316140175426,
+      "learning_rate": 3.8394408740778934e-05,
+      "loss": 1.1208,
+      "step": 135
+    },
+    {
+      "epoch": 0.35233160621761656,
+      "grad_norm": 32.35931252369273,
+      "learning_rate": 3.836054251090755e-05,
+      "loss": 1.1604,
+      "step": 136
+    },
+    {
+      "epoch": 0.3549222797927461,
+      "grad_norm": 37.90085344799495,
+      "learning_rate": 3.83263380897401e-05,
+      "loss": 1.1134,
+      "step": 137
+    },
+    {
+      "epoch": 0.35751295336787564,
+      "grad_norm": 44.49191588319939,
+      "learning_rate": 3.829179610730359e-05,
+      "loss": 1.1281,
+      "step": 138
+    },
+    {
+      "epoch": 0.3601036269430052,
+      "grad_norm": 141.98524430756757,
+      "learning_rate": 3.8256917199842715e-05,
+      "loss": 1.0928,
+      "step": 139
+    },
+    {
+      "epoch": 0.3626943005181347,
+      "grad_norm": 30.887093976524472,
+      "learning_rate": 3.822170200980815e-05,
+      "loss": 1.0936,
+      "step": 140
+    },
+    {
+      "epoch": 0.36528497409326427,
+      "grad_norm": 21.980521878837745,
+      "learning_rate": 3.818615118584472e-05,
+      "loss": 1.1368,
+      "step": 141
+    },
+    {
+      "epoch": 0.36787564766839376,
+      "grad_norm": 538.6650762618656,
+      "learning_rate": 3.815026538277943e-05,
+      "loss": 1.0918,
+      "step": 142
+    },
+    {
+      "epoch": 0.3704663212435233,
+      "grad_norm": 40.842881572203,
+      "learning_rate": 3.811404526160943e-05,
+      "loss": 1.1705,
+      "step": 143
+    },
+    {
+      "epoch": 0.37305699481865284,
+      "grad_norm": 26.891553492377298,
+      "learning_rate": 3.8077491489489835e-05,
+      "loss": 1.1468,
+      "step": 144
+    },
+    {
+      "epoch": 0.3756476683937824,
+      "grad_norm": 45.138483181178074,
+      "learning_rate": 3.8040604739721415e-05,
+      "loss": 1.1679,
+      "step": 145
+    },
+    {
+      "epoch": 0.37823834196891193,
+      "grad_norm": 35.133763086168244,
+      "learning_rate": 3.8003385691738227e-05,
+      "loss": 1.1029,
+      "step": 146
+    },
+    {
+      "epoch": 0.38082901554404147,
+      "grad_norm": 36.941250802707344,
+      "learning_rate": 3.7965835031095065e-05,
+      "loss": 1.1491,
+      "step": 147
+    },
+    {
+      "epoch": 0.38341968911917096,
+      "grad_norm": 90.1080256703095,
+      "learning_rate": 3.792795344945485e-05,
+      "loss": 1.1212,
+      "step": 148
+    },
+    {
+      "epoch": 0.3860103626943005,
+      "grad_norm": 39.70360899750413,
+      "learning_rate": 3.7889741644575914e-05,
+      "loss": 1.15,
+      "step": 149
+    },
+    {
+      "epoch": 0.38860103626943004,
+      "grad_norm": 28.229369877304094,
+      "learning_rate": 3.78512003202991e-05,
+      "loss": 1.1111,
+      "step": 150
+    },
+    {
+      "epoch": 0.3911917098445596,
+      "grad_norm": 31.611752191925987,
+      "learning_rate": 3.7812330186534815e-05,
+      "loss": 1.1366,
+      "step": 151
+    },
+    {
+      "epoch": 0.39378238341968913,
+      "grad_norm": 38.196015586772425,
+      "learning_rate": 3.777313195924998e-05,
+      "loss": 1.1433,
+      "step": 152
+    },
+    {
+      "epoch": 0.3963730569948187,
+      "grad_norm": 22.732638044547453,
+      "learning_rate": 3.773360636045481e-05,
+      "loss": 1.1125,
+      "step": 153
+    },
+    {
+      "epoch": 0.39896373056994816,
+      "grad_norm": 90.19158665385014,
+      "learning_rate": 3.7693754118189525e-05,
+      "loss": 1.1242,
+      "step": 154
+    },
+    {
+      "epoch": 0.4015544041450777,
+      "grad_norm": 42.43479974993017,
+      "learning_rate": 3.765357596651095e-05,
+      "loss": 1.1191,
+      "step": 155
+    },
+    {
+      "epoch": 0.40414507772020725,
+      "grad_norm": 88.0076735720364,
+      "learning_rate": 3.761307264547899e-05,
+      "loss": 1.1718,
+      "step": 156
+    },
+    {
+      "epoch": 0.4067357512953368,
+      "grad_norm": 30.782507703935767,
+      "learning_rate": 3.757224490114297e-05,
+      "loss": 1.109,
+      "step": 157
+    },
+    {
+      "epoch": 0.40932642487046633,
+      "grad_norm": 69.89871106113397,
+      "learning_rate": 3.7531093485527943e-05,
+      "loss": 1.1018,
+      "step": 158
+    },
+    {
+      "epoch": 0.4119170984455959,
+      "grad_norm": 37.339006645717305,
+      "learning_rate": 3.7489619156620796e-05,
+      "loss": 1.1358,
+      "step": 159
+    },
+    {
+      "epoch": 0.41450777202072536,
+      "grad_norm": 28.06388054378899,
+      "learning_rate": 3.744782267835632e-05,
+      "loss": 1.0847,
+      "step": 160
+    },
+    {
+      "epoch": 0.4170984455958549,
+      "grad_norm": 54.05874281297702,
+      "learning_rate": 3.740570482060311e-05,
+      "loss": 1.1682,
+      "step": 161
+    },
+    {
+      "epoch": 0.41968911917098445,
+      "grad_norm": 32.299093265328835,
+      "learning_rate": 3.73632663591494e-05,
+      "loss": 1.1413,
+      "step": 162
+    },
+    {
+      "epoch": 0.422279792746114,
+      "grad_norm": 31.213652090157694,
+      "learning_rate": 3.732050807568878e-05,
+      "loss": 1.1313,
+      "step": 163
+    },
+    {
+      "epoch": 0.42487046632124353,
+      "grad_norm": 40.01090035937505,
+      "learning_rate": 3.727743075780578e-05,
+      "loss": 1.1513,
+      "step": 164
+    },
+    {
+      "epoch": 0.4274611398963731,
+      "grad_norm": 47.11352577964853,
+      "learning_rate": 3.723403519896136e-05,
+      "loss": 1.2192,
+      "step": 165
+    },
+    {
+      "epoch": 0.43005181347150256,
+      "grad_norm": 28.645086506093037,
+      "learning_rate": 3.7190322198478355e-05,
+      "loss": 1.1097,
+      "step": 166
+    },
+    {
+      "epoch": 0.4326424870466321,
+      "grad_norm": 35.28541113925116,
+      "learning_rate": 3.7146292561526654e-05,
+      "loss": 1.1557,
+      "step": 167
+    },
+    {
+      "epoch": 0.43523316062176165,
+      "grad_norm": 58.30281063037669,
+      "learning_rate": 3.7101947099108425e-05,
+      "loss": 1.1829,
+      "step": 168
+    },
+    {
+      "epoch": 0.4378238341968912,
+      "grad_norm": 26.33563548968379,
+      "learning_rate": 3.70572866280432e-05,
+      "loss": 1.147,
+      "step": 169
+    },
+    {
+      "epoch": 0.44041450777202074,
+      "grad_norm": 57.00052875402651,
+      "learning_rate": 3.701231197095277e-05,
+      "loss": 1.1212,
+      "step": 170
+    },
+    {
+      "epoch": 0.4430051813471503,
+      "grad_norm": 23.672828037237174,
+      "learning_rate": 3.696702395624608e-05,
+      "loss": 1.1152,
+      "step": 171
+    },
+    {
+      "epoch": 0.44559585492227977,
+      "grad_norm": 41.1264174112964,
+      "learning_rate": 3.692142341810395e-05,
+      "loss": 1.1154,
+      "step": 172
+    },
+    {
+      "epoch": 0.4481865284974093,
+      "grad_norm": 26.72177706144361,
+      "learning_rate": 3.6875511196463715e-05,
+      "loss": 1.1725,
+      "step": 173
+    },
+    {
+      "epoch": 0.45077720207253885,
+      "grad_norm": 95.4088800585977,
+      "learning_rate": 3.682928813700375e-05,
+      "loss": 1.1339,
+      "step": 174
+    },
+    {
+      "epoch": 0.4533678756476684,
+      "grad_norm": 34.33666578349465,
+      "learning_rate": 3.678275509112788e-05,
+      "loss": 1.1867,
+      "step": 175
+    },
+    {
+      "epoch": 0.45595854922279794,
+      "grad_norm": 31.032304531003014,
+      "learning_rate": 3.6735912915949745e-05,
+      "loss": 1.1386,
+      "step": 176
+    },
+    {
+      "epoch": 0.4585492227979275,
+      "grad_norm": 55.22043313188224,
+      "learning_rate": 3.6688762474276945e-05,
+      "loss": 1.1102,
+      "step": 177
+    },
+    {
+      "epoch": 0.46113989637305697,
+      "grad_norm": 29.82713377876857,
+      "learning_rate": 3.6641304634595216e-05,
+      "loss": 1.1564,
+      "step": 178
+    },
+    {
+      "epoch": 0.4637305699481865,
+      "grad_norm": 35.71025459541737,
+      "learning_rate": 3.659354027105238e-05,
+      "loss": 1.0939,
+      "step": 179
+    },
+    {
+      "epoch": 0.46632124352331605,
+      "grad_norm": 52.41175655642653,
+      "learning_rate": 3.6545470263442265e-05,
+      "loss": 1.1578,
+      "step": 180
+    },
+    {
+      "epoch": 0.4689119170984456,
+      "grad_norm": 27.682485766528306,
+      "learning_rate": 3.649709549718849e-05,
+      "loss": 1.1875,
+      "step": 181
+    },
+    {
+      "epoch": 0.47150259067357514,
+      "grad_norm": 36.53293663303487,
+      "learning_rate": 3.6448416863328186e-05,
+      "loss": 1.1111,
+      "step": 182
+    },
+    {
+      "epoch": 0.4740932642487047,
+      "grad_norm": 31.45177998538027,
+      "learning_rate": 3.639943525849555e-05,
+      "loss": 1.113,
+      "step": 183
+    },
+    {
+      "epoch": 0.47668393782383417,
+      "grad_norm": 28.323097072885673,
+      "learning_rate": 3.635015158490533e-05,
+      "loss": 1.1159,
+      "step": 184
+    },
+    {
+      "epoch": 0.4792746113989637,
+      "grad_norm": 47.75573754341213,
+      "learning_rate": 3.6300566750336225e-05,
+      "loss": 1.1305,
+      "step": 185
+    },
+    {
+      "epoch": 0.48186528497409326,
+      "grad_norm": 21.384095061494357,
+      "learning_rate": 3.625068166811418e-05,
+      "loss": 1.1369,
+      "step": 186
+    },
+    {
+      "epoch": 0.4844559585492228,
+      "grad_norm": 30.714645036809546,
+      "learning_rate": 3.6200497257095504e-05,
+      "loss": 1.1858,
+      "step": 187
+    },
+    {
+      "epoch": 0.48704663212435234,
+      "grad_norm": 35.12161426399798,
+      "learning_rate": 3.615001444165001e-05,
+      "loss": 1.1293,
+      "step": 188
+    },
+    {
+      "epoch": 0.4896373056994819,
+      "grad_norm": 116.83443661381396,
+      "learning_rate": 3.6099234151643924e-05,
+      "loss": 1.1515,
+      "step": 189
+    },
+    {
+      "epoch": 0.49222797927461137,
+      "grad_norm": 55.47885243409044,
+      "learning_rate": 3.604815732242283e-05,
+      "loss": 1.112,
+      "step": 190
+    },
+    {
+      "epoch": 0.4948186528497409,
+      "grad_norm": 32.332747429034285,
+      "learning_rate": 3.5996784894794394e-05,
+      "loss": 1.1661,
+      "step": 191
+    },
+    {
+      "epoch": 0.49740932642487046,
+      "grad_norm": 33.039210183180046,
+      "learning_rate": 3.594511781501103e-05,
+      "loss": 1.1244,
+      "step": 192
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 21.325687337182504,
+      "learning_rate": 3.58931570347525e-05,
+      "loss": 1.1634,
+      "step": 193
+    },
+    {
+      "epoch": 0.5025906735751295,
+      "grad_norm": 51.37599478469561,
+      "learning_rate": 3.584090351110838e-05,
+      "loss": 1.2106,
+      "step": 194
+    },
+    {
+      "epoch": 0.5025906735751295,
+      "eval_loss": 1.1119717359542847,
+      "eval_runtime": 49.6027,
+      "eval_samples_per_second": 14.999,
+      "eval_steps_per_second": 0.948,
+      "step": 194
+    },
+    {
+      "epoch": 0.5051813471502591,
+      "grad_norm": 42.105169991612456,
+      "learning_rate": 3.57883582065604e-05,
+      "loss": 1.1303,
+      "step": 195
+    },
+    {
+      "epoch": 0.5077720207253886,
+      "grad_norm": 37.14457014578168,
+      "learning_rate": 3.573552208896474e-05,
+      "loss": 1.1483,
+      "step": 196
+    },
+    {
+      "epoch": 0.5103626943005182,
+      "grad_norm": 28.56241612018119,
+      "learning_rate": 3.568239613153421e-05,
+      "loss": 1.0843,
+      "step": 197
+    },
+    {
+      "epoch": 0.5129533678756477,
+      "grad_norm": 35.399304035761865,
+      "learning_rate": 3.5628981312820315e-05,
+      "loss": 1.1177,
+      "step": 198
+    },
+    {
+      "epoch": 0.5155440414507773,
+      "grad_norm": 25.91156850470446,
+      "learning_rate": 3.557527861669522e-05,
+      "loss": 1.1215,
+      "step": 199
+    },
+    {
+      "epoch": 0.5181347150259067,
+      "grad_norm": 43.509516777992324,
+      "learning_rate": 3.552128903233363e-05,
+      "loss": 1.1532,
+      "step": 200
+    },
+    {
+      "epoch": 0.5207253886010362,
+      "grad_norm": 38.18164449834795,
+      "learning_rate": 3.54670135541946e-05,
+      "loss": 1.1142,
+      "step": 201
+    },
+    {
+      "epoch": 0.5233160621761658,
+      "grad_norm": 48.576743289054534,
+      "learning_rate": 3.541245318200318e-05,
+      "loss": 1.1152,
+      "step": 202
+    },
+    {
+      "epoch": 0.5259067357512953,
+      "grad_norm": 38.65411737007163,
+      "learning_rate": 3.5357608920732e-05,
+      "loss": 1.1607,
+      "step": 203
+    },
+    {
+      "epoch": 0.5284974093264249,
+      "grad_norm": 35.663493907396834,
+      "learning_rate": 3.530248178058282e-05,
+      "loss": 1.1273,
+      "step": 204
+    },
+    {
+      "epoch": 0.5310880829015544,
+      "grad_norm": 26.829817821665976,
+      "learning_rate": 3.5247072776967805e-05,
+      "loss": 1.1174,
+      "step": 205
+    },
+    {
+      "epoch": 0.533678756476684,
+      "grad_norm": 39.79604912152638,
+      "learning_rate": 3.519138293049097e-05,
+      "loss": 1.1811,
+      "step": 206
+    },
+    {
+      "epoch": 0.5362694300518135,
+      "grad_norm": 32.26179097390416,
+      "learning_rate": 3.513541326692925e-05,
+      "loss": 1.1346,
+      "step": 207
+    },
+    {
+      "epoch": 0.538860103626943,
+      "grad_norm": 24.35769329902787,
+      "learning_rate": 3.5079164817213684e-05,
+      "loss": 1.1061,
+      "step": 208
+    },
+    {
+      "epoch": 0.5414507772020726,
+      "grad_norm": 26.645546258363844,
+      "learning_rate": 3.5022638617410396e-05,
+      "loss": 1.0514,
+      "step": 209
+    },
+    {
+      "epoch": 0.5440414507772021,
+      "grad_norm": 105.19676603444857,
+      "learning_rate": 3.496583570870152e-05,
+      "loss": 1.1474,
+      "step": 210
+    },
+    {
+      "epoch": 0.5466321243523317,
+      "grad_norm": 61.600623030405885,
+      "learning_rate": 3.4908757137366006e-05,
+      "loss": 1.104,
+      "step": 211
+    },
+    {
+      "epoch": 0.5492227979274611,
+      "grad_norm": 31.65460129853052,
+      "learning_rate": 3.485140395476038e-05,
+      "loss": 1.0737,
+      "step": 212
+    },
+    {
+      "epoch": 0.5518134715025906,
+      "grad_norm": 26.860379117211497,
+      "learning_rate": 3.4793777217299346e-05,
+      "loss": 1.1119,
+      "step": 213
+    },
+    {
+      "epoch": 0.5544041450777202,
+      "grad_norm": 39.89324262309783,
+      "learning_rate": 3.473587798643633e-05,
+      "loss": 1.1626,
+      "step": 214
+    },
+    {
+      "epoch": 0.5569948186528497,
+      "grad_norm": 39.77638257731599,
+      "learning_rate": 3.467770732864399e-05,
+      "loss": 1.1545,
+      "step": 215
+    },
+    {
+      "epoch": 0.5595854922279793,
+      "grad_norm": 30.994657564291458,
+      "learning_rate": 3.461926631539445e-05,
+      "loss": 1.1646,
+      "step": 216
+    },
+    {
+      "epoch": 0.5621761658031088,
+      "grad_norm": 51.99674092516571,
+      "learning_rate": 3.4560556023139695e-05,
+      "loss": 1.1638,
+      "step": 217
+    },
+    {
+      "epoch": 0.5647668393782384,
+      "grad_norm": 58.5132713002146,
+      "learning_rate": 3.450157753329166e-05,
+      "loss": 1.1461,
+      "step": 218
+    },
+    {
+      "epoch": 0.5673575129533679,
+      "grad_norm": 30.712469030418482,
+      "learning_rate": 3.4442331932202326e-05,
+      "loss": 1.1583,
+      "step": 219
+    },
+    {
+      "epoch": 0.5699481865284974,
+      "grad_norm": 47.00217426642832,
+      "learning_rate": 3.438282031114374e-05,
+      "loss": 1.1154,
+      "step": 220
+    },
+    {
+      "epoch": 0.572538860103627,
+      "grad_norm": 37.33927961163222,
+      "learning_rate": 3.432304376628787e-05,
+      "loss": 1.1372,
+      "step": 221
+    },
+    {
+      "epoch": 0.5751295336787565,
+      "grad_norm": 28.858636933974392,
+      "learning_rate": 3.4263003398686464e-05,
+      "loss": 1.0488,
+      "step": 222
+    },
+    {
+      "epoch": 0.5777202072538861,
+      "grad_norm": 37.842230890171486,
+      "learning_rate": 3.420270031425072e-05,
+      "loss": 1.1892,
+      "step": 223
+    },
+    {
+      "epoch": 0.5803108808290155,
+      "grad_norm": 32.65394945357516,
+      "learning_rate": 3.4142135623730954e-05,
+      "loss": 1.1218,
+      "step": 224
+    },
+    {
+      "epoch": 0.582901554404145,
+      "grad_norm": 115.22040829465772,
+      "learning_rate": 3.4081310442696114e-05,
+      "loss": 1.1546,
+      "step": 225
+    },
+    {
+      "epoch": 0.5854922279792746,
+      "grad_norm": 31.20514468446119,
+      "learning_rate": 3.402022589151325e-05,
+      "loss": 1.0969,
+      "step": 226
+    },
+    {
+      "epoch": 0.5880829015544041,
+      "grad_norm": 52.8397361926395,
+      "learning_rate": 3.395888309532687e-05,
+      "loss": 1.1218,
+      "step": 227
+    },
+    {
+      "epoch": 0.5906735751295337,
+      "grad_norm": 51.7991692917308,
+      "learning_rate": 3.3897283184038215e-05,
+      "loss": 1.1395,
+      "step": 228
+    },
+    {
+      "epoch": 0.5932642487046632,
+      "grad_norm": 33.56775233970504,
+      "learning_rate": 3.3835427292284445e-05,
+      "loss": 1.1107,
+      "step": 229
+    },
+    {
+      "epoch": 0.5958549222797928,
+      "grad_norm": 46.081120788214314,
+      "learning_rate": 3.3773316559417734e-05,
+      "loss": 1.1472,
+      "step": 230
+    },
+    {
+      "epoch": 0.5984455958549223,
+      "grad_norm": 41.72558170492288,
+      "learning_rate": 3.371095212948431e-05,
+      "loss": 1.1871,
+      "step": 231
+    },
+    {
+      "epoch": 0.6010362694300518,
+      "grad_norm": 34.27957927587091,
+      "learning_rate": 3.364833515120336e-05,
+      "loss": 1.1376,
+      "step": 232
+    },
+    {
+      "epoch": 0.6036269430051814,
+      "grad_norm": 36.58452602010953,
+      "learning_rate": 3.358546677794586e-05,
+      "loss": 1.1885,
+      "step": 233
+    },
+    {
+      "epoch": 0.6062176165803109,
+      "grad_norm": 28.010809914189192,
+      "learning_rate": 3.352234816771337e-05,
+      "loss": 1.102,
+      "step": 234
+    },
+    {
+      "epoch": 0.6088082901554405,
+      "grad_norm": 24.78419558611963,
+      "learning_rate": 3.3458980483116664e-05,
+      "loss": 1.0818,
+      "step": 235
+    },
+    {
+      "epoch": 0.6113989637305699,
+      "grad_norm": 28.12830040081226,
+      "learning_rate": 3.3395364891354316e-05,
+      "loss": 1.1862,
+      "step": 236
+    },
+    {
+      "epoch": 0.6139896373056994,
+      "grad_norm": 37.94181651161551,
+      "learning_rate": 3.333150256419127e-05,
+      "loss": 1.147,
+      "step": 237
+    },
+    {
+      "epoch": 0.616580310880829,
+      "grad_norm": 21.809518482701854,
+      "learning_rate": 3.3267394677937134e-05,
+      "loss": 1.0994,
+      "step": 238
+    },
+    {
+      "epoch": 0.6191709844559585,
+      "grad_norm": 32.12135773753589,
+      "learning_rate": 3.320304241342464e-05,
+      "loss": 1.1531,
+      "step": 239
+    },
+    {
+      "epoch": 0.6217616580310881,
+      "grad_norm": 51.959731073524054,
+      "learning_rate": 3.31384469559878e-05,
+      "loss": 1.1717,
+      "step": 240
+    },
+    {
+      "epoch": 0.6243523316062176,
+      "grad_norm": 28.045815836372345,
+      "learning_rate": 3.307360949544012e-05,
+      "loss": 1.1814,
+      "step": 241
+    },
+    {
+      "epoch": 0.6269430051813472,
+      "grad_norm": 39.55208384578746,
+      "learning_rate": 3.300853122605268e-05,
+      "loss": 1.1483,
+      "step": 242
+    },
+    {
+      "epoch": 0.6295336787564767,
+      "grad_norm": 29.799974205160808,
+      "learning_rate": 3.294321334653213e-05,
+      "loss": 1.1838,
+      "step": 243
+    },
+    {
+      "epoch": 0.6321243523316062,
+      "grad_norm": 124.31035254102245,
+      "learning_rate": 3.2877657059998584e-05,
+      "loss": 1.0698,
+      "step": 244
+    },
+    {
+      "epoch": 0.6347150259067358,
+      "grad_norm": 37.989925180187655,
+      "learning_rate": 3.281186357396351e-05,
+      "loss": 1.0984,
+      "step": 245
+    },
+    {
+      "epoch": 0.6373056994818653,
+      "grad_norm": 55.72599333657572,
+      "learning_rate": 3.274583410030745e-05,
+      "loss": 1.2333,
+      "step": 246
+    },
+    {
+      "epoch": 0.6398963730569949,
+      "grad_norm": 46.77079456439719,
+      "learning_rate": 3.267956985525774e-05,
+      "loss": 1.2157,
+      "step": 247
+    },
+    {
+      "epoch": 0.6424870466321243,
+      "grad_norm": 33.62329915252562,
+      "learning_rate": 3.261307205936603e-05,
+      "loss": 1.1752,
+      "step": 248
+    },
+    {
+      "epoch": 0.6450777202072538,
+      "grad_norm": 34.11794183225494,
+      "learning_rate": 3.2546341937485884e-05,
+      "loss": 1.1265,
+      "step": 249
+    },
+    {
+      "epoch": 0.6476683937823834,
+      "grad_norm": 36.027636323913896,
+      "learning_rate": 3.247938071875017e-05,
+      "loss": 1.103,
+      "step": 250
+    },
+    {
+      "epoch": 0.6502590673575129,
+      "grad_norm": 35.393219337329946,
+      "learning_rate": 3.2412189636548456e-05,
+      "loss": 1.1148,
+      "step": 251
+    },
+    {
+      "epoch": 0.6528497409326425,
+      "grad_norm": 31.578919022569924,
+      "learning_rate": 3.234476992850425e-05,
+      "loss": 1.1149,
+      "step": 252
+    },
+    {
+      "epoch": 0.655440414507772,
+      "grad_norm": 28.93717647736964,
+      "learning_rate": 3.227712283645224e-05,
+      "loss": 1.1425,
+      "step": 253
+    },
+    {
+      "epoch": 0.6580310880829016,
+      "grad_norm": 34.170026750703684,
+      "learning_rate": 3.2209249606415394e-05,
+      "loss": 1.1591,
+      "step": 254
+    },
+    {
+      "epoch": 0.6606217616580311,
+      "grad_norm": 27.52194954061608,
+      "learning_rate": 3.214115148858201e-05,
+      "loss": 1.1704,
+      "step": 255
+    },
+    {
+      "epoch": 0.6632124352331606,
+      "grad_norm": 81.65404753769732,
+      "learning_rate": 3.207282973728273e-05,
+      "loss": 1.161,
+      "step": 256
+    },
+    {
+      "epoch": 0.6658031088082902,
+      "grad_norm": 57.45351536522683,
+      "learning_rate": 3.200428561096737e-05,
+      "loss": 1.116,
+      "step": 257
+    },
+    {
+      "epoch": 0.6683937823834197,
+      "grad_norm": 30.968529074463714,
+      "learning_rate": 3.193552037218179e-05,
+      "loss": 1.1265,
+      "step": 258
+    },
+    {
+      "epoch": 0.6709844559585493,
+      "grad_norm": 37.8817748068655,
+      "learning_rate": 3.186653528754464e-05,
+      "loss": 1.1287,
+      "step": 259
+    },
+    {
+      "epoch": 0.6735751295336787,
+      "grad_norm": 29.197031189172545,
+      "learning_rate": 3.179733162772398e-05,
+      "loss": 1.1045,
+      "step": 260
+    },
+    {
+      "epoch": 0.6761658031088082,
+      "grad_norm": 36.56253841299107,
+      "learning_rate": 3.172791066741392e-05,
+      "loss": 1.1539,
+      "step": 261
+    },
+    {
+      "epoch": 0.6787564766839378,
+      "grad_norm": 25.799921116950998,
+      "learning_rate": 3.165827368531113e-05,
+      "loss": 1.0796,
+      "step": 262
+    },
+    {
+      "epoch": 0.6813471502590673,
+      "grad_norm": 82.81825216532526,
+      "learning_rate": 3.1588421964091276e-05,
+      "loss": 1.142,
+      "step": 263
+    },
+    {
+      "epoch": 0.6839378238341969,
+      "grad_norm": 31.100074747569124,
+      "learning_rate": 3.151835679038542e-05,
+      "loss": 1.0908,
+      "step": 264
+    },
+    {
+      "epoch": 0.6865284974093264,
+      "grad_norm": 25.57297200703221,
+      "learning_rate": 3.14480794547563e-05,
+      "loss": 1.1436,
+      "step": 265
+    },
+    {
+      "epoch": 0.689119170984456,
+      "grad_norm": 23.92492773149328,
+      "learning_rate": 3.137759125167455e-05,
+      "loss": 1.1202,
+      "step": 266
+    },
+    {
+      "epoch": 0.6917098445595855,
+      "grad_norm": 22.14274360766396,
+      "learning_rate": 3.130689347949486e-05,
+      "loss": 1.1113,
+      "step": 267
+    },
+    {
+      "epoch": 0.694300518134715,
+      "grad_norm": 26.68725288649902,
+      "learning_rate": 3.123598744043211e-05,
+      "loss": 1.1517,
+      "step": 268
+    },
+    {
+      "epoch": 0.6968911917098446,
+      "grad_norm": 25.559817524659362,
+      "learning_rate": 3.1164874440537295e-05,
+      "loss": 1.0976,
+      "step": 269
+    },
+    {
+      "epoch": 0.6994818652849741,
+      "grad_norm": 28.89996834100355,
+      "learning_rate": 3.109355578967356e-05,
+      "loss": 1.1932,
+      "step": 270
+    },
+    {
+      "epoch": 0.7020725388601037,
+      "grad_norm": 32.09658045195569,
+      "learning_rate": 3.1022032801492e-05,
+      "loss": 1.1161,
+      "step": 271
+    },
+    {
+      "epoch": 0.7046632124352331,
+      "grad_norm": 30.623705646213768,
+      "learning_rate": 3.095030679340751e-05,
+      "loss": 1.1993,
+      "step": 272
+    },
+    {
+      "epoch": 0.7072538860103627,
+      "grad_norm": 41.71263710932429,
+      "learning_rate": 3.0878379086574494e-05,
+      "loss": 1.1624,
+      "step": 273
+    },
+    {
+      "epoch": 0.7098445595854922,
+      "grad_norm": 34.68352639470226,
+      "learning_rate": 3.0806251005862535e-05,
+      "loss": 1.1156,
+      "step": 274
+    },
+    {
+      "epoch": 0.7124352331606217,
+      "grad_norm": 23.52580702428812,
+      "learning_rate": 3.073392387983202e-05,
+      "loss": 1.0963,
+      "step": 275
+    },
+    {
+      "epoch": 0.7150259067357513,
+      "grad_norm": 28.10687988214902,
+      "learning_rate": 3.0661399040709584e-05,
+      "loss": 1.1095,
+      "step": 276
+    },
+    {
+      "epoch": 0.7176165803108808,
+      "grad_norm": 66.72288729975841,
+      "learning_rate": 3.05886778243637e-05,
+      "loss": 1.0865,
+      "step": 277
+    },
+    {
+      "epoch": 0.7202072538860104,
+      "grad_norm": 25.775217430321934,
+      "learning_rate": 3.051576157027998e-05,
+      "loss": 1.1058,
+      "step": 278
+    },
+    {
+      "epoch": 0.7227979274611399,
+      "grad_norm": 36.82942099016794,
+      "learning_rate": 3.0442651621536502e-05,
+      "loss": 1.1211,
+      "step": 279
+    },
+    {
+      "epoch": 0.7253886010362695,
+      "grad_norm": 27.878820856521013,
+      "learning_rate": 3.0369349324779115e-05,
+      "loss": 1.1471,
+      "step": 280
+    },
+    {
+      "epoch": 0.727979274611399,
+      "grad_norm": 31.293156717285573,
+      "learning_rate": 3.0295856030196618e-05,
+      "loss": 1.0748,
+      "step": 281
+    },
+    {
+      "epoch": 0.7305699481865285,
+      "grad_norm": 39.315952115194435,
+      "learning_rate": 3.022217309149588e-05,
+      "loss": 1.0993,
+      "step": 282
+    },
+    {
+      "epoch": 0.7331606217616581,
+      "grad_norm": 36.79954071435495,
+      "learning_rate": 3.0148301865876913e-05,
+      "loss": 1.1045,
+      "step": 283
+    },
+    {
+      "epoch": 0.7357512953367875,
+      "grad_norm": 26.127389502147167,
+      "learning_rate": 3.0074243714007875e-05,
+      "loss": 1.1424,
+      "step": 284
+    },
+    {
+      "epoch": 0.7383419689119171,
+      "grad_norm": 25.608778060317068,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 1.1055,
+      "step": 285
+    },
+    {
+      "epoch": 0.7409326424870466,
+      "grad_norm": 36.22629669671894,
+      "learning_rate": 2.992557209138249e-05,
+      "loss": 1.0845,
+      "step": 286
+    },
+    {
+      "epoch": 0.7435233160621761,
+      "grad_norm": 35.30642111132886,
+      "learning_rate": 2.9850961359077293e-05,
+      "loss": 1.204,
+      "step": 287
+    },
+    {
+      "epoch": 0.7461139896373057,
+      "grad_norm": 29.765894622087952,
+      "learning_rate": 2.977616917737388e-05,
+      "loss": 1.168,
+      "step": 288
+    },
+    {
+      "epoch": 0.7487046632124352,
+      "grad_norm": 27.194683587397567,
+      "learning_rate": 2.9701196923903927e-05,
+      "loss": 1.1236,
+      "step": 289
+    },
+    {
+      "epoch": 0.7512953367875648,
+      "grad_norm": 63.09779240191165,
+      "learning_rate": 2.9626045979615928e-05,
+      "loss": 1.1395,
+      "step": 290
+    },
+    {
+      "epoch": 0.7538860103626943,
+      "grad_norm": 25.014233377763066,
+      "learning_rate": 2.9550717728749768e-05,
+      "loss": 1.1054,
+      "step": 291
+    },
+    {
+      "epoch": 0.7538860103626943,
+      "eval_loss": 1.0996382236480713,
+      "eval_runtime": 37.9545,
+      "eval_samples_per_second": 19.602,
+      "eval_steps_per_second": 1.238,
+      "step": 291
+    },
+    {
+      "epoch": 0.7564766839378239,
+      "grad_norm": 27.481891737318097,
+      "learning_rate": 2.947521355881122e-05,
+      "loss": 1.1252,
+      "step": 292
+    },
+    {
+      "epoch": 0.7590673575129534,
+      "grad_norm": 67.57807413949878,
+      "learning_rate": 2.9399534860546404e-05,
+      "loss": 1.1761,
+      "step": 293
+    },
+    {
+      "epoch": 0.7616580310880829,
+      "grad_norm": 65.66834495909988,
+      "learning_rate": 2.932368302791614e-05,
+      "loss": 1.0551,
+      "step": 294
+    },
+    {
+      "epoch": 0.7642487046632125,
+      "grad_norm": 30.051210942517116,
+      "learning_rate": 2.92476594580703e-05,
+      "loss": 1.138,
+      "step": 295
+    },
+    {
+      "epoch": 0.7668393782383419,
+      "grad_norm": 22.693089678510507,
+      "learning_rate": 2.917146555132206e-05,
+      "loss": 1.1495,
+      "step": 296
+    },
+    {
+      "epoch": 0.7694300518134715,
+      "grad_norm": 53.84166280540606,
+      "learning_rate": 2.909510271112212e-05,
+      "loss": 1.1409,
+      "step": 297
+    },
+    {
+      "epoch": 0.772020725388601,
+      "grad_norm": 32.69106061524578,
+      "learning_rate": 2.9018572344032823e-05,
+      "loss": 1.1709,
+      "step": 298
+    },
+    {
+      "epoch": 0.7746113989637305,
+      "grad_norm": 39.44484991312582,
+      "learning_rate": 2.8941875859702283e-05,
+      "loss": 1.1138,
+      "step": 299
+    },
+    {
+      "epoch": 0.7772020725388601,
+      "grad_norm": 31.51857596969122,
+      "learning_rate": 2.88650146708384e-05,
+      "loss": 1.1931,
+      "step": 300
+    },
+    {
+      "epoch": 0.7797927461139896,
+      "grad_norm": 70.51218412614058,
+      "learning_rate": 2.878799019318283e-05,
+      "loss": 1.155,
+      "step": 301
+    },
+    {
+      "epoch": 0.7823834196891192,
+      "grad_norm": 80.27969224752457,
+      "learning_rate": 2.8710803845484955e-05,
+      "loss": 1.1425,
+      "step": 302
+    },
+    {
+      "epoch": 0.7849740932642487,
+      "grad_norm": 28.16560857981767,
+      "learning_rate": 2.8633457049475678e-05,
+      "loss": 1.1072,
+      "step": 303
+    },
+    {
+      "epoch": 0.7875647668393783,
+      "grad_norm": 41.15138307552231,
+      "learning_rate": 2.855595122984129e-05,
+      "loss": 1.1492,
+      "step": 304
+    },
+    {
+      "epoch": 0.7901554404145078,
+      "grad_norm": 23.894217282116276,
+      "learning_rate": 2.847828781419722e-05,
+      "loss": 1.1136,
+      "step": 305
+    },
+    {
+      "epoch": 0.7927461139896373,
+      "grad_norm": 25.005501120810248,
+      "learning_rate": 2.8400468233061708e-05,
+      "loss": 1.0921,
+      "step": 306
+    },
+    {
+      "epoch": 0.7953367875647669,
+      "grad_norm": 30.91791938195468,
+      "learning_rate": 2.832249391982949e-05,
+      "loss": 1.1098,
+      "step": 307
+    },
+    {
+      "epoch": 0.7979274611398963,
+      "grad_norm": 44.776563922922726,
+      "learning_rate": 2.8244366310745398e-05,
+      "loss": 1.1845,
+      "step": 308
+    },
+    {
+      "epoch": 0.8005181347150259,
+      "grad_norm": 19.059329544784376,
+      "learning_rate": 2.816608684487787e-05,
+      "loss": 1.169,
+      "step": 309
+    },
+    {
+      "epoch": 0.8031088082901554,
+      "grad_norm": 63.97334641962602,
+      "learning_rate": 2.8087656964092472e-05,
+      "loss": 1.124,
+      "step": 310
+    },
+    {
+      "epoch": 0.805699481865285,
+      "grad_norm": 30.878848859015882,
+      "learning_rate": 2.8009078113025335e-05,
+      "loss": 1.2087,
+      "step": 311
+    },
+    {
+      "epoch": 0.8082901554404145,
+      "grad_norm": 34.63835471543836,
+      "learning_rate": 2.7930351739056533e-05,
+      "loss": 1.1338,
+      "step": 312
+    },
+    {
+      "epoch": 0.810880829015544,
+      "grad_norm": 30.03178182445718,
+      "learning_rate": 2.7851479292283442e-05,
+      "loss": 1.1321,
+      "step": 313
+    },
+    {
+      "epoch": 0.8134715025906736,
+      "grad_norm": 38.42236523356876,
+      "learning_rate": 2.7772462225494013e-05,
+      "loss": 1.1557,
+      "step": 314
+    },
+    {
+      "epoch": 0.8160621761658031,
+      "grad_norm": 39.179683790956744,
+      "learning_rate": 2.7693301994140026e-05,
+      "loss": 1.1201,
+      "step": 315
+    },
+    {
+      "epoch": 0.8186528497409327,
+      "grad_norm": 38.32243159447327,
+      "learning_rate": 2.761400005631028e-05,
+      "loss": 1.1105,
+      "step": 316
+    },
+    {
+      "epoch": 0.8212435233160622,
+      "grad_norm": 39.913808227411835,
+      "learning_rate": 2.7534557872703705e-05,
+      "loss": 1.1598,
+      "step": 317
+    },
+    {
+      "epoch": 0.8238341968911918,
+      "grad_norm": 69.73521867812421,
+      "learning_rate": 2.7454976906602513e-05,
+      "loss": 1.1145,
+      "step": 318
+    },
+    {
+      "epoch": 0.8264248704663213,
+      "grad_norm": 65.55887588207746,
+      "learning_rate": 2.7375258623845207e-05,
+      "loss": 1.1255,
+      "step": 319
+    },
+    {
+      "epoch": 0.8290155440414507,
+      "grad_norm": 30.980111545641563,
+      "learning_rate": 2.7295404492799575e-05,
+      "loss": 1.122,
+      "step": 320
+    },
+    {
+      "epoch": 0.8316062176165803,
+      "grad_norm": 30.12179911444832,
+      "learning_rate": 2.721541598433567e-05,
+      "loss": 1.113,
+      "step": 321
+    },
+    {
+      "epoch": 0.8341968911917098,
+      "grad_norm": 28.329434659508582,
+      "learning_rate": 2.7135294571798706e-05,
+      "loss": 1.0498,
+      "step": 322
+    },
+    {
+      "epoch": 0.8367875647668394,
+      "grad_norm": 25.114787597049578,
+      "learning_rate": 2.70550417309819e-05,
+      "loss": 1.0633,
+      "step": 323
+    },
+    {
+      "epoch": 0.8393782383419689,
+      "grad_norm": 27.754037709590385,
+      "learning_rate": 2.6974658940099337e-05,
+      "loss": 1.1585,
+      "step": 324
+    },
+    {
+      "epoch": 0.8419689119170984,
+      "grad_norm": 29.489888159179444,
+      "learning_rate": 2.6894147679758678e-05,
+      "loss": 1.1259,
+      "step": 325
+    },
+    {
+      "epoch": 0.844559585492228,
+      "grad_norm": 24.426102194202898,
+      "learning_rate": 2.6813509432933957e-05,
+      "loss": 1.1515,
+      "step": 326
+    },
+    {
+      "epoch": 0.8471502590673575,
+      "grad_norm": 24.75197483331429,
+      "learning_rate": 2.673274568493821e-05,
+      "loss": 1.15,
+      "step": 327
+    },
+    {
+      "epoch": 0.8497409326424871,
+      "grad_norm": 40.604864626683366,
+      "learning_rate": 2.6651857923396132e-05,
+      "loss": 1.1219,
+      "step": 328
+    },
+    {
+      "epoch": 0.8523316062176166,
+      "grad_norm": 34.694568404196026,
+      "learning_rate": 2.6570847638216698e-05,
+      "loss": 1.103,
+      "step": 329
+    },
+    {
+      "epoch": 0.8549222797927462,
+      "grad_norm": 48.715136403425035,
+      "learning_rate": 2.648971632156569e-05,
+      "loss": 1.1675,
+      "step": 330
+    },
+    {
+      "epoch": 0.8575129533678757,
+      "grad_norm": 97.77526410121799,
+      "learning_rate": 2.6408465467838225e-05,
+      "loss": 1.1502,
+      "step": 331
+    },
+    {
+      "epoch": 0.8601036269430051,
+      "grad_norm": 54.697215318949276,
+      "learning_rate": 2.632709657363124e-05,
+      "loss": 1.1446,
+      "step": 332
+    },
+    {
+      "epoch": 0.8626943005181347,
+      "grad_norm": 38.09192002041798,
+      "learning_rate": 2.6245611137715897e-05,
+      "loss": 1.1333,
+      "step": 333
+    },
+    {
+      "epoch": 0.8652849740932642,
+      "grad_norm": 46.713623556984956,
+      "learning_rate": 2.6164010661010007e-05,
+      "loss": 1.1252,
+      "step": 334
+    },
+    {
+      "epoch": 0.8678756476683938,
+      "grad_norm": 46.40552686286593,
+      "learning_rate": 2.6082296646550364e-05,
+      "loss": 1.121,
+      "step": 335
+    },
+    {
+      "epoch": 0.8704663212435233,
+      "grad_norm": 37.57424454065957,
+      "learning_rate": 2.6000470599465065e-05,
+      "loss": 1.1671,
+      "step": 336
+    },
+    {
+      "epoch": 0.8730569948186528,
+      "grad_norm": 38.580777053099204,
+      "learning_rate": 2.5918534026945787e-05,
+      "loss": 1.0849,
+      "step": 337
+    },
+    {
+      "epoch": 0.8756476683937824,
+      "grad_norm": 154.3106712010981,
+      "learning_rate": 2.5836488438220044e-05,
+      "loss": 1.0663,
+      "step": 338
+    },
+    {
+      "epoch": 0.8782383419689119,
+      "grad_norm": 34.21394067951015,
+      "learning_rate": 2.575433534452334e-05,
+      "loss": 1.0895,
+      "step": 339
+    },
+    {
+      "epoch": 0.8808290155440415,
+      "grad_norm": 36.291611242733886,
+      "learning_rate": 2.5672076259071385e-05,
+      "loss": 1.1242,
+      "step": 340
+    },
+    {
+      "epoch": 0.883419689119171,
+      "grad_norm": 29.411623389655112,
+      "learning_rate": 2.558971269703219e-05,
+      "loss": 1.1005,
+      "step": 341
+    },
+    {
+      "epoch": 0.8860103626943006,
+      "grad_norm": 30.24903086761753,
+      "learning_rate": 2.5507246175498174e-05,
+      "loss": 1.1134,
+      "step": 342
+    },
+    {
+      "epoch": 0.8886010362694301,
+      "grad_norm": 22.032293114161938,
+      "learning_rate": 2.5424678213458202e-05,
+      "loss": 1.1121,
+      "step": 343
+    },
+    {
+      "epoch": 0.8911917098445595,
+      "grad_norm": 34.997361528376956,
+      "learning_rate": 2.5342010331769635e-05,
+      "loss": 1.1341,
+      "step": 344
+    },
+    {
+      "epoch": 0.8937823834196891,
+      "grad_norm": 28.212824875732352,
+      "learning_rate": 2.5259244053130295e-05,
+      "loss": 1.0748,
+      "step": 345
+    },
+    {
+      "epoch": 0.8963730569948186,
+      "grad_norm": 23.870011592985897,
+      "learning_rate": 2.5176380902050418e-05,
+      "loss": 1.0643,
+      "step": 346
+    },
+    {
+      "epoch": 0.8989637305699482,
+      "grad_norm": 26.10018699309748,
+      "learning_rate": 2.5093422404824574e-05,
+      "loss": 1.1662,
+      "step": 347
+    },
+    {
+      "epoch": 0.9015544041450777,
+      "grad_norm": 30.191468778559166,
+      "learning_rate": 2.5010370089503578e-05,
+      "loss": 1.1023,
+      "step": 348
+    },
+    {
+      "epoch": 0.9041450777202072,
+      "grad_norm": 55.799581973427415,
+      "learning_rate": 2.4927225485866297e-05,
+      "loss": 1.1538,
+      "step": 349
+    },
+    {
+      "epoch": 0.9067357512953368,
+      "grad_norm": 35.7030284720465,
+      "learning_rate": 2.4843990125391516e-05,
+      "loss": 1.1,
+      "step": 350
+    },
+    {
+      "epoch": 0.9093264248704663,
+      "grad_norm": 28.61763302791738,
+      "learning_rate": 2.4760665541229712e-05,
+      "loss": 1.0914,
+      "step": 351
+    },
+    {
+      "epoch": 0.9119170984455959,
+      "grad_norm": 33.34233685155311,
+      "learning_rate": 2.467725326817481e-05,
+      "loss": 1.0862,
+      "step": 352
+    },
+    {
+      "epoch": 0.9145077720207254,
+      "grad_norm": 25.441052078480084,
+      "learning_rate": 2.4593754842635917e-05,
+      "loss": 1.1422,
+      "step": 353
+    },
+    {
+      "epoch": 0.917098445595855,
+      "grad_norm": 24.217974454985058,
+      "learning_rate": 2.451017180260902e-05,
+      "loss": 1.132,
+      "step": 354
+    },
+    {
+      "epoch": 0.9196891191709845,
+      "grad_norm": 57.986011465793155,
+      "learning_rate": 2.4426505687648653e-05,
+      "loss": 1.2082,
+      "step": 355
+    },
+    {
+      "epoch": 0.9222797927461139,
+      "grad_norm": 34.058264716876195,
+      "learning_rate": 2.4342758038839573e-05,
+      "loss": 1.1679,
+      "step": 356
+    },
+    {
+      "epoch": 0.9248704663212435,
+      "grad_norm": 28.621514922275253,
+      "learning_rate": 2.4258930398768317e-05,
+      "loss": 1.1319,
+      "step": 357
+    },
+    {
+      "epoch": 0.927461139896373,
+      "grad_norm": 35.33355417283227,
+      "learning_rate": 2.4175024311494835e-05,
+      "loss": 1.0705,
+      "step": 358
+    },
+    {
+      "epoch": 0.9300518134715026,
+      "grad_norm": 46.579572933583265,
+      "learning_rate": 2.4091041322524023e-05,
+      "loss": 1.0842,
+      "step": 359
+    },
+    {
+      "epoch": 0.9326424870466321,
+      "grad_norm": 35.494740787672974,
+      "learning_rate": 2.4006982978777263e-05,
+      "loss": 1.1072,
+      "step": 360
+    },
+    {
+      "epoch": 0.9352331606217616,
+      "grad_norm": 44.56606839509262,
+      "learning_rate": 2.392285082856394e-05,
+      "loss": 1.1125,
+      "step": 361
+    },
+    {
+      "epoch": 0.9378238341968912,
+      "grad_norm": 46.26363869084929,
+      "learning_rate": 2.3838646421552917e-05,
+      "loss": 1.1268,
+      "step": 362
+    },
+    {
+      "epoch": 0.9404145077720207,
+      "grad_norm": 89.17676267680146,
+      "learning_rate": 2.3754371308743975e-05,
+      "loss": 1.0893,
+      "step": 363
+    },
+    {
+      "epoch": 0.9430051813471503,
+      "grad_norm": 34.87700187494181,
+      "learning_rate": 2.367002704243927e-05,
+      "loss": 1.1203,
+      "step": 364
+    },
+    {
+      "epoch": 0.9455958549222798,
+      "grad_norm": 32.92806939217504,
+      "learning_rate": 2.3585615176214716e-05,
+      "loss": 1.1488,
+      "step": 365
+    },
+    {
+      "epoch": 0.9481865284974094,
+      "grad_norm": 27.27458755248548,
+      "learning_rate": 2.3501137264891396e-05,
+      "loss": 1.0874,
+      "step": 366
+    },
+    {
+      "epoch": 0.9507772020725389,
+      "grad_norm": 24.959123789739834,
+      "learning_rate": 2.3416594864506887e-05,
+      "loss": 1.1783,
+      "step": 367
+    },
+    {
+      "epoch": 0.9533678756476683,
+      "grad_norm": 31.838670988369724,
+      "learning_rate": 2.333198953228664e-05,
+      "loss": 1.0759,
+      "step": 368
+    },
+    {
+      "epoch": 0.9559585492227979,
+      "grad_norm": 28.112870222863155,
+      "learning_rate": 2.3247322826615276e-05,
+      "loss": 1.1481,
+      "step": 369
+    },
+    {
+      "epoch": 0.9585492227979274,
+      "grad_norm": 35.08461098450067,
+      "learning_rate": 2.316259630700787e-05,
+      "loss": 1.0953,
+      "step": 370
+    },
+    {
+      "epoch": 0.961139896373057,
+      "grad_norm": 37.80899503618479,
+      "learning_rate": 2.307781153408124e-05,
+      "loss": 1.1224,
+      "step": 371
+    },
+    {
+      "epoch": 0.9637305699481865,
+      "grad_norm": 31.644978122007387,
+      "learning_rate": 2.2992970069525202e-05,
+      "loss": 1.1608,
+      "step": 372
+    },
+    {
+      "epoch": 0.966321243523316,
+      "grad_norm": 23.51029318210938,
+      "learning_rate": 2.29080734760738e-05,
+      "loss": 1.0914,
+      "step": 373
+    },
+    {
+      "epoch": 0.9689119170984456,
+      "grad_norm": 28.97240481418573,
+      "learning_rate": 2.2823123317476522e-05,
+      "loss": 1.1117,
+      "step": 374
+    },
+    {
+      "epoch": 0.9715025906735751,
+      "grad_norm": 36.613893678320395,
+      "learning_rate": 2.273812115846951e-05,
+      "loss": 1.1118,
+      "step": 375
+    },
+    {
+      "epoch": 0.9740932642487047,
+      "grad_norm": 26.402979304578093,
+      "learning_rate": 2.2653068564746692e-05,
+      "loss": 1.13,
+      "step": 376
+    },
+    {
+      "epoch": 0.9766839378238342,
+      "grad_norm": 114.3000444613392,
+      "learning_rate": 2.2567967102931025e-05,
+      "loss": 1.1539,
+      "step": 377
+    },
+    {
+      "epoch": 0.9792746113989638,
+      "grad_norm": 26.861359932396834,
+      "learning_rate": 2.2482818340545534e-05,
+      "loss": 1.0566,
+      "step": 378
+    },
+    {
+      "epoch": 0.9818652849740933,
+      "grad_norm": 32.75509374223994,
+      "learning_rate": 2.2397623845984548e-05,
+      "loss": 1.1746,
+      "step": 379
+    },
+    {
+      "epoch": 0.9844559585492227,
+      "grad_norm": 34.11964206838379,
+      "learning_rate": 2.2312385188484718e-05,
+      "loss": 1.0834,
+      "step": 380
+    },
+    {
+      "epoch": 0.9870466321243523,
+      "grad_norm": 38.019564122226434,
+      "learning_rate": 2.2227103938096176e-05,
+      "loss": 1.1074,
+      "step": 381
+    },
+    {
+      "epoch": 0.9896373056994818,
+      "grad_norm": 39.5073811375391,
+      "learning_rate": 2.2141781665653584e-05,
+      "loss": 1.1082,
+      "step": 382
+    },
+    {
+      "epoch": 0.9922279792746114,
+      "grad_norm": 298.4258332795163,
+      "learning_rate": 2.205641994274721e-05,
+      "loss": 1.125,
+      "step": 383
+    },
+    {
+      "epoch": 0.9948186528497409,
+      "grad_norm": 36.444415670935506,
+      "learning_rate": 2.1971020341693973e-05,
+      "loss": 1.0935,
+      "step": 384
+    },
+    {
+      "epoch": 0.9974093264248705,
+      "grad_norm": 28.96533429210575,
+      "learning_rate": 2.188558443550849e-05,
+      "loss": 1.0957,
+      "step": 385
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 66.41241684127401,
+      "learning_rate": 2.180011379787411e-05,
+      "loss": 1.1335,
+      "step": 386
+    },
+    {
+      "epoch": 1.0025906735751295,
+      "grad_norm": 28.75549619538953,
+      "learning_rate": 2.1714610003113887e-05,
+      "loss": 1.1316,
+      "step": 387
+    },
+    {
+      "epoch": 1.005181347150259,
+      "grad_norm": 26.911837500852275,
+      "learning_rate": 2.1629074626161647e-05,
+      "loss": 1.1026,
+      "step": 388
+    },
+    {
+      "epoch": 1.005181347150259,
+      "eval_loss": 1.0908173322677612,
+      "eval_runtime": 37.7642,
+      "eval_samples_per_second": 19.701,
+      "eval_steps_per_second": 1.245,
+      "step": 388
+    },
+    {
+      "epoch": 1.0077720207253886,
+      "grad_norm": 34.28722746775385,
+      "learning_rate": 2.1543509242532932e-05,
+      "loss": 1.1104,
+      "step": 389
+    },
+    {
+      "epoch": 1.0103626943005182,
+      "grad_norm": 37.97709310694863,
+      "learning_rate": 2.145791542829597e-05,
+      "loss": 1.0663,
+      "step": 390
+    },
+    {
+      "epoch": 1.0129533678756477,
+      "grad_norm": 39.379668162327384,
+      "learning_rate": 2.1372294760042686e-05,
+      "loss": 1.1405,
+      "step": 391
+    },
+    {
+      "epoch": 1.0155440414507773,
+      "grad_norm": 27.136201219298698,
+      "learning_rate": 2.1286648814859636e-05,
+      "loss": 1.0963,
+      "step": 392
+    },
+    {
+      "epoch": 1.0181347150259068,
+      "grad_norm": 39.34261641469313,
+      "learning_rate": 2.120097917029897e-05,
+      "loss": 1.1276,
+      "step": 393
+    },
+    {
+      "epoch": 1.0207253886010363,
+      "grad_norm": 46.77583801285328,
+      "learning_rate": 2.1115287404349357e-05,
+      "loss": 1.1171,
+      "step": 394
+    },
+    {
+      "epoch": 1.0233160621761659,
+      "grad_norm": 55.10335066695868,
+      "learning_rate": 2.1029575095406933e-05,
+      "loss": 1.0831,
+      "step": 395
+    },
+    {
+      "epoch": 1.0259067357512954,
+      "grad_norm": 76.88533851789373,
+      "learning_rate": 2.0943843822246234e-05,
+      "loss": 1.0925,
+      "step": 396
+    },
+    {
+      "epoch": 1.028497409326425,
+      "grad_norm": 29.604569209708462,
+      "learning_rate": 2.0858095163991094e-05,
+      "loss": 1.1259,
+      "step": 397
+    },
+    {
+      "epoch": 1.0310880829015545,
+      "grad_norm": 37.71348366628868,
+      "learning_rate": 2.077233070008557e-05,
+      "loss": 1.0792,
+      "step": 398
+    },
+    {
+      "epoch": 1.0336787564766838,
+      "grad_norm": 26.866133194031644,
+      "learning_rate": 2.0686552010264872e-05,
+      "loss": 1.1649,
+      "step": 399
+    },
+    {
+      "epoch": 1.0362694300518134,
+      "grad_norm": 35.739274800620635,
+      "learning_rate": 2.060076067452622e-05,
+      "loss": 1.0837,
+      "step": 400
+    },
+    {
+      "epoch": 1.038860103626943,
+      "grad_norm": 24.479129391259896,
+      "learning_rate": 2.0514958273099778e-05,
+      "loss": 1.073,
+      "step": 401
+    },
+    {
+      "epoch": 1.0414507772020725,
+      "grad_norm": 50.49963650108008,
+      "learning_rate": 2.042914638641952e-05,
+      "loss": 1.0912,
+      "step": 402
+    },
+    {
+      "epoch": 1.044041450777202,
+      "grad_norm": 35.6875451072032,
+      "learning_rate": 2.0343326595094154e-05,
+      "loss": 1.0936,
+      "step": 403
+    },
+    {
+      "epoch": 1.0466321243523315,
+      "grad_norm": 30.212298193414487,
+      "learning_rate": 2.0257500479877965e-05,
+      "loss": 1.089,
+      "step": 404
+    },
+    {
+      "epoch": 1.049222797927461,
+      "grad_norm": 28.65828720015124,
+      "learning_rate": 2.0171669621641743e-05,
+      "loss": 1.1727,
+      "step": 405
+    },
+    {
+      "epoch": 1.0518134715025906,
+      "grad_norm": 39.2199058392425,
+      "learning_rate": 2.0085835601343627e-05,
+      "loss": 1.1493,
+      "step": 406
+    },
+    {
+      "epoch": 1.0544041450777202,
+      "grad_norm": 110.01204177059546,
+      "learning_rate": 2e-05,
+      "loss": 1.1245,
+      "step": 407
+    },
+    {
+      "epoch": 1.0569948186528497,
+      "grad_norm": 43.427381349600374,
+      "learning_rate": 1.9914164398656383e-05,
+      "loss": 1.1183,
+      "step": 408
+    },
+    {
+      "epoch": 1.0595854922279793,
+      "grad_norm": 64.78768909817894,
+      "learning_rate": 1.9828330378358264e-05,
+      "loss": 1.1528,
+      "step": 409
+    },
+    {
+      "epoch": 1.0621761658031088,
+      "grad_norm": 26.50257915912425,
+      "learning_rate": 1.974249952012204e-05,
+      "loss": 1.1568,
+      "step": 410
+    },
+    {
+      "epoch": 1.0647668393782384,
+      "grad_norm": 27.63159204178893,
+      "learning_rate": 1.9656673404905852e-05,
+      "loss": 1.1071,
+      "step": 411
+    },
+    {
+      "epoch": 1.067357512953368,
+      "grad_norm": 27.0795355533723,
+      "learning_rate": 1.957085361358049e-05,
+      "loss": 1.0809,
+      "step": 412
+    },
+    {
+      "epoch": 1.0699481865284974,
+      "grad_norm": 41.84795332660821,
+      "learning_rate": 1.9485041726900232e-05,
+      "loss": 1.0744,
+      "step": 413
+    },
+    {
+      "epoch": 1.072538860103627,
+      "grad_norm": 143.2109134427192,
+      "learning_rate": 1.939923932547379e-05,
+      "loss": 1.0905,
+      "step": 414
+    },
+    {
+      "epoch": 1.0751295336787565,
+      "grad_norm": 89.55384065946154,
+      "learning_rate": 1.931344798973513e-05,
+      "loss": 1.1012,
+      "step": 415
+    },
+    {
+      "epoch": 1.077720207253886,
+      "grad_norm": 31.072074793068015,
+      "learning_rate": 1.922766929991443e-05,
+      "loss": 1.1141,
+      "step": 416
+    },
+    {
+      "epoch": 1.0803108808290156,
+      "grad_norm": 29.82683189045969,
+      "learning_rate": 1.914190483600891e-05,
+      "loss": 1.0842,
+      "step": 417
+    },
+    {
+      "epoch": 1.0829015544041452,
+      "grad_norm": 30.09708662586305,
+      "learning_rate": 1.9056156177753776e-05,
+      "loss": 1.1088,
+      "step": 418
+    },
+    {
+      "epoch": 1.0854922279792747,
+      "grad_norm": 27.637437518920503,
+      "learning_rate": 1.897042490459307e-05,
+      "loss": 1.058,
+      "step": 419
+    },
+    {
+      "epoch": 1.0880829015544042,
+      "grad_norm": 69.34285700381683,
+      "learning_rate": 1.8884712595650653e-05,
+      "loss": 1.0314,
+      "step": 420
+    },
+    {
+      "epoch": 1.0906735751295338,
+      "grad_norm": 25.644927284592956,
+      "learning_rate": 1.8799020829701036e-05,
+      "loss": 1.0916,
+      "step": 421
+    },
+    {
+      "epoch": 1.093264248704663,
+      "grad_norm": 30.3898986852319,
+      "learning_rate": 1.871335118514037e-05,
+      "loss": 1.0797,
+      "step": 422
+    },
+    {
+      "epoch": 1.0958549222797926,
+      "grad_norm": 22.271334693423444,
+      "learning_rate": 1.862770523995732e-05,
+      "loss": 1.1134,
+      "step": 423
+    },
+    {
+      "epoch": 1.0984455958549222,
+      "grad_norm": 35.85874616678876,
+      "learning_rate": 1.854208457170404e-05,
+      "loss": 1.0927,
+      "step": 424
+    },
+    {
+      "epoch": 1.1010362694300517,
+      "grad_norm": 43.06832041948097,
+      "learning_rate": 1.8456490757467075e-05,
+      "loss": 1.093,
+      "step": 425
+    },
+    {
+      "epoch": 1.1036269430051813,
+      "grad_norm": 37.83777637993467,
+      "learning_rate": 1.8370925373838356e-05,
+      "loss": 1.1268,
+      "step": 426
+    },
+    {
+      "epoch": 1.1062176165803108,
+      "grad_norm": 23.798059023605177,
+      "learning_rate": 1.8285389996886113e-05,
+      "loss": 1.0989,
+      "step": 427
+    },
+    {
+      "epoch": 1.1088082901554404,
+      "grad_norm": 25.443104465500795,
+      "learning_rate": 1.8199886202125897e-05,
+      "loss": 1.0581,
+      "step": 428
+    },
+    {
+      "epoch": 1.11139896373057,
+      "grad_norm": 23.76241444847441,
+      "learning_rate": 1.8114415564491513e-05,
+      "loss": 1.0908,
+      "step": 429
+    },
+    {
+      "epoch": 1.1139896373056994,
+      "grad_norm": 26.5600693044426,
+      "learning_rate": 1.8028979658306033e-05,
+      "loss": 1.1321,
+      "step": 430
+    },
+    {
+      "epoch": 1.116580310880829,
+      "grad_norm": 44.854375199828986,
+      "learning_rate": 1.794358005725279e-05,
+      "loss": 1.0762,
+      "step": 431
+    },
+    {
+      "epoch": 1.1191709844559585,
+      "grad_norm": 28.05797777410846,
+      "learning_rate": 1.785821833434642e-05,
+      "loss": 1.0698,
+      "step": 432
+    },
+    {
+      "epoch": 1.121761658031088,
+      "grad_norm": 26.488479630212364,
+      "learning_rate": 1.7772896061903824e-05,
+      "loss": 1.1223,
+      "step": 433
+    },
+    {
+      "epoch": 1.1243523316062176,
+      "grad_norm": 32.77084542157883,
+      "learning_rate": 1.768761481151529e-05,
+      "loss": 1.0984,
+      "step": 434
+    },
+    {
+      "epoch": 1.1269430051813472,
+      "grad_norm": 39.13198413130026,
+      "learning_rate": 1.7602376154015456e-05,
+      "loss": 1.1551,
+      "step": 435
+    },
+    {
+      "epoch": 1.1295336787564767,
+      "grad_norm": 23.878966995283953,
+      "learning_rate": 1.751718165945447e-05,
+      "loss": 1.1133,
+      "step": 436
+    },
+    {
+      "epoch": 1.1321243523316062,
+      "grad_norm": 33.90472985566232,
+      "learning_rate": 1.743203289706898e-05,
+      "loss": 1.1219,
+      "step": 437
+    },
+    {
+      "epoch": 1.1347150259067358,
+      "grad_norm": 23.340369938533712,
+      "learning_rate": 1.734693143525331e-05,
+      "loss": 1.1244,
+      "step": 438
+    },
+    {
+      "epoch": 1.1373056994818653,
+      "grad_norm": 105.6885206147852,
+      "learning_rate": 1.7261878841530494e-05,
+      "loss": 1.0788,
+      "step": 439
+    },
+    {
+      "epoch": 1.1398963730569949,
+      "grad_norm": 28.453526076458317,
+      "learning_rate": 1.717687668252348e-05,
+      "loss": 1.1576,
+      "step": 440
+    },
+    {
+      "epoch": 1.1424870466321244,
+      "grad_norm": 36.1473991485961,
+      "learning_rate": 1.7091926523926205e-05,
+      "loss": 1.0859,
+      "step": 441
+    },
+    {
+      "epoch": 1.145077720207254,
+      "grad_norm": 27.043461146902448,
+      "learning_rate": 1.7007029930474804e-05,
+      "loss": 1.1072,
+      "step": 442
+    },
+    {
+      "epoch": 1.1476683937823835,
+      "grad_norm": 28.066170619981435,
+      "learning_rate": 1.6922188465918763e-05,
+      "loss": 1.1279,
+      "step": 443
+    },
+    {
+      "epoch": 1.150259067357513,
+      "grad_norm": 38.62445822837212,
+      "learning_rate": 1.6837403692992136e-05,
+      "loss": 1.1275,
+      "step": 444
+    },
+    {
+      "epoch": 1.1528497409326426,
+      "grad_norm": 28.077258963587767,
+      "learning_rate": 1.6752677173384734e-05,
+      "loss": 1.1004,
+      "step": 445
+    },
+    {
+      "epoch": 1.1554404145077721,
+      "grad_norm": 42.1405744301338,
+      "learning_rate": 1.6668010467713363e-05,
+      "loss": 1.1141,
+      "step": 446
+    },
+    {
+      "epoch": 1.1580310880829017,
+      "grad_norm": 26.827291684301034,
+      "learning_rate": 1.658340513549312e-05,
+      "loss": 1.1216,
+      "step": 447
+    },
+    {
+      "epoch": 1.160621761658031,
+      "grad_norm": 30.863489441619983,
+      "learning_rate": 1.649886273510861e-05,
+      "loss": 1.1898,
+      "step": 448
+    },
+    {
+      "epoch": 1.1632124352331605,
+      "grad_norm": 27.73579733476068,
+      "learning_rate": 1.641438482378529e-05,
+      "loss": 1.0971,
+      "step": 449
+    },
+    {
+      "epoch": 1.16580310880829,
+      "grad_norm": 32.84347174567353,
+      "learning_rate": 1.6329972957560736e-05,
+      "loss": 1.0579,
+      "step": 450
+    },
+    {
+      "epoch": 1.1683937823834196,
+      "grad_norm": 30.06456192962641,
+      "learning_rate": 1.6245628691256032e-05,
+      "loss": 1.1057,
+      "step": 451
+    },
+    {
+      "epoch": 1.1709844559585492,
+      "grad_norm": 36.554506394377846,
+      "learning_rate": 1.616135357844709e-05,
+      "loss": 1.1008,
+      "step": 452
+    },
+    {
+      "epoch": 1.1735751295336787,
+      "grad_norm": 27.358643056184114,
+      "learning_rate": 1.6077149171436063e-05,
+      "loss": 1.101,
+      "step": 453
+    },
+    {
+      "epoch": 1.1761658031088082,
+      "grad_norm": 111.13373813893604,
+      "learning_rate": 1.599301702122274e-05,
+      "loss": 1.0688,
+      "step": 454
+    },
+    {
+      "epoch": 1.1787564766839378,
+      "grad_norm": 33.94168250727336,
+      "learning_rate": 1.590895867747599e-05,
+      "loss": 1.0721,
+      "step": 455
+    },
+    {
+      "epoch": 1.1813471502590673,
+      "grad_norm": 53.93978395349692,
+      "learning_rate": 1.582497568850517e-05,
+      "loss": 1.0584,
+      "step": 456
+    },
+    {
+      "epoch": 1.1839378238341969,
+      "grad_norm": 29.19245794937285,
+      "learning_rate": 1.574106960123169e-05,
+      "loss": 1.067,
+      "step": 457
+    },
+    {
+      "epoch": 1.1865284974093264,
+      "grad_norm": 28.06897801999048,
+      "learning_rate": 1.5657241961160434e-05,
+      "loss": 1.0899,
+      "step": 458
+    },
+    {
+      "epoch": 1.189119170984456,
+      "grad_norm": 52.31256652964293,
+      "learning_rate": 1.557349431235135e-05,
+      "loss": 1.0925,
+      "step": 459
+    },
+    {
+      "epoch": 1.1917098445595855,
+      "grad_norm": 65.39771110845307,
+      "learning_rate": 1.5489828197390988e-05,
+      "loss": 1.1448,
+      "step": 460
+    },
+    {
+      "epoch": 1.194300518134715,
+      "grad_norm": 27.062780348557254,
+      "learning_rate": 1.5406245157364093e-05,
+      "loss": 1.0871,
+      "step": 461
+    },
+    {
+      "epoch": 1.1968911917098446,
+      "grad_norm": 41.667025056250424,
+      "learning_rate": 1.5322746731825195e-05,
+      "loss": 1.048,
+      "step": 462
+    },
+    {
+      "epoch": 1.1994818652849741,
+      "grad_norm": 24.936669803360665,
+      "learning_rate": 1.5239334458770291e-05,
+      "loss": 1.1243,
+      "step": 463
+    },
+    {
+      "epoch": 1.2020725388601037,
+      "grad_norm": 26.65392149600558,
+      "learning_rate": 1.5156009874608484e-05,
+      "loss": 1.0919,
+      "step": 464
+    },
+    {
+      "epoch": 1.2046632124352332,
+      "grad_norm": 48.57730651937978,
+      "learning_rate": 1.5072774514133708e-05,
+      "loss": 1.1259,
+      "step": 465
+    },
+    {
+      "epoch": 1.2072538860103628,
+      "grad_norm": 31.34891257114439,
+      "learning_rate": 1.4989629910496424e-05,
+      "loss": 1.0733,
+      "step": 466
+    },
+    {
+      "epoch": 1.2098445595854923,
+      "grad_norm": 24.541559850584985,
+      "learning_rate": 1.4906577595175428e-05,
+      "loss": 1.1166,
+      "step": 467
+    },
+    {
+      "epoch": 1.2124352331606219,
+      "grad_norm": 20.4345832961354,
+      "learning_rate": 1.4823619097949584e-05,
+      "loss": 1.0916,
+      "step": 468
+    },
+    {
+      "epoch": 1.2150259067357512,
+      "grad_norm": 28.860712194727487,
+      "learning_rate": 1.4740755946869708e-05,
+      "loss": 1.1043,
+      "step": 469
+    },
+    {
+      "epoch": 1.2176165803108807,
+      "grad_norm": 25.71820242946282,
+      "learning_rate": 1.4657989668230363e-05,
+      "loss": 1.0949,
+      "step": 470
+    },
+    {
+      "epoch": 1.2202072538860103,
+      "grad_norm": 51.16994773097077,
+      "learning_rate": 1.4575321786541801e-05,
+      "loss": 1.141,
+      "step": 471
+    },
+    {
+      "epoch": 1.2227979274611398,
+      "grad_norm": 32.70442309640389,
+      "learning_rate": 1.4492753824501833e-05,
+      "loss": 1.1127,
+      "step": 472
+    },
+    {
+      "epoch": 1.2253886010362693,
+      "grad_norm": 21.913285172411495,
+      "learning_rate": 1.4410287302967813e-05,
+      "loss": 1.084,
+      "step": 473
+    },
+    {
+      "epoch": 1.2279792746113989,
+      "grad_norm": 34.45727214001296,
+      "learning_rate": 1.4327923740928613e-05,
+      "loss": 1.0836,
+      "step": 474
+    },
+    {
+      "epoch": 1.2305699481865284,
+      "grad_norm": 26.768013926034776,
+      "learning_rate": 1.4245664655476663e-05,
+      "loss": 1.1264,
+      "step": 475
+    },
+    {
+      "epoch": 1.233160621761658,
+      "grad_norm": 28.401965255935572,
+      "learning_rate": 1.4163511561779956e-05,
+      "loss": 1.0805,
+      "step": 476
+    },
+    {
+      "epoch": 1.2357512953367875,
+      "grad_norm": 29.19935757288793,
+      "learning_rate": 1.4081465973054216e-05,
+      "loss": 1.0825,
+      "step": 477
+    },
+    {
+      "epoch": 1.238341968911917,
+      "grad_norm": 24.55918541541201,
+      "learning_rate": 1.3999529400534941e-05,
+      "loss": 1.1164,
+      "step": 478
+    },
+    {
+      "epoch": 1.2409326424870466,
+      "grad_norm": 25.35635406268312,
+      "learning_rate": 1.3917703353449646e-05,
+      "loss": 1.1334,
+      "step": 479
+    },
+    {
+      "epoch": 1.2435233160621761,
+      "grad_norm": 45.453901005004184,
+      "learning_rate": 1.3835989338989996e-05,
+      "loss": 1.1387,
+      "step": 480
+    },
+    {
+      "epoch": 1.2461139896373057,
+      "grad_norm": 21.67852694202104,
+      "learning_rate": 1.375438886228411e-05,
+      "loss": 1.0846,
+      "step": 481
+    },
+    {
+      "epoch": 1.2487046632124352,
+      "grad_norm": 171.2474074894732,
+      "learning_rate": 1.3672903426368773e-05,
+      "loss": 1.1388,
+      "step": 482
+    },
+    {
+      "epoch": 1.2512953367875648,
+      "grad_norm": 43.18223835070906,
+      "learning_rate": 1.3591534532161781e-05,
+      "loss": 1.1483,
+      "step": 483
+    },
+    {
+      "epoch": 1.2538860103626943,
+      "grad_norm": 29.447332565856644,
+      "learning_rate": 1.3510283678434317e-05,
+      "loss": 1.07,
+      "step": 484
+    },
+    {
+      "epoch": 1.2564766839378239,
+      "grad_norm": 28.600251051615228,
+      "learning_rate": 1.3429152361783307e-05,
+      "loss": 1.0798,
+      "step": 485
+    },
+    {
+      "epoch": 1.2564766839378239,
+      "eval_loss": 1.085669755935669,
+      "eval_runtime": 38.1134,
+      "eval_samples_per_second": 19.521,
+      "eval_steps_per_second": 1.233,
+      "step": 485
+    },
+    {
+      "epoch": 1.2590673575129534,
+      "grad_norm": 47.124643074410464,
+      "learning_rate": 1.3348142076603876e-05,
+      "loss": 1.0875,
+      "step": 486
+    },
+    {
+      "epoch": 1.261658031088083,
+      "grad_norm": 42.06019726307143,
+      "learning_rate": 1.3267254315061797e-05,
+      "loss": 1.1429,
+      "step": 487
+    },
+    {
+      "epoch": 1.2642487046632125,
+      "grad_norm": 18.950734630756962,
+      "learning_rate": 1.318649056706605e-05,
+      "loss": 1.0747,
+      "step": 488
+    },
+    {
+      "epoch": 1.266839378238342,
+      "grad_norm": 31.903949502516806,
+      "learning_rate": 1.3105852320241326e-05,
+      "loss": 1.1041,
+      "step": 489
+    },
+    {
+      "epoch": 1.2694300518134716,
+      "grad_norm": 22.957473008085927,
+      "learning_rate": 1.3025341059900675e-05,
+      "loss": 1.1046,
+      "step": 490
+    },
+    {
+      "epoch": 1.2720207253886011,
+      "grad_norm": 22.325983256563678,
+      "learning_rate": 1.2944958269018103e-05,
+      "loss": 1.0643,
+      "step": 491
+    },
+    {
+      "epoch": 1.2746113989637307,
+      "grad_norm": 29.689383331974955,
+      "learning_rate": 1.2864705428201307e-05,
+      "loss": 1.0949,
+      "step": 492
+    },
+    {
+      "epoch": 1.2772020725388602,
+      "grad_norm": 25.338298442945575,
+      "learning_rate": 1.2784584015664337e-05,
+      "loss": 1.0725,
+      "step": 493
+    },
+    {
+      "epoch": 1.2797927461139897,
+      "grad_norm": 31.591732488078588,
+      "learning_rate": 1.2704595507200435e-05,
+      "loss": 1.0347,
+      "step": 494
+    },
+    {
+      "epoch": 1.2823834196891193,
+      "grad_norm": 42.96243570696118,
+      "learning_rate": 1.26247413761548e-05,
+      "loss": 1.1196,
+      "step": 495
+    },
+    {
+      "epoch": 1.2849740932642488,
+      "grad_norm": 26.559546676266024,
+      "learning_rate": 1.254502309339749e-05,
+      "loss": 1.0187,
+      "step": 496
+    },
+    {
+      "epoch": 1.2875647668393784,
+      "grad_norm": 27.58444017584016,
+      "learning_rate": 1.2465442127296297e-05,
+      "loss": 1.0985,
+      "step": 497
+    },
+    {
+      "epoch": 1.2901554404145077,
+      "grad_norm": 36.53028730423797,
+      "learning_rate": 1.2385999943689732e-05,
+      "loss": 1.068,
+      "step": 498
+    },
+    {
+      "epoch": 1.2927461139896372,
+      "grad_norm": 38.94837307599113,
+      "learning_rate": 1.2306698005859975e-05,
+      "loss": 1.0736,
+      "step": 499
+    },
+    {
+      "epoch": 1.2953367875647668,
+      "grad_norm": 36.67208266195125,
+      "learning_rate": 1.2227537774505996e-05,
+      "loss": 1.119,
+      "step": 500
+    },
+    {
+      "epoch": 1.2979274611398963,
+      "grad_norm": 31.086410648635283,
+      "learning_rate": 1.2148520707716567e-05,
+      "loss": 1.1094,
+      "step": 501
+    },
+    {
+      "epoch": 1.3005181347150259,
+      "grad_norm": 27.96977481605826,
+      "learning_rate": 1.2069648260943473e-05,
+      "loss": 1.1345,
+      "step": 502
+    },
+    {
+      "epoch": 1.3031088082901554,
+      "grad_norm": 22.89450502840197,
+      "learning_rate": 1.1990921886974669e-05,
+      "loss": 1.12,
+      "step": 503
+    },
+    {
+      "epoch": 1.305699481865285,
+      "grad_norm": 18.54206032224653,
+      "learning_rate": 1.1912343035907535e-05,
+      "loss": 1.0929,
+      "step": 504
+    },
+    {
+      "epoch": 1.3082901554404145,
+      "grad_norm": 38.9386007237313,
+      "learning_rate": 1.1833913155122132e-05,
+      "loss": 1.1381,
+      "step": 505
+    },
+    {
+      "epoch": 1.310880829015544,
+      "grad_norm": 37.05899458809635,
+      "learning_rate": 1.1755633689254609e-05,
+      "loss": 1.0535,
+      "step": 506
+    },
+    {
+      "epoch": 1.3134715025906736,
+      "grad_norm": 27.716372794195156,
+      "learning_rate": 1.1677506080170512e-05,
+      "loss": 1.1342,
+      "step": 507
+    },
+    {
+      "epoch": 1.3160621761658031,
+      "grad_norm": 40.42306246079416,
+      "learning_rate": 1.1599531766938306e-05,
+      "loss": 1.0887,
+      "step": 508
+    },
+    {
+      "epoch": 1.3186528497409327,
+      "grad_norm": 98.56681767405578,
+      "learning_rate": 1.1521712185802789e-05,
+      "loss": 1.0954,
+      "step": 509
+    },
+    {
+      "epoch": 1.3212435233160622,
+      "grad_norm": 34.42816933350743,
+      "learning_rate": 1.1444048770158718e-05,
+      "loss": 1.0512,
+      "step": 510
+    },
+    {
+      "epoch": 1.3238341968911918,
+      "grad_norm": 52.457523653614096,
+      "learning_rate": 1.136654295052433e-05,
+      "loss": 1.1599,
+      "step": 511
+    },
+    {
+      "epoch": 1.3264248704663213,
+      "grad_norm": 26.832339531661276,
+      "learning_rate": 1.1289196154515048e-05,
+      "loss": 1.0602,
+      "step": 512
+    },
+    {
+      "epoch": 1.3290155440414508,
+      "grad_norm": 32.746047673769816,
+      "learning_rate": 1.1212009806817163e-05,
+      "loss": 1.1544,
+      "step": 513
+    },
+    {
+      "epoch": 1.3316062176165804,
+      "grad_norm": 37.44483451702055,
+      "learning_rate": 1.1134985329161608e-05,
+      "loss": 1.1421,
+      "step": 514
+    },
+    {
+      "epoch": 1.33419689119171,
+      "grad_norm": 28.625976525737606,
+      "learning_rate": 1.1058124140297718e-05,
+      "loss": 1.0858,
+      "step": 515
+    },
+    {
+      "epoch": 1.3367875647668392,
+      "grad_norm": 38.64141195246213,
+      "learning_rate": 1.0981427655967183e-05,
+      "loss": 1.0983,
+      "step": 516
+    },
+    {
+      "epoch": 1.3393782383419688,
+      "grad_norm": 29.989753893533425,
+      "learning_rate": 1.0904897288877891e-05,
+      "loss": 1.1269,
+      "step": 517
+    },
+    {
+      "epoch": 1.3419689119170983,
+      "grad_norm": 48.63990665515511,
+      "learning_rate": 1.0828534448677942e-05,
+      "loss": 1.0844,
+      "step": 518
+    },
+    {
+      "epoch": 1.3445595854922279,
+      "grad_norm": 25.477227318250847,
+      "learning_rate": 1.0752340541929711e-05,
+      "loss": 1.0742,
+      "step": 519
+    },
+    {
+      "epoch": 1.3471502590673574,
+      "grad_norm": 26.363588814537763,
+      "learning_rate": 1.0676316972083867e-05,
+      "loss": 1.0533,
+      "step": 520
+    },
+    {
+      "epoch": 1.349740932642487,
+      "grad_norm": 34.59968737708606,
+      "learning_rate": 1.060046513945361e-05,
+      "loss": 1.0983,
+      "step": 521
+    },
+    {
+      "epoch": 1.3523316062176165,
+      "grad_norm": 52.51652561846762,
+      "learning_rate": 1.0524786441188786e-05,
+      "loss": 1.1319,
+      "step": 522
+    },
+    {
+      "epoch": 1.354922279792746,
+      "grad_norm": 21.360221214301127,
+      "learning_rate": 1.0449282271250239e-05,
+      "loss": 1.0627,
+      "step": 523
+    },
+    {
+      "epoch": 1.3575129533678756,
+      "grad_norm": 37.00053933682603,
+      "learning_rate": 1.0373954020384073e-05,
+      "loss": 1.096,
+      "step": 524
+    },
+    {
+      "epoch": 1.3601036269430051,
+      "grad_norm": 39.212240822687484,
+      "learning_rate": 1.029880307609608e-05,
+      "loss": 1.0512,
+      "step": 525
+    },
+    {
+      "epoch": 1.3626943005181347,
+      "grad_norm": 24.89842378385804,
+      "learning_rate": 1.0223830822626124e-05,
+      "loss": 1.0538,
+      "step": 526
+    },
+    {
+      "epoch": 1.3652849740932642,
+      "grad_norm": 29.14416894424653,
+      "learning_rate": 1.0149038640922715e-05,
+      "loss": 1.1538,
+      "step": 527
+    },
+    {
+      "epoch": 1.3678756476683938,
+      "grad_norm": 31.688722122648855,
+      "learning_rate": 1.0074427908617515e-05,
+      "loss": 1.171,
+      "step": 528
+    },
+    {
+      "epoch": 1.3704663212435233,
+      "grad_norm": 41.918909004413734,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 1.1203,
+      "step": 529
+    },
+    {
+      "epoch": 1.3730569948186528,
+      "grad_norm": 26.70963454516576,
+      "learning_rate": 9.92575628599213e-06,
+      "loss": 1.0855,
+      "step": 530
+    },
+    {
+      "epoch": 1.3756476683937824,
+      "grad_norm": 24.819351173466824,
+      "learning_rate": 9.851698134123095e-06,
+      "loss": 1.0972,
+      "step": 531
+    },
+    {
+      "epoch": 1.378238341968912,
+      "grad_norm": 22.100465399566815,
+      "learning_rate": 9.777826908504126e-06,
+      "loss": 1.08,
+      "step": 532
+    },
+    {
+      "epoch": 1.3808290155440415,
+      "grad_norm": 29.31574709406259,
+      "learning_rate": 9.704143969803392e-06,
+      "loss": 1.0835,
+      "step": 533
+    },
+    {
+      "epoch": 1.383419689119171,
+      "grad_norm": 25.551326748473052,
+      "learning_rate": 9.630650675220892e-06,
+      "loss": 1.0396,
+      "step": 534
+    },
+    {
+      "epoch": 1.3860103626943006,
+      "grad_norm": 59.07595627892596,
+      "learning_rate": 9.557348378463503e-06,
+      "loss": 1.0814,
+      "step": 535
+    },
+    {
+      "epoch": 1.38860103626943,
+      "grad_norm": 24.96501978981908,
+      "learning_rate": 9.484238429720018e-06,
+      "loss": 1.0187,
+      "step": 536
+    },
+    {
+      "epoch": 1.3911917098445596,
+      "grad_norm": 42.530604702279234,
+      "learning_rate": 9.411322175636298e-06,
+      "loss": 1.074,
+      "step": 537
+    },
+    {
+      "epoch": 1.3937823834196892,
+      "grad_norm": 34.91129065632851,
+      "learning_rate": 9.338600959290414e-06,
+      "loss": 1.0878,
+      "step": 538
+    },
+    {
+      "epoch": 1.3963730569948187,
+      "grad_norm": 32.07525956876426,
+      "learning_rate": 9.266076120167992e-06,
+      "loss": 1.0962,
+      "step": 539
+    },
+    {
+      "epoch": 1.3989637305699483,
+      "grad_norm": 40.18387743296675,
+      "learning_rate": 9.193748994137462e-06,
+      "loss": 1.1033,
+      "step": 540
+    },
+    {
+      "epoch": 1.4015544041450778,
+      "grad_norm": 66.68031460980451,
+      "learning_rate": 9.121620913425508e-06,
+      "loss": 1.1466,
+      "step": 541
+    },
+    {
+      "epoch": 1.4041450777202074,
+      "grad_norm": 34.07506059584738,
+      "learning_rate": 9.04969320659249e-06,
+      "loss": 1.1184,
+      "step": 542
+    },
+    {
+      "epoch": 1.406735751295337,
+      "grad_norm": 17.130845779169075,
+      "learning_rate": 8.977967198508001e-06,
+      "loss": 1.0803,
+      "step": 543
+    },
+    {
+      "epoch": 1.4093264248704664,
+      "grad_norm": 22.4457025132615,
+      "learning_rate": 8.906444210326441e-06,
+      "loss": 1.0745,
+      "step": 544
+    },
+    {
+      "epoch": 1.411917098445596,
+      "grad_norm": 73.43971735356851,
+      "learning_rate": 8.83512555946271e-06,
+      "loss": 1.0717,
+      "step": 545
+    },
+    {
+      "epoch": 1.4145077720207253,
+      "grad_norm": 38.16321297719761,
+      "learning_rate": 8.764012559567899e-06,
+      "loss": 1.1371,
+      "step": 546
+    },
+    {
+      "epoch": 1.4170984455958548,
+      "grad_norm": 56.14718024907725,
+      "learning_rate": 8.693106520505147e-06,
+      "loss": 1.0185,
+      "step": 547
+    },
+    {
+      "epoch": 1.4196891191709844,
+      "grad_norm": 53.3812598790062,
+      "learning_rate": 8.622408748325461e-06,
+      "loss": 1.0859,
+      "step": 548
+    },
+    {
+      "epoch": 1.422279792746114,
+      "grad_norm": 39.69041631433326,
+      "learning_rate": 8.551920545243704e-06,
+      "loss": 1.1146,
+      "step": 549
+    },
+    {
+      "epoch": 1.4248704663212435,
+      "grad_norm": 24.099260758984773,
+      "learning_rate": 8.481643209614576e-06,
+      "loss": 1.0968,
+      "step": 550
+    },
+    {
+      "epoch": 1.427461139896373,
+      "grad_norm": 22.623850373369237,
+      "learning_rate": 8.411578035908728e-06,
+      "loss": 1.0642,
+      "step": 551
+    },
+    {
+      "epoch": 1.4300518134715026,
+      "grad_norm": 25.343746374404027,
+      "learning_rate": 8.341726314688875e-06,
+      "loss": 1.0815,
+      "step": 552
+    },
+    {
+      "epoch": 1.432642487046632,
+      "grad_norm": 35.82641011588973,
+      "learning_rate": 8.272089332586089e-06,
+      "loss": 1.1012,
+      "step": 553
+    },
+    {
+      "epoch": 1.4352331606217616,
+      "grad_norm": 24.81161215784662,
+      "learning_rate": 8.20266837227603e-06,
+      "loss": 1.1086,
+      "step": 554
+    },
+    {
+      "epoch": 1.4378238341968912,
+      "grad_norm": 54.18243481591251,
+      "learning_rate": 8.133464712455364e-06,
+      "loss": 1.0704,
+      "step": 555
+    },
+    {
+      "epoch": 1.4404145077720207,
+      "grad_norm": 23.602598217141395,
+      "learning_rate": 8.064479627818213e-06,
+      "loss": 1.1519,
+      "step": 556
+    },
+    {
+      "epoch": 1.4430051813471503,
+      "grad_norm": 31.124404868409982,
+      "learning_rate": 7.995714389032638e-06,
+      "loss": 1.0705,
+      "step": 557
+    },
+    {
+      "epoch": 1.4455958549222798,
+      "grad_norm": 24.14171016995626,
+      "learning_rate": 7.927170262717284e-06,
+      "loss": 1.1083,
+      "step": 558
+    },
+    {
+      "epoch": 1.4481865284974094,
+      "grad_norm": 47.987203109917175,
+      "learning_rate": 7.858848511417998e-06,
+      "loss": 1.0836,
+      "step": 559
+    },
+    {
+      "epoch": 1.450777202072539,
+      "grad_norm": 25.871447098066056,
+      "learning_rate": 7.790750393584616e-06,
+      "loss": 1.0787,
+      "step": 560
+    },
+    {
+      "epoch": 1.4533678756476685,
+      "grad_norm": 23.820249113937482,
+      "learning_rate": 7.72287716354776e-06,
+      "loss": 1.1165,
+      "step": 561
+    },
+    {
+      "epoch": 1.455958549222798,
+      "grad_norm": 48.04131308947624,
+      "learning_rate": 7.65523007149575e-06,
+      "loss": 1.0819,
+      "step": 562
+    },
+    {
+      "epoch": 1.4585492227979275,
+      "grad_norm": 29.273494083692352,
+      "learning_rate": 7.587810363451544e-06,
+      "loss": 1.0302,
+      "step": 563
+    },
+    {
+      "epoch": 1.4611398963730569,
+      "grad_norm": 120.01571222366722,
+      "learning_rate": 7.5206192812498345e-06,
+      "loss": 1.1291,
+      "step": 564
+    },
+    {
+      "epoch": 1.4637305699481864,
+      "grad_norm": 33.16947662083338,
+      "learning_rate": 7.4536580625141244e-06,
+      "loss": 1.0842,
+      "step": 565
+    },
+    {
+      "epoch": 1.466321243523316,
+      "grad_norm": 29.979556378166713,
+      "learning_rate": 7.386927940633981e-06,
+      "loss": 1.1116,
+      "step": 566
+    },
+    {
+      "epoch": 1.4689119170984455,
+      "grad_norm": 27.172344859281896,
+      "learning_rate": 7.32043014474227e-06,
+      "loss": 1.0676,
+      "step": 567
+    },
+    {
+      "epoch": 1.471502590673575,
+      "grad_norm": 30.208548637757318,
+      "learning_rate": 7.254165899692554e-06,
+      "loss": 1.1104,
+      "step": 568
+    },
+    {
+      "epoch": 1.4740932642487046,
+      "grad_norm": 19.385421184583773,
+      "learning_rate": 7.188136426036498e-06,
+      "loss": 1.0085,
+      "step": 569
+    },
+    {
+      "epoch": 1.4766839378238341,
+      "grad_norm": 30.350787749309685,
+      "learning_rate": 7.12234294000143e-06,
+      "loss": 1.0584,
+      "step": 570
+    },
+    {
+      "epoch": 1.4792746113989637,
+      "grad_norm": 31.520305600900198,
+      "learning_rate": 7.056786653467882e-06,
+      "loss": 1.0831,
+      "step": 571
+    },
+    {
+      "epoch": 1.4818652849740932,
+      "grad_norm": 46.13006972574487,
+      "learning_rate": 6.991468773947321e-06,
+      "loss": 1.1761,
+      "step": 572
+    },
+    {
+      "epoch": 1.4844559585492227,
+      "grad_norm": 26.72340868362835,
+      "learning_rate": 6.926390504559879e-06,
+      "loss": 1.0605,
+      "step": 573
+    },
+    {
+      "epoch": 1.4870466321243523,
+      "grad_norm": 25.992965411102556,
+      "learning_rate": 6.861553044012206e-06,
+      "loss": 1.1015,
+      "step": 574
+    },
+    {
+      "epoch": 1.4896373056994818,
+      "grad_norm": 38.60187420279626,
+      "learning_rate": 6.796957586575364e-06,
+      "loss": 1.1232,
+      "step": 575
+    },
+    {
+      "epoch": 1.4922279792746114,
+      "grad_norm": 21.7618591565717,
+      "learning_rate": 6.732605322062869e-06,
+      "loss": 1.1196,
+      "step": 576
+    },
+    {
+      "epoch": 1.494818652849741,
+      "grad_norm": 28.233093007170996,
+      "learning_rate": 6.668497435808736e-06,
+      "loss": 1.1451,
+      "step": 577
+    },
+    {
+      "epoch": 1.4974093264248705,
+      "grad_norm": 28.061514297823816,
+      "learning_rate": 6.604635108645683e-06,
+      "loss": 1.0832,
+      "step": 578
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 35.34503147975386,
+      "learning_rate": 6.5410195168833425e-06,
+      "loss": 1.118,
+      "step": 579
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 772,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 193,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0022991232499712e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}